Sarcolipin overexpression in rat slow twitch muscle inhibits sarcoplasmic reticulum Ca2+ uptake and impairs contractile function.

Sarcolipin (SLN) is an inhibitor of sarco(endo)plasmic reticulum Ca(2+)-ATPases (SERCAs) in vitro, but its function in vivo has not been defined. NF-SLN cDNA (SLN tagged N-terminally with a FLAG epitope) was introduced into rat soleus muscle in one hindlimb by plasmid injection and electrotransfer. Western blotting showed expression and co-immunoprecipitation showed physical interaction between NF-SLN and SERCA2a. Contractile properties and SERCA2a function were assessed and compared with vector-injected contralateral soleus muscles. NF-SLN reduced both peak twitch force (P(t)) (123.9 +/- 12.5 versus 69.8 +/- 8.9 millinewtons) and tetanic force (P(o)) (562.3 +/- 51.0 versus 300.7 +/- 56.9 millinewtons) and reduced both twitch and tetanic rates of contraction (+dF/dt) and relaxation (-dF/dt) significantly. Repetitive stimulation (750-ms trains at 50 Hz once every 2 s for 3 min) showed that NF-SLN increased susceptibility to fatigue. These changes in contractile function were observed in the absence of endogenous phospholamban, and NF-SLN had no effect on either SERCA2a or SERCA1a expression levels. NF-SLN also decreased maximal Ca(2+) transport activity at pCa 5 by 31% with no significant change in apparent Ca(2+) affinity (6.36 +/- 0.07 versus 6.39 +/- 0.08 pCa units). These results show that NF-SLN expression impairs muscle contractile function by inhibiting SERCA function and diminishing sarcoplasmic reticulum Ca(2+) stores.

Our initial studies indicated that SLN decreased the apparent affinity of SERCA1a for Ca 2ϩ and increased maximal transport activity at high Ca 2ϩ concentrations (6). Our more recent studies have confirmed that co-expression of NF-SLN with SERCAs in HEK-293 cells decreases the apparent affinity of both SERCA1a and SERCA2a for Ca 2ϩ but did not confirm that NF-SLN increases maximal transport activity of SERCA1a in high Ca 2ϩ (7). The apparent increase in maximal transport activity of SERCA1a observed earlier (6) could be explained by an underestimate in the enzyme-linked immunosorbant assay used in that study to measure the amount of SERCA1a (7). NF-SLN was found to be a more effective inhibitor of SERCA2a than of SERCA1a, being almost equal to PLN in its ability to decrease apparent Ca 2ϩ affinity and, in contrast to PLN, even decreasing maximal transport activity at pCa 5 (7).
PLN is a well characterized regulator of SERCA2a activity in cardiac muscle (8). Inhibitory interactions between PLN and SERCA2a result in a decrease in the apparent affinity of SERCA2a for Ca 2ϩ (9) with no effect on the maximal rate of Ca 2ϩ uptake by SERCA2a (7,10,11). Inhibitory interactions can be reversed by elevation of cytosolic Ca 2ϩ or by phosphorylation of Ser 16 and Thr 17 in PLN cytosolic domain 1A (12). Several studies have shown that PLN is a major regulator of left ventricular basal contractile parameters and their responses to ␤-agonists (13)(14)(15)(16). Ablation of PLN is associated with significant increases in cardiac contractility and left ventricular systolic function (14,15), whereas the 2-fold overexpression of wild-type PLN in transgenic mice had an inhibitory effect on both the kinetics of Ca 2ϩ transients and contractile parameters in ventricular myocytes and impaired basal left ventricular systolic function in vivo (16). Structural similarities between SLN and PLN genes and SLN and PLN protein sequences indicate that the two genes are members of a family (4).
The physiological function of SLN has not been evaluated, either in skeletal muscle or in cardiac muscle. Because NF-SLN and PLN affect SERCA2a function in a similar fashion in vitro, it is likely that SLN and PLN would affect muscle contractility in a similar fashion. In this study, we expressed NF-SLN in rat soleus muscle by intramuscular injection and electrotransfer of rabbit NF-SLN cDNA to explore the possibility that SLN can regulate SERCA2a activity and slow twitch soleus muscle contractility just as PLN can regulate cardiac contractility. We found that NF-SLN reduces peak isometric force, slows the rates of contraction and relaxation, and increases susceptibility to fatigue. Ca 2ϩ uptake in postnuclear homogenates from these muscles was also impaired. We propose that NF-SLN can impair muscle contractile function indirectly by inhibiting SERCA function and thus lowering basal Ca 2ϩ stores in the sarcoplasmic reticulum.

EXPERIMENTAL PROCEDURES
Materials-Enzymes for DNA manipulation were from New England Biolabs and Pharmacia Corp., and the expression vector pcDNA3.1 was from Invitrogen. Protein G-Sepharose and a chemiluminescence kit for measurement of co-immunoprecipitation and immunoblotting were from Pierce. FLAG antibody M2 and the monoclonal antibody 5C5 against ␣-sarcomeric actin were from Sigma; the anti-PLN antibody 1D11 (17) was a gift from Dr. Robert Johnson (Merck Research Laboratories); the monoclonal antibody 2A7-A1 against SERCA2a and SERCA2b was from Affinity Bioreagents Inc. The A52 monoclonal antibody against SERCA1 was produced in our laboratory (18).
Reverse Transcription-PCR for Estimation of SLN Expression in Total mRNA-Reverse transcription-PCR was performed on total RNA isolated from soleus and heart using guanidium-isothiocyanate-phenolchloroform extraction (19), the latter being included as a positive marker for SLN because SLN is expressed highly in rat heart (5). RNA quality and quantity were determined by absorbance at 260 and 280 nm.
A 5-g aliquot of total RNA was reverse transcribed in 20 l of a reaction mix (Sigma Enhanced Avian HS reverse transcription-PCR kit) containing 20 units of eAMV reverse transcriptase at 42°C in the presence of RNase inhibitor and 1 M specific SLN reverse primer (SLN, 5Ј-GGG AGT GAC TGC TGT GTG CCC T-3Ј). The polymerase chain reaction was carried out in a total volume of 50 l with JumpStart AccuTaq LA DNA polymerase mix (Sigma), with 200 M dNTPs, 0.4 M each of the appropriate 5Ј and 3Ј primers (SLN, forward, 5Ј-GGT GTG CAC TCA GAA GTC CTC CT-3Ј and reverse, 5Ј-GGG AGT GAC TGC TGT GTG CCC T-3Ј) and 5 l of first strand cDNA. Denaturation was for 1 min at 94°C, annealing was at 63°C for 1 min, and extension for 35 cycles was at 72°C for 1 min, followed by a final extension at 72°C for 10 min. The amplified products were separated on 1.2% agarose gel containing ethidium bromide.
Cell Culture and Heterologous Expression-The culture of HEK-293 cells, their transfection with cDNAs encoding SERCA1a, SERCA2a, and NF-SLN, and the isolation of microsomal fractions from transfected cells expressing these proteins have been described in earlier publications (6,20,21).
Construction of NF-SLN-The preparation of the rabbit NF-SLN cDNA construct was described previously (6). NF-SLN is a fusion protein of SLN with a FLAG epitope at its N terminus that does not alter its function (6). The NF-SLN construct was cloned into the XbaI and XhoI sites of the expression vector pcDNA3.1.
DNA Injection and Electric Pulse Delivery-Injection and electric pulse delivery of the NF-SLN cDNA construct into rat soleus muscle was achieved using the protocol of Mir et al. (22). Male Sprague-Dawley rats weighing 339 Ϯ 12 g (n ϭ 8) were fully anesthetized with a mixture of ketamine (50 -100 mg/kg) and xylazine (5-10 mg/kg). After achieving full anesthesia, each soleus muscle was exposed and injected with a total of 150 g of DNA dissolved in 0.9% sterile NaCl (0.5 g/l) using a 26G 0.5 stainless steel needle. For each animal, the soleus muscle from the experimental hindlimb was injected with NF-SLN, and the soleus muscle from the contralateral control hindlimb was injected with an equal concentration of the expression vector pcDNA3.1. Approximately 1 min after DNA injection, electric pulses (9 pulses, 2 Hz, 200 V/cm, 20 ms/pulse) were applied by two implanted stainless steel needle electrodes (21G 1.5) connected to a Grass stimulator (Grass S88). Following electroporation, incisions in the skin directly covering each soleus muscle were closed using surgical staples, and the animals were allowed to recover under a heat lamp.
Electrical Stimulation and Muscle Contractile Measurements-Three days after DNA injection, electrically evoked muscle force was assessed in situ from the soleus muscle from each hindlimb across a range of stimulation frequencies from 1 to 70 Hz. The rats were anesthetized initially using a mixture of ketamine (50 -100 mg/kg) and xylazine (5-10 mg/kg) and given maintenance intraperitoneal injections of pentobarbital sodium (40 mg/kg), as required. All other details pertaining to the surgical preparation and the animal stimulation apparatus have been described elsewhere (23,24). Soleus muscles contracted isometrically (i.e. length remained constant), and both twitch (1 Hz) and tetanic (10 -70 Hz) force were obtained via direct muscle stimulation, with the use of stainless steel electrodes and with a single 0.2-ms pulse at 70 V. Muscle stimulation was performed using a Grass S88 stimulator, and force data were collected on-line using a 640A signal interface (Aurora Scientific Inc.) connected to a National Instruments 16-bit A/D card and analyzed using the Dynamic Muscle Control and Data Acquisition (DMC) and Dynamic Muscle Analysis (DMA) Software (Aurora Scientific Inc.). An independent calibration was performed daily for each force transducer (Grass, model FT 10). At least 5 min before commencing data collection, optimal length (L o ) for peak twitch force (P t ) was established. Peak tetanic force (P o ) occurred at a stimulation frequency of 50 Hz, and only P t and P o data are reported here, including peak force amplitude and peak rates of contraction (ϩdF/dt) and relaxation (ϪdF/ dt). To assess the effects of NF-SLN on soleus susceptibility to fatigue, each muscle was stimulated repeatedly for 3 min using a fatigue protocol consisting of 700-ms contractions at 50 Hz, once every 2 s. Fatigue data are expressed as percentages of resting data. Experimental protocols were approved by the Animal Care Committee of the University of Toronto.
Preparation of Postnuclear Homogenates-Immediately after muscle contractile measurements, soleus muscles were excised, weighed, and diluted 10-fold (w/v) in homogenizing buffer containing 5 mM HEPES, pH 7.5, 250 mM sucrose, 0.2 mM phenylmethylsulfonyl fluoride, and 1 mM dithiothreitol and homogenized with a Tissumizer (Tekmar Company, TR-10) at 90% maximal power with two 30-s bursts. The homogenates were frozen immediately in liquid nitrogen for later analysis. Just prior to analysis, the homogenates were thawed on ice and centrifuged at 1,700 ϫ g in an Eppendorf 5415C centrifuge (Brinkman Instruments Ltd.) for 5 min to sediment nuclei and cell debris. The postnuclear homogenate was assessed for protein concentration using the Bio-Rad method with bovine serum albumin as a standard and was used for all subsequent analyses, unless otherwise stated.
Western Blot Analysis-Western blotting was performed to determine the relative expression levels of NF-SLN, PLN, SERCA2a, and SERCA1a (25) in rat soleus muscles that were either injected with NF-SLN or the expression vector pcDNA3.1. After assuring linearity of band density, 20 l (20 g) of the postnuclear homogenate were applied to 12.5% (NF-SLN and PLN) or 8% (SERCA2a and SERCA1a) acrylamide gels, separated by SDS-PAGE, and transferred to nitrocellulose membranes. After blocking with 5% skim milk, the membranes were incubated with both anti-␣-sarcomeric actin antibody 5C5 to control for protein loading and either anti-FLAG antibody M2, anti-PLN antibody 1D11, anti-SERCA2a antibody 2A7-A1, or anti-SERCA1a antibody A52 for 1 h at room temperature and then washed with Tris-buffered saline with 0.1% Tween 20. They were then treated with horseradish peroxidase-conjugated goat anti-mouse secondary antibody (Promega) and stained with an enhanced chemiluminescence kit (Pierce Super Signal). Densitometric analysis was performed using NIH Image 16.1 software. Both samples for a given animal were run in duplicate on separate gels along with a standard. Microsomes prepared from HEK-293 cells that had been transfected with either NF-SLN or PLN cDNAs (7) served as positive controls and molecular weight standards for NF-SLN and PLN, respectively.
Comparison of NF-SLN Synthesis in HEK-293 Cells and in Soleus Muscle-Western blotting was performed on post-nuclear homogenates from soleus muscles that expressed NF-SLN and microsomes prepared from HEK-293 cells that expressed both NF-SLN and either SERCA1a or SERCA2a. Soleus and HEK-293 samples were applied to the same 8% gel, separated by SDS-PAGE, transferred to nitrocellulose, and stained for NF-SLN and either SERCA1a or SERCA2a using the appropriate monoclonal antibodies. Densitometric analysis was performed to determine the relative densitometric ratios of SERCA1a and SERCA2a to NF-SLN in both rat soleus and HEK-293 cells.
Co-immunoprecipitation from Postnuclear Homogenates-To confirm association between SERCA2a, the predominant SERCA isoform expressed in rat soleus (26), and NF-SLN, co-immunoprecipitation of SERCA2a and NF-SLN was carried out as described previously (21) using aliquots of postnuclear homogenates at 1 mg protein/ml and a protein G-Sepharose/NF-SLN FLAG antibody M2 complex. The presence of SERCA2a associated with NF-SLN in the samples was then detected by Western blotting and detection procedures using the anti-SERCA2a monoclonal antibody 2A7-A1.
Ca 2ϩ Transport Assay-Ca 2ϩ transport activity in postnuclear homogenates at 1 mg protein/ml was assayed in 150 l of a reaction mixture containing 20 mM MOPS/Tris-HCl, pH 6.8, 100 mM KCl, 5 mM MgCl 2 , 5 mM ATP, 5 mM potassium oxalate, and about 10 g of protein as described previously (20,21). The data were analyzed by nonlinear regression using the Sigma Plot Scientific graph system obtained from Jandel Scientific. K 0.5 values were calculated using an equation for a general cooperative model for substrate activation. The values for maximal transport activity that occurred at pCa 5.5 to pCa 5 were taken directly from the experimental data and normalized for total protein concentration. The maximal transport activity values are reported as percentages of control (pcDNA), which was set to 100%. On a given analytical day, the samples from all conditions were analyzed in duplicate.
Statistical Analysis-The data are presented as the means Ϯ S.E. Statistical analysis was performed by t test between control (pcDNA only) and soleus muscles injected with NF-SLN. p Ͻ 0.05 was considered statistically significant.

Endogenous SLN Expression in Rat Soleus and Heart-Prep-
arations of total RNA from rat soleus muscle and rat heart were used in reverse transcription-PCR reactions to determine whether SLN is expressed endogenously in these tissues. We did not attempt to quantify SLN expression in either rat soleus or cardiac muscles, but we were able to confirm endogenous SLN expression in both of these tissues (data not shown).
Expression and Localization of NF-SLN cDNA in Rat Soleus-Rabbit NF-SLN cDNA was inserted into the pcDNA3.1 expression vector under regulation by the cytomegalovirus immediate-early promoter contained in the vector. To confirm the presence of NF-SLN in rat soleus muscles injected with NF-SLN cDNA, we performed Western blotting on postnuclear homogenates from experimental muscles injected with the NF-SLN cDNA and control muscles injected with vector pcDNA3.1. NF-SLN was detected only in muscles that were injected with NF-SLN cDNA (Fig. 1A). Although NF-SLN expression was consistent across muscles, it was variable from experiment to experiment with standard deviations ranging from 1.7 S.D. above the mean to 1.1 S.D. below the mean (Fig. 1B).
To confirm that NF-SLN was localized with SERCA2a in the same microsomal membrane fraction, co-immunoprecipitation of SERCA2a with NF-SLN was carried out using the M2 antibody against NF-SLN. Fig. 2 shows that physical interactions occurred between SERCA2a and NF-SLN in soleus muscles injected with NF-SLN cDNA.

Expression of PLN, SERCA2a, and SERCA1a in Rat Soleus
Muscles-The relative expression of both SERCA2a (Fig. 3A) and SERCA1a (Fig. 3B) were compared between soleus muscles injected with NF-SLN cDNA and controls. Overall, there were no differences (p Ͼ 0.05) in either SERCA2a or SERCA1a expression levels between NF-SLN and control muscles (Fig.  3C). We confirmed an earlier report (27) in which immunoblot and Northern blot analysis were used to show that PLN is not expressed in rat soleus muscle (Fig. 3D).
It was of interest to obtain a measure of the amount of NF-SLN synthesis that occurred in soleus muscle following injection of NF-SLN cDNA. It was not feasible to determine absolute amounts, because purified NF-SLN was not available to us. It was, however, possible to determine the relative level of NF-SLN in relation to the amount of SERCA in soleus muscle and to compare the results with the ratio between NF-SLN and SERCA expressed in HEK-293 cells under conditions where optimal physiological effects are obtained.
Western blotting was carried out on post-nuclear homogenates from soleus muscles that expressed NF-SLN and on microsomes prepared from HEK-293 cells that expressed NF-SLN together with either SERCA1a or SERCA2a. Soleus and HEK-293 samples were applied to the same 8% gel, separated by SDS-PAGE, transferred to nitrocellulose, and stained for NF-SLN and either SERCA1a or SERCA2a using the appropriate monoclonal antibodies. Densitometric analysis was performed to determine the ratios of SERCA1a and SERCA2a to NF-SLN in the transfected rat soleus homogenate. Densitometric analysis was also used to determine ratios of SERCA1a to NF-SLN and SERCA2a to NF-SLN in HEK-293 cells that expressed NF-SLN and either SERCA1a or SERCA2a.
Analysis of all eight soleus samples that expressed NF-SLN distinguished three samples that expressed the highest amount of NF-SLN relative to SERCA1a and SERCA2a (Fig.  1B, lanes 1, 5, and 8). If we set the average SERCA:NF-SLN ratio in HEK-293 cells at 1, then we could calculate that the average densitometric ratio of SERCA2a to NF-SLN expression in soleus muscle was 3.75, and the ratio of SERCA1a to NF-SLN was 8.5. If we assume that SERCA1a and SERCA2a each account for about 50% of the total SERCA in soleus muscle (28), then the average densitometric ratio of SERCA1a plus SERCA2a to NF-SLN would be about 6. On the basis of endogenous SERCA expression in soleus muscle, we are expressing only about 1 ⁄6 of the amount of NF-SLN that has optimal effects on SERCA function in HEK-293 cells. However, SLN is expressed endogenously in rat soleus muscle, presumably at an optimal level. Thus, in our experiments, it is the effect of NF-SLN overexpression on top of an optimal base line of SLN expression that we are measuring in our physiological studies of transfected soleus muscle. We present the analysis of muscle Antibody M2 against NF-SLN was used to co-immunoprecipitate SERCA2a from postnuclear homogenates from experimental and control soleus muscles. The co-immunoprecipitates were separated by SDS-PAGE, transferred to nitrocellulose, and stained with antibody 2A7-A1 against SERCA2a.

Physiological Effects of NF-SLN Expression
contractile function and SERCA function for the three experiments in which NF-SLN was most highly expressed.
Effects of NF-SLN Expression on Slow Twitch Muscle Contractility and Susceptibility to Fatigue-To evaluate the physiological effects of NF-SLN expression in slow twitch skeletal muscle, isometric contractile characteristics, including both twitch (Fig. 4A) and tetanus (Fig. 5A), from soleus muscles injected with NF-SLN cDNA were measured in situ and compared with contractile characteristics assessed simultaneously in controls. NF-SLN reduced (p Ͻ 0.05) peak twitch force (P t ) by 44% compared with control (Fig. 4B). The maximal twitch rate of contraction (ϩdF/dt) was reduced (p Ͻ 0.05) by 34%, and the maximal twitch rate of relaxation (ϪdF/dt) was reduced (p Ͻ 0.05) by 25% with NF-SLN expression (Fig. 4C). Maximum tetanic force (P o ), which occurred at a stimulation frequency of 50 Hz, was reduced (p Ͻ 0.05) by 47% (Fig. 5B), tetanic ϩdF/dt was reduced (p Ͻ 0.05) by 50%, and tetanic ϪdF/dt was reduced (p Ͻ 0.05) by 40% with NF-SLN expression (Fig. 5C).
Soleus susceptibility to fatigue was assessed by repetitive stimulation of the muscle using the protocol outlined under "Experimental Procedures." The force loss and slowing of contraction that are characteristic of fatigue were greater (p Ͻ 0.05) in soleus muscles expressing NF-SLN compared with control. When expressed as a percentage of normal resting values, P o was reduced (p Ͻ 0.05) by 71 Ϯ 3.9%, and ϩdF/dt was reduced by 45 Ϯ 6.6% in NF-SLN-expressing muscles compared with reductions in P o (p Ͻ 0.05) of 65 Ϯ 2.8 and reductions in ϩdF/dt of 32 Ϯ 7.6% in controls. In contrast, the peak relaxation rate was slowed with fatigue to a similar extent: 63 Ϯ 5.1% in NF-SLN muscle and 71 Ϯ 7.3% in control muscle.
Ca 2ϩ Uptake by Sarcoplasmic Reticulum-Postnuclear homogenates were adjusted to a concentration of 1 mg protein/ml and assayed for Ca 2ϩ dependence of Ca 2ϩ uptake to assess the effects of NF-SLN expression on SERCA function. NF-SLN reduced absolute Ca 2ϩ uptake across a range of free Ca 2ϩ levels from pCa 7 to pCa 5 (Fig. 6). NF-SLN reduced Ca 2ϩ uptake at pCa 5.5 to pCa 5 such that maximal transport activity was reduced (p Ͻ 0.05) by ϳ31% compared with control. There was no significant change, however, in apparent Ca 2ϩ affinity, expressed as K Ca in pCa units (6.36 Ϯ 0.07 versus 6.39 Ϯ 0.08, mean Ϯ S.E.).
These data differ from those obtained in studies of co-expression of NF-SLN with SERCA1a or SERCA2a in HEK-293 cells, where a significant change was observed in apparent Ca 2ϩ affinity (7). This difference is most likely due to differences in the molar ratios of NF-SLN expression compared with the expression of SERCA. In our studies with HEK-293 cells (7), we noted changes in ⌬K Ca of Ϫ0.17 to Ϫ0.34 pCa units when the molar ratios of NF-SLN to SERCA1a or SERCA2a cDNAs used in transfection were 4:1. However, if the ratios were reduced to 1:1, then no differences in ⌬K Ca were observed in the presence and absence of NF-SLN expression. It is probable that the ratio of SERCA to SLN remains high, even after overexpression of NF-SLN. FIG. 3. Western blot analysis of SERCA2a, SERCA1a, and PLN. Postnuclear homogenates were prepared from soleus muscles injected with NF-SLN and contralateral control muscles injected with vector only. A, postnuclear homogenates (20 g) from all eight rats were separated by SDS-PAGE on 8% acrylamide gels, transferred to nitrocellulose, and stained with antibody 2A7-A1 against SERCA2a and antibody 5C5 against ␣-actin. B, postnuclear homogenates (20 g) from all eight rats were separated by SDS-PAGE on 8% acrylamide gels, transferred to nitrocellulose, and stained with antibody A52 against SERCA1 and antibody 5C5 against ␣-actin. C, when corrected relative to ␣-actin content, there were no differences (p Ͼ 0.05) in SERCA1a or SERCA2a expression levels between control (pcDNA) and NF-SLN samples. D, samples representing a soleus muscle that expressed NF-SLN and a paired control sample were incubated with antibody 1D11 against PLN. PLN was not expressed in these muscles. Right lane, 20 g of microsomal protein from HEK-293 cells transfected with PLN was used as a positive control and M r standard for PLN. Bands indicating the pentamer (p) and monomer (m) forms of PLN are shown. std, standard.

DISCUSSION
To assess the physiological function of SLN in vivo, we expressed NF-SLN in rat soleus muscles from one hindlimb while the contralateral limb served as a control. Our results confirm our most recent findings from in vitro studies that SLN acts as an inhibitor of SERCA function over a concentration range from pCa 7 to pCa 5, resulting in a decreased apparent Ca 2ϩ affinity and lower maximal transport activity (7). This decrease in Ca 2ϩ uptake is reflected in muscle contractile performance,

FIG. 4. Effect of NF-SLN on soleus isometric twitch properties.
Three days following injection and electrotransfer of NF-SLN into rat soleus, isometric twitch properties were measured in situ as described under "Experimental Procedures" and compared with twitch properties assessed simultaneously in contralateral control muscles, which were injected with the expression vector pcDNA3.1. Twitch contractile properties from three muscles that expressed the highest levels of NF-SLN and paired control muscles are shown. A, a typical twitch force record for a soleus muscle that was injected with NF-SLN and a matched control from one animal. B, peak twitch amplitude (P t ) in control (filled bar) and soleus muscles that expressed NF-SLN (open bar). C, maximal twitch rate of contraction (ϩdF/dt) and rate of relaxation (ϪdF/dt) in control (filled bar) and soleus muscles that expressed NF-SLN (open bar). *, p Ͻ 0.05 versus control.

FIG. 5. Effect of NF-SLN on soleus isometric tetanic properties.
Three days following injection and electrotransfer of NF-SLN into rat soleus, isometric tetanic properties (50 Hz) were measured in situ as described under "Experimental Procedures" and compared with tetanic properties assessed simultaneously in contralateral control muscles, which were injected with the expression vector pcDNA3.1. Tetanic contractile properties from three muscles that expressed the highest levels of NF-SLN and paired control muscles are shown. A, a typical tetanic (50 Hz) force record for a soleus muscle that was injected with NF-SLN and a matched control from one animal. B, peak tetanic amplitude (P o ) measured at 50 Hz in control (filled bar) and soleus muscles that expressed NF-SLN (open bar). C, maximal tetanic rate of contraction (ϩdF/dt) and rate of relaxation (ϪdF/dt) in control (filled bar) and soleus muscles that expressed NF-SLN (open bar). *, p Ͻ 0.05 versus control.
which is impaired following expression of NF-SLN in slow twitch skeletal muscle. In resting soleus, both the kinetics and amplitude of contraction were reduced with NF-SLN expression, and when soleus muscles were stimulated repeatedly, fatigue was more pronounced in muscles expressing NF-SLN compared with controls. Thus, NF-SLN expression in slow twitch skeletal muscle is similar to a model of PLN overexpression in the heart, at least in terms of its effects on muscle contractility.
We were able to express NF-SLN in rat soleus using a protocol of injection and electrotransfer of plasmid DNA. Electroporation combined with cDNA injection was necessary to achieve NF-SLN expression, because we were unable to detect NF-SLN in rat soleus muscles that were injected with NF-SLN cDNA, without electroporation (data not shown). This finding is in agreement with other studies that employed in vivo electroporation to transfer plasmid DNA into skeletal muscles (22, 29 -31). Our co-immunoprecipitation experiments and Ca 2ϩ transport assays in postnuclear homogenates indicate that expressed NF-SLN interacts both functionally and physically with SERCA2a in rat soleus sarcoplasmic reticulum membranes.
We confirmed that rat soleus does not express PLN, which is important given that co-expression of PLN and NF-SLN with either SERCA2a or SERCA1a is superinhibitory for SERCA function (7). We have also shown that rat soleus is a useful model system in which to assess the function of SLN co-expressed with SERCA2a and SERCA1a. The design of this study incorporated an internal paired control, and because muscle function and sampling occurred just 3 days following DNA injection, at the earliest time point where peak expression of a typical introduced cDNA is observed (22), compensatory changes in the expression of other functionally related Ca 2ϩ regulatory proteins would be minimal. In fact, we did not find any significant changes in the expression levels of either SERCA2a or SERCA1a with NF-SLN expression.
A limitation in this study was that we were unable to quantify SLN and NF-SLN expression in each muscle. Despite several attempts, an antibody against SLN has not been generated to date, so we could not determine total SLN protein levels or compare expression between control and muscles expressing NF-SLN. The lack of highly purified NF-SLN also made it difficult to carry out measurements of NF-SLN through enzyme-linked immunosorbant assay. SLN mRNA is abundant in rat soleus muscle, suggesting that SLN is also expressed in this muscle.
Despite these limitations, we could show that NF-SLN expression relative to SERCA expression in soleus muscles was only 1 ⁄6 of the level that was optimal for regulation of SERCA in HEK-293 cells. Nevertheless, the mechanical and functional differences between control and NF-SLN expressing soleus muscles were considerable in terms of contractile function and Ca 2ϩ transport, permitting evaluation of the effect of exogenous overexpression of NF-SLN. It is probable that endogenous SLN protein levels are optimal for physiological function and that only small increases in SLN expression can result in impaired function.
The main finding of this study was that NF-SLN expression can act as an inhibitor of SERCA function in vivo and, thereby, impair soleus contractile function. This finding confirms the similarity between NF-SLN and PLN because PLN overexpression can inhibit SERCA function to impair cardiac contractility (16) and, to a lesser extent, skeletal muscle contractility (32). The effects of NF-SLN on force amplitude, rate of contraction, and rate of relaxation observed in this study are similar to the effects of wild-type PLN overexpression in transgenic mice (16) and isolated adult rat ventricular myocytes (33) and to the overexpression of the monomeric, dominant-acting, superinhibitory L37A and I40A mutant forms of PLN in mouse heart (34). In all of these studies, maximum contraction amplitudes and maximum rates of shortening and relengthening measured in isolated myocytes were impaired.
The effects of PLN overexpression on cardiac myocyte contractile function were undoubtedly due to parallel changes in Ca 2ϩ transient amplitudes and kinetics that were observed (16,33). These changes in Ca 2ϩ transient and muscle contractile parameters are all expected to result from reduced sarcoplasmic reticulum Ca 2ϩ uptake and release following impairment of SERCA2a function and lowering of sarcoplasmic reticulum Ca 2ϩ stores (16,33,34). Indeed, the amount of Ca 2ϩ in the sarcoplasmic reticulum stores from adult rat myocytes with 2-fold overexpression of PLN was reduced by 21% (33).
As support for the view that a similar mechanism is responsible for the changes in contractile function that resulted from expression of NF-SLN in rat soleus muscle, the maximal Ca 2ϩ transport activity was reduced by 31% in postnuclear homogenates from muscle expressing NF-SLN compared with control. The calculated K Ca in this study was unaffected by NF-SLN, but the absolute Ca 2ϩ uptake rate was lower with NF-SLN expression compared with control over a range of low to high Ca 2ϩ concentrations. A lower rate of SERCA-mediated Ca 2ϩ removal from the cytoplasm in the presence of NF-SLN would reduce the size of the Ca 2ϩ store in the sarcoplasmic reticulum, FIG. 6. Effect of NF-SLN on SERCA Ca 2؉ transport activity. Ca 2ϩ dependence of Ca 2ϩ transport activity was assessed in post-nuclear homogenates that were prepared from soleus muscles injected with NF-SLN (q) and controls injected with vector only (f). Mean data from three muscles that expressed the highest levels of NF-SLN and paired control muscles are shown. and a reduced Ca 2ϩ store could alter excitation-contraction coupling in skeletal muscle (35), accounting for the overall negative effect of NF-SLN on isometric twitch and tetanic contractile properties observed in this study. Because SLN is normally expressed in the heart, an obvious implication from this study and an earlier study (7) is that overexpression of SLN in the heart would have the potential to impair cardiac contractile function.
In a comparable study, the overexpression of PLN in mouse fast twitch skeletal muscle impaired only the relaxation rate with no effect on the rate of contraction or peak twitch amplitude (32), in contrast to the effects of overexpression of PLN in the heart (16). This is surprising, because mouse fast twitch skeletal muscle should also express SLN, and we have shown that co-expression of NF-SLN and PLN is superinhibitory for SERCA1a function (7), the predominant isoform expressed in fast twitch skeletal muscle (28). In our study it is not possible to define the percentage of SERCA molecules that were inhibited by the sum of NF-SLN plus endogenous SLN, nor was it possible to define the percentage of SERCA molecules that were inhibited by the sum of ectopically expressed PLN plus endogenously expressed SLN in the study of Slack et al. (32). SERCA pumps are much more abundant in fast twitch muscle compared with slow twitch skeletal muscle (36). Therefore, a limitation in both studies is that the systems were not manipulated to achieve optimal ratios of SERCA to inhibitor.
We also assessed the effects of NF-SLN expression on soleus susceptibility to fatigue, an important characteristic that distinguishes slow twitch from fast twitch fibers. Compared with control, resting tetanic force was 47% lower in soleus muscles that expressed NF-SLN. Nevertheless, our fatigue protocol indicated that there was a relatively greater loss of force and reduced rate of contraction in soleus muscles expressing NF-SLN. However, the relative change in relaxation rate was similar between soleus muscles that expressed NF-SLN and control, suggesting that there is no interaction between NF-SLN expression and fatigue in relation to the slowing of relaxation that normally occurs with fatigue (37). These results suggest that NF-SLN expression was only indirectly responsible for the observed differences in susceptibility to fatigue. In the relationship between force and pCa, there is a range over which small changes in pCa lead to large changes in force (38). It is possible that resting soleus muscles expressing NF-SLN, as opposed to control, were already in that sensitive range where small reductions in Ca 2ϩ release would lead to reductions in force, especially given that basal sarcoplasmic reticulum Ca 2ϩ stores and Ca 2ϩ release are suspected to be lower in soleus muscles that express NF-SLN.
Brody disease is an inherited disorder of skeletal muscle function characterized by exercise-induced impairment of muscle relaxation (39). We have associated mutations in the ATP2A1 gene encoding SERCA1 with autosomal recessive inheritance of Brody disease but not with autosomal dominant inheritance (4,40). A search for mutations in the SLN gene in five Brody families that were not linked to ATP2A1 has not revealed any alterations in coding, splice junction, or promoter sequences in the SLN gene (4). On the basis of the results of this study, however, abnormal SLN expression levels should not be ruled out in evaluating SLN as a candidate gene for Brody disease.
In summary, we have found that expression of NF-SLN in rat soleus results in a significant depression in muscle contractility and increased susceptibility to fatigue. Ca 2ϩ uptake in postnuclear homogenates from these muscles was also depressed, confirming that NF-SLN acts as an inhibitor of SERCA function in vivo. These results imply that overexpres-sion of SLN has the potential to impair skeletal muscle relaxation.