The Presence of Sarcolipin Results in Increased Heat Production by Ca2+-ATPase*

Skeletal muscle sarcoplasmic reticulum of large mammals such as rabbit contains sarcolipin (SLN), a small peptide with a single transmembrane α-helix. When reconstituted with the Ca2+-ATPase from skeletal muscle sarcoplasmic reticulum into sealed vesicles, the presence of SLN leads to a reduced level of accumulation of Ca2+. Heats of reaction of the reconstituted Ca2+-ATPase with ATP were measured using isothermal calorimetry. The heat released increased linearly with time over 30 min and increased with increasing SLN content. Rates ATP hydrolysis by the reconstituted Ca2+-ATPase were constant over a 30-min time period and were the same when measured in the presence or absence of an ATP-regenerating system. The calculated values of heat released per mol of ATP hydrolyzed increased with increasing SLN content and fitted to a simple binding equation with a dissociation constant for the SLN·ATPase complex of 6.9 × 10–4 ± 2.9 × 10–4 in units of mol fraction per monolayer. It is suggested that the interaction between Ca2+-ATPase and SLN in the sarcoplasmic reticulum could be important in thermogenesis by the sarcoplasmic reticulum.

involves binding of ATP and two Ca 2ϩ ions from the cytoplasm to the E1 conformation of the Ca 2ϩ -ATPase followed by phosphorylation of the Ca 2ϩ -ATPase on Asp-351 (steps 1-3 in Scheme 1). On phosphorylation of the Ca 2ϩ -ATPase, the two Ca 2ϩ binding sites change to a state in which they are of low affinity and facing the lumen, so that Ca 2ϩ dissociates from the phosphorylated Ca 2ϩ -ATPase into the SR lumen (step 4). Dephosphorylation of the Ca 2ϩ -ATPase then allows recycling to E1 (steps 5-7).
A number of processes compete with the transport process to reduce the net level of accumulation of Ca 2ϩ below two per ATP molecule hydrolyzed. At high lumenal and low cytoplasmic concentrations of Ca 2ϩ , when the concentration of ADP is high, the reaction cycle can be driven backwards, ADP binding to the phosphorylated ATPase, reforming ATP at the expense of moving two Ca 2ϩ ions back across the membrane (steps 4 to 2) (4). The rate of this process is low when the concentration of ADP is low and when the concentration of Ca 2ϩ in the cytoplasm is high (4). Two other processes can also reduce the efficiency of transport. The first is passive leak of Ca 2ϩ out of the SR lumen down its concentration gradient (steps 8 -10). Passive leak of Ca 2ϩ has been observed from Ca 2ϩ -loaded SR vesicles (8) and from vesicles reconstituted from purified Ca 2ϩ -ATPase (9,10), but the rate of this passive leak is slow when the cytoplasmic concentration of Ca 2ϩ is sufficient to saturate the cytoplasmic binding sites for Ca 2ϩ on the ATPase (10,11). The second process is slippage, in which the phosphorylated, Ca 2ϩ -bound intermediate (E2PCa 2 in Scheme 1) releases its two Ca 2ϩ ions to the cytoplasmic side of the membrane rather than to the lumenal side (step 11) (8,12,13). Because during slippage there is no transport of Ca 2ϩ , all the energy derived from ATP hydrolysis will be converted into heat. Conditions favoring maximum heat production by the Ca 2ϩ -ATPase, therefore, correspond to those that favor slippage (5).
The SR of fast-and slow-twitch skeletal muscles of large mammals such as rabbit contain a 31-residue transmembrane peptide sarcolipin (SLN), which is absent from atrial muscle (14). In contrast, in small mammals such as rat, SLN is present in the atria but absent from skeletal muscles (14). In reconstitution experiments with Ca 2ϩ -ATPase from rabbit skeletal fast-twitch muscle, the presence of SLN led to a reduced level of accumulation of Ca 2ϩ (7), and expression of SLN in rat slowtwitch muscle, a muscle that normally lacks SLN, also led to a decreased level of accumulation of Ca 2ϩ by the SR (15). The effect of SLN on Ca 2ϩ accumulation in reconstituted vesicles was consistent with interaction between SLN and the Ca 2ϩ -ATPase, leading to an increased rate of slippage on the Ca 2ϩ -ATPase (7). This suggests that the presence of SLN might increase heat production by the Ca 2ϩ -ATPase; here we show that this is indeed the case.

EXPERIMENTAL PROCEDURES
Materials-Dioleoylphosphatidylcholine (DOPC) and dioleoylphosphatidic acid (DOPA) were obtained from Avanti Polar Lipids, and ␤-D-octyl glucoside and octa(ethylene glycol)n-dodecyl ether (C 12 E 8 ) were obtained from Sigma and Calbiochem, respectively. SR was prepared from rabbit skeletal muscle as described in Dalton et al. (13). Concentrations of ATPase were estimated using a specific absorption coefficient of 1. Reconstitution into Sealed Vesicles-SLN was dissolved in trifluoroethanol, and lipids were dissolved in chloroform. The solutions were mixed in the desired proportion, dried, and resuspended in buffer A (10 mM Pipes, 100 mM K 2 SO 4 , pH 7.1) containing 40 mM ␤-D-octyl glucoside to give a final lipid concentration of 7.5 mM. The sample was then sonicated to optical clarity in a bath sonicator. SR was solubilized in buffer A containing C 12 E 8 (6 mg/ml) and 0.1 mM CaCl 2 . The solubilized SR was mixed with the lipid sample to give a 5000:1 molar ratio of lipid:ATPase. Detergent was removed by the addition of 80 mg of washed SM2 Bio-Beads (mesh size 20 -50) followed after 1 h by a second addition of 80 mg of Bio-Beads. After a further hour the sample of reconstituted vesicles was removed from the Bio-Beads and kept on ice until use.
Assay of Ca 2ϩ Uptake-Accumulation of Ca 2ϩ by the reconstituted vesicles was measured at 35°C using Antipyrylazo III to monitor the external Ca 2ϩ concentration (13). The absorption difference 720 -790 nm was recorded using an SLM Aminco dual wavelength spectrophotometer. The assay buffer used was 10 mM Pipes, 100 mM K 2 SO 4 , 5 mM MgSO 4 , pH 7.1, containing 70 M Antipyrylazo III with a protein concentration of 0.026 mg/ml. Carbonyl cyanide p-(trifluoromethoxy)phenylhydrazone (FCCP) was added to a concentration of 0.25 M to make the vesicles permeable to H ϩ . The system was calibrated by the incremental addition of Ca 2ϩ before the addition of ATP to initiate uptake.
Assays of ATPase Activity-ATPase activities were determined at 35°C over a 2-min period using a coupled enzyme assay in a medium containing 40 mM Hepes, pH 7.2, 100 mM KCl, 5 mM MgSO 4 , 2.1 mM ATP, 1.1 mM EGTA, 0.41 mM phosphoenolpyruvate, 0.15 mM NADH, 7.5 units of pyruvate kinase, and 18 units of lactate dehydrogenase with 10 g of ATPase. The reaction was initiated by the addition of an aliquot of a 25 mM CaCl 2 solution to a cuvette containing the ATPase and the other reagents to give the required concentration of free Ca 2ϩ , typically 80 M. The same coupled enzyme assay system was used to measure activities over a 30-min time period but with 2 g of ATPase.
ATPase activities were also measured using a phosphate release assay using the Biomol green phosphate assay reagent from BioMol Research labs. Activities were measured in medium containing 40 mM Hepes, pH 7.2, 100 mM KCl, 5 mM MgSO 4 , 2.1 mM ATP, 1 mM EGTA, and 0.86 mM Ca 2ϩ , corresponding to a concentration of free Ca 2ϩ of 80 M. Assays were started by the addition of 5 g of protein in a total volume of 2.5 ml. At the chosen time intervals, 20-l samples were taken, added to 100 l of 100 mM EGTA to stop the reaction, and stored on ice for up to 60 min. Phosphate levels were then determined by mixing with the Biomol reagent, and after a 30-min incubation at 25°C, the absorbance at 620 nm was read. Free concentrations of Ca 2ϩ were calculated by using the binding constants of Ca 2ϩ , Mg 2ϩ , and H ϩ for EGTA given by Godt (16).
Heats of Reaction-Heats of reaction were measured using an OMEGA isothermal titration calorimeter from Microcal Inc. using a procedure similar to that used by de Meis (4). The 1.5-ml calorimeter cell contained the reaction medium (40 mM Hepes, 100 mM KCl, 2.1 mM ATP, 5 mM MgCl 2 , 1 mM EGTA, 0.86 mM Ca 2ϩ , pH 7.1, and a free Ca 2ϩ concentration of 80 M), and the reference cell contained water. After equilibration at 35°C for 20 min, the reaction was started by injecting vesicles (5 g protein) into the reaction cell, and the heat change was recorded for 30 min. As described by de Meis (4), the heat change observed in the first 2 min of reaction was discarded since this includes a variety of artifacts, including heats of sample dilution. It was confirmed that vesicles were fully active after equilibration at 35°C under the conditions used for the calorimetric measurements.

RESULTS
ATPase Activities of Sealed Vesicles-Vesicles of sarcoplasmic reticulum were prepared as described and contained typically 75% protein as Ca 2ϩ -ATPase, the remainder being largely the lumenal protein calsequestrin (7). The Ca 2ϩ -ATPase was reconstituted into sealed vesicles containing SLN by mixing lipid and protein in detergent solution followed by removal of detergent with Bio-Beads (7). It was reported previously that levels of accumulation of Ca 2ϩ by vesicles containing Ca 2ϩ -ATPase were relatively low if the vesicles contained DOPC as the only lipid but increased considerably of the vesicles contained 10 mol % of an ionic lipid (13). Because the concentration of anionic lipid in the native SR membrane is also ϳ10 mol % (17), experiments were performed with vesicles containing 10 mol % of the anionic phospholipid DOPA.
The rate of ATP hydrolysis was measured over 3 min using a coupled enzyme assay at saturating concentrations of Ca 2ϩ . As shown in Table 1, activities of sealed vesicles approximately doubled on the addition of the detergent octa(ethylene glycol)n-dodecyl ether to make the vesicles leaky to ATP. As described previously, this can be attributed to a close to random insertion of the Ca 2ϩ -ATPase into the sealed vesicles, so that approximately half the ATPase molecules will be in the wrong orientation in the membrane to bind to ATP from the external SCHEME 1 medium unless detergent is added to make the vesicles leaky to ATP (13,18). Small variations in the sidedness of insertion of the Ca 2ϩ -ATPase into the vesicles probably explain the variation in activity observed between different preparations of sealed vesicles. Studies with unsealed membrane preparations suggest that SLN has no effect on ATPase activity measured at saturating concentrations of Ca 2ϩ (7).
ATPase activities were also measured for sealed vesicles using a phosphate release assay under the conditions used for the heat output measurements. ATPase activities were found to be linear over 60 min both in the absence and presence of SLN (Fig. 1). The measured rates were very similar to those measured using the coupled enzyme assay over a 30-min period (Fig.  1, Table 1). Using the phosphate release assay, the concentration of ADP in the medium increased as a function of time as a result of ATP hydrolysis, whereas in the coupled enzyme assay ADP was reconverted to ATP. Thus, the fact that similar rates were observed in the two assays shows that the level of ADP generated as a result of ATPase activity in the absence of a regenerating system is not sufficient to affect the rate of ATP hydrolysis.
Effects of SLN on Accumulation of Ca 2ϩ -As reported previously at 25°C (7), the presence of SLN at 35°C leads to a decreased level of accumulation of Ca 2ϩ with the effect increasing with increasing molar ratios of SLN to ATPase (Fig. 2). Similar effects are seen for vesicles containing DOPC as the only phospholipid (data not shown). In these experiments FCCP was present to make the vesicles permeable to H ϩ ; in the absence of FCCP the presence of SLN again reduced the level of accumulation of Ca 2ϩ , but all the levels of accumulation of Ca 2ϩ were lower than in the presence of FCCP (7).
Heat Production by Ca 2ϩ -ATPase-The rate of heat release resulting from hydrolysis of ATP was measured for the Ca 2ϩ -ATPase reconstituted into sealed vesicles containing 10 mol % DOPA as a function of the molar ratio of SLN to ATPase (Fig.  3). In all cases the heat released increased linearly with time, and the heat released increased with increasing SLN content. The measured heat released was found to be the same within 10% in buffer containing 0.86 mM Ca 2ϩ and 1 mM EGTA, corresponding to a free Ca 2ϩ concentration of 80 M, and in buffer containing 30 or 100 M Ca 2ϩ in the absence of EGTA. Table 2 lists the ATPase activities of the reconstituted samples whose heat releases are shown in Fig. 3 together with the calculated heat released per mol of ATP hydrolyzed. The measurements of ATPase activity and heat produced were both performed in the absence of FCCP. Measurements were also made in the presence of FCCP to make the vesicles permeable to H ϩ . It was found that the addition of FCCP resulted in changes in ATPase activity and heat produced of less than 10% (data not shown), showing that the observed effects of SLN could not be attributed to an effect on the permeability of the vesicles to H ϩ . The heat released from reconstituted vesicles containing DOPC as the only phospholipid also increased linearly with time, giving calculated heats released per mol ATP hydrolyzed comparable with those observed for vesicles containing 10 mol % of DOPA (Table 2).

DISCUSSION
Of the 31 amino acid residues in SLN, about 22 are required to form a transmembrane ␣-helix (19). The structure of the transmembrane region of SLN is very similar to that of phospholamban (PLN), and SLN and PLN appear to bind to the same site on the Ca 2ϩ -ATPase, located in a groove between transmembrane ␣-helices M2, M4, and M6 (19,20). Binding of The assay conditions are given under "Experimental Procedures." Coupled enzyme assays of ATP hydrolysis were performed over either a 3-min time period or over a 30-min time period. ATPase activity was also determined using a phosphate release assay over 60 min. Where added, the concentration of C 12 E 8 was 0.8 mg/ml. The vesicles contained a 1:9 molar ratio of DOPA to DOPC and a molar ratio of total phospholipid:ATPase of 5000:1. The   Table 1. DECEMBER 1, 2006 • VOLUME 281 • NUMBER 48 the transmembrane domain of PLN to the Ca 2ϩ -ATPase results in a reduction in the apparent affinity of the Ca 2ϩ -ATPase for Ca 2ϩ (21,22). The presence of SLN has also been reported under some conditions to result in a modest decrease in the apparent affinity of the Ca 2ϩ -ATPase for Ca 2ϩ (15,19,23), although under other conditions no significant effect on apparent affinity was observed (7,15). A major difference between SLN and PLN concerns the ways in which their interaction with the Ca 2ϩ -ATPase can be modulated. Phosphorylation of PLN by protein kinases leads to a reversal of the effect of PLN on the Ca 2ϩ -ATPase, therefore, linking Ca 2ϩ -ATPase function to adrenergic activation (24). This kind of control is not possible for SLN because SLN cannot be phosphorylated (25). In muscles such as the atria of small mammals, where SLN is found together with PLN (20), com-plex interactions between SLN, PLN, and the Ca 2ϩ -ATPase are possible because SLN and PLN can form heterodimers that interact with the Ca 2ϩ -ATPase, thus linking the effects of SLN on the Ca 2ϩ -ATPase to phosphorylation of PLN (20). However, such interactions are clearly not possible in fast-twitch skeletal muscle that lacks PLN (26), and the role of SLN in such muscles is not clear.

Sarcolipin and Thermogenesis
The possibility explored here is that the presence of SLN in fast-twitch skeletal muscle of large mammals is related to the importance of SR in heat generation during nonshivering thermogenesis in animals lacking brown adipose tissue. de Meis et al. (3)(4)(5)(6) has shown that ATP is hydrolyzed by SR Ca 2ϩ -ATPase both to transport Ca 2ϩ and to generate heat (3)(4)(5)(6). In the process of transport, two Ca 2ϩ ions are pumped from the cytoplasm to the lumen of the SR for each molecule of ATP hydrolyzed (Scheme 1), and the expected ratio of two Ca 2ϩ ions accumulated per ATP molecule hydrolyzed was observed during the first reaction cycle of the ATPase before the lumenal concentration of Ca 2ϩ had increased to a high level (27). However, when the lumenal concentration of Ca 2ϩ reached millimolar concentrations, the level of Ca 2ϩ accumulated became less than 2:1 with respect to ATP hydrolysis (8, 28) because high concentrations of lumenal Ca 2ϩ led to a build up of E2PCa 2 , and E2PCa 2 can dephosphorylate and release its bound Ca 2ϩ ions back to the cytoplasmic side of the membrane in a process of slippage (step 11 in Scheme 1). The process of slippage generates maximum heat from ATP hydrolysis since it corresponds to ATP hydrolysis without performing work. Thus, switching between transport and slippage by the Ca 2ϩ -ATPase is "controlled" by the state of the system itself. When the concentration of Ca 2ϩ in the lumen of the SR is low, the level of E2PCa 2 will be low, and there will be little slippage; ATP will be used mostly for transporting Ca 2ϩ into the SR lumen. The concentration of Ca 2ϩ in the SR lumen will build up as a result of this transport, leading to a build up of E2PCa 2 and, thus, of slippage and heat production. What this means physiologically is as follows. After the release of lumenal Ca 2ϩ to cause muscle contraction, the level of lumenal Ca 2ϩ will be low, and the Ca 2ϩ -   Table 2. Table 2 also lists the ATPase activities for the samples used in these experiments.

TABLE 2 Heat release by reconstituted vesicles
Heat release from reconstituted vesicles was measured at 2.1 mM ATP and a free Ca 2ϩ concentration of 80 M in buffer (40 mM Hepes, 100 mM KCl, 5 mM MgCl 2 ) at 35°C as described in the legend to Fig. 3. ATPase activities for the vesicles used for the heat release measurements were determined using a coupled enzyme assay. Vesicles contained either a 1:9 molar ratio of DOPA to DOPC or DOPC alone at a molar ratio of total phospholipid:ATPase of 5000:1 and the given molar ratios of SLN:ATPase. ATPase will be used mainly to pump Ca 2ϩ back into the lumen, lowering the cytoplasmic concentration of Ca 2ϩ and leading to muscle relaxation. After the Ca 2ϩ has been pumped back into the lumen, the lumenal Ca 2ϩ concentration will be high, and now the Ca 2ϩ -ATPase can be used for heat production since it is no longer required for muscle relaxation. Thus, switching of the Ca 2ϩ -ATPase between pumping and heat production is controlled by the level of Ca 2ϩ in the lumen of the SR. Any factor that increases the rate of slippage of the Ca 2ϩ -ATPase should increase heat generation by the Ca 2ϩ -ATPase. In previous studies it was shown that incorporation of SLN into reconstituted vesicles of the Ca 2ϩ -ATPase led to reduced levels of accumulation of Ca 2ϩ at 25°C, the effects of SLN being consistent with an increase in the rate of slippage of the Ca 2ϩ -ATPase (7). Here we show that the presence of SLN also reduces the level of accumulation of Ca 2ϩ at 35°C (Fig. 2). As shown in Fig. 1, ATPase activities in the absence or presence of SLN are constant for up to 60 min, suggesting that concentrations of Ca 2ϩ within the vesicle do not reach values high enough to significantly inhibit the Ca 2ϩ -ATPase, since otherwise activities would decrease with increasing time. Activities measured in the presence and absence of an ATP-regenerating system are the same (Fig. 1, Table 1), showing that under the conditions of these experiments the concentrations of ADP generated in the absence of a regenerating system are not sufficient to inhibit ATP hydrolysis.
Isothermal calorimetry was used to measure directly the heat generated by the Ca 2ϩ -ATPase as a result of ATP hydrolysis. As shown in Fig. 3 and Table 2, the presence of SLN did indeed result in increased heat production, the effect increasing with increasing SLN content. From the Ca 2ϩ uptake data shown in Fig. 2, it can be estimated that the external free concentration of Ca 2ϩ is greater than 10 M at the end of the experiment, this being sufficiently high to inhibit passive leak of Ca 2ϩ from the vesicles (10,11) and to inhibit the back reaction of ADP with the phosphorylated ATPase (steps 4 to 2 in Scheme 1) (4). Thus, the heat produced per mol of ATP hydrolyzed can be calculated directly from the heat produced/mg of protein/min and the rate of ATP hydrolysis/mg of protein/min, giving the values listed in Table 2. A plot of heat produced per mol of ATP hydrolyzed against concentration of SLN, expressed as mole fraction per lipid monolayer to account for the fact that the lipids are arranged as a bilayer, fits to a simple binding equation, giving a value for the dissociation constant for the SLN⅐ATPase complex of 6.9 ϫ 10 Ϫ4 Ϯ 2.9 ϫ 10 Ϫ4 in units of mole faction (Fig. 4). The dissociation constant can also be expressed as 1.7 Ϯ 0.7 in units of molar ratio of SLN to ATPase at a molar ratio of 5000 lipids:ATPase, but in molar ratio units the dissociation constant will vary with the molar ratio of lipid:protein in the membrane.
Although high molar ratios of SLN to ATPase are required for maximal effects of SLN in the reconstituted system (Fig. 4), this is simply the result of the high lipid content of the reconstituted system. The molar ratio of SLN:ATPase in rabbit extensor digitorum longus muscle (fast-twitch skeletal muscle) has been estimated by Vangheluwe et al. (14) to be ϳ0.4:1. The molar ratio of lipid:ATPase in skeletal muscle SR is ϳ90:1 (29), giving a mole fraction of SLN per lipid monolayer in the SR of ϳ0.01, so that with a dissociation constant of 6.9 ϫ 10 Ϫ4 mole fraction for the SLN⅐ATPase complex, most of the SLN present in the native membrane will be bound to Ca 2ϩ -ATPase.
It is not yet possible to explain the effect of SLN on slippage in molecular terms. The crystal structure of the Ca 2ϩ -ATPase shows the two Ca 2ϩ ions to be transported bound to two sites in the transmembrane region of the ATPase (30,31). The structure of the protein around these two sites will determine the relative rates at which Ca 2ϩ ions are released to the two sides of the membrane after phosphorylation of the ATPase. Mutation of Tyr-763 to Gly or Lys-758 to Ile, two residues located "above" the Ca 2ϩ binding sites on the cytoplasmic side of the membrane, results in fully uncoupled mutants (32) presumably after the slippage pathway in Scheme 1 (33). Thus, the relative rates of slippage and transport could be affected by other factors affecting the packing of the ATPase around the Ca 2ϩ binding sites, including the binding of SLN between transmembrane helices M2, M4, and M6.