Abstract

Objective: Annexins are Ca2+-dependent phospholipid binding proteins. Externalized annexin A5 has been recently suggested to have a proapoptotic effect. Our aim was to determine whether annexin A5, which is intracellular in cardiomyocytes, could be translocated and/or externalized and play a role during the apoptotic process.

Methods: Apoptosis was induced in rat cardiomyocytes by continuous incubation with staurosporine or 30 min treatment with H2O2 and was measured by phosphatidylserine (PS) externalization, TUNEL staining and DNA ladder. Immunofluorescence labeling of annexin A5 was performed on permeabilized or nonpermeabilized cardiomyocytes.

Results: Staurosporine or H2O2 treatment of neonatal cardiomyocytes resulted in significant increases of apoptosis at 24 h, but H2O2 treatment led to a faster and higher PS externalization than that observed with ST. In both neonatal and adult cardiomyocytes, annexin A5 was intracellular in control conditions but was found at the external face of sarcolemma during apoptosis. Furthermore, neonatal cardiomyocytes with externalized annexin A5 have apoptotic characteristics and their number increased with time. Interestingly, immediately after H2O2 induction, the number of annexin A5-positive cells was higher than that of PS-positive cells (p≤0.05) and colabeling showed that half annexin A5-positive cells were PS-negative. We further demonstrated by immunoblots that free annexin A5 was absent from the media and could not be released from cardiomyocytes by washes at 1.8 mM Ca2+. Removing annexin A5 by Ca2+-free washes 15 or 30 min after H2O2 treatment or blocking externalized annexin A5 by antibodies lead to a significant decrease of apoptotic cardiomyocytes, cytochrome c release and caspase 3 activity.

Conclusion: This study indicated for the first time that annexin A5 was externalized at a very early stage of apoptosis and could have a proapoptotic effect in cardiomyocytes.

Introduction

Annexins are a family of proteins capable of binding in a Ca2+-dependent manner to negatively charged phospholipids [1]. Among annexins, annexin A5 exhibited a high affinity for phosphatidylserine (PS) and has been reported to detect in vitro and in vivo the externalization of PS, hallmark of cells undergoing apoptosis [2–5]. However, the properties of endogenous, cellular, annexin A5 are still largely unknown. Annexin A5 which was more often considered as a major intracellular Ca2+-binding protein [6] has been reported to be translocated from cytosol to plasma or nuclear membranes after a rise in intracellular Ca2+ concentration [7–9]. It is known for inducing Ca2+ influx by forming Ca2+ channels [10] and for inhibiting enzyme activities linked to Ca2+ activation such as phospholipase A2 and protein kinase C [11,12]. Annexin A5 has been thus suggested to mediate Ca2+ signalization, cell cycle regulation, signal transduction, membrane trafficking and organization [13,14]. Furthermore, recent lines of evidence indicated that cellular annexin A5 could be a participant in modulating apoptosis. Indeed, Hawkins et al. [15] demonstrated that DT40 cells lacking annexin A5 are resistant to Ca2+-dependent apoptosis and Wang et al. [16] reported that inhibition of the annexin A5 mediated Ca2+ influx reduced apoptosis in chondrocytes.

In normal myocardium, we and others have previously shown that annexin A5 was found in cardiomyocytes, at the level of sarcolemma, intercalated discs and T-tubules [17–20]. However, in failing myocardium, a pathological situation associated with increased cardiomyocyte apoptosis, we have reported an increased expression and a translocation of endogenous annexin A5 to the interstitial tissue [20]. Therefore, according to the central role for Ca2+ in the signaling pathway of apoptosis [21,22] and the annexin A5 properties, the question arises whether annexin A5 translocation or externalization could occur and play a role during the apoptotic process in cardiomyocytes. To test this hypothesis, we first analyzed the cellular distribution of endogenous annexin A5 during apoptosis of isolated rat cardiomyocytes. We reported for the first time that in neonatal and adult cardiomyocytes annexin A5 was translocated and externally expressed during the early phase of apoptosis. Furthermore, we demonstrated that washout of externalized annexin A5 or antiannexin A5 antibodies were able to reduce apoptosis at a stage upstream of mitochondrial activation. Taken together, these findings indicated that externalization of annexin A5 is likely modulating apoptosis mechanism in cardiomyocytes.

Materials and methods

The investigation conforms with the Guide for the care and use of laboratory animals published by the US National Institutes of Health (NIH publication No. 85-23, revised 1996).

Isolation of cardiomyocytes

Monolayer cultures of neonatal rat cardiac cells were prepared as described [23]. Briefly, hearts from 1- to 2-day-old Wistar rats were removed and ventricular cells were dispersed by digestion with collagenase A (0.45 mg/ml) (Boehringer Mannheim) and pancreatin (0.05 mg/ml) (GIBCO) in Ads buffer (116 mM NaCl, 20 mM HEPES, 1 mM NaH2PO4, 5.4 mM KCl, 5.5 mM glucose, 0.8 mM MgSO4, pH 7.35). Cardiomyocytes were purified on a discontinuous Percoll gradient (1.059/1.085) collected by centrifugation at 3000×g and resuspended (0.33×106 cells/ml) in Dulbecco's modified Eagle medium (DMEM, 1.8 mM Ca2+), 17% Medium 199 (GIBCO), 10% horse serum, 5% newborn calf serum, 1% penicillin and 1% streptomycin. After 24 h, horse serum was decreased to 5% and newborn calf serum to 0.5% for 48 h then removed for 24 h.

Adult ventricular cardiomyocytes were isolated from rat (200–250 g) hearts subjected to retrograde perfusion in a Langendorff apparatus with Krebs' solution containing collagenase (Boehringer Mannheim), and hyaluronidase (Sigma-Aldrich) for approximately 15 min [24]. After several centrifugation (10×g, 3 min) and sedimentation (5 min) steps, cardiomyocytes were plated at 0.15×106 cells/ml. The culture medium was renewed on day 1 and cells were subjected to apoptosis stimulation.

Induction of apoptosis and protection assays

Cardiomyocytes from neonatal rats were incubated in serum-free medium (SFM) with (1) 0.6 or 3 μM staurosporine (ST) for 30 min to 24 h [25]; (2) 0.075 mM H2O2+0.1 mM FeSO4 for 30 min, washed and incubated for 24 h in SFM [26]. Cardiomyocytes from adult rats were incubated with (1) 0.075 mM H2O2+0.1 mM FeSO4 for 30 min, washed and incubated for 24 h in SFM; (2) 50 μM C2-ceramide for 3 h [24]; (3) 0.6 μM ST.

Protection assays were performed on neonatal cardiomyocytes by washout of externalized annexin A5 with Ca2+-free washes (2×3 min) at 15 and 30 min after H2O2 treatment or by treatment with antiannexin A5 antibody added to the medium during apoptosis development.

Cell counts

Approximately 2000 cells were counted per dish (one field containing 200–250 neonatal cardiomyocytes and 80–130 adult cardiomyocytes). Cells were counterstained with Hoechst 33342 (1 μg/ml, Sigma) to detect condensation and chromatin fragmentation of the nuclei and to count the cell number. Propidium iodide exclusion assay (PI, 1 μg/ml) was used to estimate membrane leakiness. Necrotic cells that have PI-positive nuclei were not scored; they never exceeded 1% total cell number (data not shown). Immunofluorescent images were obtained using a DMR P Leica microscope with a JVC color video camera KY-F50, and were counted by coupling to an imaging analysis system (Histolab software, Microvision, France).

Detection of apoptosis

PS externalization

Externalized PS staining was performed with annexin A5-FITC [3] (Annexin A5-Fluos, 1/50, Boehringer Mannheim) according to the manufacturer's protocol.

TUNEL staining

Terminal deoxyribonucleotide transferase (TdT)-mediated dUTP nick end-labeling was performed according to the manufacturer's protocol (In Situ Cell Death Detection Kit, fluorescein; Roche diagnostics). Positive controls were obtained after DNase treatment of the cultured cells.

DNA ladder

DNA was isolated from two dishes according to Hermann et al. [27] and electrophoresis was carried out in 1.5% agarose gels containing ethidium bromide.

Caspase 3 activity

Caspase 3 activity was measured in cell lysate according to the protocol of the commercially available kit BIOMOL Quantizyme assay System Caspase-3 cellular activity). Results were expressed as fold change over control condition.

Immunodetection of cytochrome c

Cytochrome c detection was performed according to Vander Heiden et al. [28]. The cell culture was incubated on ice for 10 min with 0.07% saponin to permeabilize the sarcolemma, scraped then homogenized by five strokes in a Dounce homogenizer. Subcellular fractions were prepared by differential centrifugations as previously described. The release of cytochrome c from mitochondria was measured on concentrated cytosolic extracts by immunoblotting with anticytochrome c rabbit polyclonal antibody (Santa Cruz). Subunit VIb of cytochrome c oxidase (COX; a mitochondrial marker protein) was detected by immunoblotting with anti-COX VIb monoclonal antibody (Molecular Probes). Results were expressed as fold change over control condition.

Localization of annexin A5

Endogeneous annexin A5 labeling in permeabilized cells

Cells were fixed, permeabilized and sequentially incubated with 5% BSA in PBS for 30 min at RT, with polyclonal antibodies specific for annexin A5 [20] and with anti α-actinin monoclonal antibody (1/20 in PBS-2% BSA for 30 min at 37°C). Cardiomyocytes were visualized using a BioRad MRC 600 confocal microscope with a Nikon ×40 oil immersion objective.

Externalized annexin A5 labeling of intact cells

Cells were sequentially incubated with antiannexin A5 antibody (1/20 in SFM containing 2% BSA and 1.8 mM Ca2+ for 20 min at 37 °C), fixed with 2.5% PFA for 15 min at RT and incubated with antirabbit IgG antibody conjugated to FITC. To attest the membrane integrity, propidium iodide was added with the first antibody. No labeling was detected if antiannexin A5 was omitted.

Double labeling of annexin A5 and PS of intact cells

Cells were sequentially incubated with antiannexin A5 antibody and annexin A5-FITC as stated above. After a short washing, cells were fixed with 2.5% PFA and incubated with antirabbit IgG antibody conjugated to Texas Red (1/50 in PBS containing 1/25 serum of rat). To assure that each binding was specific and that the cell membranes kept their integrity under these conditions, controls were performed in which all the incubations and washes were maintained except that either antiannexin A5 antibody or annexin A5-FITC was omitted.

Ca2+ sensitivity of externalized annexin A5

The conditioned media and washes with Ca2+-free medium or 1.8 mM Ca2+ HEPES buffer were centrifuged (400×g, 5 min) and supernatants concentrated 100 times in presence of 4 mM EDTA, 1 mM EGTA and inhibitors of proteases, using microconcentrators (Microsep 10 KD, Filtron Technology Corporation, 7500g). They were subjected to 10% SDS-PAGE, transferred to PVDF membrane (Amersham Pharmacia Biotech) and incubated with antiannexin A5 antibody (1/2500) for 2 h and with horseradish peroxidase-conjugated antirabbit antisera (Amersham) (1/5000). Peroxidase activity was detected with ECL+Plus Detection System (Amersham) according to the manufacturer's instruction.

Statistical analysis

All values are presented as mean±S.E.M. of at least seven independent experiments unless stated otherwise. Comparisons between means were made using ANOVA analysis and Student's t-test with the Bonferroni correction for multiple comparisons when appropriate. A value of p<0.05 was considered to be statistically significant.

Results

Cardiomyocyte apoptosis

To determine whether endogenous annexin A5 was externalized during the apoptotic process and at what phase, we performed a time course study of early apoptosis events and studied annexin A5 localization in cardiomyocytes. Most of the study has been carried out on neonatal cardiomyocytes after apoptosis induction with either ST or H2O2 as a representative oxidant [25,26]. These apoptotic stimuli were chosen because they both engaged the intrinsic mitochondrial pathway involving ER Ca2+ gateway [21] but H2O2 treatment differs by a limited induction period of 30 min.

As reported previously [25,26], ST (0.6 μM) or H2O2 (0.075 mM) treatment induced a typical apoptosis pattern in neonatal cardiomyocytes with externalization of PS and punctated labeling and blebbing of the membrane associated with a fragmented chromatin (Fig. 1A). Both ST and H2O2 led to time-dependent increases of PS-positive cells and TUNEL-positive cells (Fig. 1B). Likewise, specific DNA ladders were seen only in ST- and H2O2-treated cardiomyocytes (Fig. 1C). Important to note was the faster PS externalization with H2O2 than with ST and the delayed DNA fragmentation.

Fig. 1

Time course of apoptosis induced with staurosporine or H2O2. Apoptosis was induced in neonatal cardiomyocytes with 0.6 μM staurosporine or with 0.075 mM H202 for 30 min. (A) Double labeling of externalized PS (a, b, c) and Hoechst 33342 (d, e, f) of cardiomyocytes in control condition (a and d) and after 4 h of ST (b and e) or H202 (c and f) treatment. Typical simultaneous externalization of PS and nuclear chromatin condensation or fragmentation in apoptotic cardiomyocytes was observed. Bar: 20 μm. (B) Time course of PS externalization (●) and TUNEL-positive cardiomyocytes (▄) after ST or H202 treatment (n≤11 wells obtained from five separate experiments for each assay). *p≤0.05, ***p≤0.001 significant from PS control; $p≤0.05, $$$p≤0.001 significant from TUNEL control. Percentages of spontaneous PS-positive or TUNEL-positive cardiomyocytes under control conditions remained similar whatever the incubation time. They were represented at the zero time. (C) DNA ladder in control cells (C) and after ST or H202 treatment. Molecular weight markers (MW).

Fig. 1

Time course of apoptosis induced with staurosporine or H2O2. Apoptosis was induced in neonatal cardiomyocytes with 0.6 μM staurosporine or with 0.075 mM H202 for 30 min. (A) Double labeling of externalized PS (a, b, c) and Hoechst 33342 (d, e, f) of cardiomyocytes in control condition (a and d) and after 4 h of ST (b and e) or H202 (c and f) treatment. Typical simultaneous externalization of PS and nuclear chromatin condensation or fragmentation in apoptotic cardiomyocytes was observed. Bar: 20 μm. (B) Time course of PS externalization (●) and TUNEL-positive cardiomyocytes (▄) after ST or H202 treatment (n≤11 wells obtained from five separate experiments for each assay). *p≤0.05, ***p≤0.001 significant from PS control; $p≤0.05, $$$p≤0.001 significant from TUNEL control. Percentages of spontaneous PS-positive or TUNEL-positive cardiomyocytes under control conditions remained similar whatever the incubation time. They were represented at the zero time. (C) DNA ladder in control cells (C) and after ST or H202 treatment. Molecular weight markers (MW).

Besides, we determined that (1) the number of adherent cells was similar in apoptosis and control conditions and (2) the percentage of spontaneous apoptotic cells under control conditions remained similar whatever the incubation time (data not shown).

Localization of endogenous annexin A5 in permeabilized cells

Confocal microscopy analysis showed that annexin A5 was present in neonatal cardiomyocytes identified by sarcomeric α-actinin labeling. Under control conditions, annexin A5 exhibited mainly a cytosolic distribution whereas sarcomeric α-actinin showed an expected cross-striation (Fig. 2a and b, respectively). Under ST treatment, both endogenous annexin A5 and sarcomeric α-actinin were found to be condensed in cells showing the classic morphologic changes of apoptosis such as cell shrinkage, or membrane blebbing (Fig. 2c and d, respectively). Interestingly, as shown in the inset, we have found a specific sarcolemmal relocalization of annexin A5 restricted to ST-treated cardiomyocytes.

Fig. 2

Confocal microscopy images of annexin A5 and sarcomeric α-actinin. Immunofluorescent labeling of annexin A5 (a and c) and α-actinin (b and d) in control conditions (a and b) and in 0.6 μM ST-treated cells at 4 h (c and d). Typical fields showing translocation of annexin A5 to the membrane in ST-treated cardiomyocytes. Insets correspond to 2.8-fold magnification of the rectangle zones. Bar: 20 μm.

Fig. 2

Confocal microscopy images of annexin A5 and sarcomeric α-actinin. Immunofluorescent labeling of annexin A5 (a and c) and α-actinin (b and d) in control conditions (a and b) and in 0.6 μM ST-treated cells at 4 h (c and d). Typical fields showing translocation of annexin A5 to the membrane in ST-treated cardiomyocytes. Insets correspond to 2.8-fold magnification of the rectangle zones. Bar: 20 μm.

Externalization of annexin A5 in nonpermeabilized cells

To investigate whether the sarcolemmal localization of annexin A5 observed above corresponded to externalization and was an overall feature independent of the proapoptotic trigger or cell maturity, annexin A5 immunolabeling of nonpermeabilized cells was performed on neonatal and adult cardiomyocytes after ST, H2O2, or ceramide treatment. After ST or H2O2 treatment of neonatal cardiomyocytes, externalized annexin A5 was concentrated in dots and blebbing of the sarcolemma in cells with or without chromatin fragmentation (Fig. 3A and B). Interestingly, externalized annexin A5 was also present in adult rat cardiomyocytes after H2O2, ceramide (Fig. 3C) or staurosporine (not shown) treatments. The number of neonatal cardiomyocytes with externalized annexin A5 was significantly increased as early as 30 min after ST addition (6.6±0.6% vs. 3.6±0.4% in control cells; p≤0.01) and reached 25±4.2% after 18 h (Fig. 3D). This increase was both faster and higher after H2O2 induction (21.1±3.6% after 30 min and 30±3.7% after 18 h) (Fig. 3D). In adult cardiomyocytes, we also found that annexin A5 externalization after H2O2 induction was faster and higher than with staurosporine (12.5±0.2% and 9.1±1.9% at 1 h, respectively, p≤0.05 vs. control; 30.2±2.12%, vs. 18.4±0.5% at 4 h, respectively; p≤0.05).

Fig. 3

Externalization of annexin A5 in neonatal and adult cardiomyocytes during apoptosis. Intact cells were incubated with antiannexin A5, Hoechst 33342 and propidium iodide before fixation then incubated with secondary antibody. (A) Typical fields showing external localization of annexin A5 in neonatal control cardiomyocytes (a), after 30 min (b), and 4 h (c) of 0.6 μM ST treatment and corresponding nuclei (d, e, f). Bar: 50 μm. (B) Typical fields showing external localization of annexin A5 in neonatal control cardiomyocytes (a), after 30 min (b), and 4 h (c) of H2O2 treatment and corresponding nuclei (d, e, f). Bar: 50 μm. (C) Typical fields showing external localization of annexin A5 in adult control cardiomyocytes (a), after 4 h of H2O2 treatment (b), and 3 h of ceramide treatment (c) and corresponding nuclei (d, e, f). Bar: 20 μm. (D) Time course of annexin A5 externalization after ST or H202 treatment (n=7 wells obtained from five separate experiments for each condition). Percentages of spontaneous annexin A5-positive cardiomyocytes under control conditions remained similar whatever the incubation time. They were represented at the zero time. **p≤0.01, ***p≤0.001 significant from control.

Fig. 3

Externalization of annexin A5 in neonatal and adult cardiomyocytes during apoptosis. Intact cells were incubated with antiannexin A5, Hoechst 33342 and propidium iodide before fixation then incubated with secondary antibody. (A) Typical fields showing external localization of annexin A5 in neonatal control cardiomyocytes (a), after 30 min (b), and 4 h (c) of 0.6 μM ST treatment and corresponding nuclei (d, e, f). Bar: 50 μm. (B) Typical fields showing external localization of annexin A5 in neonatal control cardiomyocytes (a), after 30 min (b), and 4 h (c) of H2O2 treatment and corresponding nuclei (d, e, f). Bar: 50 μm. (C) Typical fields showing external localization of annexin A5 in adult control cardiomyocytes (a), after 4 h of H2O2 treatment (b), and 3 h of ceramide treatment (c) and corresponding nuclei (d, e, f). Bar: 20 μm. (D) Time course of annexin A5 externalization after ST or H202 treatment (n=7 wells obtained from five separate experiments for each condition). Percentages of spontaneous annexin A5-positive cardiomyocytes under control conditions remained similar whatever the incubation time. They were represented at the zero time. **p≤0.01, ***p≤0.001 significant from control.

Moreover, it is important to note that immediately after H2O2 induction the number of neonatal cardiomyocytes with externalized annexin A5 was significantly higher than that of PS-positive cells (21.1±3.6% and 15.6±1.8% at 30 min, p≤0.05).

Externalization of annexin A5 in apoptotic cardiomyocytes

To determine whether externalization of annexin A5 was restricted to apoptotic cardiomyocytes, annexin A5 and PS were sequentially colabeled on nonpermeabilized cells during apoptosis process induced with either ST or H2O2. At 2 h of ST treatment (Fig. 4A) or 4 h after H2O2 treatment (Fig. 4B), all PS-positive neonatal cardiomyocytes exhibited annexin A5 staining. The percentage of double-labeled cells was similar to that previously described for individual labeling of annexin A5 and PS. Interestingly, as illustrated in the merge image, annexin A5 and PS localizations were either superimposed or distinct within the same cell suggesting that annexin A5 binding was not exclusively occurring on PS. This assumption was verified by colabeling experiments performed 30 min after H2O2 induction (Fig. 4C). At this early phase of apoptosis, annexin A5 was found externalized in PS-negative cardiomyocytes, in agreement with the significant higher number of neonatal cardiomyocytes with externalized annexin A5 than with externalized PS, as reported above. It is worthy to note that this single annexin A5 labeling was barely observed in ST-treated cardiomyocytes, likely because of the different kinetics of the apoptotic process (Fig. 1B).

Fig. 4

Colabeling of annexin A5 and PS in neonatal cardiomyocytes after apoptosis induction with ST or H2O2. Intact cells were successively incubated with antiannexin A5, Hoechst 33342 and propidium iodide then with annexin V-FITC, fixed and treated as indicated in experimental procedures. Typical fields showed colabeling of externalized annexin A5 in red (a), PS (b) Hoechst 33342 (c) and merge (d). Propidium iodide labeling was negative and not shown. Assays were performed after 2 h of ST treatment (A), 4 h after H2O2 induction (B), and 30 min after H2O2 induction (C). Bar: 20 μm.

Fig. 4

Colabeling of annexin A5 and PS in neonatal cardiomyocytes after apoptosis induction with ST or H2O2. Intact cells were successively incubated with antiannexin A5, Hoechst 33342 and propidium iodide then with annexin V-FITC, fixed and treated as indicated in experimental procedures. Typical fields showed colabeling of externalized annexin A5 in red (a), PS (b) Hoechst 33342 (c) and merge (d). Propidium iodide labeling was negative and not shown. Assays were performed after 2 h of ST treatment (A), 4 h after H2O2 induction (B), and 30 min after H2O2 induction (C). Bar: 20 μm.

Origin of the external annexin A5 and its binding sensitivity to Ca2+

It is well known that annexin A5 binding to phospholipids is a Ca2+-dependent mechanism with half-maximum binding at 1 μM [14]. Therefore, under the conditions of this study (1.8 mM Ca2+) annexin A5 must remained bound to membrane phospholipids. To investigate this assumption, 3 μM ST incubations were interrupted at 2 h by Ca2+-free or 1.8 mM Ca2+ washes (Fig. 5A). As shown by immunoblotting (Fig. 5B), annexin A5 was almost undetectable in the conditionned medium from control and ST-treated cardiomyocytes or in 1.8 mM. Ca2+ washes, whereas only the absence of Ca2+ in the washes can force the release of externalized annexin A5, particularly evidenced in ST-treated cells. Interestingly, the apoptotic indexes at 2 and 4 h were not altered by the presence or the absence of Ca2+ in washes (Fig. 5C). In contrast, the number of cells with externalized annexin A5 decreased to basal level immediately after the Ca2+-free washes whereas it was not modified by 1.8 mM Ca2+ washes. At 4 h, likely due to the presence of ST in the medium, annexin A5 externalization resumed but remained lower than that of PS. In agreement with this data, colabeling of PS and annexin A5 showed that half of the PS-positive cells were annexin A5-negative at 4 h (Fig 5D). These results represent strong arguments against the presence of free annexin A5 in the medium and for an externalization process of annexin A5 independent from that of PS.

Fig. 5

Effects of Ca2+-free washes on externalized annexin A5 and apoptosis induced by ST. (A) Time course protocol: incubations for 4 h in control or 3 μM ST media were interrupted after 2 h by washes performed in presence or absence of Ca2+. (B) Western blot analysis of annexin A5 in control medium (CM), 3 μM ST medium (STM), 1.8 mM Ca2+ or Ca2+-free washes of control cells (C) and of ST treated cells (ST) after 2 hours incubation. Representative results from five independent experiments. (C) Percentages of PS-positive and externalized annexin A5-positive cells according to panel A. Washes were performed in the presence of 1.8 mM (open bars) or absence (dark bars) of Ca2+. Dashed lines represented the control values obtained after Ca2+-free washes. For clarity, statistical symbols have been omitted except those representing Ca2+-free washes versus washes in presence of Ca2+ (*p<0.05) and externalized annexin A5 versus PS-positive cells after Ca2+-free washes ($p<0.05) for n=4 independent experiments. (D) Annexin A5 (a) and PS (b) colabeling of cardiomyocytes showed that externalized annexin A5 could be absent from PS-positive cardiomyocytes at the end of protocol A, 2 h after Ca2+-free washes. Hoechst 33342 (c). Merge (d). Representative experiment from four independent assays. Bar: 20 μm.

Fig. 5

Effects of Ca2+-free washes on externalized annexin A5 and apoptosis induced by ST. (A) Time course protocol: incubations for 4 h in control or 3 μM ST media were interrupted after 2 h by washes performed in presence or absence of Ca2+. (B) Western blot analysis of annexin A5 in control medium (CM), 3 μM ST medium (STM), 1.8 mM Ca2+ or Ca2+-free washes of control cells (C) and of ST treated cells (ST) after 2 hours incubation. Representative results from five independent experiments. (C) Percentages of PS-positive and externalized annexin A5-positive cells according to panel A. Washes were performed in the presence of 1.8 mM (open bars) or absence (dark bars) of Ca2+. Dashed lines represented the control values obtained after Ca2+-free washes. For clarity, statistical symbols have been omitted except those representing Ca2+-free washes versus washes in presence of Ca2+ (*p<0.05) and externalized annexin A5 versus PS-positive cells after Ca2+-free washes ($p<0.05) for n=4 independent experiments. (D) Annexin A5 (a) and PS (b) colabeling of cardiomyocytes showed that externalized annexin A5 could be absent from PS-positive cardiomyocytes at the end of protocol A, 2 h after Ca2+-free washes. Hoechst 33342 (c). Merge (d). Representative experiment from four independent assays. Bar: 20 μm.

Externalized annexin A5 involvement in apoptosis

To investigate whether externalization of annexin A5 could participate to apoptosis we removed externalized annexin A5 by Ca2+-free washes and followed the apoptotic process during 24 h of culture (Fig. 6A). We used H2O2 because of its short induction window and fast annexin A5 externalization. We showed that annexin A5 was removed by Ca2+-free washes performed 15 or 30 min after induction (Fig. 6B) and that the number of TUNEL-positive cells after 24 h was significantly decreased to the control level (Fig. 6C) suggesting that removing annexin A5 immediately after externalization inhibits apoptosis. By contrast, washes at 1.8 mM Ca2+ had no effect on annexin A5 release and on the number of TUNEL-positive cells.

Fig. 6

Effects of Ca2+-free washes on externalized annexin A5 and apoptosis induced by H2O2. (A) Time course protocol: after H2O2 induction for 30 min, incubations for 24 h in SFM were interrupted after 15 or 30 min by washes performed in presence or absence of Ca2+. C: control. (B) Western blot analysis of annexin A5 in washes performed in presence or absence of Ca2+. Representative results from four independent experiments. (C) Percentages of TUNEL-positive cardiomyocytes at 24 h in control (C) and after H2O2 apoptosis induction (H2O2); washes at 15 or 30 min were performed in presence (open bars) or absence (dark bars) of Ca2+. A significant protection against apoptosis was observed after washes performed in absence of Ca2+ (n≤5 wells obtained from five separate experiments for each condition). *p≤0.05, ***p≤0.001 significant from control; $$p≤0.01, $p≤0.001 significant from H2O2.

Fig. 6

Effects of Ca2+-free washes on externalized annexin A5 and apoptosis induced by H2O2. (A) Time course protocol: after H2O2 induction for 30 min, incubations for 24 h in SFM were interrupted after 15 or 30 min by washes performed in presence or absence of Ca2+. C: control. (B) Western blot analysis of annexin A5 in washes performed in presence or absence of Ca2+. Representative results from four independent experiments. (C) Percentages of TUNEL-positive cardiomyocytes at 24 h in control (C) and after H2O2 apoptosis induction (H2O2); washes at 15 or 30 min were performed in presence (open bars) or absence (dark bars) of Ca2+. A significant protection against apoptosis was observed after washes performed in absence of Ca2+ (n≤5 wells obtained from five separate experiments for each condition). *p≤0.05, ***p≤0.001 significant from control; $$p≤0.01, $p≤0.001 significant from H2O2.

To further investigate whether blocking annexin A5 could have an antiapoptotic effect, cardiomyocytes were incubated with antiannexin A5 during ST- or H2O2-induced apoptosis. We found that the number of TUNEL-positive cells and caspase 3 activity were decreased, whereas incubation with a nonimmune serum had no effect (Table 1). Moreover, we showed that in cytosolic extracts devoided of COX VIb, cytochrome c was increased after ST or H2O2 treatment and decreased after blocking annexin A5 by antiannexin A5 (Fig. 7A and B). Accordingly, these results suggested that externalization of annexin A5 appears as a modulating mechanism of myocardial cell apoptosis upstream of mitochondrial signaling pathway.

Fig. 7

Inhibition of cytochrome c release by antiannexin A5. Representative Western blot experiments of cytochrome c release in cytosol. Antiannexin A5 antibody was added to the culture medium of control cells and cells treated with staurosporine or after induction with H2O2. After 7 h, mitochondrial and cytosolic extracts were prepared for cytochrome c and COX VIb subunit detection. (A) Staurosporine treatment (n=4). (B) H2O2 treatment (n=6). Lanes 1, 2, 3: cytosolic extracts (5 μg); lanes 4, 5, 6: mitochondrial fractions (2 μg); lanes 7, 8, 9: homogenates (2 μg). Lanes 1, 4, 7: staurosporine or H2O2 treatment; lanes 2, 5, 8: staurosporine or H2O2 treatment+antiannexin A5 antibody; lanes 3, 6, 9: control conditions.

Fig. 7

Inhibition of cytochrome c release by antiannexin A5. Representative Western blot experiments of cytochrome c release in cytosol. Antiannexin A5 antibody was added to the culture medium of control cells and cells treated with staurosporine or after induction with H2O2. After 7 h, mitochondrial and cytosolic extracts were prepared for cytochrome c and COX VIb subunit detection. (A) Staurosporine treatment (n=4). (B) H2O2 treatment (n=6). Lanes 1, 2, 3: cytosolic extracts (5 μg); lanes 4, 5, 6: mitochondrial fractions (2 μg); lanes 7, 8, 9: homogenates (2 μg). Lanes 1, 4, 7: staurosporine or H2O2 treatment; lanes 2, 5, 8: staurosporine or H2O2 treatment+antiannexin A5 antibody; lanes 3, 6, 9: control conditions.

Table 1

Protective effect of antiannexin A5 on apoptosis induced by staurosporine or H2O2

 Control (n=19) Control+antiannexin A5 (1/100) (n=5) ST-treated cells (n=5) ST-treated cells+antiannexin A5 (1/100) (n=5) H2O2-treated cells (n=19) H2O2-treated cells+antiannexin A5 (1/100) (n=5) 
Caspase 3 activity (fold increase) 1±0.01 0.8±0.01 4.3±0.2** 2.5±0.2††   
Cytochrome c release (%) 31±9 29±7 100±4.1*** 42±7††† 100±8.2*** 49±6.3 
TUNEL-positive myocytes 6.9±1.2 7.8±1.7 28.1±3.8** 9.1±1.9†† 33.1±1.4*** 18.6±2.1 
 Control (n=19) Control+antiannexin A5 (1/100) (n=5) ST-treated cells (n=5) ST-treated cells+antiannexin A5 (1/100) (n=5) H2O2-treated cells (n=19) H2O2-treated cells+antiannexin A5 (1/100) (n=5) 
Caspase 3 activity (fold increase) 1±0.01 0.8±0.01 4.3±0.2** 2.5±0.2††   
Cytochrome c release (%) 31±9 29±7 100±4.1*** 42±7††† 100±8.2*** 49±6.3 
TUNEL-positive myocytes 6.9±1.2 7.8±1.7 28.1±3.8** 9.1±1.9†† 33.1±1.4*** 18.6±2.1 

Caspase 3 activity after 4 h and cytochrome c released in the cytosolic fraction were determined after 7 h incubation and number of TUNEL-positive cells after 24 h.

‡‡p≤0.01 versus H2O2-treated cells.

**

p≤0.01 versus control.

***

p<0.01 versus ST-treated cells.

††

p<0.01 versus ST-treated cells.

†††

p<0.001 versus ST-treated cells.

p≤0.001 versus H2O2-treated cells.

Discussion

In this study, we investigated the contribution of cellular annexin A5 to apoptosis of cardiomyocytes. We demonstrated for the first time the translocation and externalization of endogenous annexin A5 during the early phase of apoptosis in cardiomyocytes. Moreover, protection against apoptosis by blocking externalized annexin A5 with antibodies or its removal early after externalization suggested a proapoptotic role for endogenous annexin A5.

As previously reported, H2O2 or ST treatments of neonatal cardiomyocytes induced reproducible apoptosis engaging the ER Ca2+gateway and mitochondrial pathway in a time-dependent process [21,25,26]. Interestingly, the H2O2 treatment induced a faster PS externalization and a slower DNA fragmentation than ST. In these models, we demonstrated that annexin A5 externalization was a time-dependent phenomenon occurring in both neonatal and adult cardiomyocytes whatever the apoptotic trigger (staurosporine, H2O2 or ceramide). Furthermore, annexin A5 externalization in neonatal and adult cardiomyocytes was faster and higher after H2O2 treatment than after ST treatment. It is worthy to note that these differences between ST or H2O2 induction were likely due not only to the trigger itself but to the induction process that was continued with ST and that occurred in a short window with H2O2.

Although annexin A5 and PS externalization seemed closely time related, our results are in favor of independent processes. First, kinetics of annexin A5 externalization was very similar to that of PS externalization following either ST or H2O2 induction. However, likely owing to the coordinated process of apoptosis induction with H2O2, it is worthy to note that at a very early time (15–30 min) annexin A5 externalization could precede PS externalization (Figs. 1 and 3). In line with our study, formation of Ca2+ channel by annexin A5 during apoptosis of chondrocytes has been reported to occur in the first 30 min after H2O2 treatment. Second, later on during apoptosis, we observed also independent localization of annexin A5 and PS in the washout experiments performed in absence of Ca2+ and after prolonged incubation with ST (Fig. 5C and D). Taken together, these results strongly suggested that annexin A5 externalization was occurring very early during apoptosis process and was independent from that of PS.

Dependent on an increase in [Ca2+]i, translocation of annexin A5 to nuclear and/or cellular membrane has been reported in various cell types [7,8] and in cardiomyocytes during metabolic inhibition or heart failure [13,20,29]. Beyond translocation to membranes this study showed for the first time that annexin A5 was externalized in cardiomyocytes during apoptosis. Interestingly, an increase in intracellular Ca2+ and a pH acidification have been reported during apoptosis [30,31]. Because such conditions have been shown to favor in vitro annexin insertion in membranes [9], we suggest that they could in turn trigger membrane insertion and externalization of annexin A5 in cardiomyocytes during apoptosis. Interestingly, Arur et al. [32] recently reported that annexin A1 could also be externalized during apoptosis.

Of importance is our data indicating that apoptosis was decreased by using antiannexin A5 or annexin A5 washout 15 or 30 min after H2O2 induction, suggesting that annexin A5 externalization could exert a proapoptotic role per se in cardiomyocytes. These results are in agreement with the study of DT40 cells lacking annexin A5 and of chondrocytes suggesting a proapoptotic role for annexin A5 by acting as a Ca2+ channel during the first 30 min [15,16]. Therefore, beyond translocation, membrane insertion and externalization of annexin A5 and antiapoptotic effect of antiannexin A5 or annexin A5 release at the early beginning of the apoptotic process, and assuming that annexin A5 role were likely similar in chondrocytes and cardiomyocytes we suggested that annexin A5 developed a proapoptotic role by acting as a calcium channel.

In conclusion, regarding apoptosis in the myocardium as a degenerative process leading to heart failure [33–35] such an externalization of endogenous annexin A5 might favor apoptosis and play a role in transition to heart failure.

Acknowledgments

The authors thank J.L. Samuel and C. Delcayre for helpful and constructive discussions. We are indebted to Dr. R. Sercombe for expert technical assistance in confocal microscopy experiments. Dr. Belikova was supported by a research fellowship from “Region Ile de France” (poste vert INSERM) and from Fondation de la Recherche Médicale. Dr. Charlemagne was a recipient from CNRS.

References

[1]
Raynal
P.
Pollard
H.B.
Annexins: the problem of assessing the biological role for a gene family of multifunctional calcium- and phospholipid-binding proteins
Biochim. Biophys. Acta
 
1994
1197
63
93
[2]
Reutelingsperger
C.P.
van Heerde
W.L.
Annexin
V.
The regulator of phosphatidylserine-catalyzed inflammation and coagulation during apoptosis
Cell. Mol. Life Sci.
 
1997
53
527
532
[3]
van Engeland
M.
Nieland
L.J.
Ramaekers
F.C.
Schutte
B.
Reutelingsperger
C.P.
Annexin V-affinity assay: a review on an apoptosis detection system based on phosphatidylserine exposure
Cytometry
 
1998
31
1
9
[4]
Stuart
M.C.
Damoiseaux
J.G.
Frederik
P.M.
Arends
J.W.
Reutelingsperger
C.P.
Surface exposure of phosphatidylserine during apoptosis of rat thymocytes precedes nuclear changes
Eur. J. Cell Biol.
 
1998
76
77
83
[5]
Dumont
E.A.
Hofstra
L.
van Heerde
W.L.
van den Eijnde
S.
Doevendans
P.A.
DeMuinck
E.
et al
Cardiomyocyte death induced by myocardial ischemia and reperfusion: measurement with recombinant human annexin-V in a mouse model
Circulation
 
2000
102
1564
1568
[6]
Mollenhauer
J.
Bee
J.A.
Lizarbe
M.A.
von der Mark
K.
Role of anchorin CII, a 31,000-mol-wt membrane protein, in the interaction of chondrocytes with type II collagen
J. Cell Biol.
 
1984
98
1572
1579
[7]
Raynal
P.
Kuijpers
G.
Rojas
E.
Pollard
H.B.
A rise in nuclear calcium translocates annexins IV and V to the nuclear envelope
FEBS Lett.
 
1996
392
263
268
[8]
Barwise
J.L.
Walker
J.H.
Annexins II, IV, V and VI relocate in response to rises in intracellular calcium in human foreskin fibroblasts
J. Cell Sci.
 
1996
109
247
255
[9]
Langen
R.
Isas
J.M.
Hubbell
W.L.
Haigler
H.T.
A transmembrane form of annexin XII detected by site-directed spin labeling
Proc. Natl. Acad. Sci. U. S. A.
 
1998
95
14060
14065
[10]
Burger
A.
Voges
D.
Demange
P.
Perez
C.R.
Huber
R.
Berendes
R.
Structural and electrophysiological analysis of annexin V mutants. Mutagenesis of human annexin V, an in vitro voltage-gated calcium channel, provides information about the structural features of the ion pathway, the voltage sensor and the ion selectivity filter
J. Mol. Biol.
 
1994
237
479
499
[11]
Mira
J.P.
Dubois
T.
Oudinet
J.P.
Lukowski
S.
Russo-Marie
F.
Geny
B.
Inhibition of cytosolic phospholipase A2 by annexin V in differentiated permeabilized HL-60 cells. Evidence of crucial importance of domain I type II Ca2+-binding site in the mechanism of inhibition
J. Biol. Chem.
 
1997
272
10474
10482
[12]
Dubois
T.
Mira
J.P.
Feliers
D.
Solito
E.
Russo-Marie
F.
Oudinet
J.P.
Annexin V inhibits protein kinase C activity via a mechanism of phospholipid sequestration
Biochem. J.
 
1998
330
1277
1282
[13]
Gerke
V.
Moss
S.E.
Annexins: from structure to function
Physiol. Rev.
 
2002
82
331
371
[14]
Babiychuk
E.B.
Draeger
A.
Annexins in cell membrane dynamics. Ca(2+)-regulated association of lipid microdomains
J. Cell Biol.
 
2000
150
1113
1124
[15]
Hawkins
T.E.
Das
D.
Young
B.
Moss
S.E.
DT40 cells lacking the Ca2+-binding protein annexin 5 are resistant to Ca2+-dependent apoptosis
Proc. Natl. Acad. Sci. U. S. A.
 
2002
99
8054
8059
[16]
Wang
W.
Xu
J.
Kirsch
T.
Annexin-mediated Ca2+influx regulates growth plate chondrocyte maturation and apoptosis
J. Biol. Chem.
 
2003
278
3762
3769
[17]
Pula
G.
Bianchi
R.
Ceccarelli
P.
Giambanco
I.
Donato
R.
Characterization of mammalian heart annexins with special reference to CaBP33 (annexin V)
FEBS Lett.
 
1990
277
53
58
[18]
Luckcuck
T.
Trotter
P.J.
Walker
J.H.
Localization of annexin V in the adult and neonatal heart
Biochem. Biophys. Res. Commun.
 
1997
238
622
628
[19]
Trouve
P.
Legot
S.
Belikova
I.
Marotte
F.
Benevolensky
D.
Russo-Marie
F.
et al
Localization and quantitation of cardiac annexins II, V, and VI in hypertensive guinea pigs
Am. J. Physiol.
 
1999
276
H1159
H1166
[20]
Benevolensky
D.
Belikova
Y.
Mohammadzadeh
R.
Trouve
P.
Marotte
F.
Russo-Marie
F.
et al
Expression and localization of the annexins II, V, and VI in myocardium from patients with end-stage heart failure
Lab. Invest.
 
2000
80
123
133
[21]
Scorrano
L.
Oakes
S.A.
Opferman
J.T.
Cheng
E.H.
Sorcinelli
M.D.
Pozzan
T.
et al
BAX and BAK regulation of endoplasmic reticulum Ca2+: a control point for apoptosis
Science
 
2003
300
135
139
[22]
Demaurex
N.
Distelhorst
C.
Cell biology. Apoptosis—the calcium connection
Science
 
2003
300
65
67
[23]
Knowlton
K.U.
Baracchini
E.
Ross
R.S.
Harris
A.N.
Henderson
S.A.
Evans
S.M.
et al
Co-regulation of the atrial natriuretic factor and cardiac myosin light chain-2 genes during alpha-adrenergic stimulation of neonatal rat ventricular cells. Identification of cis sequences within an embryonic and a constitutive contractile protein gene which mediate inducible expression
J. Biol. Chem.
 
1991
266
7759
7768
[24]
Henaff
M.
Antoine
S.
Mercadier
J.J.
Coulombe
A.
Hatem
S.N.
The voltage-independent B-type Ca2+channel modulates the apoptosis of cardiac myocytes
FASEB J.
 
2001
29
29
[25]
Yue
T.L.
Wang
C.
Romanic
A.M.
Kikly
K.
Keller
P.
DeWolf
W.E.
Jr.
et al
Staurosporine-induced apoptosis in cardiomyocytes: a potential role of caspase-3
J. Mol. Cell. Cardiol.
 
1998
30
495
507
[26]
von Harsdorf
R.
Li
P.F.
Dietz
R.
Signaling pathways in reactive oxygen species-induced cardiomyocyte apoptosis
Circulation
 
1999
99
2934
2941
[27]
Herrmann
M.
Lorenz
H.M.
Voll
R.
Grunke
M.
Woith
W.
Kalden
J.R.
A rapid and simple method for the isolation of apoptotic DNA fragments
Nucleic Acids Res.
 
1994
22
5506
5507
[28]
Vander Heiden
M.G.
Chandel
N.S.
Williamson
E.K.
Schumacker
P.T.
Thompson
C.B.
Bcl-xL regulates the membrane potential and volume homeostasis of mitochondria
Cell
 
1997
91
627
637
[29]
Jans
S.W.
Willems
J.
van Bilsen
M.
Reutelingsperger
C.P.
van der Vusse
G.J.
Phospholipid degradation in energy-deprived cardiac myocytes: does annexin V play a role?
J. Mol. Cell. Cardiol.
 
1997
29
1401
1410
[30]
Webster
K.A.
Discher
D.J.
Kaiser
S.
Hernandez
O.
Sato
B.
Bishopric
N.H.
Hypoxia-activated apoptosis of cardiac myocytes requires reoxygenation or a pH shift and is independent of p53
J. Clin. Invest.
 
1999
104
239
252
[31]
Frasch
S.C.
Henson
P.M.
Kailey
J.M.
Richter
D.A.
Janes
M.S.
Fadok
V.A.
et al
Regulation of phospholipid scramblase activity during apoptosis and cell activation by protein kinase Cdelta
J. Biol. Chem.
 
2000
275
23065
23073
[32]
Arur
S.
Uche
U.E.
Rezaul
K.
Fong
M.
Scranton
V.
Cowan
A.E.
et al
Annexin I is an endogenous ligand that mediates apoptotic cell engulfment
Dev. Cell
 
2003
4
4
587
598
[33]
Haunstetter
A.
Izumo
S.
Apoptosis: basic mechanisms and implications for cardiovascular disease
Circ. Res.
 
1998
82
1111
1129
[34]
Kang
P.M.
Izumo
S.
Apoptosis and heart failure: a critical review of the literature
Circ. Res.
 
2000
86
1107
1113
[35]
Nadal-Ginard
B.
Kajstura
J.
Leri
A.
Anversa
P.
Myocyte death, growth, and regeneration in cardiac hypertrophy and failure
Circ. Res.
 
2003
92
139
150

Author notes

1
The two first authors equally participated in this paper.
Time for primary review 29 days