Absence of the β subunit (cchb1) of the skeletal muscle dihydropyridine receptor alters expression of the α₁ subunit and eliminates excitation-contraction coupling

Abstract

The multisubunit (α_1S, α₂/δ, β₁, and γ) skeletal muscle dihydropyridine receptor transduces transverse tubule membrane depolarization into release of Ca²⁺ from the sarcoplasmic reticulum, and also acts as an L-type Ca²⁺ channel. The α_1S subunit contains the voltage sensor and channel pore, the kinetics of which are modified by the other subunits. To determine the role of the β₁ subunit in channel activity and excitation-contraction coupling we have used gene targeting to inactivate the β₁ gene. β₁-null mice die at birth from asphyxia. Electrical stimulation of β₁-null muscle fails to induce twitches, however, contractures are induced by caffeine. In isolated β₁-null myotubes, action potentials are normal, but fail to elicit a Ca²⁺ transient. L-type Ca²⁺ current is decreased 10- to 20-fold in the β₁-null cells compared with littermate controls. Immunohistochemistry of cultured myotubes shows that not only is the β₁ subunit absent, but the amount of α_1S in the membrane also is undetectable. In contrast, the β₁ subunit is localized appropriately in dysgenic, mdg/mdg, (α_1S-null) cells. Therefore, the β₁ subunit may not only play an important role in the transport/insertion of the α_1S subunit into the membrane, but may be vital for the targeting of the muscle dihydropyridine receptor complex to the transverse tubule/sarcoplasmic reticulum junction.

Voltage-dependent Ca²⁺ channels are key factors in the control of Ca²⁺-linked cellular functions including muscle contraction (1). Mammalian skeletal muscle contains a dihydropyridine sensitive voltage-dependent Ca²⁺ channel, that is called the dihydropyridine receptor (DHPR), and it is localized to the transverse tubules (T tubules) in adult muscle. The DHPR, in addition to acting as an L-type Ca²⁺ channel, also couples T tubule depolarization to release of Ca²⁺ from the sarcoplasmic reticulum (SR), by triggering the opening of the SR Ca²⁺ release channel also called the ryanodine receptor (RyR-1). This process is referred to as excitation-contraction (E–C) coupling. The DHPRs are present in clusters that are juxtaposed to the junctions between the T tubules and SR, so-called diads, triads, and, in immature cells, peripheral couplings (2). Freeze-fracture analyses have shown that the DHPR complex is present in tetrads (groups of four DHPRs) that are present as arrays, that are arranged opposite an ordered array of ryanodine receptors (2).

The skeletal muscle DHPR is comprised of four subunits, α_1S, α₂/δ, β₁, and γ (3). The α_1S subunit is able to direct the formation of a functional channel when introduced into heterologous expression (4, 5, 6). However, the gating kinetics of the channel are considerably different from the native channel. These differences appear to be due to the absence of the other (α₂/δ, β, and γ) subunits that make up the normal DHPR. Coexpression of the α₁ subunits with combinations of the other subunits in heterologous expression systems can return the activity and kinetics of the channel to values more representative of those seen for the native channel. However, the particular effect observed is dependent on both the expression system used and the specific subunits under study. The skeletal muscle β subunit has a large effect on the functional expression of the α_1S subunit from skeletal muscle (7). Coexpression of the β subunit with α₁ subunits increases the Ca²⁺ current and accelerates activation and inactivation more than tenfold (8, 9, 10). Features of the primary structure of the rabbit skeletal β₁ polypeptide (11) and biochemical data suggested that it is a peripheral membrane protein that interacts with an intracellular domain of the α₁ subunit (12). Subsequently, a β binding site on the α₁ subunits (13), and an α₁ binding site on the β subunits (14) have been identified. The β subunit also has been shown to facilitate insertion of the α₁ subunit into the membrane of HEK293 cells (15).

The β₁ subunit is expressed at high levels in skeletal muscle and brain and at lower levels in spleen and heart (11, 16, 17). Alternate splicing produces three isoforms, one of which is expressed exclusively in skeletal muscle (17). To determine directly the role of the β₁ subunit in DHPR kinetics and E–C coupling we used gene targeting to inactivate the β₁ gene. The absence of the β₁ subunit eliminates E–C coupling and modifies the Ca²⁺ currents in skeletal muscle cells. In addition, the absence of the β₁ subunit results in a failure of the α₁ subunit to be transported to and/or inserted appropriately into the T tubule membrane.

MATERIALS AND METHODS

Gene Targeting.

A bacteriophage clone containing a portion of the β₁ gene was isolated from a mouse 129Sv genomic library (Stratagene no. 946305). A 6.5-kb genomic DNA fragment was subcloned into a pBluescript-derived vector that contains a herpes simplex virus thymidine kinase expression cassette. To inactivate the β₁ gene, an internal 1.5-kb fragment, containing exons 6 (used in all transcripts), 7a (used in muscle), and 7b (used in brain), was replaced by a 1.6-kb fragment containing the neo gene linked to a thymidine kinase promoter and a polyadenylylation signal (provided by R. Behringer, M. D. Anderson Cancer Center, Houston). The targeting vector also contains the herpes simplex virus thymidine kinase gene to allow for negative selection of nonrecombinants. The targeting vector retains 2.5 kb of identity with the native β₁ gene upstream and 2.1 kb of identity downstream of the neo cassette. A unique BamHI restriction site was used to linearize the plasmid prior to its introduction into mouse embryonic stem (ES) cells by electroporation. Colony selection and identification of targeted clones using probe 2 was as described (18). 5 μg of the targeting vector was electroporated into 5 × 10⁶ AB1 ES cells. Ninety-two G418 and 1-(2′-deoxy-2′-fluoro-β-d-arabinofuranosyl)-5-iodouracil- resistant clones were analyzed and 19 showed specific targeting with probe 2. Three were expanded and analyzed with probe 1, and used to produce germ-line chimeras.

Histology.

For light microscopy, tissues were fixed in formalin in saline, embedded in paraffin and 7-μm sections prepared and stained with hematoxylin and eosin. For electron microscopy, the hindlimbs were immersed in Karnovsky’s fixative, post fixed in osmium tetroxide, dehydrated, and embedded in resin. Sections were cut and stained with uranyl acetate and lead citrate.

Contraction Measurements.

Experiments were done essentially as described (19). Embryonic day 16 (E16) fetuses were used for these experiments because bone calcification in E18 fetuses prevented forelimb contraction. The skinned forelimb was attached by a silk filament at the wrist and the humerus, to a force transducer, and the other end to a fixed wire. The preparation was immersed in Krebs–Ringer solution (136 mM NaCl/5 mM KCl/2 mM CaCl₂/1 mM MgCl₂/10 mM Hepes, pH 7.2) at 23°C. Electrical field stimulation was provided by platinum electrodes placed on each side of the preparation. Caffeine was added by changing the bath solution.

Action Potentials, Ca²⁺ Currents and Ca²⁺ Transients.

Ribcages of E18 fetuses were dissected in Krebs–Ringer solution. The tissue was incubated at 37°C for 10 min in PBS containing 3 mg/ml collagenase (Sigma, type I) and 1 mg/ml trypsin (Sigma, type III) to release myotubes. Cells were held under either current or voltage clamp with pipettes that had a tip resistance of 2–7 MΩ when filled with the internal solution. Action potentials were recorded in Krebs–Ringer. The pipette solution was 140 mM KCl, 5 mM MgCl₂, 10 mM Hepes·Tris, pH 7.3. For Ca²⁺ current recordings the external solution was 130 mM tetraethylammonium-methanesulfonate, 10 mM CaCl₂, 1 mM MgCl₂, 2 μM tetrodotoxin, and 10 mM Hepes, pH 7.2, and the pipette solution was 140 mM Cs-aspartate, 5 mM MgCl₂, 5 mM EGTA, 10 mM Mops, pH 7.2. Intracellular Ca²⁺ was measured in a photomultiplier-based microfluorimeter. Cells were loaded with 0.1 μg/ml fura-2AM (Molecular Probes) for 20–30 min. A 30 × 50-μm slit was centered on the cell held under current clamp. Fluorescence intensity ratios, F340/F380, were taken at a rate of 200 ratios/sec. Fluorescence intensity ratios during stimulation were divided by the ratio at rest.

Immunofluorescence Labeling of Cultured Cells.

Primary muscle cultures were derived from myoblasts isolated from E18 fetuses. Cultures were grown and fixed on gelatin-coated primaria culture dishes (Becton Dickinson) and processed for immunohistochemistry as described (20). Monoclonal antibodies to DHPR subunits α_1S (Chemicon) and β₁ (Upstate Biotechnology, Lake Placid, NY) were used at dilutions of 1:200 and 1:10 respectively. Antibodies to RyR-1 (Upstate Biotechnology) were used at 1:100. Primary antibodies have been shown to be specific to the various proteins by Western blot analysis. Secondary antibodies used were fluorescein conjugated polyclonal goat anti mouse IgG (Cappel) at a dilution of 1:200.

RESULTS

Targeted Disruption of the β₁ Subunit Gene (cchb1) Produces Perinatal Lethality.

Gene targeting was done as described in Materials and Methods (Fig. 1). Three independent targeted ES cell clones were injected into C57BL/6 blastocysts and chimeras obtained. All three ES clones were transmitted through the germ line, and all showed the same phenotype when the mutation was homozygous. Heterozygous mutant mice (+/cchb1^tm1) were indistinguishable from wild type (+/+) animals. Mice homozygous for the targeted mutation (cchb1^tm1/cchb1^tm1), hereafter called β₁-null, failed to survive the perinatal period. However, analyses of E18 fetuses revealed that ≈25% were homozygous for the modified β₁ allele. The β₁-null fetuses fail to show any movement, have flexed necks and extremities, curvature of the spine, and thin limbs. Compared with control littermates the β₁-null fetuses show a drastic reduction in muscle mass (Fig. 2 a and b). Muscle from β₁-null fetuses contains myofibrils with well-defined Z-bands, but the thick and thin filaments are disorganized compared with controls (either +/+ or +/cchb1^tm1, Fig. 2 d and e). T tubule and SR junctional complexes were identified in β₁-null fetal muscle (Fig. 2c). Both the gross morphology and the histology of the β₁-null fetuses are similar to the muscular dysgenesis (21) and RyR-1 knockout (skrr^m1) mutants (19).

Figure 1.

Gene targeting of the β₁ subunit gene. (a) Portion of the β₁ gene showing relevant restriction enzyme sites and gene structure. (b) Targeting vector. (c) Modified β₁ allele expected after homologous recombination between the endogenous locus (a) and the targeting vector (b). Exons are represented by solid boxes. The neo gene is shown by an open box. X, XmnI; B, BamHI; H, HindIII; S, SpeI; E, EcoRV.

Figure 2.

Morphology of E18 β₁-null fetuses shows a reduction in muscle mass and disorganization of the myofibrils. (a and b) Transverse sections through the hindlimbs of a control (a) and β₁-null (b) fetus. (Bar = 0.3 mm.) Sections were cut at approximately the same position and the tibia (T) and fibula (F) can be seen. The amount of muscle in the β₁-null fetus is markedly decreased. (c) Electron micrograph of a triad from a β₁-null fetus. (Bar = 0.1 μm.) (d and e) Electron micrograph of muscle from control (d) and β₁-null (e) fetus. (Bar = 1.2 μm.)

β₁-Null Fetuses Lack E–C Coupling.

The absence of any movement by the β₁-null fetuses indicated that E–C coupling may be impaired. This was examined by measuring the twitch contractions of the forelimb of E16 day control and β₁-null fetuses. Electrical field stimulation-induced twitches in forelimbs from control fetuses (Fig. 3a), but failed to induce twitches in forelimbs from β₁-null fetuses (Fig. 3b). Exposure of both preparations to caffeine produced a slow increase in tension (Fig. 3 c and d), although the force produced by the β₁-null forelimb is reduced 20–30-fold compared with controls. This is probably due to the decrease in muscle mass, because Ca²⁺ transients in single cells, while highly variable in magnitude, are similar to those in control cells (data not shown). Tetrodotoxin-sensitive action potentials can be elicited in both control and β₁-null myotubes (Fig. 3 e and f) indicating membrane excitability is normal. In control myotubes, the induced action potential produced a Ca²⁺ transient (Fig. 3g). However, in β₁-null myotubes the action potential failed to elicit a Ca²⁺ transient (Fig. 3h). These data show that membrane excitability and the ability of the SR to accumulate Ca²⁺ are normal in β₁-null myotubes. The lack of E–C coupling indicates that the β₁ subunit has an important role in either transport and/or assembly or function of the DHPR.

Figure 3.

Contractile properties, action potentials and Ca²⁺ transients in control and β₁-null tissues. (a) Electrical field stimulation induced contractions in forelimb from control fetus. Contraction is seen in response to each stimulus. (b) as in a except from β₁-null fetus. Stimulus fails to produce contraction. (c and d) Caffeine (25 mM) addition to the bath solution induced contraction in control (c) and β₁-null (d) tissue. The response in the β₁-null tissue is 20-fold lower than the control tissue. (e and f) Action potentials can be induced in both control (e) and β₁-null (f) myotubes and are tetrodotoxin sensitive. (g and h) Action potentials (Inset) induce a Ca²⁺ transient in control myotubes (g), but fail to induce a Ca²⁺ transient in myotubes from β₁-null (h) fetuses.

L-Type Ca²⁺ Currents are Decreased in the β₁-Null Myotubes.

Whole cell patch-clamp was used to characterize the Ca²⁺ currents in freshly dissociated intercostal myotubes. Both control and β₁-null myotubes contain a transient, rapidly inactivating (T-type) current that can be observed using a test potential of −20 mV (Fig. 4 a and b). T-type currents in control and mutant myotubes had similar peak amplitudes (legend to Fig. 4). At positive test potentials a large sustained slowly inactivating (L-type) inward Ca²⁺ current is observed in control cells (Fig. 4a). In contrast, the L-type current present in the β₁-null myotubes at positive test potentials, while present, is decreased ≈10- to 20-fold (Fig. 4b). Current-voltage relationships of the L-type currents were determined 1 ms before the end of a 300-ms test pulse, and are shown in Fig. 4c. The L-type current density in the β₁-null cells is 16-fold lower and the peak current is shifted 10 mV to more positive potentials when compared with control myotubes. In the current-voltage curves of β₁-null cells the peak at −20 mV represents residual, approximately 10% of maximum, T-type current that is only detectable in the records at negative test potentials. A similar amount of T-type current is present in control cells at negative test potentials, however this does not appear on the curves because of the scale difference for the two curves. The peak at +25 mV represents the L-type current and contains no T-type current.

Figure 4.

Ca²⁺ currents and current-voltage curves for control and β₁-null myotubes. (a and b) Ca²⁺ currents from control (a) and β₁-null (b) myotubes. A holding potential of −80 mV was used. Test potentials are indicated next to each trace and lasted 300 ms. The control myotubes show a T-type current activated by weak depolarizations [I_(peak) = −1.62 ± 0.19 pA/pF at −20 mV; n = 14] and larger L-type currents activated by the strong depolarizations [I_{(end of pulse)} = −5.20 ± 0.38 pA/pF at +20 mV; n = 14]. In the β₁-null myotube, the T-type current is unaltered [I_(peak) = −1.59 ± 0.17 pA/pF at −20 mV; n = 19], however the L-type current is reduced 10-fold [I_{(end of pulse)} = −0.40 ± 0.04 pA/pF at +20 mV; n = 19]. c, Current-voltage relationships for control and β₁-null myotubes. Current amplitude was determined at the end of a 300-ms test pulse. Note difference in I_Ca scales for the control (•) and β₁-null (▴) myotubes.

The β₁ Subunit Alters Expression of α₁ Subunit.

Our observation that the absence of the β₁ subunit drastically reduces the L-type current could arise from two alternative mechanisms. First, the β₁ subunit may be required to produce a functional channel. Second, the β₁ subunit may be required for correct insertion and/or localization of the α₁ subunit into the T tubule membrane. To examine this latter possibility, we used immunohistochemistry to determine the distribution of the α₁ and β₁ subunits in control, β₁-null, and dysgenic myotubes. In normal myotubes there is a punctate expression pattern for both the β₁ (Fig. 5a) and α_1S (Fig. 5b) subunits. The punctate staining pattern is sometimes found in a random array or in curled patterns (arrows in Fig. 5 a and b) indicative of early T tubule/SR junctions. The pattern of staining for α_1S and β₁ are equivalent in control myotubes, however the staining intensity is not always the same because of the different affinities of the two antibodies. Double labeling of the α_1S subunit and the RyR-1, reveals that they colocalize in control mouse myotubes (22). In the β₁-null myotubes, there is no β₁ subunit (Fig. 5c) (as would be expected), and also no detectable α_1S subunit (Fig. 5d). In addition, there is no evidence that the α_1S subunit is present elsewhere in the cell. Because the absence of the α_1S subunit was unexpected, we routinely stained dysgenic myotubes as negative controls and normal cultures as a positive controls. In all experiments the staining for the α_1S subunit in the β₁-null and dysgenic cells were essentially identical. Because the α_1S subunit is absent in the dysgenic myotubes (20) these data indicate that the α_1S subunit is in fact either absent, or present at undetectable levels, in the β₁-null myotubes.

Figure 5.

Immunofluorescence of control, β₁-null and dysgenic, α_1S-null, myotubes with antibodies to β₁ and α_1S DHPR subunits and RyR-1. (a and b) Control myotubes labeled with either anti-β₁ (a) or anti-α_1S (b). Both show a punctate pattern (arrows) reflecting aggregates of the respective DHPR subunits at putative coupling junctions (triads, diads, pentads, and/or peripheral couplings) (20). (c and d) β₁-Null myotubes labeled with either anti-β₁ (c) or anti-α_1S (d). (e) Dysgenic myotubes labeled with anti-β show a weak, but punctate pattern (arrows). (f) β₁-Null myotubes labeled with anti-RyR-1 in a punctate pattern (arrows). The diffuse background staining in the myotube may represent RyR-1 that is not membrane localized. (Bar = 20 μM.)

The β subunit is present in dysgenic myotubes and shows a punctate and diffuse expression pattern (Fig. 5e). Similarly, RyR-1 is present in β₁-null myotubes as both punctate and diffuse staining (Fig. 5f), a pattern reminiscent of the staining for RyR-1 in dysgenic cells (22). The punctate staining is most likely from T tubule/SR junctions and the diffuse staining from mistargeted or nonclustered proteins. Staining of dysgenic myotubes for the α₂ subunit produces a diffuse pattern and no punctate staining, indicating it is not clustered at T tubule/SR junctions (20, 23).

The absence of either diffuse or punctate staining for the α_1S subunit in the β₁-null myotubes suggests that the β₁ subunit is required for either correct transport, insertion, or localization of the α_1S subunit into the T tubules. In contrast, the α₁ subunit is not required for the appropriate localization of the β₁ subunit as indicated by punctate staining for the β₁ subunit in dysgenic myotubes.

DISCUSSION

We used gene targeting to inactivate the β₁ subunit of the multisubunit skeletal muscle L-type Ca²⁺ channel in mice. Homozygous mutant fetuses have a phenotype that is very similar to that seen in mice with mutations in either the α_1S subunit, muscular dysgenic (21), or RyR-1, skrr (19). All three mutants lack E–C coupling, and so it is likely that the gross phenotype is due to the effects of the absence of E–C coupling rather than the absence of the respective subunit. Based on heterologous expression experiments, the β₁ subunit was thought to modulate channel kinetics and possibly increase the expression of the α₁ subunit and so the complete lack of E–C coupling in the β₁-null mice was somewhat unexpected. Analyses of skeletal muscle myotubes from the β₁-null mice demonstrates that the L-type Ca²⁺ current is reduced 10-20-fold. Given our data show that the α_1S subunit is absent or extremely low, the residual L-type current in the β₁-null cells could be the same as I_dys, present in dysgenic cells (24). However, confirmation that this is true will require careful analyses of the residual currents in both dysgenic and β₁-null myotubes.

There is considerable evidence that β subunits modulate channel kinetics (25). In addition, evidence is accumulating that β subunits play an important role in increasing the expression of α₁ subunits (26). Most recently, Chien et al. (15) demonstrated that the β subunit was involved in the transport and/or insertion of the α_1C subunit into the membrane of HEK293 cells. In contrast, other studies using charge movement as a measure of the amount of α₁ subunit present in the membrane, showed that coexpression of a β_2A subunit had little effect on the expression of the α_1C subunit (27). These differing results could reflect the differences in the expression system used. We demonstrate, using immunohistochemistry, that the β subunit is important for clustering of the α₁ subunit at the triad in vitro. It is possible that the α₁ subunit is dispersed throughout the cell membrane and therefore difficult to detect, however, the 60% decrease in charge movement and 70% reduction in PN200-110 binding in the β₁-null cells (28) would not support this as the only explanation.

The mechanism by which expression of the α₁ subunit is controlled by the β₁ subunit is unknown, but it is possible that the failure of the formation of the α₁/β complex results in the degradation of the α₁ subunit. Many channel complexes are assembled at the endoplasmic reticulum and the interaction between newly synthesized subunits may be important to prevent rapid intracellular degradation (29). A recent study of potassium channel assembly showed that the KVβ2 subunit binds to the KV1.2 α subunit in the endoplasmic reticulum. This binding facilitates the posttranslational modification of KV1.2, and its subsequent transport, probably as a KVβ2-KV1.2 complex, to the plasma membrane (30). Chien et al. (15) proposed a similar role for the β subunits of the voltage-dependent Ca²⁺ channels.

Our studies indicate that the transport and/or insertion of the α_1S subunit into the membrane is dependent on the presence of the β₁ subunit, but that the converse is not true. We show that in skeletal muscle cells, the β₁ subunit is clustered, most likely at T tubule/SR junctions in the absence of the α_1S subunit. The insertion of the α₂ subunit into the membrane also is independent of the expression of the α₁ subunit, although it is not localized appropriately as indicated by diffuse staining in dysgenic myotubes (20). Whether the α₂ subunit is inserted into and clustered in the membrane of the β₁-null myotubes is currently under investigation. At this time the role of the γ subunit in any of these processes is unknown.

The skeletal muscle DHPR occupies a unique niche among Ca²⁺ channels in that its primary function, to transduce T tubule membrane depolarization into opening of the SR Ca²⁺ release channel or RyR-1, is independent of its role as a Ca²⁺ channel. The biogenesis of the T tubule/SR junctional complex containing the DHPRs and RyR-1 has received considerable attention. Franzini-Armstrong and colleagues (31) have shown that the junctional regions in skeletal muscle myotubes are formed in the absence of the RyR-1 and DHPR, but that the insertion of RyR-1 and DHPR into the SR and T tubules, respectively, serves to expand and stabilize the tertiary structure. The insertion of RyR-1 is independent of the insertion of the DHPR α₁ and β₁ subunits because in both dysgenic (20, 23) and β₁-null myotubes (Fig. 5f), the immunofluorescence staining shows a punctate pattern for RyR-1 similar to that seen in control myotubes. Studies on myotubes from mice that lack RyR-1 (skrr1) indicate that the normal tetrad arrays of DHPRs in the T tubule are not formed (31), even though biophysical data indicate that the DHPRs are present in the membrane, however they do not act as Ca²⁺ channels presumably because they fail to form tetrads (32). Therefore, unlike RyR-1 that is positioned appropriately in the T tubule/SR junctional region in the absence of DHPRs, the localization and/or organization of the DHPRs in this region appears dependent on interactions with RyR-1. The punctate distribution of the β subunit in the absence of the α₁ subunit in dysgenic myotubes (Fig. 5e), indicates that the β subunit can be transported to the T tubule/SR junction in the absence of the α₁ subunit. Presumably the β subunit binds to an as yet unidentified component of the T tubule/SR junction. The availability of β₁-null mice will provide valuable tools with which to dissect the multifunctional nature of the skeletal muscle β₁ subunit. These functions include, modulation of channel kinetics, transport, and/or insertion of the α₁ subunit into the membrane and possibly the targeting of the DHPRs to the T tubule/SR junction. Whether β subunits in other tissues are involved in appropriate cellular or subcellular localization of voltage-dependent Ca²⁺ channels remains to be determined. However, this possibility is intriguing given that many tissues express more than one of the four β subunit genes and that each gene produces splice variants. The availability of β₁-null mutant mice will provide a valuable resource for the complete dissection of this multifunctional protein.