Altered mechanical properties and intracellular calcium signaling in cardiomyocytes from annexin 6 null-mutant mice

The FASEB Journal express article 10.1096/fj.01-0892fje. Published online February 12, 2002. Altered mechanical properties and intracellular calcium signaling in cardiomyocytes from annexin 6 null-mutant mice Guojie Song,* Sian E. Harding,� Michael R. Duchen,� Richard Tunwell,� Peter O'Gara,� Tim E. Hawkins,� and Stephen E. Moss* *Division of Cell Biology, Institute of Ophthalmology, University College London, 11-43 Bath Street, London EC1V 9EL, UK; �Cardiac Medicine, NHLI, Imperial College School of Medicine, Dovehouse Street, London SW3 6LY, UK; �Department of Physiology, University College London, Gower Street, London WC1E 6BT, UK Corresponding author: Stephen E. Moss, Division of Cell Biology, Institute of Ophthalmology, University College London, 11-43 Bath Street, London EC1V 9EL. E-mail: [email protected] ABSTRACT Annexin 6 is one of a widely expressed family of calcium-binding proteins found in most mammalian tissues, including the heart. Several studies have implicated annexin 6 in the regulation of intracellular Ca2+ signaling, and it has been shown in vitro to act as a modulator of the sarcoplasmic reticulum Ca2+-release channel, cardiac L-type calcium channel, and Na+/Ca2+ exchanger. To investigate the role of annexin 6 in intact cardiomyocytes, we used mice containing a targeted disruption of the annexin 6 gene. Compared with controls, the myocytes of annexin 6 null-mutant mice demonstrated a significant increase in the rates of shortening and relengthening. Intracellular Ca2+ transients in fura-2-loaded cardiomyocytes induced by caffeine showed a normal baseline and amplitude, whereas the rate of decay was doubled in annexin 6�/� myocytes compared with control mice. These results show that annexin 6 knockout in the mouse leads to an increase in myocyte contractility and faster diastolic Ca2+ removal from the cytoplasm. In light of published findings showing annexin 6 to be down-regulated in end-stage heart failure, these results are consistent with a role for annexin 6 as a negative inotropic factor in the regulation of cardiomyocyte mechanics. Key words: heart failure • homeostasis • knockout • stimulus-response coupling

T

he annexins are an extensively studied family of Ca2+-dependent phospholipid-binding proteins expressed in most eukaryotic phyla, but their functions are poorly understood (1). Annexins have multiple tandem repeats of a highly conserved ~70 amino acid motif that

contains the Ca2+-binding site and have variable N-terminal domains that are probably important in defining functional specificity. Annexin 6 is unique within the family because it has eight conserved repeats (all other annexins have four), which in the crystal form, fold as two similarly structured lobes connected by a flexible linker (2). Theories abound about the function of annexin 6, for example, as an inhibitor of blood coagulation (3), an inhibitor of protein kinase C (4), and in the processing and intracellular trafficking of endosomal vesicles (5). Annexin 6 also

has been proposed to be required for the budding of clathrin-coated pits in vitro (6), although transferrin receptors are processed normally through this route in A431 cells that lack annexin 6, arguing that this is probably not a general role for this protein (7). Other functional ideas suggest a role for annexin 6 in the regulation of intracellular Ca2+ signaling. The first indication of such a role came from in vitro experiments showing that annexin 6 increases the mean open time and opening probability of sarcoplasmic reticulum (SR) Ca2+ release channels (8, 9). However, these experiments were performed before it was known that annexin 6 possesses Ca2+ channel activity in artificial lipid bilayers (2), so the effects observed may not have been due to any direct modulatory effect of annexin 6 on SR Ca2+ channels. Transgenic studies in which annexin 6 overexpression was targeted to cardiomyocytes suggested a role for annexin 6 in the regulation of Ca2+ signaling and contractility (10). Thus, resting Ca2+ levels were diminished, together with the amplitude of electrically generated Ca2+ spikes, with a concomittant loss of contractile function. The theoretical corollary of these observations is that down-regulation of annexin 6 in cardiomyocytes could lead to increased contractility and that this might occur in conjunction with changes in Ca2+ handling. This notion is supported by studies showing that annexin 6 is indeed down-regulated in human end-stage heart failure (11). To resolve these issues and gain insight into the role of annexin 6 in cardiomyocyte function, we generated mice containing a targeted disruption of the ANX6 gene (12). First, we examined the consequences of annexin 6 loss on the contractile properties of isolated cardiomyocytes. Second, we investigated the relationship between Ca2+ signaling and contractility in these cardiomyocytes. Third, we examined the expression levels of certain Ca2+ regulatory proteins because up- or down-regulation of these molecules could influence Ca2+ release and uptake profiles during excitation-contraction coupling. MATERIALS AND METHODS Isolation of cardiomyocytes Ventricular myocytes were isolated by enzymatic dissociation as previously described (13). Mice were killed by cervical dislocation, and the heart was quickly removed and placed into ice-cold Krebs-Henseleit (KH) solution (119 mM NaCl, 4.7 mM KCl, 0.94 mM MgSO4, 1.2 mM KH2PO4, 25 mM NaHCO3, 11.5 mM glucose, 1 mM CaCl2) previously bubbled with 95% O2/5% CO2 to bring the pH to 7.4. The heart was then mounted on a needle canal (23G 1 1/4 in.) attached to a Langendorff perfusion apparatus and equilibrated with KH solution containing 1mM Ca2+ for 5 min at 37°C. The perfusate was then changed to a low-calcium medium (120 mM NaCl, 5.4 mM KCl, 5 mM MgSO4, 5 mM pyruvate, 20 mM glucose, 20 mM taurine, 10 mM HEPES, 5 mM nitrilotriacetic acid) containing 12−15 mM calcium (measured with a calcium electrode at 37°C) and bubbled with 100% O2 at pH 6.95 for 5 min. This was followed by 1 min perfusion with the same low-calcium solution (with nitrilotriacetic acid omitted) at pH 7.4 with 200 µM calcium and 2 I.U. protease type XXIV (P-8038, Sigma, Poole, UK) added. This enzyme solution was then switched to one that did not contain added protease but did contain1.0 mg/ml collagenase (Worthington 4376 CLS-2) and 0.6 mg/ml hyaluronidase (H-3506, Sigma). Incubation continued for an additional 10 min, after which the ventricles were chopped

and incubated at 35°C in the same enzyme solution twice for 10 min each time. The medium was shaken gently throughout this incubation and was kept under an atmosphere of 100% O2. The dispersed cells were then strained through gauze of mesh size 300 µm, and supernatant was centrifuged at 40g for 1 min at room temperature. Cells were washed and resuspended in KH solution at room temperature until they were used. Measurement of cell contraction The contraction of single ventricular myocytes was recorded as previously described (14). Myocytes were placed in a Perspex cell bath on the stage of an inverted microscope (Nikon or Zeiss IM, Welwyn Garden City, UK) and superfused continuously with KH solution (32°C) equilibrated with 95% O2/5% CO2 at 1.8 to 2.0 ml/min. Cells were stimulated at a basal rate of 1 Hz by bipolar pulses through platinum electrodes placed along the bath. Unloaded cell shortening was measured with a video camera/length-detection system, with spatial resolution at 1 to 512 and time resolution at 10 ms. After electrical stimulation for at least 10 min at 1mM Ca2+, contraction was measured in up to 32 cells from each preparation. Each of these myocytes was then allowed a 1-min rest period, and the amplitude of the first post rest beat was recorded. In one cell from each preparation, stimulation frequency was increased progressively from 0.5 Hz to 1, 2, 3, 4, and 5 Hz, with equilibration times of 20�30 s at each frequency. Statistical analysis Results from up to 32 myocytes from each preparation were pooled; therefore, n values for statistical analysis refer to the number of animals. Calcium measurements Cells were loaded with fura-2 by incubation with 5 µM fura-2 acetoxymethyl ester (AM) and 0.02% pluronic for 30 min at room temperature. The cells were washed and placed in a home-built chamber on the stage of an epifluorescence microscope as previously described (15) before stimulation with caffeine (2 mM). Light from a xenon arc lamp was transmitted through a filter wheel (Cairn, Faversham, UK) equipped with bandpass filters centered at 340, 360, and 380 nm, spinning at 12 Hz. Emitted light was collected by a photomultiplier tube after passing through a band pass filter at 530 nm. The signal was digitized (Tecmar [Cambridge, MA] labmaster board under control of LabTech Notebook) and processed off line. Background subtraction, ratioing, and curve fitting was done in Origin (MicroCal, Northampton, MA). Caffeine (2 mM) was used to release SR Ca2+ in order to explore the dynamics of [Ca2+]c signaling in these cells and was applied locally by pressure ejection from a micropipette placed close to the cell. Western blotting Six hearts (~0.8 g) from both annexin 6�/� or wild-type mice were collected, and microsomes were prepared at 4°C. Equal protein loadings (40 µg/lane) were electrophoresed through 5% or 10% polyacrylamide gels and transferred to Immobilon P for immunoblotting. The type 2 ryanodine receptor (RyR2) blot was performed with a 1:500 dilution of a rabbit antipeptide antibody specific to all known RyR2s (16). The pan ryanodine receptor (RyR) blot was

performed using a 1:500 dilution of a rabbit antipeptide raised against a peptide common to all three isoforms of mouse RyRs. The SR ATPase (SERCA) blot was performed using Y/1F4; a mAb generated from a mouse immunized with purified ATPase in the laboratories of Dr. Malcolm East and Professor Anthony Lee at University of Southampton (UK). The epitope has been mapped to KMFVK, a peptide sequence common in most types of SERCAs (17). The actin blot was performed using a rabbit polyclonal antibody (A-5060, Sigma) to the N-terminal of actin. Antigen detection was completed using Western Blue (Promega, Southampton, UK). RESULTS AND DISCUSSION The annexins are a family of widely expressed calcium/phospholipid-binding proteins with proposed functions in diverse cellular activities such as vesicle motility, anticoagulation, phospholipase A2 inhibition, and cell growth regulation (1). Biochemical and immunohistochemical studies have shown that cardiac tissues are particularly enriched in two members of the family: annexins 5 and 6 (18�21). Annexin 6 is localized around individual myofibrils and is associated with the terminal cisternae of isolated SR membranes. In this context, it has been suggested that annexin 6 is involved in atrial secretary granule binding (21) and the cycle of calcium release/uptake in the SR (9, 22, 23). Studies on the effect of annexin 6 on the ryanodine-sensitive Ca2+ release channel in the SR of skeletal muscle cell revealed that annexin 6 modified the gating properties of the Ca2+ release channel (reconstituted into artificial bilayers) by increasing the open probability and mean open time (8). Annexin 6 was also shown to completely reverse the inhibition of Na+/Ca2+ exchange activity of cardiac sarcolemma vesicles consequent to EGTA-treatment, suggesting a modulation of Na+/Ca2+ exchange activity in myocardium by this protein (18). Cardiomyocytes isolated from transgenic mice with cardiac-specific overexpression of annexin 6 display impaired myocyte contractility with depression of intracellular Ca2+ transients (10). These investigations indicate that annexin 6 has the potential to influence ion fluxes vital to heart function. Consistent with this, annexin 6 mRNA and protein levels were found to be decreased in end-stage congestive heart failure in humans, whereas the expression levels of annexins 2 and 5 were increased (11). Although these studies highlight the potential importance of annexin 6 in cardiac physiology, the role of annexin 6 in cardiac myocytes is poorly understood. Here, we used isolated myocytes from mice containing a targeted disruption of the annexin 6 gene (12) and compared them with 12-wk-old wild-type control mice to provide further information on the influence of annexin 6 on myocyte contractility and relaxation and on intracellular calcium dynamics. Morphological investigation of isolated ventricular myocytes revealed that cell lengths were not significantly different between annexin 6�/� mice (133±5 µm, mean ±SE, n=6 animals [133 cells]) and controls (122±3 µm, n=6 animals [132 cells]). Contraction was monitored at a calcium concentration of 1 mM and a steady-state (SS) stimulation rate of 1 Hz (see Fig. 1A). Figure 1A shows averaged data from six pairs of mice; amplitudes obtained from up to 30 cells for each mouse have been pooled to give a single value per animal. Steady-state amplitude was increased by 83% in the annexin 6�/� mice (P<0.01). Mouse ventricular myocytes have a strong post rest potentiation of contraction, and amplitude after 1 min of rest is close to the maximum. Figure 1A

shows the amplitude of the first beat after 1 min of rest (PR B1), which was also increased, in this case by 53% (P<0.01). As is normal for this species, spontaneous contractions were observed during rest in a proportion of myocytes. Of 111 myocytes from control animals, 42 contracted spontaneously during rest, compared with 68 of 121 for annexin 6�/� mice (P<0.005, χ-squared test). The average frequency of spontaneous contraction was 8.8/min for control and 6.6/min for annexin 6�/� mice, although these figures were very variable. In cells in which contraction occurred during the rest period, it is possible that the post rest amplitude was submaximal. This may explain why the increase in amplitude in annexin 6�/� mice was slightly lower for the first beat after 1 min of rest than steady-state values and the variation slightly greater. Velocities of contraction (�dL/dT) and relaxation (+dL/dT) (dL is change in length; dT is change in time) also increased significantly (Fig. 1B), but largely in proportion to the increase in amplitude. Consistent with this, time-to-peak contraction (TTP) and time-to-50% relaxation (R50) were unaffected (Fig. 1C). However, the late phase of relaxation, defined by the time-to-90% relaxation (R90), was significantly accelerated in the annexin 6�/� animals. Responses to increasing stimulation frequency from rested-state beats (1 min rest) to 5 Hz are shown in Figure 2. The difference in contraction amplitude was more pronounced at lower stimulation frequencies, with significant differences up to 2 Hz. To investigate whether the changes in contractility in the annexin 6 null-mutant animals could be explained by alterations in Ca2+ handling, we loaded isolated cardiomyocytes with fura-2 AM and performed ratiometric analysis of Ca2+ fluxes during the contraction-relaxation cycle (Fig. 3). The diastolic intracellular Ca2+ level, shown as the baseline [Ca2+] signal, in myocytes isolated from annexin 6�/� mice was identical to that in controls under our experimental conditions (0.96±0.04 vs. 0.95±0.04, knockout vs. wild type). In addition, the amplitude of the [Ca2+]c signal in response to emptying SR stores with caffeine reached similar levels in the two groups (the increases in fura-2 ratio (∆R) were 0.51±0.198 vs. 0.56± 015, knockout vs. wild type, n=15 and 13, respectively). No significant difference occurred in the time-to-peak of the calcium signal in the two experimental groups, demonstrating that the velocity of calcium release from the SR to cytoplasm is similar under these experimental conditions. However, the rate of recovery of the calcium signal was significantly faster in the cells from annexin 6�/� mice, with a mean time constant of decay some twofold faster than the controls (time constant of 5.13±0.49 [n=15, SE] vs. 10.35±0.79 sec�1 [n=13] for knockout vs. wild type, P<0.05). The difference between the two genotypes was also observed at the 90% decline point of the Ca2+ signal (T90) and the 50% decline point (T50). To examine whether the anomalies in calcium handling might be explained by changes in the levels of expression of major Ca2+ regulatory proteins, we immunoblotted cardiomyocyte extracts for RyRs and SERCA pumps (Fig. 3D). The use of extracts pooled from several mice in each case (n=6) provides an averaged but nonquantitative assessment of protein expression, revealing in this case no obvious up- or down-regulation in expression of the proteins under investigation. Thus, the accelerated clearance of cytosolic Ca2+ cannot be explained by simple elevation of expression of extrusive intracellular Ca2+ pumps.

In considering the mechanistic effects of annexin 6 in the light of published data, one might hypothesize that the absence of annexin 6 in myocytes leads to a decrease in the opening probability and the mean open time of SR ryanodine-sensitive calcium release channels (8, 9, 22). Alternatively, loss of annexin 6 in myocytes may result in amplified Na+/Ca2+ exchange activity, inducing an augmentation of Ca2+ efflux and uptake (18, 24, 25). Further investigations using annexin 6 antibody or inhibitors of cardiomyocyte Ca2+ channels, pumps, and exchangers are required to determine the specific target(s) of annexin 6 that are responsible for the regulation of intracellular Ca2+. In summary, our observations indicate an important role for annexin 6 in excitation-contraction coupling in mouse cardiomyocytes. Loss of annexin 6 does not affect intracellular calcium levels, but the temporal anomalies in Ca2+ clearance and the amplified contractile properties suggest a negative role for annexin 6 in the regulation of intracellular calcium and cardiomyocyte mechanical function. ACKNOWLEDGMENTS This work was supported by grants from the British Heart Foundation and the Wellcome Trust. REFERENCES 1. Gerke, V., and Moss, S. E. (1997) Annexins and membrane dynamics. Biochim. Biophys.

Acta 1357, 129�154 2. Benz, J., Bergner, A., Hofmann, A., Demange, P., Göttig, P., Liemann, S., Huber, R., and

Voges, D. (1996) The structure of recombinant human annexin VI in crystals and membrane-bound. J. Mol. Biol. 260, 638�643

3. Iwasaki, A., Suda, M., Watanabe, M., Nakao, H., Hattori, Y., Nagoya, T., Saino, Y.,

Shidara, Y., and Maki, M. (1989) Structure and expression of cDNA for calphobindin II, a human placental coagulation inhibitor. J. Biochem. 106, 43�49

4. Shibata, S., Sato, H., and Maki, M. (1992) Calphobindins (placental annexins) inhibit

protein kinase C. J. Biochem. 112, 552�556 5. Ortega, D., Pol, A., Biermer, M., Jäckle, S., and Enrich, C. (1998) Annexin VI defines an

apical endocytic compartment in rat liver hepatocytes. J. Cell Sci. 111, 261�269 6. Lin, H. C., Sudhof, T. C., and Anderson, R. G. (1992) Annexin VI is required for

budding of clathrin-coated pits. Cell 70, 283�291 7. Smythe, E., Smith, P. D., Jacob, S. M., Theobald, J., and Moss, S. E. (1994) Endocytosis

occurs independently of annexin VI in human A431 cells. J. Cell Biol. 124, 301�306

8. Diaz-Munoz, M., Hamilton, S. L., Kaetzel, M. A., Hazarika, P., and Dedman, J. R. (1990) Modulation of Ca2+ release channel activity from sarcoplasmic reticulum by Annexin VI (67-kDa Calcimedin). J. Biol. Chem. 265, 15894�15899

9. Hazarika, P., Sheldon, A., Kaetzel, M. A., Diaz-Munoz, M., Hamilton, S. L., and

Dedman, J. R. (1991) Regulation of the sarcoplasmic reticulum Ca2+-release channel requires intact annexin VI. J. Cell. Biochem. 46, 86�93

10. Gunteski-Hamblin, A. E., Song, G., Walsh, R. A., Frenzke, M., Dorn II, G. W., Kaetzel,

M. A., Horseman, N. D., and Dedman, J. R. (1996) Annexin 6 overexpression in transgenic mouse heart alters cardiac myocyte contractile function, calcium dynamics and heart pathology. Am. J. Physiol. 270, H1091�H1100

11. Song, G., Campos, B., Wagoner, L. E., Dedman, J. R., and Walsh, R. A. (1998) Altered

cardiac annexin mRNA and protein levels in the left ventricle of patients with end-stage heart failure. J. Mol. Cell. Cardiol. 30, 443�451

12. Hawkins. T. E., Roes, J., Rees, D., Monkhouse, J., and Moss, S. E. (1999) Immunological

development and cardiovascular function are normal in annexin VI null mutant mice. Mol. Cell Biol. 19, 8028�8032

13. Harding, S. E., Vescovo, G., Kirby, M., Jones, S. M., Gurden, J., and Poole-Wilson, P. A.

(1998) Contractile responses of isolated rat and rabbit myocytes to isoproterenol and calcium. J. Mol. Cell. Cardiol. 20, 635�647

14. Heubach, J. F., Trebe, I., Wettwer, E., Himmel, H. M., Michel, M. C., Kaumann, A. J.,

Koch, W. J., Harding, S. E., and Ravens, U. (1999) L-type calcium current and contractility in ventricular myocytes from mice overexpressing the cardiac �2-adrenoceptor. Cardiovasc. Res. 42, 173�182

15. Nowicky, A. V., and Duchen, M. R. (1998) Changes in [Ca2+]i and membrane currents

during impaired mitochondrial metabolism in dissociated rat hippocampal neurons. J. Physiol. 507, 131�145

16. Tunwell, R. E. A., Wickenden, C., Bertrand B. M. A., Shevchenko, V., Walsh, M. B.,

Allen, P. D., and Lai, F. A. (1996) The human cardiac ryanodine receptor-calcium release channel: identification, primary structure and topological analysis. Biochem. J. 318, 477�487

17. Tunwell, R. E. A., Conlan, W., Matthews, I., East, J. M., and Lee, A. G. (1991)

Definition of surface exposed epitopes on the (Ca2+-Mg2+)-ATPase of sarcoplasmic reticulum. Biochem. J. 279, 203�213

18. Sobota, A., Cusinato, F., and Luciani, S. (1990) Identification and purification of

calpactins from cardiac muscle and their effect on Na+/Ca2+ exchange activity. Biochem. Biophys. Res. Comm. 1723, 1067�1072

19. Doubell, A. F., Lazure, C., Charbonneau, C., and Thibault, G. (1993) Identification and

immunolocalisation of annexins V and VI, the major cardiac annexins, in rat heart. Cardiovasc. Res. 27, 1359�1367

20. Pula, G., Bianchi, R., Ceccarelli, P., Giambanco, I., and Donato, R. (1990)

Characterization of mammalian heart annexins with special reference to CaBP33 (annexin 5). FEBS Lett. 277, 53�58

21. Doubell, A. F., Baster, A. J., and Thibault, G. (1991) Annexin V and VI: major calcium-

dependent atrial secretory granule-binding proteins. Hypertension 18, 648�656 22. Hazarika, P., Kaetzel, M. A., Sheldon, A., Karin, N. J., Fleischer, S., Nelson, T. E., and

Dedman, J. R. (1991) Annexin 6 is associated with calcium-sequestering organelles. J. Cell. Biochem. 46, 78�85

23. Hawkins, T. E., Merrifield, C. J., and Moss, S. E. (2000) Calcium signaling and annexins.

Cell. Biochem. Biophys. 33, 275�296 24. Song, G., Kirkpatrick, D. L., Gunteski-Hamblin, A. M., Dorn II, G. W., Dedman, J. R.,

and Walsh, R. A. (1995) Abnormal Na+/Ca2+ exchanger activity in transgenic mice with cardiac-specific overexpression of annexin VI. Circulation 92, I697�I698

25. Naciff, J. M., Behbehani, M. M., Kaetzel, M. A., and Dedman, J. R. (1996) Annexin VI

modulates Ca2+ and K+ conductances of spinal cord and dorsal root ganglion neurons. Am. J. Physiol. 271, C2004�C2015

Received Dec. 6, 2001; revised January 2, 2002.

Fig. 1

Figure 1. Altered contractile dynamics in cardiomyocytes lacking annexin 6. A) Contraction amplitude of isolated ventricular myocytes (percent shortening) at 1-Hz steady state (SS) and at 1 min post rest (PR B1). B) Contraction (–dL/dT) and relaxation (+dL/dT) velocities at 1 Hz (dL is change in length, and dT is change in time). C) Times to peak contraction (TTP), to 50% relaxation (R50), and to 90% relaxation (R90). (Mean ± SE, n = 6 animals, *P<0.05, **P<0.01 vs. control.)

Fig. 2

Figure 2. Contraction amplitude of isolated ventricular myocytes. A) Percent shortening at increasing stimulation frequency, primary data recordings. B) Statistical analysis of cardiomyocyte contractility (0=beat 1 min post rest); n = 6 for control, n = 7 for annexin 6–/– animals (*P<0.05, **P<0.01 vs. control).

Fig. 3

Figure 3. Statistical analysis of cardiomyocyte intracellular Ca2+ dynamics. Caffeine (2 mM) was applied locally to fura-2-loaded cells to release sarcoplasmic reticulum Ca2+, thus allowing the exploration of the dynamics of [Ca2+]c signaling (as indicated by the bar). Analysis of the resulting [Ca2+]c transients revealed that the mean baseline, time to peak, and peak amplitudes were not significantly different in the two populations of cells. However, the time constant of recovery to baseline was significantly shorter in the annexin 6–/– cells. Examples of individual responses from the control (A) and knock-out (B) are shown (the x axis indicates units in fura-2 ratio for baseline and peak signals and sec–1 for the time constant (τ), and the data are summarized (C). Immunoblots of cardiomyocyte microsomal extracts from control (+/+) and knockout (–/–) mice probed with antibodies specific to protein components of the Ca2+-regulating system are shown (D).