Organization of Calcium Channel β1a Subunits in Triad Junctions in Skeletal Muscle*

In skeletal muscle, dihydropyridine receptors (DHPRs) in the plasma membrane interact with the type 1 ryanodine receptor (RyR1) at junctions with the sarcoplasmic reticulum. This interaction organizes junctional DHPRs into groups of four termed tetrads. In addition to the principle α1S subunit, the β1a subunit of the DHPR is also important for the interaction with RyR1. To probe this interaction, we measured fluorescence resonance energy transfer (FRET) of β1a subunits labeled with cyan fluorescent protein (CFP) and/or yellow fluorescent protein (YFP). Expressed in dysgenic (α1S-null) myotubes, YFP-β1a-CFP and CFP-β1a-YFP were diffusely distributed in the cytoplasm and highly mobile as indicated by fluorescence recovery after photobleaching. Thus, β1a does not appear to bind to other cellular proteins in the absence of α1S. FRET efficiencies for these cytoplasmic β1a subunits were ∼6-7%, consistent with the idea that <10 nm separates the N and C termini. After coexpression with unlabeled α1S (in dysgenic or β1-null myotubes), both constructs produced discrete fluorescent puncta, which correspond to assembled DHPRs in junctions and that did not recover after photobleaching. In β1-null myotubes, FRET efficiencies of doubly labeled β1a in puncta were similar to those of the same constructs diffusely distributed in the cytoplasm and appeared to arise intramolecularly, since no FRET was measured when mixtures of singly labeled β1a (CFP or YFP at the N or C terminus) were expressed in β1-null myotubes. Thus, DHPRs in tetrads may be arranged such that the N and C termini of adjacent β1a subunits are located >10 nm from one another.

In skeletal muscle, dihydropyridine receptors (DHPRs) in the plasma membrane interact with the type 1 ryanodine receptor (RyR1) at junctions with the sarcoplasmic reticulum. This interaction organizes junctional DHPRs into groups of four termed tetrads. In addition to the principle ␣ 1S subunit, the ␤ 1a subunit of the DHPR is also important for the interaction with RyR1. To probe this interaction, we measured fluorescence resonance energy transfer (FRET) of ␤ 1a subunits labeled with cyan fluorescent protein (CFP) and/or yellow fluorescent protein (YFP). Expressed in dysgenic (␣ 1S -null) myotubes, YFP-␤ 1a -CFP and CFP-␤ 1a -YFP were diffusely distributed in the cytoplasm and highly mobile as indicated by fluorescence recovery after photobleaching. Thus, ␤ 1a does not appear to bind to other cellular proteins in the absence of ␣ 1S . FRET efficiencies for these cytoplasmic ␤ 1a subunits were ϳ6 -7%, consistent with the idea that <10 nm separates the N and C termini. After coexpression with unlabeled ␣ 1S (in dysgenic or ␤ 1 -null myotubes), both constructs produced discrete fluorescent puncta, which correspond to assembled DHPRs in junctions and that did not recover after photobleaching. In ␤ 1 -null myotubes, FRET efficiencies of doubly labeled ␤ 1a in puncta were similar to those of the same constructs diffusely distributed in the cytoplasm and appeared to arise intramolecularly, since no FRET was measured when mixtures of singly labeled ␤ 1a (CFP or YFP at the N or C terminus) were expressed in ␤ 1 -null myotubes. Thus, DHPRs in tetrads may be arranged such that the N and C termini of adjacent ␤ 1a subunits are located >10 nm from one another.
Excitation-contraction (EC) 2 coupling in skeletal muscle involves conformational coupling between the dihydropyridine receptor (DHPR), a voltage-gated calcium channel present in the plasma membrane, and the type 1 ryanodine receptor (RyR1) in the sarcoplasmic reticulum (SR). Specifically, the DHPR responds to membrane depolarization by inducing the opening of RyR1 and the release of Ca 2ϩ from the SR. Although the DHPR can function as a calcium channel, skeletal type EC coupling does not require the entry of extracellular Ca 2ϩ , leading to the idea that DHPRs and RyR1 interact mechanically. Indeed, DHPRs in skeletal muscle are arranged in groups of four, termed tetrads, such that each DHPR within a tetrad is apposed to one of the four, identical subunits of RyR1 (1).
The DHPR consists of a principal ␣ 1S subunit, which contains the ion-conducting pathway and voltage sensing structures, together with auxiliary ␤ 1a , ␣ 2 ␦-1, and ␥ 1 subunits (for review, see Ref. 2). A significant body of evidence indicates that the cytoplasmic loop connecting ␣ 1S homology repeats II and III (the II-III loop) plays an important role in conformational coupling between the DHPR and RyR1. The auxiliary ␤ 1a subunit appears to be similarly important. One essential function of all calcium channel ␤ subunits (currently known to be encoded by four different genes) is that of facilitating trafficking to the plasma membrane, which occurs largely as a consequence of ␤ binding to the ␣ 1 I-II loop. Consequently, knock-out of the gene encoding the predominant form in skeletal muscle (␤ 1 ) causes a nearly complete absence of DHPRs in the plasma membrane (3) and a resultant loss of EC coupling (4). Significantly, truncating or altering the sequence of the ␤ 1a C-terminal preserves membrane expression but suppresses EC coupling (5,6), indicating that this region is important for communication between the DHPR and RyR1.
All ␤ subunits can be subdivided into five regions: highly variable Nand C-terminal regions flanking a core segment with three regions. Crystal structures have been recently reported for this core segment (7)(8)(9), which confers the principal functional properties on full-length ␤ subunits. Within the core segment, the first region is highly conserved among ␤ subunits and is homologous to canonical Src homology 3 domains. The third region of the core segment is also highly conserved and is guanylate kinase-like. The Src homology 3-and guanylate kinaselike domains are linked by a more variable, flexible loop so that the core segment as a whole shares structural features with the membrane-associated guanylate kinases, a protein family that organizes signaling components near membranes (10).
Based on sequence alignment of ␤ 1a with ␤ 2a (GenBank TM numbers: M25514 and M80545), and on the crystal structure of the ␤ 2a core segment, leucine 74 and threonine 463 of ␤ 1a are likely separated by 42 Å. However there is little basis for predicting the structure of the ␤ 1a segments N-terminal (73 residues) and C-terminal (61 residues) to the core. Thus, one goal of the present experiments was to obtain a rough idea about the separation of the N and C termini of ␤ 1a and about whether or not there were large changes in the relative arrangements of the N and C termini as a consequence of the binding of ␤ 1a to ␣ 1S . A second goal was to obtain information about the arrangement of ␤ 1a subunits within tetrads. Toward these ends, we measured fluorescence resonance energy transfer (FRET) with cyan fluorescent protein (CFP) and yellow fluorescent protein (YFP) used as the donor and acceptor, respectively (11). The ␤ 1a subunits were singly labeled on either the N or C terminus with either CFP or YFP or doubly labeled with CFP on one terminal and YFP on the other. These fluorescently labeled ␤ 1a subunits were expressed in dysgenic (␣ 1S -null) or ␤ 1 -null myotubes. We found that the ␤ 1a subunit was incorporated into fluorescent puncta at the cell surface (indicative of junctional targeting) only when the ␣ 1S subunit was also present. Based on intramolecular FRET, it appears that the N and C termini of ␤ 1a are Ͻ10 nm apart. Additionally, the absence of intermolecular FRET between singly labeled subunits is consistent with the idea that the N and C termini of ␤ 1a subunits are directed away from the center of tetrads. Fig. 1 illustrates the ␤ 1a constructs that were used for this study. The coding sequence for rabbit skeletal muscle ␤ 1a (GenBank TM number M25514) was isolated by using PCR to introduce EcoRI and SalI sites immediately before the ATG start codon and immediately after the last codon, respectively. The EcoRI-SalI segment containing the ␤ 1a coding sequence was then ligated to EcoRI/SalI-digested pEYFP-C1, pEYFP-N1, pECFP-C1, or pECFP-N1 (Clontech, Palo Alto, CA) to obtain the constructs YFP-␤ 1a , ␤ 1a -YFP, CFP-␤ 1a , and ␤ 1a -CFP, respectively. To construct CFP-␤ 1a -YFP, CFP-␤ 1a and ␤ 1a -YFP were individually digested with MfeI, and the 5903-bp fragment obtained from CFP-␤ 1a and the 1128-bp fragment from ␤ 1a -YFP were ligated together. For the YFP-␤ 1a -CFP construct, YFP-␤ 1a and ␤ 1a -CFP were individually digested with MfeI, and the 5903-bp fragment from YFP-␤ 1a and the 1128-bp fragment from ␤ 1a -CFP were ligated together. Unlabeled ␤ 1a was obtained by digesting ␤ 1a -YFP with Acc651 and BsrGI, discarding the small fragment containing the YFP sequence, and re-ligating the large fragment containing the ␤ 1a coding sequence. Unlabeled ␣ 1S was obtained as described previously (12).

Generation of the Constructs Expressed in Myotubes-
Cell Culture and cDNA Microinjection-Primary cultures of dysgenic (13) and ␤ 1 -null (14) myotubes were prepared from newborn mice as described previously (15). Briefly, myoblasts were plated into 35-mm culture dishes (MatTek, Ashland, MA) with glass coverslip bottoms that had been coated with ECL (Upstate Biotechnology, Lake Placid, NY) and grown for 6 -7 days in a humidified 37°C incubator with 5% CO 2 . The culture medium, Dulbecco's modified Eagle's medium supplemented with 20% fetal bovine serum, was then replaced by differentiation medium (Dulbecco's modified Eagle's medium supplemented with 2% horse serum), and after 2-4 days single nuclei of myotubes were microinjected with a small amount of plasmid DNA in water. The DNA concentration in the injection solution was 10 -20 ng/l for both ␤ 1a constructs and 40 ng/l for ␣ 1S .
The ability of the ␣ 1S and ␤ 1a constructs to support EC coupling in dysgenic or ␤ 1 -null myotubes, respectively, was assayed by the presence of spontaneous contractions and/or evoked contractions. To test for evoked contractions, the myotubes were placed in physiological saline (containing in mM: 140 NaCl, 1.5 KCl, 1 MgCl 2 , 2.5 CaCl 2 , 11 glucose, 10 HEPES, pH 7.4 with NaOH) and stimulated with 100-V, 15-30-ms pulses applied via a patch pipette that contained physiological saline and was positioned near the myotube. Images of these myotubes were acquired at a rate of ϳ11 Hz. The contractions were quantified by measuring the movement of an identifiable portion of a myotube across the visual field.
Measurement of Ionic Currents-Macroscopic Ca 2ϩ currents were measured using the whole-cell patch clamp method. Patch pipettes of borosilicate glass had resistances of 1.8 -2.2 M⍀ when filled with an intracellular solution containing (mM): 110 CsCl, 3 MgCl 2 , 10 Cs 2 EGTA, 10 HEPES, 3 Mg-ATP, and 0.6 GTP, pH ϭ 7.4 with CsOH. The external bath solution contained (mM): 10 CaCl 2 , 145 tetraethylammonium-Cl, 0.003 tetrodotoxin, 0.02 N-benzyl-p-toluene sulfonamide, and 10 HEPES, pH ϭ 7.4 with tetraethylammonium-OH. To measure L-type current, myotubes were stepped from the holding potential (Ϫ80 mV) to Ϫ30 mV for 1 s (to inactivate endogenous T-type current), repolarized to Ϫ50 mV for 30 -50 ms, depolarized to varying test potentials (V test ) for 200 ms, and then returned to the holding potential. Cell capacitance was determined by integration of the capacity transient resulting from 30-mV hyperpolarizations from the holding potential and was used to normalize current amplitudes (pA/pF) obtained from different myotubes. Current-voltage curves were fitted using the Boltzmann expression: where I is the current for the test potential V, V rev is the reversal potential, G max is the maximum Ca 2ϩ channel conductance, V1 ⁄ 2 is the halfmaximal activation potential and k G is the slope factor.
Measurement of FRET and Fluorescence Recovery after Photobleaching-Measurement of FRET was as described previously (12). Briefly, intact fluorescent myotubes were examined 48 h after cDNA microinjection using an LSM 510 META laser scanning confocal microscope (Zeiss, Thornwood, NY). In a typical experiment, an image of the cell region of interest was taken using standard spectroscopic settings: CFP and YFP were excited with separate sweeps of the 458 and 514 nm lines, respectively, of an argon laser (30 milliwatt maximum output, operated at 50% or 6.3 A) attenuated to 10 and 2%, respectively, and directed to the cell via a 458/514 nm dual dichroic mirror. The emitted fluorescence was split via a 515 nm dichroic mirror and for CFP was directed to a photomultiplier equipped with a 465-495 nm bandpass filter (Chroma, Rockingham, Vermont) and for YFP was directed to a photomultiplier equipped with a 530-nm-long pass filter. Confocal fluorescence intensity data (I CFPpre and I YFPpre ) were recorded as the average of four line scans per pixel and digitized at 8 bits. Repeated scans (20 -60) with unattenuated 514 nm illumination were used to photobleach YFP, which required ϳ30 -120 s at maximal scan rates. After completion of YFP bleaching, fluorescence intensity (I CFPpost and I YFPpost ) was measured using the identical parameters as before bleaching. FRET efficiency (E) in percent was calculated as E ϭ (I CFPpost Ϫ FIGURE 1. Schematic diagrams of the ␤ 1a constructs used in these experiments. The position of the fluorescent protein, CFP and/or YFP, is indicated for each construct, in relationship to the native ␤ 1a subunit. The linker sequences connecting CFP and YFP to the ␤ 1a subunit are indicated by wavy lines labeled with the number of linker residues. I CFPpre /I CFTpost ) ϫ 100%, where I CFPpre and I CFPpost are the backgroundcorrected CFP fluorescence intensities before and after photobleaching YFP, respectively. Data are reported as mean Ϯ S.D. Statistical significance was tested with an unpaired Student's t test.
To measure the recovery of yellow fluorescence after photobleaching of YFP, myotubes were first imaged for cyan and yellow fluorescence as already described above. Within this image, a smaller region-of-interest was designated for photobleaching of YFP, which was accomplished with repeated scans of non-attenuated 514 nm excitation. Complete bleaching required about 70 s (similar sized regions of interest were used for all these experiments). Images of cyan and yellow fluorescence were then obtained as above every ϳ30 s, beginning immediately after, and continuing until ϳ400 s after, completion of the bleaching.

RESULTS
␤ 1a Subunit Constructs Are Targeted to the Membrane in the Presence of ␣ 1S and Restore EC Coupling and L-type Ca 2ϩ Current-To analyze the arrangement of ␤ 1a subunits at plasma membrane/SR junctions, both singly labeled (CFP-␤ 1a , YFP-␤ 1a , ␤ 1a -CFP, and ␤ 1a -YFP) and doubly labeled (CFP-␤ 1a -YFP and YFP-␤ 1a -CFP) constructs were examined. Confocal microscopy and functional measurements were used to verify that these fusions of one or two fluorescent proteins to the ␤ 1a subunit did not affect targeting or the ability of the DHPR to function as a voltage sensor for EC coupling. Fig. 2 illustrates confocal fluorescence sections of myotubes expressing CFP-␤ 1a -YFP. After expression in dysgenic myotubes (lacking ␣ 1S ), CFP-␤ 1a -YFP was diffusely distributed within the cytoplasm of the cell (Fig. 2a). When the same construct was coexpressed with unlabeled ␣ 1S in dysgenic myotubes, it was arranged in fluorescent puncta, many of which were close to the surface of the cell (Fig. 2b). Similar, fluorescent puncta were also observed when CFP-␤ 1a -YFP was expressed in b 1 -null myotubes, which endogenously express ␣ 1S (Fig. 2c). Patterns of fluorescence like those illustrated for CFP-␤ 1a -YFP in Fig. 2 were also seen for YFP-␤ 1a -CFP (data not shown). The diffuse fluorescence of ␤ 1a in absence of ␣ 1S , and the punctate fluorescence in the presence of ␣ 1S , agrees with previous results with CFP-YFP-tagged ␤ 1a (12), as well as with results from GFP-tagged ␤ 1a (16). Furthermore, in the latter study, the puncta of GFP fluorescence colocalized with puncta of ␣ 1S and RyR1 as revealed by immunolabeling. Thus, the diffuse fluorescence appears to represent the ␤ 1a subunit free in the myoplasm of the cell, whereas the fluorescent puncta appear to correspond to DHPRs inserted into junctional regions of the plasma membrane.
Previously, we demonstrated that the presence of a CFP-YFP tandem on either the N or C terminus of ␤ 1a did not affect function of the DHPR as either a voltage sensor for EC coupling or as an L-type Ca 2ϩ channel (12). To determine whether the addition of fluorescent proteins to both the N and C termini affected function of ␤ 1a , we measured evoked contractions and L-type Ca 2ϩ currents in ␤ 1 -null myotubes expressing CFP-␤ 1a -YFP. Such myotubes produced robust contractions in response to focal extracellular stimulation ( Fig. 4A and Table 1) and also produced large, whole-cell L-type Ca 2ϩ currents (Fig. 4, B and C). Thus, the function of ␤ 1a did not appear to be altered by the attachment of fluorescent proteins.
Photobleaching of Labeled ␤ 1a Subunits-To further compare labeled ␤ 1a subunits in the absence and presence of ␣ 1S , subregions of cells were When expressed alone in a dysgenic myotube (a), CFP-␤ 1a -YFP produced a diffuse fluorescence that filled the entire cell except for the nuclei, whereas when it was coexpressed in a dysgenic myotube with ␣ 1S (b) it localized in fluorescent puncta. Similar fluorescent puncta were also observed after expression of CFP-␤ 1a -YFP in ␤ 1 -null myotubes, which endogenously express ␣ 1S . Bars, 5 m.  FEBRUARY 10, 2006 • VOLUME 281 • NUMBER 6 subjected to intense illumination at 514 nm to bleach YFP. Fig. 5 illustrates this kind of experiment for YFP-␤ 1a -CFP expressed in either a dysgenic myotube without ␣ 1S (A) or coexpressed in a dysgenic myotube together with unlabeled ␣ 1S (B). In the absence of ␣ 1S , both the yellow and cyan fluorescence had a diffuse appearance. The ␤ 1a subunits producing this diffuse pattern appeared to be freely mobile, since immediately after the bleaching episode (total duration of 70 s), there was a partial loss of yellow fluorescence outside the portion of the myotube exposed to bleaching illumination, and within the bleached region the intensity of yellow fluorescence gradually increased as the distance to the non-bleached portion of the myotube decreased (Fig. 5A, panel b). Moreover, by 400 s after the end of the bleaching episode, there was a restoration of uniform pattern of yellow fluorescence observed prior to bleaching (Fig. 5A, panel c). The time course of the yellow fluorescence intensity in Fig. 5C clearly shows recovery within the bleached region relative to the non-bleached region, while the cyan fluorescence remains relatively stable. Note the small increase in cyan fluorescence immediately after the bleaching of YFP, indicative of the FRET from CFP to YFP prior to bleaching. These results appear to indicate that there is negligible binding of the labeled ␤ 1a subunits to cellular structures. By contrast, the majority of ␤ 1a subunits in the presence of ␣ 1S appeared to be immobile because of anchoring within junctional membranes. Thus, there was a sharp demarcation in yellow fluorescence between bleached and non-bleached regions of the myotube and no loss of yellow fluorescence in the non-bleached regions (Fig. 5B, panel b). Furthermore, even 400 s after the completion of bleaching (Fig. 5B,  panel c), there was essentially no recovery of yellow fluorescence within puncta inside the bleached region, although a small amount of diffuse fluorescence did recover within the bleached region. The corresponding time course of YFP fluorescence intensity within (YELLOW 2) as well as outside (YELLOW 1) the bleached area is shown in Fig. 5D. The small continuous decrease of YELLOW 1 can be attributed to bleaching due to repetitive scanning whereas the slow recovery of YELLOW 2 may reflect diffusional invasion of ␤ 1a subunits not associated with ␣ 1S subunits. Results similar to those illustrated in Fig. 5A were obtained from a total of 6 dysgenic myotubes without ␣ 1S and similar to those illustrated in Fig. 5B from a total of 3 dysgenic myotubes plus unlabeled ␣ 1S and a total of 2 ␤ 1 -null myotubes.
One way to explain the results in Fig. 6 is to suppose that (i) intramolecular FRET is lower for doubly labeled ␤ 1a in junctions compared with that in cytoplasm and (ii) the measured FRET efficiency for doubly labeled ␤ 1a in junctions contains contributions from both intra-and intermolecular FRET. Accordingly, the presence of unlabeled ␤ 1a would reduce the contribution of the inter-molecular component. If this explanation were correct, one would expect to be able to measure intermolecular FRET with singly labeled constructs. This was tested by coexpression in ␤ 1 -null myotubes of 1:1 mixtures of CFP-␤ 1a ϩ YFP-␤ 1a , ␤ 1a -CFP ϩ ␤ 1a -YFP, or CFP-␤ 1a ϩ ␤ 1a -YFP. For none of these mixtures was detectable FRET observed (Table 2), which argues against a contribution of intermolecular FRET for the doubly labeled ␤ 1a constructs in ␤ 1 -null myotubes.

DISCUSSION
The four major findings of the present work are as follows. First, the addition of fluorescent protein to the N terminus, C terminus, or both termini of ␤ 1a does not interfere with its targeting or function. In the FIGURE 4. Restoration of excitation-contraction coupling and calcium currents in ␤ 1 -null myotubes expressing CFP-␤ 1a -YFP. A, electrically evoked contraction in a ␤ 1 -null myotube expressing CFP-␤ 1a -YFP (vertical scale: arbitrary units). B, representative whole-cell calcium currents, for the indicated test potentials, from a ␤ 1 -null myotube expressing CFP-␤ 1a -YFP. C, average peak current versus voltage relationship for ␤ 1 -null myotubes expressing CFP-␤ 1a -YFP (n ϭ 8). The smooth curve represents the best fit of the Boltzmann expression, which yielded the values G max ϭ 0.24 nS/nF, V rev ϭ 82.63 mV, V1 ⁄2 ϭ 12.23 mV, k G ϭ 5.15 mV. presence of ␣ 1S (in ␤ 1 -null myotubes or coexpressed with ␣ 1S in dysgenic myotubes), the ␤ 1a constructs are incorporated into DHPRs that are present in punctate regions of the plasma membrane that form junctions with the SR. The DHPRs containing these labeled ␤ 1a subunits retain the ability to mediate excitation-contraction coupling. Second, in the absence of ␣ 1S , there were no foci of ␤ 1a indicative of binding to other cellular structures and these ␤ 1a subunits appeared to be freely diffusible in the cytoplasm. Third, in the cytoplasm: the 6 -7% intramolecular FRET efficiency of the doubly labeled constructs (CFP-␤ 1a -YFP or YFP-␤ 1a -CFP) expressed in dysgenic myotubes indicates that the N and C termini of cytoplasmic ␤ 1a subunits are likely separated by Ͻ10 nm. Fourth, within junctions: there was no measurable intermolecular FRET for 1:1 mixtures of CFP-␤ 1a ϩ YFP-␤ 1a , ␤ 1a -CFP ϩ ␤ 1a -YFP, or CFP-␤ 1a ϩ ␤ 1a -YFP expressed in ␤-null myotubes. This result is consistent with the idea that then N and C termini of adjacent ␤ 1a subunits are Ͼ10 nm apart, which would occur if they were oriented away from the center of tetrads. After expression in dysgenic myotubes lacking ␣ 1S , doubly labeled ␤ 1a was diffusely distributed (Figs. 2a and 5A). Such a diffuse distribution would seem to indicate that in the absence of ␣ 1S , the ␤ 1a subunits do not bind to immobile cellular structures. Consistent with this idea, yellow fluorescence at the ends of dysgenic myotubes showed substantial recovery within several minutes following the selective bleaching of YFP within YFP-␤ 1a -CFP or CFP-␤ 1a -YFP. The observation here that doubly labeled ␤ 1a is diffusely distributed in the absence of ␣ 1S is consistent with earlier results showing a diffuse distribution after expression in dysgenic  . Average FRET efficiencies for doubly labeled ␤ 1a expressed in dysgenic (i, ii) or ␤ 1 -null (iii, iv) myotubes. The doubly labeled ␤ 1a constructs were expressed in dysgenic myotubes either without (i) or with (ii) unlabeled ␣ 1S and in ␤ 1 -null myotubes either without (iii) or with (iv) unlabeled ␤ 1a . The labeled ␤ 1a subunits had a cytoplasmic distribution in the absence of ␣ 1S (i) and a punctate (junctional) distribution in the presence of exogenous (ii) or endogenous (iii, iv) ␣ 1S . Error bars indicate ϮS.D. For each of the conditions (i-iv), FRET efficiencies were not significantly different between CFP-␤ 1a -YFP and YFP-␤ 1a -CFP (p Ͼ 0.24). FRET efficiencies were significantly different between conditions i and ii (p Ͻ 0.0036) and between conditions iii and iv (p Ͻ 0.043). FEBRUARY 10, 2006 • VOLUME 281 • NUMBER 6 myotubes of either ␤ 1a subunits tagged with a CFP-YFP tandem (12) or with GFP (16). However, small, weak puncta on a diffuse background were observed following immunostaining for endogenous ␤ 1a in dysgenic myotubes, from which it was concluded that the isolated ␤ 1a subunits could bind to some component of plasma membrane/SR junctions (14). One possible explanation for the discrepancy between the results with immunostaining and expressed fluorescent protein-labeled ␤ 1a is that immunostaining puncta in dysgenic myotubes reflect nonspecific sites of anti-␤ 1a antibody binding. Alternatively, the presence of fluorescent protein might interfere with the ability of isolated ␤ 1a to bind to other cellular proteins, although if this occurred it would be difficult to explain why fluorescent protein labeling had no observable effect on ␤ 1a function (12,16).

Calcium Channel ␤ 1a Subunits in Skeletal Muscle
Within junctions, the FRET efficiency (E) for the doubly labeled ␤ 1a was 6 -7% in ␤ 1 -null myotubes (where no unlabeled ␤ 1a was present) and decreased to about 3% when unlabeled ␤ 1a was present (expression in dysgenic myotubes or expression together with unlabeled ␤ 1a in ␤ 1 -null myotubes). The interpretation of this result requires one to consider the possibility that measured FRET efficiency within junctions ("jun") contained contributions from both intra-and intermolecular FRET (E jun ϭ a jun ϩ e), where a and e are the intra-and intermolecular FRET efficiencies, respectively). For doubly labeled ␤ 1a in the cytoplasm ("cyt"), only intramolecular FRET would occur (E cyt ϭ a cyt ). One could then imagine a number of possibilities for why the presence of unlabeled ␤ 1a would cause a decrease in E.
One possibility is that the only effect of unlabeled ␤ 1a was to reduce the contribution of intermolecular FRET in junctions (because the unlabeled ␤ 1a reduced the number of adjacent doubly labeled ␤ 1a subunits within tetrads). If this were the case, one would have to conclude that e Ն4% and that a jun Յ 3%, which is substantially less than a cyt (6 -7%). This change in intramolecular FRET could be a consequence of a conformational change in ␤ 1a that occurred upon binding to ␣ 1S (for example, an increase in separation between the N and C termini). However, x-ray crystallography demonstrates that the structure of the ␤ subunit core is little affected by binding to the AID of the ␣ 1 I-II loop (9). Moreover, this interpretation suffers from the weakness that intermolecular FRET was not detected for combinations of singly labeled ␤ 1a subunits ( Table 2).
A second possibility is that e ϭ 0 and that the conformation of the doubly labeled ␤ 1a subunit depends upon whether or not its neighboring ␤ 1a subunit had attached fluorescent proteins. Specifically, one could imagine that the presence of two fluorescent proteins on all four ␤ 1a subunits within a tetrad produces "molecular crowding" that results in a decreased distance between the N and C termini of each of the individual ␤ 1a subunits. The presence of one or more unlabeled ␤ 1a subunits could then allow a relaxation of the doubly labeled ␤ 1a subunits to the native conformation for ␤ 1a subunits bound to ␣ 1S . An argument against molecular crowding is that previous work has shown that addition of two fluorescent proteins per ␤ 1a subunit (tagging with CFP-YFP tandems) has no discernible affect on DHPR function as channel or voltage sensor for EC coupling (12).
A third possibility is that e ϭ 0, the observed FRET signal from puncta is entirely intramolecular, and the effect of unlabeled ␤ 1a can be attributed to errors in measurement of cyan intensity. In particular, the contrast between the cyan fluorescence signal and cellular autofluorescence can sometimes be low. The presence of unlabeled ␤ 1a within the puncta would decrease the cyan fluorescent signal and cause a further loss of contrast. Because the autofluorescence is unaffected by bleaching of YFP, its presence would cause an underestimate of FRET efficiency, which would become worse as the contrast between cyan fluorescence and autofluoresence decreased as a result of unlabeled ␤ 1a .
Whatever the explanation for the effect of unlabeled ␤ 1a on the FRET signal produced by doubly labeled ␤ 1a , it remains the case that we were unable to measure directly any intermolecular FRET from N to N terminus, from C to C terminus or from N to C terminus of adjacent ␤ 1a subunits in tetrads. Taken together with the presence of N-to C-terminal intramolecular FRET for cytoplasmic and junctional ␤ 1a , our results are consistent with the model presented in Fig. 7. Individual, ␤ 1a subunits are sufficiently compact (Ͻ10 nm) that FRET occurs between fluorescent proteins on the N and C termini. In the absence of ␣ 1S , the ␤ 1a subunits do not appear to bind to other cellular structures and are randomly distributed in the cytoplasm at an average distance too large to allow intermolecular FRET (Fig. 7A). In the presence of ␣ 1S , the ␤ 1a subunits are assembled into DHPRs, which are organized into groups of four (tetrads), where each tetrad is associated with a single RyR1 (Fig.  7B). Within tetrads, intramolecular FRET occurs within individual doubly labeled ␤ 1a subunits, but adjacent subunits are separated by too great a distance to allow FRET. The specific model illustrated in Fig. 7B is based on the arrangement proposed by Wolf et al. (17), but our data are also consistent with the tetradic arrangement of DHPRs suggested by Paolini et al. (18). Independent of any particular arrangement of DHPRs into tetrads, our data are most easily explained by the hypothesis that the N and C termini of ␤ 1a lie outside a 10-nm diameter circle at the center of each tetrad. An important goal of future work will be to refine
our knowledge of the spatial orientation of defined DHPR regions with respect to tetrads and with respect to defined regions of RyR1.