Functional Expression and Characterization of Skeletal Muscle Dihydropyridine Receptors in Xenopus Oocytes*

Dihydropyridine receptors in vertebrate skeletal muscle serve a dual role: as voltage sensors for excitation-contraction coupling and as voltage-activated calcium channels. Although they were the first of six classes of calcium channels to be cloned, skeletal muscle dihydropyridine receptors remain the only ones not functionally expressed as calcium channels in Xenopus oocytes, leading to the hypothesis that an interacting component is missing. Using β1b, an isoform previously found in brain, we have for the first time reconstituted skeletal muscle calcium channel function in Xenopus oocytes. We show that this β subunit is necessary for functional expression and that the α2δ subunit significantly enhances the expressed current. The majority of the α1 subunit in skeletal muscle is a truncated form. Here we show that both the full-length and truncated forms produce functional calcium channels in Xenopus oocytes, but the truncated form gives significantly larger currents. In addition, we show that the β1b transcript is expressed in rat skeletal muscle, although at a much lower level than the abundant β1a isoform.

Dihydropyridine receptors in vertebrate skeletal muscle serve a dual role: as voltage sensors for excitationcontraction coupling and as voltage-activated calcium channels. Although they were the first of six classes of calcium channels to be cloned, skeletal muscle dihydropyridine receptors remain the only ones not functionally expressed as calcium channels in Xenopus oocytes, leading to the hypothesis that an interacting component is missing. Using ␤ 1b , an isoform previously found in brain, we have for the first time reconstituted skeletal muscle calcium channel function in Xenopus oocytes. We show that this ␤ subunit is necessary for functional expression and that the ␣ 2 ␦ subunit significantly enhances the expressed current. The majority of the ␣ 1 subunit in skeletal muscle is a truncated form. Here we show that both the full-length and truncated forms produce functional calcium channels in Xenopus oocytes, but the truncated form gives significantly larger currents. In addition, we show that the ␤ 1b transcript is expressed in rat skeletal muscle, although at a much lower level than the abundant ␤ 1a isoform.
Dihydropyridine receptors (DHPRs) 1 in skeletal muscle serve two functions: as voltage sensors for excitation-contraction coupling and as slow voltage-gated calcium channels (1). They are composed of ␣ 1 , ␣ 2 ␦, ␤, and ␥ subunits (2,3). The ␣ 1 subunit is the main pore-forming subunit; the other subunits, especially ␤, play important roles modulating channel properties (2,3). Although they were the first of six classes of calcium channel ␣ 1 subunits to be cloned, the skeletal muscle ␣ 1S subunit from several sources (4, 5) has not been expressed in Xenopus oocytes despite efforts to do so (6 -9). The rabbit skeletal muscle ␣ 1S subunit expresses poorly in L-cells (10 -12) where coexpression with ␤ 1a speeds up channel kinetics without stimulating current levels (11,12). However, in contrast to the sensitivity shown by skeletal muscle channels in vivo (13), in L-cells coexpression with the ␤ 1a subunit, especially at high levels, makes the expressed channel insensitive to the calcium channel dihydropyridine agonist Bay K 8644 (12,14). Although coexpression of ␤ 1a and ␣ 2 ␦ subunits with the cardiac ␣ 1C subunit increases current dramatically (15), these subunits do not stimulate skeletal muscle ␣ 1S currents. However, the expressed ␣ 1S subunit protein in membrane is increased (16). Recently, an ␣ 1S was expressed in human embryonic kidney cells, but currents were still not detected in most cells in the absence of the agonist Bay K 8644 (17).
The difficulty of reconstituting skeletal muscle calcium channels in nonmuscle cells has led to the hypothesis that some interacting component is missing (18). We report here the successful functional expression and characterization of skeletal calcium channels in Xenopus oocytes. We show that one particular splice variant of the ␤ 1 subunit is especially important for this expression. We also demonstrate for the first time a functional difference between the full-length and truncated ␣ 1S forms found in skeletal muscle.
RT-PCR-Reverse transcription reactions using 1 g of poly(A) ϩselected RNA and a ratio of 1 g primer/g RNA were done according to Promega Technical Bulletin 502. The 50-l PCR reaction mixture contained: 0.2 mM each dNTP, 20 mM Tris buffer, pH 8.4, 50 mM KCl, 1.5 mM MgCl 2 , 0.1 M each primer, 1.25 units of Taq DNA polymerase, and 0.1 l of reverse transcription reaction. Following an initial treatment for 2 min at 94°C, the following cycle was repeated 35 times: 1 min at 94°C, 1 min at 50°C, and 1 min at 72°C. Final extension was 10 min at 72°C.

Functional Expression of Skeletal Muscle Calcium Channels in Xenopus
Oocytes-␤ 1a and ␤ 1b are alternatively spliced isoforms of the ␤ 1 subunit. ␤ 1a is found predominantly in skeletal muscle (3), whereas ␤ 1b is found in brain (3). As a positive control, both isoforms are shown to stimulate calcium channel currents when the cardiac ␣ 1C⌬N60 subunit is expressed in Xenopus oocytes (Fig. 1A). As reported previously, no significant barium currents are detected when the skeletal muscle ␣ 1S subunit is coexpressed with ␤ 1a and the other subunits (␣ 2 ␦ and ␥) (Fig. 1B). To determine whether interaction with other ␤ isoforms stimulates skeletal muscle ␣ 1S current expression, we coexpressed ␣ 1S , ␣ 2 ␦, and ␥ with ␤ 1b , the brain ␤ isoform. Surprisingly, appreciable barium currents (293 Ϯ 24 nA, n ϭ 15, V test ϭ 30 mV) were detected 3-5 days after cRNA injection (Fig. 1C). This current activates slowly ( ϭ 33.7 Ϯ 3.3 ms, n ϭ 15, V test ϭ 30 mV), which is typical for skeletal muscle L-type calcium channels (31). In contrast, endogenous currents (recorded from cells injected with ␤ 1b , ␣ 2 ␦, and ␥ but not ␣ 1 ) are much smaller and activate much faster (Fig. 1, B and C), making it unlikely that the detected current from the ␣ 1S /␤ 1b / ␣ 2 ␦/␥ combination is a contamination. Like skeletal muscle calcium channel currents (17), the expressed current is sensitive to 1 M (Ϫ) Bay K 8644, a calcium channel agonist (data not shown). Thus, we have defined, for the first time, conditions for recording calcium channel currents from skeletal muscle DHPRs expressed in Xenopus oocytes and have shown a differential effect of ␤ 1 splice isoforms on this DHPR.
Functional Characterization of the Truncated Skeletal Calcium Channel ␣ 1 Subunit-In skeletal muscle, the majority of ␣ 1S is a truncated form, presumably formed by post-translational proteolytic processing (1,32,33). To determine whether the truncated form also expresses calcium channel currents in Xenopus oocytes, we constructed ␣ 1S⌬C1698 , truncated at amino acid 1698, where the proteolytic processing most likely occurs in vivo (33). When the truncated form is coexpressed with ␣ 2 ␦, ␥, and ␤ 1a , little or no current (Ͻ100 nA) is detected (Fig. 2, A  and B). However, in the presence of ␤ 1b , currents are readily detectable 3 days after injection (Fig. 2, A and B). The maximum currents from the truncated ␣ 1S⌬C1698 (2.2 Ϯ 0.2 A, n ϭ 37) are significantly larger than those from the full-length form (0.3 Ϯ 0.02 A, n ϭ 15). With the truncated form, currents as large as 6 A are found 7 days after cRNA injection (data not shown). Consistent with the pharmacological properties of skeletal muscle calcium channels, the expressed currents are blocked by 10 M nifedipine, a dihydropyridine antagonist (data not shown).
Role of Auxiliary Subunits on Skeletal Muscle ␣ 1S Expression in Xenopus Oocytes-Skeletal muscle calcium channels are composed of the main pore-forming ␣ 1 subunit and the auxiliary subunits ␣ 2 ␦, ␤, and ␥ (35). We recorded from oocytes expressing different subunit combinations to test if ␣ 1S currents are enhanced by other auxiliary subunits. Our results (Fig. 3) show that in addition to the ␤ subunit, the ␣ 2 ␦ subunit also significantly enhances the expressed currents (cf., Fig. 3, A  and C). The ␥ subunit does not have much effect on the expressed current level (Figs. 3, A and B). currents. Representative barium current traces are shown for test potentials of Ϫ40, 0, and ϩ20 mV. B and C, ␤ 1b , but not ␤ 1a , stimulates rabbit skeletal muscle ␣ 1S . The scale bar in C is also for B. B, representative current traces from oocytes injected with ␣ 1S /␤ 1a /␣ 2 ␦/␥ (left) and oocytes injected with ␤ 1b /␣ 2 ␦/␥ (right) are shown. Test potentials are Ϫ40 mV and ϩ30 mV. C, representative current traces from oocytes injected with ␣ 1S /␤ 1b /␣ 2 ␦/␥ are shown (left panel). The right panel shows peak inward current versus test potential (I/V) curves for the ␣ 1S /␤ 1b / ␣ 2 ␦/␥ combination (q, n ϭ 15) and for ␤ 1b /␣ 2 ␦/␥ (E, n ϭ 9). The error bars show S.E.
Detection of the ␤ 1b Isoform in Skeletal Muscle-Our data show that ␤ 1b is unique in its ability to stimulate functional expression of skeletal calcium channel ␣ 1 subunits expressed in Xenopus oocytes. Does ␤ 1b play a role in vivo in expression of calcium currents in skeletal muscle? If so, then it should be expressed in that tissue. Previous reports showed that ␤ 1a is the major ␤ isoform in skeletal muscle (3). To determine whether ␤ 1b is also present, we probed a Northern blot from rat skeletal muscle with probe N (Fig. 4A) that recognizes both ␤ 1a and ␤ 1b . Consistent with previous reports (24,26), a 1.6-kilobase transcript (the size of ␤ 1a ) is detected in skeletal muscle and a 2.9-kilobase transcript (the size of ␤ 1b ) is found in brain when short exposure times are used (lanes 1:SkM and 3:Br in Fig. 4B). After a longer exposure, a band the size of ␤ 1b (2.9 kilobase) appears in skeletal muscle mRNA (Fig. 4B, lane 4). Another probe specific for ␤ 1b (C in Fig. 4A) also recognizes the 2.9-kilobase ␤ 1b message in skeletal muscle (lane 5, Fig. 4B). (The additional bands seen at 2.5 and 4.4 kilobase presumably represent additional ␤ 1 splice variants not characterized in this study.) We calculate by phosphoimaging that the relative abundance of the ␤ 1b message in skeletal muscle mRNA is 2.64 Ϯ 0.05% (mean Ϯ S.D., n ϭ 2) that of ␤ 1a . To test further for the presence of ␤ 1b in skeletal muscle, we did RT-PCR using mRNA from skeletal muscle or brain as template and primers (PF and PR, Fig. 4A) specific for ␤ 1b . A PCR product of the size predicted for ␤ 1b (716 base pairs) is detected in rat skeletal muscle (Fig. 4C,  lane 2), although it is significantly less abundant than the product produced with brain mRNA (Fig. 4C, lane 4) as the template. DISCUSSION We have reconstituted mammalian skeletal muscle calcium channel function in Xenopus oocytes. Because the slow calcium channel current is significantly reduced in the dyspedic mouse lacking ryanodine receptors (18,37), it was hypothesized that the difficulty in functionally reconstituting the skeletal muscle calcium channels in nonmuscle cells was due to a missing interaction between dihydropyridine receptors and ryanodine receptors. Although we cannot rule out a possible interaction of our expressed DHPRs with Xenopus ryanodine receptors, our results point to an essential role for the ␤ subunit. Our finding that a ␤ subunit dramatically increases the expressed current is similar to its effect on cardiac calcium channels (15) but seems to contradict previous expression studies in L-cells where coexpression of the ␤ 1a subunit did not increase the current even though drug binding and ␣ 1S membrane protein was increased (11,12,16). The major difference between our study and the one in L-cells (11,12,16) is the ␤ 1 subunit isoform used. We used ␤ 1b , whereas the L-cell study used ␤ 1a . Our demonstration of the importance of the ␤ subunit in skeletal muscle calcium channel function agrees with results in knock-out mice where disruption of the ␤ 1 subunit gene led to a dramatic reduction of both the ␣ 1S membrane protein and the slow calcium channel current (38,39).
In skeletal muscle we found that the ratio of ␤1a to ␤1b mRNAs is 100:2.6. The ratio of these two proteins in skeletal  N (lanes 1-4) or C (lane 5). The final two washes for the blots were: 0.1 ϫ SSC, 0.1% SDS at 65°C for 30 min. Lane 1, rat skeletal muscle; lane 2, negative control (yeast RNA); lane 3, rat brain. Lanes 1-3 were exposed for 5 h. Lane 4 was the same as lane 1, but exposed for 40 h. Lane 5 was rat skeletal muscle probed with C and exposed for 10 days. C, RT-PCR using primers PF and PR with poly(A) ϩ RNA from rat skeletal muscle (lane 2), brain (lane 4), or water (lane 3). Lane 1 contains a 100-base pair DNA ladder (Life Technologies, Inc.). muscle and their relative affinities for ␣ 1S are unknown. Our experiments describe the effects of ␤1a and ␤1b individually on expression of skeletal muscle calcium channel currents in oocytes. Conditions for current expression in actual muscle may differ from those we describe for oocytes. Along these lines it will be interesting to determine whether ␤1b can outcompete ␤1a for interaction with ␣ 1S . Such information may aid in understanding the importance of ␤1 isoforms in the in vivo function of dihydropyridine receptors as voltage sensors and as voltage-activated calcium channels.
The ␤ 1a and ␤ 1b used in this study differ at many sites. There are major sequence differences in the middle region (which separates the two highly conserved domains I and II, Fig. 4A) and also in the C terminus (24,26) beginning ϳ32 amino acids after the end of Domain II. In addition, even within the conserved regions there are isolated, single amino acid differences (indicated by the small arrows under ␤ 1a in Fig. 4A). There are two amino acid differences in the N terminus, three in Domain I, two in the conserved part of the C terminus immediately adjacent to Domain II, and five in Domain II itself (four of which lie at the very beginning of Domain II in or near the beta interaction domain (40)). Studies are underway to determine the molecular site(s) responsible for the differential effects of the ␤ subunit isoforms on ␣ 1S expression.
The skeletal muscle ␣ 1S has two forms presumably generated by proteolytic processing: the less abundant full-length form and the more abundant truncated one (33). The difference in function of these two forms has been elusive. The injection of cDNA constructs encoding the full-length form or a truncated form into dysgenic myocytes lacking ␣ 1S restores both the excitation-contraction coupling and slow calcium channel currents (41). This rescue experiment fails to reveal any functional differences between the two forms, presumably because the protein product from the full-length form can be cleaved into the truncated form in the myocytes. Our study in Xenopus oocytes, which may not have the same proteolytic cleavage machinery found in skeletal muscle, suggests a significant functional difference between these two forms, with the truncated form yielding much larger currents. This is reminiscent of cardiac ␣ 1C where deletion of up to 70% of the C terminus increases the channel current (42). With three cAMP-dependent protein kinase phosphorylation consensus sites in the C terminus removed in the truncated form of ␣ 1S , the two isoforms have different phosphorylation patterns (43). The ability to reconstitute both forms in Xenopus oocytes will allow a determination of whether their function is differentially regulated by protein phosphorylation.