To study PLB (phospholamban) inhibition of the cardiac Ca2+ pump [SERCA2a (sarcoplasmic/endoplasmic reticulum Ca2+-ATPase 2a)], a fusion protein (SER-20G-PLB) was engineered by tethering SERCA2a with PLB through a 20-glycine residue chain, allowing the PLB tether to either bind to or dissociate from the inhibition site on SERCA2a. When expressed in insect cells, SER-20G-PLB produced active Ca2+ uptake, which was stimulated by the anti-PLB antibody, both similar to that which occurred with the control sample co-expressing WT (wild-type)-SERCA2a and WT-PLB. The KCa values of Ca2+-dependent ATPase were similar for SER-20G-PLB (0.29±0.02 μM) and for the control sample (0.30±0.02 μM), both greater than 0.17±0.01 μM for WT-SERCA2a expressed alone. Thus SER-20G-PLB retains a fully active Ca2+ pump, but its apparent Ca2+ affinity was decreased intrinsically by tethered PLB at a 1:1 molar stoichiometry. Like WT-PLB, SER-20G-PLB ran as both monomers and homo-pentamers on SDS/PAGE. As Ca2+ concentrations increase from 0 to the micromolar range, the proportion of non-inhibiting pentamers increased from 32% to 52%, suggesting that Ca2+ activation of the pump completely dissociates the PLB tether from the inhibition site on SERCA2a, with concurrent association of PLB pentamers. Collectively, the regulation of SERCA2a is achieved through the Ca2+-dependent equilibria involving PLB association and dissociation from SERCA2a, and assembling and disassembling of SER-20G-PLB pentamers.
- cardiac Ca2+-ATPase
- enzyme regulation
- protein–protein interaction
The cardiac muscle Ca2+-ATPase [SERCA2a (sarcoplasmic/endoplasmic reticulum Ca2+-ATPase 2a) isoform] is a 100 kDa membrane protein that pumps two Ca2+ ions into the lumen of the SR (sarcoplasmic reticulum) per ATP molecule hydrolysed [1,2]. SERCA2a is naturally regulated by PLB (phospholamban), a 52 amino acid SR membrane protein that provides key modulation of myocardial contraction [3,4]. PLB inhibits SERCA2a, decreasing the apparent affinity of the pump for Ca2+. Phosphorylation of PLB by PKA (protein kinase A) at Ser16 or by calmodulin kinase II at Thr17 during the β-adrenergic stimulation of the heart reverses PLB inhibition. This inhibition can also be fully reversed by the monoclonal anti-PLB antibody 2D12, which decreases the binding of PLB to SERCA2a [5,6]. Owing to its significant contribution to the Ca2+ kinetics, the PLB/SERCA2a system remains an important target for understanding cardiac function in physiological and pathophysiological conditions, and for treatment of cardiac diseases .
PLB exists at an equilibrium between monomers and homo-pentamers in SR membranes [8,9]. Disturbing this equilibrium, e.g. by introducing mutations (L37A) to destabilize PLB pentamer formation, causes enhanced SERCA2a inhibition [10,11]. Thus it is suggested that PLB monomers are responsible for binding to SERCA2a, inhibiting the enzyme, whereas PLB pentamers are inactive species [10–12]. Interestingly, several other mutations of PLB, e.g. N27A or L31A, increase or decrease PLB inhibitory function respectively, with little effect on pentamer stability [11,13]. Two recent studies suggested that pentamers of PLB might play an active role in facilitating PLB monomer binding to SERCA2a [14,15].
It is controversial whether Ca2+ binding to SERCA2a dissociates the PLB monomer from the enzyme inhibition site [16–20]. One model suggests that binding of PLB and Ca2+ to SERCA2a are mutually exclusive: PLB binds to the Ca2+-free, E2 conformation of SERCA2a, inhibiting the pump, whereas Ca2+ binding to the high Ca2+ affinity, E1 conformation of SERCA2a completely dissociates PLB, activating the enzyme [16,20,21]. This model was developed based on measurement of cross-linking of PLB to SERCA2a at several sites in both cytoplasmic and transmembrane PLB domains with use of cross-linking agents of variable length. The cross-linking between PLB and SERCA2a at all of these sites was achieved only in the Ca2+-free condition, but was completely inhibited by Ca2+ at micromolar concentrations or by the SERCA2a inhibitor TG (thapsigargin) [16,19,20,22,23]. On the other hand, fluorescence labels on PLB revealed an equilibrium between pentamers and monomers of PLB , but reported association of PLB with SERCA2a in the presence of Ca2+ [17,18], although rapid exchange between PLB monomers and SERCA2a was noted . Thus an alternative model was proposed suggesting that a PLB monomer remains associated with SERCA2a during the entire kinetic cycle and that conformational changes of PLB (e.g. phosphorylation of PLB) control SERCA2a activation or inactivation [17,18,25,26]. Both models rely on results reporting local interactions by fluorescence probes at specific sites [17,18] or cross-linking of mutated residues between the two proteins [16,19,20,22,23]. It is not clear whether the intact WT (wild-type)-PLB molecule dissociates from its inhibition site on SERCA2a when the enzyme binds Ca2+.
The stoichiometry of PLB and SERCA2a is important for proper PLB inhibition . Abnormal PLB expression in hearts, e.g. caused by a naturally occurring L39Stop truncation of PLB in patients, has been directly linked to human heart diseases [27,28]. Although several studies suggest that full PLB inhibition in the heart requires a molar expression ratio of PLB to SERCA2a to be greater than 1 [29–31], the ratio at 1:1 appears to be sufficient for SERCA2a regulation [32,33]. Nonetheless, PLB and SERCA2a can be co-expressed at a greater than 1:1 molar ratio to provide mechanistic insights into their interactions [10–12,16–18,20–22,26].
A new experimental system is needed to provide a more precise control of the molar expression ratio; one that also promotes functional efficient coupling between the two proteins. The fusion protein system has been widely used to study the molecular interaction between two proteins . Since the C-terminus of SERCA2a and the N-terminus of PLB are both on the cytoplasmic side of the SR membrane, construction of SERCA2a–PLB fusion proteins is structurally feasible. The present paper reports the development and initial characterization of a fully functional, catalytically regulatable SERCA2a–PLB fusion protein.
MATERIALS AND METHODS
SER-5G-PLB and SER-20G-PLB are two fusion proteins containing canine SERCA2a, followed by a five glycine residue or 20 glycine residue linker chain respectively, and canine PLB. To make fusion protein constructs in the pVl1393 vector, two PCR inserts were first created. The PCR constructs contained either 15 or 60 bp encoding five or 20 glycine residues respectively, the entire coding region of PLB, and the stop codon, with a cohesive BglII site at both ends. Both PCR constructs were then individually inserted into the C-terminus of SERCA2a in the pVl1393 vector, in which the stop codon of SERCA2a was replaced by a BglII site. Mutations of PLB, including N27A, L31A, L37A and L39Stop, were introduced to SER-20G-PLB using the QuikChange® XL-Gold system (Stratagene). All fusion DNA vectors were individually co-transfected with BaculoGold™ (Pharmingen) linearized-baculovirus DNA into Sf21 insect cells to produce baculoviruses encoding each fusion protein. All constructs were confirmed by DNA sequencing.
Protein expression and detection
All fusion proteins were expressed individually in Sf21 insect cells as described previously [10,16,20,21,35]. Control samples were insect cell microsomes with co-expression of WT-SERCA2a and PLB. The microsomes of Sf21 cells were harvested after 60 h of infection and stored in sucrose in small aliquots at −40°C. SDS/PAGE (5% gels) was used to separate pentamers and monomers of fusion proteins. After a 3 h transfer on to nitrocellulose membranes, the proteins on the blots were routinely probed with the anti-PLB monoclonal antibody 2D12, or the anti-SERCA2a monoclonal antibody 2A7-A1. Antibody-binding protein bands were visualized by 125I-labelled Protein A and autoradiography, and quantified with a Bio-Rad Personal Fx phosphorimager.
Ca2+ transport and ATPase assays
Insect cell microsomes expressing individual fusion proteins, or control sample of WT-SERCA2a with co-expression of PLB, were incubated in a buffer containing 50 mM Mops (pH 7.0), 3 mM MgCl2, 100 mM KCl, 5 mM sodium azide and 3 mM ATP. For the active Ca2+ transport assays, 10 mM potassium oxalate and 0.1 mM CaCl2/1 mM EGTA containing trace amounts of 45Ca were included in the buffer. 45Ca uptake by the 11 μg of microsomes, in the absence or presence of 5.5 μg of the affinity-purified anti-PLB monoclonal antibody 2D12, was retained by Millipore glass filters and counted . Ca2+-dependent ATPase activities of 5 μg of membranes was determined by an enzyme-coupled continuous assay in the presence of 3 μg/ml of the Ca2+ ionophore A23187. ATP hydrolysis was determined by measuring the rate of NADH decay of samples at 340 nm at 37°C in variable CaCl2 (0–1 mM) and 1 mM EGTA [21,35]. In some experiments, the Ca2+-ATPase activity assays were performed after pre-incubation of the microsomes with 2.5 μg of the affinity-purified anti-PLB monoclonal antibody 2D12 for 20 min on ice, or with prior phosphorylation of PLB at Ser16 by PKA (catalytic subunit of PKA; Sigma) [6,21,35].
Construction and expression of SER-5G-PLB and SER-20G-PLB fusion proteins in insect cell microsomes
We made two SERCA2a–PLB fusion constructs in which the N-terminus of PLB (residue Met1) was tethered to the C-terminus of SERCA2a (residue Glu997) through a five glycine residue linker chain (SER-5G-PLB) or 20 glycine residue linker chain (SER-20G-PLB). SER-5G-PLB has a short linker chain so that PLB is kept in close contact with SERCA2a at all times. Alternatively, SER-20G-PLB employs a relatively long linker chain that should give the PLB tether a greater freedom, which allows it either to bind to SERCA2a at the binding site or to move away from the enzyme inhibitory site, permitting enzyme activation. SER-5G-PLB or SER-20G-PLB, together with the control sample of insect cell microsomes co-expressing WT-SERCA2a and WT-PLB, was expressed in insect cells and analysed in duplicate with immunoblotting. As shown in Figure 1(A), for samples expressing SER-5G-PLB (lanes 1 and 5) and SER-20G-PLB (lanes 3 and 7), both the monoclonal anti-SERCA2a antibody 2A7-A1 and the monoclonal anti-PLB antibody 2D12 detected two identical bands at the same positions, demonstrating that SERCA2a and PLB were expressed in the same fusion proteins. The lower 110 kDa bands were the monomeric form of fusion proteins, indicated by a PLB tether inducing an upward mobility shift from that of WT-SERCA2a (lane 9). The upper 550 kDa bands were five times the molecular mass of one fusion protein molecule. When the samples were boiled prior to SDS/PAGE, all 550 kDa bands disappeared and dissociated into the lower 110 kDa bands (lanes 2, 4, 6 and 8). This behaviour in response to boiling in SDS is a classic feature of WT-PLB, where PLB pentamers dissociate into monomers (lane 10) . Thus the 550 kDa bands were probably pentamers of fusion proteins, formed by leucine zippers of the PLB tether  which remained stable in SDS, in a similar manner to free WT-PLB which ran at 5 kDa and 25 kDa (lane 9). These results suggest that expression of SER-5G-PLB or SER-20G-PLB in insect cells was biochemically similar to that of control samples co-expressing WT-SERCA2a and WT-PLB. Furthermore, even when covalently linked to SERCA2a, the PLB tether, identical with the free diffusing WT-PLB, retained its ability to form pentamers.
Enzyme activity of fusion proteins
A series of experiments was carried out to determine whether fusion proteins expressed in insect cells maintained enzymatic activity and, importantly, whether the PLB tether remained able to regulate the pump. The 45Ca-uptake assay revealed that SER-5G-PLB cannot transport Ca2+ in the absence and presence of the antibody 2D12 (Figure 1B, open and closed circles). In contrast, in a very similar fashion to 45Ca accumulation brought about by co-expression of WT-SERCA2a and WT-PLB, SER-20G-PLB increased 45Ca accumulation inside membrane vesicles (Figure 1B, compare open squares with open triangles), indicating that SER-20G-PLB actively transported Ca2+. Furthermore, as occurred with the control sample (Figure 1B, closed triangles), the antibody 2D12 also created an approximately 2-fold stimulation of 45Ca uptake by SER-20G-PLB (Figure 1B, closed squares). Therefore, compared with co-expression of WT-SERCA2a and WT-PLB, expression of SER-20G-PLB produced similar Ca2+-uptake profiles, suggesting that SER-20G-PLB possessed an equal efficiency in generating and regulating active Ca2+ transport. Antibody 2D12 binding to the PLB tether augmented Ca2+ transport by SER-20G-PLB, strongly suggesting that the PLB tether was fully capable of inhibiting SERCA2a.
Although SER-5G-PLB did not have any Ca2+-dependent ATPase activity (results not shown), for samples expressing SER-20G-PLB, ATP hydrolysis occurred in a similar Ca2+-dependent manner (Figure 1C, compare open squares with open triangles), confirming that the enzyme is active. Compared with the curve for WT-SERCA2a expressed alone (Figure 1C, broken line), SER-20G-PLB shifted the Ca2+-dependent ATPase curve to the right, demonstrating that the apparent Ca2+-binding affinity of SER-20G-PLB was decreased. The Ca2+ concentration of half-maximal activation (KCa value) for SER-20G-PLB was 0.29 μM, which was similar to the KCa value of 0.28 μM for the control sample, both larger than the KCa value of 0.17 μM for WT-SERCA2a expressed alone (Table 1). Moreover, phosphorylation of the PLB tether by PKA shifted the curve to the left (Figure 1C, half solid squares), partially reversing enzyme inhibition, yielding a KCa value of 0.22 μM, compared with a KCa value of 0.20 μM for the phosphorylation of the control sample (half solid triangles). Furthermore, incubation of antibody 2D12 with SER-20G-PLB shifted the curve to the left (solid squares), nearly completely reversing enzyme inhibition, yielding a KCa value of 0.17 μM, again similar to the KCa value of 0.17 μM for the control sample (solid triangles). Both phosphorylation of PLB and addition of 2D12 had little effect on the maximal enzyme velocity at Ca2+ concentrations in the micromolar range. Finally, the Ca2+-ATPase activity at the 10 μM Ca2+ concentration of the control sample was normalized to WT-SERCA2a protein expression (measured by the intensity of the antibody 2A7-A1 signal on the immunoblot), and arbitrarily defined as 1. This ratio was 0.92±0.13 (n=5) for the sample expressing SER-20G-PLB. Thus, compared with co-expression of WT-SERCA2a and WT-PLB, these results demonstrated that SER-20G-PLB maintained full catalytic activity, and was adequately inhibited by the covalently attached PLB, suggesting that the interactions between the SERCA2a anchor and its PLB tether are functionally intact.
Effect of mutations of PLB on SER-20G-PLB enzyme activity
PLB mutational analysis was used to examine the specificity of the PLB tether to inhibit SERCA2a. Studies using a co-expression system revealed multiple mutations of PLB that affected its inhibitory function [10,11,13,27]. Several of these mutations, including N27A, L31A, L37A and L39Stop, were introduced to the PLB tether of SER-20G-PLB and the mutants were individually expressed in insect cell microsomes. As expected, mutants SER-20G-N27A-PLB and SER-20G-L31A-PLB maintained the ability to form pentamers and monomers, whereas mutants SER-20G-L37A-PLB and SER-20G-L39Stop-PLB existed predominantly as monomers in SDS (Figure 1D). In addition, the shifts of the Ca2+-dependent ATPase activity curves revealed that the PLB tether with the L31A or L39Stop mutation lost inhibitory function, whereas the PLB tether with the N27A or L37A mutation enhanced SERCA2a inhibition (Figure 1E). Furthermore, KCa values of the Ca2+- dependent ATPase activity for these mutants, measured in controls, with prior phosphorylation of PLB by PKA, or in the presence of the anti-PLB antibody 2D12, were similar to that obtained for the co-expression of their counterpart PLB mutants with WT-SERCA2a (Table 1). All of these results using mutations of the PLB tether confirmed the previous findings obtained in the co-expression system, further supporting the idea that the Ca2+-dependent ATPase activity of SER-20G-PLB is inhibited specifically by its PLB tether.
Ca2+-dependent distribution of monomers and pentamers of SER-20G-PLB
The next set of experiments examined whether Ca2+ affected the distribution of pentamers and monomers of SER-20G-PLB. As shown in the immunoblot, addition of Ca2+ or TG to samples pre-incubated in EGTA increased the amount of pentamers and decreased monomers of SER-20G-PLB (Figure 2A, left-hand panel). The average percentage of pentamers in SER-20G-PLB was significantly increased from 32±2.0% in the Ca2+-free conditions to 54±2% or 52±2% after addition of Ca2+ or TG respectively (Figure 2B). In control experiments using samples expressing SER-5G-PLB, addition of Ca2+ or TG did not change the distribution of pentamers and monomers established in EGTA (Figures 2A and 2B, right-hand panels). In another control experiment, incubation in EGTA, Ca2+ or TG had no effect on the distribution of pentamers and monomers of WT-PLB expressed alone (results not shown). SDS is known to preserve the pentameric structure of PLB [35,36]. Even so, after dissociation from the SERCA2a-binding site, the monomeric PLB tethers did not form pentamers in SDS, as indicated by the fact that boiled samples did not have any pentamers of fusion proteins. Therefore these additional pentamers must be formed in the membrane by the PLB tether monomers just dissociated from the inhibitory site on SERCA2a after addition of Ca2+ or TG (see the Discussion).
Correlation between Ca2+ activation of ATPase and pentamers of fusion proteins
The Ca2+-dependency of enzyme activity and pentamer formation for SER-20G-PLB was studied in the steady-state at an identical series of Ca2+ concentrations at 37°C. As concentrations of free Ca2+ increased from 0 to micromolar ranges, the immunoblot analysis showed that there was an increase in pentamers of SER-20G-PLB, coupled with a decrease in monomers (Figure 2C, left-hand panel). Compared with the Ca2+-dependent ATPase activity, the curve of the percentage of pentamers of SER-20G-PLB exhibited a similar sigmoidal shape in a Ca2+-dependent manner (Figure 2D). The minimum and maximum pentamer formation occurred at the Ca2+-free and micromolar Ca2+ concentrations respectively, as did the Ca2+-ATPase activity. In addition, the KCa value for pentamer formation (Ca2+ concentration for a half-maximal increase of the percentage of pentamers) was 0.20±0.05 μM, in agreement with the KCa value of 0.29±0.02 μM for the Ca2+-ATPase activity. Thus Ca2+ activation of SERCA2a correlated well with an increase in pentamers of SER-20G-PLB, suggesting that SER-20G-PLB pentamers represent active enzymes.
Distribution of monomers and pentamers of mutants SER-20G-N27A-PLB and SER-20G-L31A-PLB were also examined for their Ca2+-dependency. SER-20G-L31A-PLB did not significantly affect SERCA2a activity (Figure 1E), nor did it exhibit any significant Ca2+-dependent changes in distribution of monomers and pentamers (Figure 2C). The percentage of pentamers of SER-20G-L31A-PLB remained at approximately 60% at all Ca2+ concentrations tested (Figure 2E, circles). Alternatively, the distribution of monomers and pentamers of SER-20G-N27A-PLB was Ca2+-dependent. Pentamers of the mutant increased from 25% in the Ca2+-free conditions to 49% at high micromolar Ca2+ concentrations. Furthermore, compared with that for SER-20G-PLB, SER-20G-N27A-PLB shifted the curve of Ca2+-dependent pentamer formation to the right, increasing the KCa value to 0.85±0.10 μM (Figure 2E, triangles). This rightward shift also connected well to the rightward shift which occurred in the Ca2+-dependent ATPase activity for SER-20G-N27A-PLB. Therefore, at the same Ca2+ concentration, there were fewer pentamers and a smaller Ca2+-ATPase activity for SER-20G-N27A-PLB than that for SER-20G-PLB, suggesting that increased SERCA2a inhibition by the N27A mutation of PLB corresponded to a decreased amount of pentamers.
In the present study, a SERCA2a–PLB fusion protein SER-20G-PLB was generated and characterized to investigate interactions between PLB and SERCA2a. When expressed in insect cell microsomes and compared with the control sample with co-expression of WT-SERCA2a and WT-PLB, SER-20G-PLB is not only fully functional, as demonstrated by its ability to hydrolyse ATP and transport Ca2+, but also intrinsically possesses the hallmark of PLB regulation, including lowered apparent affinity of the enzyme for Ca2+ and reversal of SERCA2a inhibition by PKA phosphorylation of the PLB tether or by the anti-PLB antibody 2D12. Furthermore, compared with that in the co-expression system, several well-characterized mutations of PLB, including the loss-of-function L31A or L39Stop mutations, and the gain-of-function N27A or L37A mutations, generated similar effects on the inhibitory function of the PLB tether. Therefore as a functional unit containing a fully active pump and a well coupled, flexibly anchored PLB tether at a built-in 1:1 molar expression ratio, SER-20G-PLB provides a useful new system for studying molecular interactions between PLB and SERCA2a.
Pentamer formation of SER-20G-PLB measured on SDS/PAGE is Ca2+-dependent. It was shown previously that PLB pentamers in SDS are positively correlated with their stability in the membrane [35,36]. Furthermore, monomers of the PLB tether did not re-form pentamers in SDS under the experimental conditions (see the boiled samples in Figure 2A). Thus the amount of pentamers of SER-20G-PLB measured on SDS/PAGE should qualitatively reflect their distribution in the membrane, which is Ca2+-dependent. Furthermore, the oligomeric state of PLB was previously shown to be altered by binding of PLB monomers to SERCA2a , but was not analysed for the Ca2+ effect. Since the distribution of pentamers and monomers of PLB itself is not affected by Ca2+, the Ca2+-dependent association and dissociation of PLB from SERCA2a must occur in the membrane, arguing against the model that PLB remains attached to SERCA2a throughout the entire kinetic cycle [17,18].
The present study established that the increased SER-20G-PLB pentamer formation and SERCA2a enzyme activation are directly associated in a similar Ca2+-dependent fashion (Figure 2). Together with a recent finding showing the strong correlation between the amount of SERCA2a cross-linked with a PLB monomer and enzyme inactivation , the results of the present study strongly suggest that PLB monomers binding to SERCA2a inactivates the enzyme, whereas Ca2+ activation of the enzyme dissociates PLB monomers [16,19–21]. In the present study, a 20 glycine residue linker chain appears to give adequate freedom for the PLB tether to reach and bind to the inhibitory binding site, thus efficiently regulating SER-20G-PLB. For example, in the presence of EGTA and ATP, as a result of the favourable binding of the PLB tether monomers to the inhibition site of the SERCA2a anchor at the Ca2+-free E2·ATP conformation, SER-20G-PLB molecules exist in minimal pentamers and maximal monomers. Addition of Ca2+ allowed the enzyme transition from the E2 state to the E1·Ca state, and concurrently increased pentamers of SER-20G-PLB. Moreover, TG, which locks SERCA2a into a dead-end state (E2·TG) that does not interact with PLB [16,20,23], also increased SER-20G-PLB pentamers to a similar level as that which occurred at the micromolar Ca2+ concentrations. Note that only monomers, but not inactive pentamers, of PLB [9–11] can fit into the binding site to inhibit SERCA2a, and that Ca2+ or TG has no direct effect on stability of PLB pentamers. Thus for the amount of pentamers to increase in both conditions, the 20 glycine residue linker chain must allow monomers of the PLB tether, even covalently anchored to SERCA2a and possessing high binding affinity for SERCA2a, first to dissociate completely from the binding site of SERCA2a in the E1·Ca or E2·TG state. Finally, in contrast with the addition of fluorescence probes at specific sites [17,18] or introducing mutations to the two proteins [16,19,20,23], pentamer formations were global assembling of intact PLB monomers. Thus the whole PLB tether, behaving like freely diffusing PLB molecules unanchored to SERCA2a, remains mobile in the membranes, favouring association with the inhibitory site of SERCA2a at the Ca2+-free conditions and complete dissociation from the binding site on SERCA2a at micromolar Ca2+ concentrations, which permitted full activation of the enzyme.
The association and dissociation of the PLB tether from SERCA2a establishes Ca2+-dependent equilibria. A simple model with only three states was proposed to describe the equilibria (Figure 3). Pentamers of SER-20G-PLB are a portion of active SERCA2a without PLB at the inhibitory site. The monomeric form of SER-20G-PLB consists of at least two populations: the PLB-‘on’ state, with PLB in the binding groove of SERCA2a; and the PLB-‘off’ state, with PLB unbound from the inhibitory site, enabled by the 20 glycine residue linker chain to make active SERCA2a. In a similar fashion to the equilibria in co-expression of free diffusing PLB and SERCA2a, SER-20G-PLB reaches two equilibria among these three populations. In particular, the first equilibrium is reached between PLB-‘off’ monomers and pentamers, which is not Ca2+-dependent. The second equilibrium is reached among PLB-‘off’ monomers, active SERCA2a (PLB-‘off’ and pentamers) and PLB-‘on’ SERCA2a, which is Ca2+-dependent.
The competition of Ca2+ and PLB binding to E1 and E2 respectively, affects the overall equilibrium in a steady-state, causing different amounts of active SERCA2a and PLB-bound inactive SERCA2a. In the Ca2+-free conditions favouring PLB binding to E2, the equilibria are shifted to the left to yield maximal PLB monomers bound to the inhibitory site of SERCA2a (PLB-‘on’). As a result, there is a minimal amount of active SERCA2a, reflected as only approximately 32% pentamers on SDS/PAGE. Increased Ca2+ concentration favours the formation of the E1•Ca conformation and subsequent PLB dissociation to yield additional PLB-‘off’ enzyme, shifting the equilibria to the right. Accordingly, monomers in the PLB-‘off’ state were increased and consequently associated into pentamers, which increased to 52% at micromolar Ca2+ concentrations on SDS/PAGE. In any steady-state after the equilibria are reached, there are two populations of pumps: active pumps corresponding to the amount of PLB-‘free’ enzyme, and inactive pumps corresponding to the PLB-‘on’ SERCA2a  (Figure 3). The distribution of these populations affects activity of the pump, reflected as PLB reduction of the apparent affinity of the pump for Ca2+.
The equilibria can also be used to explain the effect of the loss-of-function L31A and gain-of-function N27A mutations of SER-20G-PLB. Akin et al.  recently showed that gain-of-function PLB mutants have increased binding affinity for SERCA2a, thus binding and inhibiting more pumps than that for WT-PLB at a given Ca2+ concentration. Hence, at any given Ca2+ concentration, owing to increased binding affinity of the mutant N27A-PLB-tether for SERCA2a, relative to WT-PLB, SER-20G-N27A-PLB had fewer active pentamers (active SERCA2a) (Figure 2E). As a result, gain-of-function N27A-PLB-tether further lowered SERCA2a activity. Alternatively, at any given Ca2+ concentration, there were more active pentamers (Figure 2E) and higher enzyme activity (Figure 1E) for SER-20G-L31A-PLB than that occurring for SER-20G-PLB. This suggests that the loss-of-function PLB mutants (e.g. L31A) must have a decreased binding affinity for SERCA2a. Therefore these results further demonstrated that this direct competition between PLB and Ca2+ to bind to SERCA2a, reflected in the present study as Ca2+-dependent formation of pentameric active enzyme, controls the equilibria for SERCA2a regulation.
Regarding stoichiometry interactions between PLB and SERCA2a, since the KCa values of Ca2+-dependent ATPase activity were similar for SER-20G-PLB and for the co-expression of SERCA2a and WT-PLB, 1 mole of PLB tether provides sufficient inhibition of 1 mole of SERCA2a, consistent with previous findings [32,33]. Nonetheless, at a given steady-state, the presence of non-inhibiting pentamers indicates that only a portion of SERCA2a molecules have a PLB tether in the inhibitory site, yet the enzyme was sufficiently regulated. Although the PLB tether has an intrinsic higher affinity for SERCA2a than WT-PLB in the co-expression system, PLB pentamers may also play an active role in regulation of SERCA2a [14,15]. Studies using cross-linking or fluorescent probes did not reveal direct interactions between PLB pentamers and SERCA2a. However, in a two-dimensional crystal, pentamers of PLB are located at a possible non-inhibiting site of SERCA2a near M3, away from the binding groove form by M2, M4 and M9 . It is not clear whether pentamers of PLB attach to SERCA2a under physiological conditions. In the present study, the linker chain ensures that the pentamers of the PLB tether remain loosely attached to SERCA2a. In this respect, the future structural determination of SER-20G-PLB should be informative to address whether pentamers of PLB are relatively far away (or off) from SERCA2a or may bind to a non-inhibiting site (e.g. near M3 of SERCA2a). Further investigations are also needed to understand the nature of the equilibria, which might involve a fast Ca2+-dependent association and dissociation of the PLB tether from SERCA2a, and a fast assemble and disassemble of pentamers of the PLB tether.
Finally, an additional attractive feature of SER-20G-PLB may allow the testing of interactions of two species of PLB with SERCA2a stoichiometrically. Co-expression of a PLB mutant with the fusion protein can be established and determined for the molar expression ratio. These future studies might reveal new information on how the interactions between PLB and SERCA2a might be affected by a competition between two species of PLB to bind to the pump.
This work was supported by a Showalter Cardiovascular Research Grant and the Krannert Institute of Cardiology.
Abbreviations: PKA, protein kinase A; PLB, phospholamban; SERCA2a, sarcoplasmic/endoplasmic reticulum Ca2+-ATPase 2a; SR, sarcoplasmic reticulum; TG, thapsigargin; WT, wild-type
- © The Authors Journal compilation © 2011 Biochemical Society