The actin-binding proteins dystrophin and α-actinin are members of a family of actin-binding proteins that may link the cytoskeleton to membrane proteins such as ion channels. Previous work demonstrated that the activity of Ca2+ channels can be regulated by agents that disrupt or stabilize the cytoskeleton. In the present study, we employed immunohistochemical and electrophysiological techniques to investigate the potential regulation of cardiac L-type Ca2+channel activity by dystrophin and α-actinin in cardiac myocytes and in heterologous cells. Both actin-binding proteins were found to colocalize with the Ca2+ channel in mouse cardiac myocytes and to modulate channel function. Inactivation of the Ca2+channel in cardiac myocytes from mice lacking dystrophin (mdx mice) was reduced compared with that in wild-type myocytes, voltage dependence of activation was shifted by 5 mV to more positive potentials, and stimulation by the β-adrenergic pathway and the dihydropyridine agonist BAY K 8644 was increased. Furthermore, heterologous coexpression of the Ca2+ channel with muscle, but not nonmuscle, forms of α-actinin was also found to reduce inactivation. As might be predicted from a reduction of Ca2+ channel inactivation, a prolonging of the mouse electrocardiogram QT was observed in mdx mice. These results suggest a combined role for dystrophin and α-actinin in regulating the activity of the cardiac L-type Ca2+ channel and a potential mechanism for cardiac dysfunction in Duchenne and Becker muscular dystrophies.
- muscular dystrophy
- intracellular regulation
voltage-dependent ca2+ channels play an essential role in the regulation of many cellular processes, including gene transcription, muscle contraction, cell division, and exocytosis, by transducing a voltage signal into an elevation of intracellular Ca2+ (14). The cardiac voltage-gated L-type Ca2+ channel (CaV1.2) mediates the Ca2+ current that is responsible for the plateau phase of the cardiac action potential and for initiating contraction. Previous work suggests that the actin-based cytoskeleton contributes to the regulation of both voltage- and ligand-gated ion channels (16), and L-type Ca2+ channels in cardiac and smooth muscle have been reported as being regulated by actin filament organization (8, 18, 21, 26).
Two actin-binding proteins (ABPs), dystrophin and α-actinin, are known to link membrane-associated elements to the cytoskeleton (5). Dystrophin, a member of the spectrinlike superfamily of actin-binding proteins, acts as a link between the actin cytoskeleton, the plasmalemma, and the surrounding basal lamina. Disruption of this link through deletion of the dystrophin gene results in Duchenne and Becker muscular dystrophies (DMD, BMD), characterized by progressive weakness and wasting of skeletal muscles, nonprogressive cognitive impairment, and failure of the electrical conduction system in the heart. In the mouse model of DMD, themdx mouse, the upregulation of a related protein, utrophin, effectively prolongs the life span of these mice and decreases muscle atrophy, but signs of skeletal and cardiac myopathy are still present (11). Dystrophin and its associated proteins have been implicated in playing a role in receptor/channel localization (10, 22).
α-Actinin functions to anchor parallel filaments of F-actin throughout the cytoskeleton in all tissue types, as well as to anchor antiparallel actin filaments in muscle tissue, forming the Z disk in skeletal and cardiac muscle and the cytoplasmic-dense bodies in smooth muscle. Recent data have suggested that α-actinin may also associate with several types of ion channels. The voltage-gated K+channel Kv1.5, expressed in the cardiovascular system and brain, binds to α-actinin-2 and colocalizes at the membrane in transfected human embryonic kidney (HEK) cells (23). This colocalization of α-actinin-2 with voltage-gated K+ channels was found to modulate channel gating and current density (6, 23). α-Actinin has also been found to bind to the NR1 and NR2B subunits of the N-methyl-D-aspartate (NMDA)-type glutamate receptor and influence channel regulation (7). Krupp et al. (20) found that overexpressed muscle isoforms of α-actinin competed with calmodulin for binding to the NMDA receptor.
We examined colocalization and regulation of the Ca2+channel by the actin-binding proteins α-actinin and dystrophin using immunofluorescence and whole cell patch-clamp techniques. Within adult mouse cardiac tissue, the Ca2+ channel was found to colocalize with dystrophin at both the Z and M lines, whereas α-actinin was found to colocalize only over regions of the Z line. This colocalization with dystrophin and α-actinin indicates the potential for channel regulation by these proteins. In cardiac myocytes from mdx mice, which lack dystrophin, inactivation was reduced by a positive shift in the voltage dependence of activation, a likely allosteric change in channel conformation that made it more sensitive to stimulation by the β-adrenergic agonistl-isoproterenol and the dihydropyridine agonistl-BAY K 8644. This change in channel kinetics may have led to a slowing of ventricular repolarization, observed as an increase in the QT interval in mdx mice. Coexpression of the Ca2+ channel with muscle isoforms of α-actinin in heterologous cells resulted in a similar slowing of channel inactivation. Together, these data suggest a regulatory role for the ABPs α-actinin and dystrophin, acting on the Ca2+ channel by either direct or indirect interactions.
MATERIALS AND METHODS
Mice used for immunofluorescence studies and electrocardiograms were 2- to 3-mo-old C57BL/10SnJ (wild type) and C57BL/10ScSn-Dmdmdx (mdx), both from Jackson Laboratories (Bar Harbor, ME), weighing 20–30 g. Cardiac myocytes were prepared from mice between 1 and 4 days old. The animals were killed in accordance with the guidelines of the Institutional Animal Care and Use Committee at the University of Connecticut.
Wild-type mice were injected with heparin (5000 U/kg ip), anesthetized 30 min later with xylazine-ketamine (7 mg/kg xylazine, 80 mg/kg ketamine ip), and perfused with 10 ml of PBS and then with 10 ml of 4% formaldehyde in PBS. Hearts and brains were removed and postfixed in 4% PBS-buffered formaldehyde overnight at 4°C. After being washed in PBS for 30 min (3 times), the tissues were successively sunk in 30% sucrose in PBS for 4 h and 30% sucrose in PBS-OCT mounting medium (1:1; Tissue-Tek 4583; Sakura Finetek) overnight at 4°C. Tissue was then immediately frozen in OCT mounting medium, and 10-μm-thick sections were cut on a cryostat (model HM 500 OM; Carl Zeiss, Thornwood, NY). Sections were washed and hydrated twice for 10 min in PBS, once for 30 min in PBS containing 2% BSA, and once for 15 min in PBS containing 2% BSA or 3% goat serum and 0.3% Triton X100. Sections were incubated in PBS containing 2% BSA or 3% goat serum overnight at 4°C. Sections were labeled in incubation buffer containing 0.3% Triton X-100 with rabbit polyclonal anti-Ca2+ channel α1C antibody and/or mouse monoclonal anti-α-actinin antibody overnight at 4°C. Sections were first labeled with polyclonal antibody against L-type Ca2+channel α1-subunit, followed by a monoclonal antibody against α-actinin, monoclonal antibody against dystrophin, or fluorescent phalloidin to label F-actin. An average of four to eight cryosectioned tissues from different animals were transferred to gelatin-coated glass slides and processed for immunolabeling. Antibodies used in this study included 1) affinity-purified rabbit polyclonal antibody against the Ca2+ channel α1-subunit (13); secondary antibody for this was goat anti-rabbit conjugated with Texas red (Vector Laboratories);2) mouse monoclonal antibody against the dystrophin rod domain (Novocastra Laboratories); secondary antibody was goat anti-mouse conjugated with FITC (Sigma-Aldrich, St. Louis, MO); and3) mouse monoclonal antibody against sarcomeric α-actinin; secondary antibody was goat anti-mouse conjugated with FITC (Sigma-Aldrich). F-actin was labeled with fluorescent phalloidin. Labeling of all of these structures was specific, because it was not obtained with nonimmune rabbit antibodies or any secondary antibody that failed to bind rabbit IgG. Cells were imaged using fluorescent microscopy.
Cell preparation and electrophysiology.
TsA-201 cells (large T antigen-transformed human embryonic kidney cells, HEK-293) were maintained at 37°C in Dulbecco's modified Eagle's medium (DMEM) supplemented with 10% fetal calf serum and 100 units of penicillin and streptomycin (GIBCO-Life Technologies) in a humidified atmosphere containing 5% carbon dioxide. Calcium phosphate coprecipitation method was used for transfections (17). All cultures were transiently transfected with the following Ca2+ channel subunits: rabbit α1C(32), β2A (27), and rat α2δ (30). Cotransfected α-actinin isoforms were human skeletal muscle α-actinin-2 (1, 33), chicken smooth muscle α-actinin, chicken nonmuscle α-actinin containing only the spectrin domain, or full-length chicken nonmuscle α-actinin (20, 31). CD8 reporter plasmid was used for identification of transfected cells (EBO-pCD-Leu2, American Type Culture Collection). Before patch clamping, cell cultures were incubated with CD8 antibody-coated beads (M-450 Dynabeads, Dynal Biotech) for 1 min.
Cardiac myocytes from 1- to 4-day-old mouse neonates were dissociated as described (Worthington Biomedical, Freehold, NJ). Hearts were excised and minced in ice-cold Hanks' balanced salt solution. Minced hearts were incubated with trypsin (50 μg/ml) at 4°C overnight. Trypsinized cells were incubated with trypsin inhibitor and incubated with collagenase at 37°C for 45 min. After trituration, isolated cells were transferred to a new petri dish containing DMEM supplemented with 10% fetal bovine serum (Hyclone, Logen, UT) and incubated for 45 min at 37°C. Cells were then replated to dissociate cardiac myocytes from fibroblasts and incubated in 5% carbon dioxide at 37°C until use.
Ca2+ channel current was recorded using the whole cell patch-clamp configuration. Patch pipettes were pulled from VWR micropipettes and fire-polished to produce an inner diameter of 2–4 μm. Currents were recorded with a List EPC-9 patch amplifier and filtered at 2 kHz. Data were acquired using HEKA software (HEKA Electronik). Voltage-dependent currents were corrected for leakage using an online P/4 subtraction paradigm. Extracellular saline for HEK cell cultures contained 150 mM Tris, 1 mM MgCl2, and 10 mM BaCl2. pH was adjusted to 7.4 with methanesulfonic acid (MSA). Extracellular saline for cardiac myocyte cultures contained 145 mM tetraethylammonium chloride (TEA-Cl), 10 mM BaCl2, 1 mM MgCl2, and 10 mM HEPES. pH was adjusted to 7.4 with Tris-OH. Intracellular patch pipette saline for both cultures contained 130 mM N-methyl-D-glucamine, 10 mM EGTA, 60 mM HEPES, 2 mM MgATP, 1 mM MgCl, and MSA (used to adjust pH to 7.3). All experiments were performed at room temperature (20–23°C).
Wild-type and mdx mice, 4–8 wk old, were anesthetized with xylazine-ketamine (7 mg/kg xylazine, 80 mg/kg ketamine ip). Two fine needle electrodes were inserted subcutaneously on the shoulder at the junction between the chest and the left and right forelimbs (lead I). Electrodes were connected to an electrocardiogram (ECG) amplifier (World Precision Instruments) and digitized at 5–200 Hz. An average of 4–10 successive cycles were used to determine heart rate, QT, PR, QRS, and ST intervals, and amplitude of S, T, and R waves. Mean heart rate was calculated from the R-R intervals of the ECG. Heart rate-corrected QT values (QTc) were obtained as described (25). All measurements and recordings were made using the HEKA EPC-9 system (HEKA Electronik). Data are presented as means ± SE.
Colocalization of actin-binding proteins α-actinin and dystrophin with the Ca2+ channel.
To determine whether the Ca2+ channel colocalizes with actin or ABPs, we employed immunofluorescent labeling of the Ca2+ channel, actin, α-actinin, and dystrophin in cryosections of cardiac and neural tissues. The Ca2+channel was at highest density over Z lines (Fig.1, A,D, and G), which confirms previous reports on the distribution of the L-type Ca2+ channel in the cardiac myocyte (9). Z and M lines could be distinguished from each other because structures over Z lines are usually wider than those over M lines. Labeling of actin (Fig. 1 E), which is absent in the H zone of sarcomeres, also helped to distinguish the striated banding pattern. α-Actinin (Fig. 1 B), a major protein forming Z disks, was distributed over the Z line of costameres as expected. Figure 1 C shows colocalization of α-actinin and the Ca2+ channel over Z lines. However, the absence of cortical α-actinin near the M line suggests that the Ca2+channel may be associated with another binding protein or isoform of α-actinin that links it to the cortical actin cytoskeleton. As shown in Fig. 1 H, immunofluorescent labeling of dystrophin revealed a pattern that is consistent with previous findings of dystrophin localization over Z and M lines in striated muscle (29). Dystrophin colocalized intensely with the Ca2+ channel over Z lines (Fig. 1 I). Labeling of dystrophin over the M line was similar to that found for the Ca2+ channel (compare G and H, Fig.1).
We next asked whether colocalization of the L-type Ca2+channel with ABPs might occur in other tissues and looked in the brain, where Ca2+ channel subtypes have been shown to be localized to distinct cellular structures (13). Cryosections of mouse cerebral cortex (Fig. 1 J) exhibited a punctate pattern of L-type Ca2+ channel staining. α-Actinin colocalized with the Ca2+ channel in an overlapping punctate pattern (Fig. 1 K). Taken together, these observations suggest that, within cardiac muscle and brain, at least two different ABPs belonging to the spectrin family are in the right place to interact with the Ca2+ channel at specific sites.
Effect of dystrophin deletion on Ca2+channel inactivation and voltage dependence.
Cardiac myocytes isolated from mdx mice could not be distinguished by gross morphology from those isolated from wild-type mice. Cells were of similar size as estimated from cell capacitance (wild type: 30 ± 3 pF, n = 33; mdx: 28 ± 2 pF, n = 29) and showed no signs of hypo- or hypertrophy. Ca2+ channel expression and peak activity were also similar, as determined from current density (wild type: 22 ± 2 pA/pF; mdx: 19 ± 2 pA/pF,n = 34). The absence of dystrophin did not appear to affect the ability of the channel to localize within the membrane and did not result in any visible cell changes.
After opening in response to a membrane depolarization, most voltage-gated ion channels turn themselves off in a process known as inactivation. This process occurs despite continued membrane depolarization and prevents the cell from becoming overloaded with Ca2+. Inactivation occurs by multiple mechanisms, two of which are known for Ca2+ channels: a slow, voltage-dependent mechanism and a faster, Ca2+-dependent mechanism (12). Alterations in the cytoskeleton are known to influence both types of inactivation (16). To study primarily the voltage-dependent component, Ba2+ was used as the charge carrier because it does not substitute for Ca2+as effectively in the ion-dependent inactivation mechanism (12).
When we examined the time course of channel inactivation inmdx mice, we found that there was a significant slowing. Figure 2 compares the time course of inactivation in mdx and wild-type cardiac myocytes during 1-s depolarizations (−80 to 0 mV). Inactivation was found to be slowed 63% in mdx mice compared with wild type. Even at a potential that activated the maximal number of Ca2+channels, such a slowing of inactivation might increase Ca2+ influx, leading to cell damage.
Examination of channel voltage dependence of activation revealed that it was shifted 5.2 mV more positive in mdx cardiac myocytes (wild type: V 1/2 = −10.5 ± 1.4 mV,n = 14; mdx:V 1/2 = −5.3 ± 0.7 mV,n = 14; P < 0.01). This is shown in Fig. 3 B by plotting the fraction of channels activated (measured from tail current) vs. membrane potential. This positive shift in channel voltage dependence was also evident in the standard current-voltage relationships shown in Fig. 3 B, inset. There was no change in the steepness (slope factor) of activation (wild type: k =6.7 ± 0.4 mV; mdx, k = 7.3 ± 0.3 mV), and the voltage dependence of steady-state inactivation did not appear to show the same positive shift as activation (wild type:V 1/2 = −29.5 ± 0.8 mV;mdx, V 1/2 = −28.4 ± 1.3 mV).
To determine whether the reduction in channel inactivation was due to the positive shift in activation potential, we examined inactivation of Ca2+ channels in wild-type and mdx cardiac myocytes at varying potentials (−10 to 40 mV in 10-mV increments). By superimposing 200-ms current traces at similar membrane potentials (Fig. 4 A), the difference between inactivation in mdx and wild-type myocytes could be seen. The reduction of inactivation was largest at less positive membrane potentials (Fig. 4 B), and the two curves converged at 40 mV. The parallel positive shifts of channel voltage dependence of activation and of the kinetics of inactivation could be seen by superimposing current traces from wild-type cardiac myocytes at 0 mV and mdx traces at 10 mV (the smallest voltage increment tested). These results suggest that Ca2+ channel interaction with dystrophin or a closely associated protein influences channel gating.
Influence of dystrophin on β-adrenergic and dihydropyridine modulation of the Ca2+ channel.
Adrenergic agonists stimulate cAMP production and activate cAMP-dependent protein kinase (PKA) in cardiac myocytes, leading to increased phosphorylation of the Ca2+ channel, enhancement of channel current, and an increase in the force of contraction (24). PKA is localized near the L-type Ca2+channel by the binding of its regulatory subunit to proteins known as A-kinase anchoring proteins (AKAPs) (16). Because of the importance of PKA anchoring near the channel, disruption of submembrane structure could result in aberrant regulation of the channel. We undertook experiments on β-adrenergic modulation of the L-type Ca2+ channel in cardiac myocytes with the hypothesis that the absence of dystrophin in the mdx mouse might disrupt the interaction between the Ca2+ channel and PKA and that the positive shift in channel voltage dependence and slowing of inactivation could be explained by a reduction in tonic channel phosphorylation by PKA.
Instead, when we compared β-adrenergic upregulation of the Ca2+ channel in wild-type and mdx cardiac myocytes, we found a significant enhancement of upregulation (Fig.5, A–C). Ca2+channel current was monitored by using a voltage ramp protocol (−60 to 60 mV over 200 ms) applied at 5-s intervals. Peak currents measured during the voltage ramp protocol were increased 97.6 ± 12% by 100 nM isoproterenol in wild-type (n = 9) compared with a 181 ± 22% increase in mdx myocytes (n = 14, P < 0.01). Given the sensitivity of ramp currents to differences in inactivation, measurement of changes in channel voltage dependence could not be obtained from this data. Representative currents in the same cell before and after application of isoproterenol are shown in Fig. 5,A and B. The difference in β-adrenergic modulation between wild type and mdx could be explained by a decreased basal level of phosphorylation (leaving more channels to be phosphorylated during stimulation), a decrease in phosphatase and phosphodiesterase activity that opposes phosphorylation (19), or an enhancement in the coupling between phosphorylation and stimulation of channel opening.
To choose among these possibilities, we examined channel inactivation after β-adrenergic stimulation and activation after direct stimulation of the channel by the dihydropyridine agonist BAY K 8644. β-Adrenergic stimulation of the Ca2+ channel caused a negative shift in channel voltage dependence, accompanied by an increase in inactivation. Channel inactivation was enhanced to the same extent in both wild-type and mdx myocytes (wild type: 52 ± 5% increase in inactivation at the end of a 1-s depolarization, n = 10; mdx: 52 ± 1% increase, n = 8), and Ca2+ channels inmdx myocytes still inactivated less than in wild-type myocytes after a maximal stimulation with isoproterenol (wild type: 80 ± 3%, n = 10, mdx: 70 ± 3%,n = 8, P < 0.05). This observation offers evidence against a decrease in basal PKA activity and suggests that the channel in mdx mice is modified in another manner.
To look for functional differences between the Ca2+channels themselves, currents were compared before and after a maximally stimulating concentration (500 nM) of the dihydropyridine agonist BAY K 8644 was applied. Dihydropyridines bind to residues in the pore-forming α-subunit of the L-type Ca2+ channel and allosterically inhibit or enhance channel opening. Ca2+channels in mdx myocytes were stimulated 2.3-fold more by BAY K 8644 than channels in wild-type myocytes (Fig. 5, Dand E), and this stimulation was similar to the 1.9-fold increase in stimulation obtained with isoproterenol. These findings suggest that the enhanced β-adrenergic upregulation results from allosteric changes in the channel itself rather than from changes in the β-adrenergic pathway.
Alterations in the mdx mouse ECG.
In the heart, a slowing of Ca2+ channel inactivation as observed in mdx myocytes might be predicted to prolong ventricular depolarization (prolong QT interval) and could also lead to cardiac arrhythmia. Dystrophin-deficient mice (mdx) were found to have heart rates similar to those of wild-type mice (Fig.6 A, 237 beats/min compared with 248 beats/min). No significant difference was found in PR and QRS intervals or among amplitudes of S, T, or R waves. However, QT (112 ms compared with 79 ms, P < 0.003) and QTc intervals (70 ms compared with 50 ms, P < 0.003) were significantly prolonged in mdx mice (Fig. 6, B andC). Figure 6, D–I, shows characteristic ECG records for wild-type and mdx mice. We found notched QRS complexes in ∼20% of ECGs from mdx mice (Fig.6 H) but not in any ECGs from wild-type mice. QRS notches were also observed in patients with DMD (15). Abnormal regulation of the Ca2+ channel or other ion channels inmdx cardiac myocytes may be manifest at the organ level as a prolonging of the QT interval.
Regulation of Ca2+ channel voltage-dependent inactivation by muscle isoforms of α-actinin.
We next examined whether α-actinin could modulate the activity of the Ca2+ channel in heterologous cells. Channel subunits (α1C, β2A, and α2δ) were coexpressed with muscle and nonmuscle isoforms of α-actinin in the TsA-201 cell line, and channel activity was compared with that in control cells expressing only Ca2+ channel subunits. Current density, a measure of active channel density in the membrane, showed a trend (although not statistically significant) toward smaller currents in the presence of muscle isoforms (control: 21 ± 3 pA/pF, n = 7; smooth muscle: 15 ± 3,n = 6; skeletal muscle: 16 ± 3, n= 5) and larger currents in the presence of nonmuscle isoforms (nonmuscle: 30 ± 6 pA/pF, n = 5; spectrin repeat: 28 ± 5 pA/pF, n = 10).
We next analyzed inactivation and voltage dependence of the Ca2+ channel in the presence and absence of α-actinin. Using 1-s-long depolarizations to induce voltage-dependent inactivation, we found that nonmuscle isoforms of α-actinin had no effect on inactivation (control: 28 ± 5%, n = 13; nonmuscle α-actinin: 27 ± 2%), whereas the muscle isoform significantly slowed inactivation (control: 28 ± 5%,n = 13; α-actinin: 16 ± 3%, n= 9; P < 0.05) (Fig. 7). This result is similar to both that of Krupp et al. (20), in which coexpression of muscle isoforms of α-actinin with the NMDA-type glutamate receptor caused a decrease in inactivation and the reduction in inactivation observed in mdx myocytes that lack dystrophin (Fig. 2).
Given that the reduction in channel inactivation in mdx mice was caused by a shift in voltage dependence, we were surprised to find that current-voltage relationships in the presence of muscle α-actinin did not show any significant changes. Half-activation potentials (V 1/2), the membrane potential at which 50% of the channels are activated, showed no consistent difference between conditions (control: −14.9 ± 0.7 mV, n = 7; skeletal muscle: −17.2 ± 1.2 mV,n = 5; smooth muscle: −11.3 ± 1.5 mV,n = 6; nonmuscle: −6.7 ± 4.3 mV,n = 5; spectrin repeat: −13.4 ± 2.4,n = 10). These data suggest that muscle isoforms of α-actinin regulate Ca2+ channel inactivation at a point past the voltage-dependent step.
In the current study, we found that alterations in the expression of either α-actinin or dystrophin lead to changes in Ca2+channel voltage dependence and kinetics of inactivation. These actin-binding proteins colocalize with the Ca2+ channel at the Z line of cardiac myocytes and the Ca2+ channel and α-actinin colocalize in the brain. At the cardiac myocyte M line where α-actinin is absent, we found that dystrophin colocalizes with the channel. Dystrophin-associated proteins such as syntrophin interact with other ion channels and therefore may also be involved in localization and regulation of the Ca2+ channel in these regions. Our results are in accord with those of Gao et al. (9), who noted an association between the L-type Ca2+ channel and α-actinin in rabbit cardiac myocytes at the Z line.
The absence of dystrophin in mdx myocytes was found to slow inactivation of the L-type Ca2+ channel by shifting the voltage dependence of the channel to more positive potentials, and coexpression of muscle forms of α-actinin with the Ca2+channel in heterologous cells reduced Ca2+ channel inactivation. One interpretation of these results is that α-actinin may replace dystrophin when it is missing. Coming from the same family, these proteins bear many homologous regions and therefore may both be able to bind to the Ca2+ channel. The positive shift in channel voltage dependence in mdx cardiac myocytes was not apparent in these heterologous expression experiments, suggesting that dystrophin and α-actinin modulate the Ca2+ channel by different molecular mechanisms. One possible explanation of the differing results is that α-actinin might modulate channel inactivation, whereas dystrophin might regulate channel voltage dependence. Disruption of dystrophin in mdx myocytes would then cause changes in both properties by altering the channel's interaction with α-actinin, whereas only inactivation is reduced in heterologous cells because dystrophin expression is unchanged.
The difference between the two isoforms of α-actinin (muscle and nonmuscle) lies primarily in the putative Ca2+-binding regions known as EF-hand domains, located near the carboxy terminus. The nonmuscle isoform has two EF-hand domains that are believed to be Ca2+ sensitive, whereas the muscle isoform may have only one EF-hand domain that may not be capable of binding Ca2+. Cotransfection of these isoforms with the NMDA-type glutamate receptor gave similar results (20), whereas inactivation of the receptor was slowed by the muscle isoform alone.
An unexpected result in these experiments was that channel modulation by the β-adrenergic pathway was enhanced in mdx compared with wild-type cardiac myocytes. Similarly, BAY K 8644 increased Ca2+ channel activity in mdx cardiac myocytes more than in wild-type myocytes, suggesting that the channel itself may be allosterically modified such that it is more sensitive to upregulation. Ca2+ channel voltage dependence and inactivation (properties of the pore-forming α1-subunit) have been found to be modified by interactions with channel auxiliary subunits, G proteins, and intracellular synaptic proteins, making it likely that any interaction the channel has with the cytoskeleton would have similar consequences. Dystrophin itself or one of the proteins in the dystrophin-associated protein complex (DPC) is likely a partner. This could include an interaction with auxiliary subunits such as the β-subunit, or a direct allosteric effect on channel gating.
These results have implications for the mechanism of cardiac and central nervous system dysfunction in BMD and DMD. Progressive weakness and wasting of skeletal muscles, nonprogressive cognitive impairment, and failure of the electrical conduction system in the heart are associated with defects or deficiencies in dystrophin (2-4,34). Increased sensitivity of the channel to sympathetic stimulation may cause cardiac tissue to become more susceptible to damage from Ca2+ loading. In addition, cytoskeletal disruption appears to alter Ca2+ channel kinetics, producing a late Ca2+ current that might prolong ventricular depolarization. The significant prolongation of the QT interval in the mdx ECG suggests that Ca2+channel abnormalities play a role in the potential for arrhythmias in these muscular dystrophies. The importance of Ca2+ channels in the development and functioning of the brain may help to explain some of the differences observed in the central nervous system (4, 28).
We thank the LoTurco laboratory for help with immunolabeling and Chu Ngo for preparing cultures, cryosectioning, and other technical assistance.
This work was supported by an American Heart Association Scientist Development Grant (B. D. Johnson) and by start-up funds from the University of Connecticut Research Foundation.
Address for reprint requests and other correspondence: B. D. Johnson, Dept. of Pharmacology and Therapeutics, Univ. of British Columbia, 3650 Wesbrook Mall, Vancouver, Canada BC V6S 2L2 (E-mail:).
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked “advertisement” in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
First published February 13, 2002;10.1152/ajpcell.00435.2001
- Copyright © 2002 the American Physiological Society