Ectopic Expression of Phospholamban in Fast-Twitch Skeletal Muscle Alters Sarcoplasmic Reticulum Ca2+ Transport and Muscle Relaxation*

There are three isoforms of the sarcoplasmic reticulum Ca2+-ATPase; they are known as SERCA1, SERCA2, and SERCA3. Phospholamban is present in tissues that express the SERCA2 isoform and is an inhibitor of the affinity of SERCA2 for calcium. In vitro reconstitution and cell culture expression studies have shown that phospholamban can also regulate SERCA1, the fast-twitch skeletal muscle isoform. To determine whether regulation of SERCA1 by phospholamban can be of physiological relevance, we generated transgenic mice that ectopically express phospholamban in fast-twitch skeletal muscle, a tissue normally devoid of phospholamban. Ectopic expression of phospholamban was associated with a decrease in the affinity of SERCA1 for calcium. Assessment of isometric twitch contractions of intact fast-twitch skeletal muscles revealed depressed rates of relaxation in transgenic mice compared with wild-type cohorts. Furthermore, the prolongation of muscle relaxation appeared to correlate with the levels of phospholamban expressed in two transgenic mouse lines. These findings indicate that ectopic expression of phospholamban in fast-twitch skeletal muscle is associated with inhibition of SERCA1 activity and decreased relaxation rates of this muscle.

The sarcoplasmic reticulum (SR) 1 is an internal membrane system that plays an important role in the initiation of muscle relaxation via the reduction of cytosolic calcium levels. The translocation of calcium from the cytosol into the SR is mediated by the SR Ca 2ϩ -ATPase enzyme. There are currently three known isoforms of the SR Ca 2ϩ -ATPase, which are the products of separate genes (1)(2)(3)(4)(5). The SERCA1 gene is expressed exclusively in adult fast-twitch skeletal muscle SR (1,2,5). The SERCA2 gene has two alternatively spliced isoforms, SERCA2a and SERCA2b. SERCA2a is expressed in cardiac, slow-twitch, and fetal fast-twitch skeletal muscles, while SERCA2b can be found in smooth muscle and nonmuscle tissues (4,6,7). The SERCA3 gene is expressed in a number of muscle and nonmuscle tissues and is considered the endoplasmic reticulum Ca 2ϩ -ATPase isoform (3).
SERCA2a is the major isoform responsible for SR calcium transport in cardiac and slow-twitch skeletal muscles; it is regulated by phospholamban (8,9). Dephosphorylated phospholamban is an inhibitor of the affinity of cardiac SR Ca 2ϩ -ATPase for Ca 2ϩ , and phosphorylation relieves this inhibition (9). In vitro studies have demonstrated phosphorylation of phospholamban at distinct sites by cAMP-dependent, Ca 2ϩcalmodulin-dependent, and Ca 2ϩ -phospholipid-dependent protein kinases (10). In vivo studies of isolated heart preparations have also shown that phospholamban can be phosphorylated during ␤-adrenergic stimulation and that this phosphorylation is associated with increases in the SR Ca 2ϩ -ATPase activity and enhanced rates of cardiac relaxation (11)(12)(13)(14). Furthermore, pretreatment of SR vesicles with an anti-phospholamban monoclonal antibody could relieve the inhibitory effects of phospholamban on the Ca 2ϩ affinity of SERCA2 and enhance EϳP formation, while this antibody had no effect on SERCA1 activity (15). SERCA1 mediates relaxation in fast-twitch skeletal muscle and although phospholamban is absent from this tissue, several in vitro studies suggest that SERCA1 can be regulated by phospholamban (8, 9, 16 -18): (a) cross-linking of phospholamban to SERCA revealed a putative phospholamban-binding domain that is present in both SERCA1 and SERCA2 (8, 9, 16 -18); (b) reconstitution of phospholamban with SERCA1 or SERCA2 in lipid bilayers was associated with similar inhibition of Ca 2ϩ uptake rates using either isoform (9,17); (c) co-expression studies of phospholamban with either SERCA1 or SERCA2 in COS-1 cells demonstrated that phospholamban could regulate either the SERCA1 or SERCA2 enzyme (8); and (d) stable expression of phospholamban in C 2 C 12 cells, a fasttwitch skeletal muscle cell line that expresses SERCA1, resulted in an inhibition of SERCA1 activity in its native SR environment (16).
While evidence obtained in these in vitro studies suggests that phospholamban can regulate the SERCA1 enzyme, it is not presently clear whether this regulation would lead to alterations of skeletal muscle kinetics. The advent of transgenesis allows for tissue-specific expression of a protein of interest and assessment of its physiological consequences in that particular tissue. Thus, transgenic mice provide ideal systems for testing phospholamban-mediated regulation of SERCA1 activity in vivo and its effects on fast-twitch skeletal muscle function. In this study, we generated a transgenic mouse model, which ectopically expresses phospholamban exclusively in fasttwitch skeletal muscle under the control of the rat myosin light chain 1f (MLC1f) promoter and enhancer elements. Introduc-tion of phospholamban into fast-twitch skeletal muscle resulted in inhibition of SERCA1 activity in its native lipid environment and in alterations in muscle relaxation kinetics.

Generation of Transgenic Mice by Pronuclear Injection-The
MLC1f/ PLB (Fig. 1A) construct was generated by combining the 5Ј-flanking promoter (1.5 kilobases (kb)) and the 920-base pair (bp) enhancer regions of the rat myosin light chain 1/3 locus, isolated from NRCAT900 (19), with an 853-bp cassette that included the mouse PLB cDNA (603-bp) fragment as well as a 250-bp fragment containing the SV40 polyadenylation signal. The 603-bp phospholamban cDNA fragment, isolated from the ␣-MHC 5.5 /PLB transgene construct (20), encompassed 13 bp of the phospholamban 5Ј-untranslated region, the entire phospholamban-coding region, as well as 434 bp of the 3Ј-untranslated region including the first polyadenylation signal. The resulting 3.2-kb construct, containing the MLC1f promoter and enhancer elements, the entire phospholamban cDNA, and the SV40 polyadenylation signal were released from the plasmid backbone as an EcoRI/KpnI fragment and used for pronuclear microinjection of fertilized mouse eggs to generate transgenic mice.
Transgenic mice carrying the MLC1f-driven phospholamban transgene were identified using polymerase chain reaction and Southern analyses of genomic DNA isolated from tail biopsies (20). The transgene copy number was determined relative to the phospholamban gene (copy number ϭ 2), using the PhosporImager and ImageQuant analysis system (Molecular Dynamics, Sunnyvale, CA). Mice positive for the transgene were bred, and expression of the transgene was determined using Northern blot analysis (21).
Quantitative Immunoblot Analysis-Quantitative immunoblotting of transgenic fast-twitch skeletal homogenates was carried out as described previously (22). The transgenic phospholamban was detected using anti-phospholamban polyclonal antisera (1:25 dilution) isolated from rabbits that had been immunized with a peptide corresponding to amino acids 1-25 of phospholamban (22). The SR Ca 2ϩ -ATPase levels were determined using a 1:500 dilution of an IgG polyclonal antiserum fraction (9,22), which was raised against a conserved region present in all SERCA isoforms. Briefly, muscles were homogenized at 4°C in a 10-fold (w/v) volume of 10 mM imidazole (pH 7.0), 300 mM sucrose, 1 mM dithiothreitol, and 1 mM sodium metabisulfite, according to methods previously described (22). Protein concentrations were determined by the Bio-Rad method using bovine serum albumin as a standard. Homogenates were analyzed using SDS-PAGE, and cardiac homogenates were included as a standard for comparison of protein expression, similar to methods previously described (20). An alkaline phosphataseconjugated secondary antibody (Organon Teknika, Durham, NC) was used to visualize both the phospholamban and the SR Ca 2ϩ -ATPase proteins, which were quantitated using ImageQuant Software, as described previously (23).
Detection of Phosphorylated Phospholamban-Homogenates were prepared from extensor digitorum longus (EDL) muscles that were exposed to maximal isoproterenol stimulation (6 M) and subjected to Western blot analysis as described above. In addition, homogenates from Langendorff-perfused hearts that had been maximally stimulated with isoproterenol (1 M) were included for comparison. Following transfer, the membranes were then incubated with polyclonal antibodies that specifically recognize either phosphoserine or phosphothreonine residues in phospholamban (24,25). An alkaline phosphataseconjugated secondary antibody (Organon Teknika) was used to visualize the phosphoserine and phosphothreonine residues in phospholamban.
Immunofluorescence Co-localization-Indirect immunofluorescence double-labeling of skeletal muscle was used to visualize phospholamban and SERCA proteins in the same tissue section, similar to methods previously described (26). Due to the small size of the tibialis anterior and EDL muscles, we used the large gastrocnemius muscle for these studies. Briefly, the muscles were excised, cut into 5-mm strips, and longitudinally mounted on cork specimen mounts with gum tragacanth. The tissue samples were flash-frozen and cut into 5-mm-thick frozen sections. The sections were fixed in 4% paraformaldehyde in phosphatebuffered saline, washed, and then blocked in 1% bovine serum albumin in phosphate-buffered saline. The blocked samples were subsequently incubated with an anti-PLB monoclonal antibody (Upstate Biotechnology Inc., Lake Placid, NY; 1:1000 dilution). The anti-PLB monoclonal antibody was then reacted with a fluorescein-labeled anti-mouse IgG secondary antibody (Jackson ImmunoResearch, West Grove, PA). The samples were then incubated with an anti-Ca 2ϩ -ATPase polyclonal antibody (9) that was detected with a rhodamine-labeled anti-rabbit secondary antibody (Jackson ImmunoResearch, West Grove, PA.). Single and double stained specimens were examined with an inverted Nikon Diaphot fluorescence microscope. Laser scanning confocal microscope images were collected, digitized, and analyzed using CoMOS software (Bio-Rad Laboratories, Cambridge, MA). Fluorescein and rhodamine channels were merged (superimposed) to determine how much overlap was present in the staining patterns of the different antibodies. The regions where the staining was not superimposed appeared either red (rhodamine) or green (fluorescein). However, the regions in which the staining patterns of both antibodies overlapped appeared yellow, indicating that the two antigens were co-localized. The merged images were printed with a Kodak color printer, rephotographed, and printed from color negatives.
Determination of Muscle Twitch Dynamics-Mice of either sex were anesthetized intraperitoneally with sodium pentobarbital (30 mg/kg body weight) and mechanically respirated via tracheotomy. Intact EDL muscles were removed from the anesthetized mice, and contractile parameters of isometric twitch contractions were measured as described previously (27). The EDL muscle was utilized for these studies, since its small size would allow for complete oxygenation of the entire muscle. Twitch contractions were recorded using a Grass P7 polygraph chart recorder and the developed force (dF, in mg) and its first derivative (ϩdF/dt and ϪdF/dt, in mg/s) were continuously displayed. The data were further analyzed using the Biopac MP100 computer acquisition system (Biopac Systems Inc., Santa Barbara, CA) to determine the contraction time (in ms) and the half-relaxation time (in ms) of these muscles. After establishing L max , the muscles were exposed to continuous, cumulative doses (1 nM-10 M) of isoproterenol. At the end of the experiment, isoproterenol was washed out, and the muscles were removed from the bath, blotted dry, and weighed.
Calcium Uptake Assays-Skeletal muscles from 12-week-old mice were excised, rinsed in ice-cold phosphate-buffered saline, frozen in liquid nitrogen, and stored at Ϫ80°C. The frozen muscles were homogenized at 4°C, and Ca 2ϩ uptake measurements were made as described previously (28,29).
Statistical Analysis-Data are presented as mean Ϯ S.E. The number (n) of mice used is indicated. Statistical differences between the means were determined using analysis of variance and the Student Newman-Keuls test for multiple comparisons. Values with p Ͻ 0.05 were considered statistically significant.

Generation of Transgenic Mice-The
MLC1f/PLB transgenic construct ( Fig. 1A) was generated to direct ectopic expression of the mouse phospholamban cDNA in fast-twitch skeletal muscle. Sixteen transgenic lines were identified by polymerase chain reaction and Southern blot analyses of genomic DNA from tail biopsies. Transgene expression was assessed by Northern blot analysis of total RNA from gastrocnemius, tibialis anterior (TA) and EDL muscles. Abundant expression of the 1-kb phospholamban transcript was detected in all three muscles and in two separate lines, 75 and 76 (Fig. 1B), consistent with previous observations (30). Southern blot analysis of genomic DNA indicated that the transgene copy number was 12 and 7 in lines 75 and 76, respectively. Both of these lines were propagated for further characterization. The tibialis anterior and EDL muscles, which are composed of 98% fast-twitch fibers (31), were used for the biochemical and physiological studies, respectively. However, due to the limited size of these muscles, the gastrocnemius, which is 80% fast-twitch (31), was used for immunofluorescence studies to examine transgene insertion into the SR membranes. Western blot analysis of these three muscles using monoclonal antibodies specific for either SERCA1 or SERCA2 indicated that the TA and EDL muscles express only SERCA1, while the gastrocnemius expresses SERCA1 and nominal (ϳ15%) levels of SERCA2 (data not shown).
Quantitative Immunoblot Analysis of Transgene Expression-To determine the levels of phospholamban expression relative to the levels of the SR Ca 2ϩ -ATPase, we performed quantitative immunoblots using homogenates of tibialis anterior and EDL muscles isolated from line 75-derived mice. Cardiac homogenates were also processed in parallel to compare the relative levels of phospholamban expression in the transgenic skeletal muscles. We observed phospholamban expression levels that were ϳ70% of those present in the heart ( Fig.  2A), while the SR Ca 2ϩ -ATPase levels were ϳ2-fold higher in the skeletal homogenates than those in the heart (Fig. 2B). Thus, the relative ratio of phospholamban to SR Ca 2ϩ -ATPase was ϳ1:3 in the transgenic fast-twitch skeletal muscles, when compared with a relative value of 1:1 in cardiac muscle (Fig.  2C). The actual ratio of phospholamban to SR Ca 2ϩ -ATPase in cardiac muscle is currently unclear, but in this study it was set as 1:1. Examination of the phospholamban expression levels in animals from line 76 indicated that they were ϳ40% of those present in the heart (data not shown). Furthermore, there was no phospholamban detected in the gastrocnemius, tibialis anterior or EDL muscles from the wild-type, nontransgenic cohorts (data not shown).
Immunofluorescence Co-Localization of Phospholamban and the SR Ca 2ϩ -ATPase-To determine whether the transgenic phospholamban protein was localized in the SR membranes, we utilized indirect immunofluorescence labeling of both phospholamban and the SR Ca 2ϩ -ATPase in the same cryosections of transgenic muscles. Gastrocnemius muscles were excised from wild-type and transgenic animals, sectioned using a cryotome, and mounted on glass slides. The cryosections were then reacted with antisera to either phospholamban or the SR Ca 2ϩ -ATPase. The binding of antisera to phospholamban was detected using a rhodamine-conjugated secondary antibody, while binding to the SR Ca 2ϩ -ATPase was detected with a fluorescein-conjugated secondary antibody. The use of differentially labeled secondary antibodies allows for the detection of both proteins in the same cryosection. Similar staining pat-terns for the SR Ca 2ϩ -ATPase were observed in wild-type and transgenic muscles. However, we could detect staining for phospholamban expression only in transgenic muscles (Figs. 3  and 4). The lack of phospholamban detection in the wild-type muscles, which are composed of 80% slow-twitch fibers (31), is likely due to the low levels of phospholamban expression in slow-twitch fibers (27). To determine the extent of overlap present in the staining patterns for the two different antibodies, the images obtained in the transgenic muscles for both phospholamban and the SR Ca 2ϩ -ATPase were then digitized and electronically superimposed. In regions of the superimposed images where the staining patterns were not overlapping, the color was either red or green, representing rhodamine or fluorescein staining, respectively (Fig. 4A). The regions that appeared yellow represent areas in which the staining for both antibodies was superimposed, indicating that the two antigens were co-localized. As shown in Fig. 4B, merging of images for phospholamban and the SR Ca 2ϩ -ATPase resulted in a pattern that was predominantly yellow, suggesting that the transgenic phospholamban was co-distributed with the Ca 2ϩ -ATPase in the SR membranes.
Effects of Ectopic Phospholamban Expression on SR Ca 2ϩ Uptake-The effects of ectopic phospholamban expression on the apparent initial rates of ATP-dependent, oxalate-facilitated SR Ca 2ϩ -uptake were assessed in whole muscle homogenates using tibialis anterior muscles from wild-type and transgenic animals (line 75). The advantages of using homogenates rather than SR vesicles and the validity of this approach have been previously defined (32,33). SR Ca 2ϩ -uptake was determined over a wide range of calcium concentrations in each preparation. Measurements of Ca 2ϩ -uptake at each calcium concentration were taken at time intervals of 0.5, 1.0, and 1.5 min to Total homogenates (5, 10, and 20 g) of EDL, TA, and heart tissues were electrophoresed on a 13% SDS-polyacrylamide gel and electroblotted onto a nitrocellulose membrane. The membrane was hybridized with an anti-phospholamban polyclonal antibody, followed by an alkaline phosphatase-conjugated anti-rabbit secondary antibody (A), or with an anti-SR Ca 2ϩ -ATPase polyclonal antibody, followed by an alkaline phosphatase-conjugated anti-rabbit secondary antibody (B). C, relative ratio of phospholamban to the SR Ca 2ϩ -ATPase (SERCA) protein levels in transgenic TA muscle homogenates (PLB-TG) in reference to cardiac muscle homogenates, where the ratio was set as 1.0. Values represent the mean Ϯ S.E. of 4 muscles, each examined separately in triplicate. determine initial rates. We observed very high rates of Ca 2ϩuptake at the 0.5-min interval but these rates decreased at subsequent time points (data not shown). This phenomenon has been previously observed in studies of fast-twitch skeletal SR membranes by Feher et al. (34), who suggested that the nonlinearity observed beyond 0.7 min was due to rupture of the SR vesicles. On the basis of these findings, we chose to determine the apparent initial rates of SR Ca 2ϩ -uptake at the 0.5min interval. Ectopic expression of PLB was associated with a significant decrease in SR Ca 2ϩ uptake rates at low Ca 2ϩ , with no apparent change in the maximal velocity (V max ) of Ca 2ϩ uptake (Fig. 5). Analysis of these data indicated that the EC 50 values of the SR Ca 2ϩ uptake for Ca 2ϩ were significantly (p Ͻ 0.05) higher in transgenic muscles (0.284 M Ϯ 0.01; n ϭ 5) than in wild-type muscles (0.189 M Ϯ 0.01; n ϭ 5). Thus, expression of phospholamban in fast-twitch skeletal muscle resulted in a decrease in the affinity of the SR Ca 2ϩ -ATPase for calcium, consistent with its inhibitory effects previously reported in cardiac muscle (32,33).
Isometric Twitch Contractions in Isolated Skeletal Muscles-To determine whether the decreased affinity of the SR Ca 2ϩ -uptake for Ca 2ϩ was associated with alterations of muscle function, isometric twitch contractions were assessed in EDL muscles from wild-type and transgenic animals (line 75). The EDL muscle was utilized for these studies since its small size allows for complete oxygenation of the muscle, which is critical for maintenance of muscle function. Although the transgenic mice appeared phenotypically indistinguishable from their wild-type cohorts at the gross level, the transgenic EDL muscles exhibited a significantly altered isometric twitch contraction profile, which was associated with a prolonged relaxation phase (Fig. 6A). Quantitative assessment of isometric twitch contractions of EDL muscles at L max revealed that ectopic phospholamban expression in the EDL muscle resulted in a significant increase in the half-relaxation times (12.5 Ϯ 0.6 ms, n ϭ 7 in transgenics versus 7.3 Ϯ 0.9 ms, n ϭ 4 in wild types, p Ͻ 0.05) without a change in the contraction times (19.8 Ϯ 0.3 ms, n ϭ 7 in transgenics versus 18.9 Ϯ 0.4 ms, n ϭ 4 in wild types) of transgenic EDL muscles compared with wild-type EDL muscles. We also characterized the contractile parameters of EDL muscles from another transgenic line (line 76), which expressed lower levels (40% versus heart; Fig. 1B) of phospholamban than those present in line 75 (70% versus heart). Consistent with the results obtained in line 75, we observed a prolonged half-relaxation time (9.7 Ϯ 0.97 ms; n ϭ 3) but no effect on contraction time (18.8 Ϯ 0.2 ms; n ϭ 3) in EDL muscles from line 76. When the relative levels of phospholamban in the EDL muscles from wild-type (0%), line 76 (40%), and line 75 (70%) were plotted against the half-relaxation times in these muscles, a close linear correlation was observed (Fig. 6B), indicating that phospholamban is a major regulator of relaxation in these muscles.
To determine whether the prolonged relaxation observed in the transgenic EDL muscles could be reversed by ␤- adrenergic   FIG. 3. Double immunofluorescence labeling of the SR Ca 2؉ -ATPase but not phospholamban in the same wild-type gastrocnemius cryosection. The SR Ca 2ϩ -ATPase was detected using an anti-SR Ca 2ϩ -ATPase polyclonal antibody in conjunction with a fluorescein-conjugated anti-rabbit secondary antibody, while phospholamban was detected using an anti-phospholamban monoclonal antibody in conjunction with a rhodamine-conjugated anti-mouse secondary antibody. Images were produced using laser confocal scanning microscopy. Bar, 10 m.

FIG. 4. Double immunofluorescence labeling of phospholamban and SR Ca 2؉ -ATPase in the same transgenic gastrocnemius cryosection.
In A, phospholamban was detected using an anti-phospholamban monoclonal antibody in conjunction with a rhodamine-conjugated anti-mouse secondary antibody and appears red, while the SR Ca 2ϩ -ATPase was detected using an anti-SR Ca 2ϩ -ATPase polyclonal antibody in conjunction with a fluorescein-conjugated anti-rabbit secondary antibody and appears green. B, digital overlay of staining patterns for both proteins. Regions, where staining for both phospholamban and the SR Ca 2ϩ -ATPase are superimposed, appear yellow and indicate that the two antigens are co-distributed, at the limits of resolution of light microscopy. Regions in which they are not superimposed appear either red or green. Control experiments were performed in which the primary antibody was omitted and no fluorescence was detected. Images were produced using laser confocal scanning microscopy. Bar, 10 m. stimulation, we exposed both wild-type and transgenic muscles to cumulative doses of isoproterenol. Isoproterenol stimulated force production by 30% in both wild-type and transgenic muscles (data not shown) but had no effect on either the contraction or the half-relaxation times of wild-type and transgenic EDL muscles (Fig. 7). Thus, isoproterenol could not reverse the depressed rates of relaxation observed in the transgenic EDL muscles on phospholamban expression. To determine whether this lack of isoproterenol stimulation was due to defects in the transgenic protein, which prevented its phosphorylation by cAMP-dependent protein kinase, transgenic homogenates were phosphorylated in vitro with the catalytic subunit of protein kinase A. Phosphorylation of transgenic phospholamban was associated with alterations in its electrophoretic mobility in SDS-polyacrylamide gels (data not shown), consistent with previous reports (35), indicating that there were no inherent defects that precluded phosphorylation of phospholamban in vivo.
Since isoproterenol is an activator of both ␤ 1 -and ␤ 2 -adrenergic receptors (36), we investigated whether the isoproterenol-elicited increases in developed force in the EDL muscle were due to ␤ 1 -or ␤ 2 -adrenergic receptor activation. Isolated, intact EDL muscles were preincubated with either CGP-210712A or ICI-118551, which are specific ␤ 1 -and ␤ 2 -adrenergic receptor antagonists, respectively (35). Pretreatment with CGP-210712A had no effect, while ICI-118551 completely abolished the isoproterenol-induced increases in developed force (data not shown). These findings suggest that the effects of isoproterenol in the mouse fast-twitch skeletal muscle are mediated through the ␤ 2 -adrenergic receptor pathway and that its inability to reverse the depressed rates of relaxation in the transgenic muscles may be due to the lack of phospholamban phosphorylation.
To determine whether phospholamban is phosphorylated during isoproterenol stimulation, we exposed isolated transgenic EDL muscles to 6 M isoproterenol and then subjected the muscle homogenates to Western blot analysis. Two sitespecific polyclonal antibodies, which recognize phosphorylated serine-16 or threonine-17 residues in phospholamban (24,25), were used to determine the phosphorylation status of phospholamban. We could not detect phosphorylation of either serine or threonine residues in phospholamban after isoproterenol stimulation in the transgenic skeletal muscles (Fig. 8). However, in parallel studies both phosphorylated residues of endogenous phospholamban could be identified after isoproterenol stimulation of mouse hearts (Fig. 8). These findings suggest that activation of ␤ 2 -adrenergic signaling pathways in mouse EDL muscles does not promote phospholamban phosphorylation, in agreement with previous observations in rat cardiomyocytes (35). DISCUSSION This study presents the first evidence that SERCA1 activity can be regulated by phospholamban in vivo, resulting in depressed rates of fast-twitch skeletal muscle relaxation. A transgenic mouse model was generated, which ectopically expressed phospholamban exclusively in fast-twitch skeletal muscle under the control of the rat myosin light chain 1f promoter and enhancer elements. Indirect immunofluorescence labeling of phospholamban and SR Ca 2ϩ -ATPase in transgenic gastrocnemius tissues indicated that the two proteins were co-distributed, suggesting that the transgenic phospholamban was incorporated into the SR membranes. This ectopic phospholamban expression was associated with decreases in the affinity of the SR Ca 2ϩ pump for Ca 2ϩ , in agreement with previous studies involving phospholipid bilayer reconstitution (9,18) or cell culture co-expression studies (8, 16) of phospholamban and SERCA1. Although we cannot presently exclude the possibility that modulation of SR Ca 2ϩ uptake activity occurs through the FIG. 6. Typical isometric twitch contractions of isolated wildtype and transgenic extensor digitorum longus muscles maintained at L max and relationship between half-relaxation time and the levels of ectopic phospholamban expression. A, polygraph chart recorder tracings obtained at a linear chart speed of 100 mm/s (1 mm ϭ 10 ms). B, relationship between half-relaxation time and the levels of ectopic phospholamban expression. The half-relaxation times for isometric twitch contractions of EDL muscles were plotted against the relative phospholamban levels in wild-type muscles and two transgenic lines. Wild-type (n ϭ 4) muscles express no phospholamban, while lines 75 (n ϭ 7) and 76 (n ϭ 3) exhibit phospholamban expression levels that are ϳ70 and 40% of those in cardiac muscle, respectively. formation of Ca 2ϩ -selective phospholamban channels (37), our findings indicate that the affinity of the SR transport system for calcium was altered without any effects on its maximal velocity.
The generation of this ectopic expression model allowed us to extend our previous studies on phospholamban-mediated regulation of SERCA1 activity in vitro (9,16,18) and to examine the functional implications of this modulation in the intact muscle. The inhibitory effects of phospholamban expression on SR Ca 2ϩ uptake reflected alterations in the contractile properties of isolated fast-twitch skeletal muscles. Examination of isometric twitch dynamics of transgenic EDL muscles revealed a prolongation in the half-relaxation time, without alterations in the contraction time, as compared with the wild-type EDL muscles. The observed changes in relaxation rates were not due to a reduction in SR Ca 2ϩ -ATPase expression levels, since the V max of the Ca 2ϩ -ATPase activity was similar in transgenic and nontransgenic muscles. Furthermore, the decreases in the rates of relaxation appeared to correlate with the levels of phospholamban expressed in the transgenic fast-twitch skeletal muscles. However, it is not currently known whether the observed physiological alterations were only due to phospholamban expression or whether an unknown compensatory response also occurred in the transgenic mice, which led to an increase in passive muscle stiffness.
Since the inhibitory effects of phospholamban in cardiac and slow-twitch skeletal muscles can be relieved during ␤-adrenergic stimulation (11,20,27,28,32), we exposed the wild-type and transgenic EDL muscles to cumulative doses of isoproterenol, a mixed ␤-adrenergic agonist. Isoproterenol stimulated force production by 30% but had no effect on either the contraction or half-relaxation times in wild-type EDL muscles, in agreement with previous findings (38). The observed stimulation of force production by isoproterenol in this study may be due to cAMP-dependent and/or Ca 2ϩ -calmodulin-dependent phosphorylation of key calcium-handling proteins such as the sarcolemmal calcium channel (39), the ryanodine receptor (40,41), and the SR Ca 2ϩ -ATPase (42). While the functional implications of these phosphorylations is currently unclear, it is possible that isoproterenol stimulation may alter the activities of these proteins, thereby increasing the cytosolic calcium available for contraction and leading to enhanced force production. In the transgenic EDL muscles, isoproterenol stimulation was also associated with a 30% increase in developed force and had no effect on the contraction or half-relaxation times, consistent with its effects in the wild-type EDL muscles. Thus, isoproterenol could not reverse the prolonged half-relaxation time associated with phospholamban expression in this muscle. These findings in fast-twitch skeletal muscle appear to be in contrast with our recent observations in the murine soleus, which suggested that phospholamban was a key modulator of the effects of isoproterenol on the half-relaxation time of mouse slow-twitch skeletal muscle (27). Therefore, we postulated that the failure of isoproterenol to reverse the inhibitory effects of phospholamban might be due to the lack of phospholamban phosphorylation in the transgenic muscles. Actually, previous studies in rat cardiomyocytes suggested that phosphorylation of phospholamban may occur during ␤ 1 -but not ␤ 2 -adrenergic receptor activation (36). To determine which receptor subtype mediates the effects of isoproterenol in the EDL muscle, we used specific ␤ 1 -and ␤ 2 -adrenergic receptor antagonists. Our results indicate that the effects of isoproterenol in mouse fasttwitch skeletal muscle are mediated via the ␤ 2 -adrenergic receptor cascade. In addition, Western blot analysis of transgenic muscles, using antibodies that specifically recognize phosphoserine-16 or phosphothreonine-17 residues in phospholamban (24,25), revealed that isoproterenol stimulation was not associated with any detectable phosphorylation of phospholamban. Together, these findings suggest that the effects of isoproterenol in mouse fast-twitch skeletal muscle occur through activation of ␤ 2 -adrenergic receptors, which are not coupled to phospholamban phosphorylation. These results in fast-twitch skeletal muscle are consistent with those of Xiao et al. (35), who reported that specific activation of ␤ 2 -adrenergic receptors in rat cardiac myocytes was not associated with changes in Ca 2ϩ dynamics, contractility, and phospholamban phosphorylation.
In summary, our findings indicate that the fast-twitch skel- FIG. 8. Phospholamban phosphorylation during isoproterenol stimulation. Transgenic EDL muscles were exposed to 6 M isoproterenol for 5 min and then immediately frozen in liquid nitrogen during the peak of the isoproterenol response (EDL ϩ Iso). In parallel studies, mouse hearts were Langendorff-perfused, stimulated with isoproterenol (1 M), and frozen at the peak of the inotropic response. Transgenic muscles that had not been exposed to isoproterenol (EDL Ϫ Iso) were also processed in parallel. The muscles were then homogenized in homogenization buffer containing 1 mM okadaic acid to inhibit endogenous phosphatase activity, and 10, 20, or 30 g of the homogenates were gel electrophoresed and subjected to Western blot analysis as described under "Experimental Procedures." The membranes were reacted with polyclonal antibodies that specifically recognize either: phosphoserine-16 of phospholamban (A) or phosphothreonine-17 of phospholamban (B) and then reacted with an alkaline phosphatase-conjugated anti-rabbit secondary antibody. FIG. 7. Effects of isoproterenol stimulation on contraction (A) and half-relaxation (B) times in wild-type and transgenic extensor digitorum longus muscles. After equilibration and adjustment to L max , the muscles were exposed to cumulative doses of isoproterenol (1 nM-10 M). Each dose was incubated for 5 min; data represent the mean Ϯ S.E. for 3 muscles from 3 wild-type and 3 transgenic animals.
etal muscle isoform of the SR Ca 2ϩ -ATPase may be regulated by phospholamban in vivo and is associated with significant decreases in the rate of relaxation of transgenic muscles. Thus, induction of phospholamban expression in fast-twitch skeletal muscle would result in some functional alterations resembling its slow-twitch counterpart and therefore compromise the physiological responsiveness of fast-twitch fibers especially in "fight or flight" situations. Future studies designed to elucidate the mechanisms involved in dictating tissue-specific expression of phospholamban may identify the factors/elements missing in fast-twitch skeletal muscle, which prevent expression of this important regulatory protein in vivo.