CASQ (calsequestrin) is a Ca2+-buffering protein localized in the muscle SR (sarcoplasmic reticulum); however, it is unknown whether Ca2+ binding to CASQ2 is due to its location inside the SR rich in Ca2+ or due to its preference for Ca2+ over other ions. Therefore a major aim of the present study was to determine how CASQ2 selects Ca2+ over other metal ions by studying monomer folding and subsequent aggregation upon exposure to alkali (monovalent), alkaline earth (divalent) and transition (polyvalent) metals. We additionally investigated how CPVT (catecholaminergic polymorphic ventricular tachycardia) mutations affect CASQ2 structure and its molecular behaviour when exposed to different metal ions. Our results show that alkali and alkaline earth metals can initiate similar molecular compaction (folding), but only Ca2+ can promote CASQ2 to aggregate, suggesting that CASQ2 has a preferential binding to Ca2+ over all other metals. We additionally found that transition metals (having higher co-ordinated bonding ability than Ca2+) can also initiate folding and promote aggregation of CASQ2. These studies led us to suggest that folding and formation of higher-order structures depends on cationic properties such as co-ordinate bonding ability and ionic radius. Among the CPVT mutants studied, the L167H mutation disrupts the Ca2+-dependent folding and, when folding is achieved by Mn2+, L167H can undergo aggregation in a Ca2+-dependent manner. Interestingly, domain III mutants (D307H and P308L) lost their selectivity to Ca2+ and could be aggregated in the presence of Mg2+. In conclusion, these studies suggest that CPVT mutations modify CASQ2 behaviour, including folding, aggregation/polymerization and selectivity towards Ca2+.
The interaction between metal ions and proteins plays an essential role in cellular physiology and is vital for the co-ordinated functioning of many organs, including cardiac rhythmicity [1–5]. In cardiac muscle, Ca2+ plays a central role in the contraction/relaxation cycle of the heart [1,6]. Ca2+ cycling is facilitated by cyclical release of stored Ca2+ from the SR (sarcoplasmic reticulum) via RyR2 (ryanodine receptor 2) and re-uptake of Ca2+ by SERCA (sarcoplasmic/endoplasmic reticulum Ca2+-ATPase) . The ability to store a large amount of Ca2+ inside the SR requires a selective Ca2+-buffering protein and is accomplished mainly by a high-capacity Ca2+-binding protein called CASQ2 (calsequestrin 2) [8,9]. Although the SR lumen is highly enriched with Ca2+, studies have shown the presence of other cations including Mg2+, suggesting a role for luminal Mg2+ [10–12]. In addition, there are findings to suggest that the relative ratio of free Zn2+ and Mg2+ to Ca2+ on both sides of the SR membrane could regulate RyR2 opening and contribute to cardiac arrhythmia [10,13–17].
The ability of CASQ2 to selectively buffer Ca2+ plays an essential role in maintaining the SR Ca2+ gradient . The crystal structure shows that the CASQ2 monomer is a mixed α-β protein composed of three globular domains, each of which is formed by five β-strands sandwiched by two α-helices on each side . Each domain consists of a hydrophobic core with the acidic residues on the surface making the surface highly electro-negative . It has been suggested that these negative surface charges can be neutralized by Ca2+ and bring about linear polymerization of CASQ2 in a Ca2+-dependent manner and that protein folding and polymerization of CASQ2 is closely integrated with its Ca2+ interaction [20,21]. Unlike calmodulin, CASQ2 does not have any EF-hand motifs for tight Ca2+ binding, therefore it is not known whether CASQ2 has a preferential selectivity towards Ca2+ or its interaction is merely a consequence of its location in the SR lumen, a Ca2+-rich medium. The absence of an EF-hand motif and the presence of a negatively charged surface suggest that the interaction between CASQ2 and Ca2+ should be very weak. This would allow dynamic association and dissociation of Ca2+ and the rapid release of bound Ca2+ (within ms) to support increased heart rates during high physiological demand [22,23]. It has been shown that CASQ1 (the skeletal isoform of CASQ) has the ability to bind many cations other than Ca2+ , but it is not known whether CASQ2 can also bind other ions.
It is presently unknown whether Ca2+ binding to CASQ2 is due to its location inside the SR rich in Ca2+ or due to its preference to Ca2+ over other ions. Studies have shown that several point mutations in human CASQ2 can cause a form of arrhythmia namely CPVT (catecholaminergic polymorphic ventricular tachycardia). The CPVT mutations (R33Q, L167H, D307H and P308L) occur in all three domains of CASQ2 and could potentially disrupt charge clusters in each domain. R33Q is located in domain I, affecting polymerization and response to Ca2+ [23,25,26]; L167H, located in domain II, inhibits back-to-back dimer formation [27,28]; D307H, occurring in domain III, impairs Ca2+ buffering and polymerization dynamics [29–31]; and, lastly, P308L, reported to cause CPVT, lies adjacent to residue Asp307 . We hypothesize that these mutations should affect cationic selectivity and the ability to buffer Ca2+. Therefore, in the present study, we determined cation-induced folding and subsequent aggregation using CASQ2 purified to homogeneity in the presence of various metal ions. We additionally explored how the individual CPVT mutations impair local structure and/or surface-charge distribution using computational studies to better understand molecular behaviour of the CPVT mutants.
Mutagenesis and purification of mutant CASQ2
The cloning, mutagenesis and purification of rat WT (wild-type) CASQ2, and R33Q and D307H mutants have been reported previously by us [23,29]. L167H and P308L were generated using a site-directed mutagenesis kit (Stratagene) following the manufacturer's protocol. The pair of primers used for the generation of the L167H mutation are: 5′-GAGGACCAGATCAAACACCTTGGCTTTTTCAA-3′ and 5′-TTGAAAAAGCCAAGGTGTTTGATCTGGTCCTC-3′. The primer set used for the generation of the P308L mutation is: 5′-GAGCATCTTGTGGATTGACCTAGACGACTTTCC-3′ and 5′-GGAAAGTCGTCTAGGTCAATCCACAAGATGCTC-3′. The L167H and P308L mutant proteins were overexpressed and purified using methods published previously with minor optimization to ensure a maximum yield of the protein . All of the proteins used in the present study were used at equivalent purity. The protein concentration was estimated using Bradford reagent (Bio-Rad), according to the manufacturer's protocol, with BSA as the standard.
CD spectra were acquired using our recently described protocol  in 20 mM Tris/HCl buffer (pH 7.5) containing 20 mM NaCl. For recording CD spectra with 300 mM NaCl and 300 mM KCl, the proteins were extensively dialysed against 20 mM Tris/HCl buffer (pH 7.5) containing 300 mM NaCl or KCl. In order to examine the effect of Mg2+/Ca2+/Sr2+ on the CD spectra of proteins, 0.2 mM of the salt was added from a stock solution (100 mM MgCl2/CaCl2/SrCl2) to the protein solution, and was mixed and incubated for at least 5 min before recording the spectra. Addition of MgCl2/CaCl2/SrCl2 (0.2 mM) was repeated until the CD spectra overlapped with the spectra obtained at 300 mM NaCl. In order to record CD spectra in presence of Ni2+, Zn2+ and Mn2+, 0.1 or 0.2 mM of the chloride salt was added from a 100 mM stock and was repeated as above. Ellipticity is expressed as observed ellipticity in mdeg.
CASQ2 protein polymerization/aggregation was monitored by adding aliquots (2–5 μl) of stock solutions of metals (0.1–2 M) to a solution of 2.5 μM CASQ2 in 20 mM Tris/HCl buffer (pH 7.5) containing 20 mM NaCl using a Genesys spectrophotometer. After cation addition, protein samples were mixed and incubated to equilibrate for at least 5 min, and the absorbance was then recorded at 350 nm. Absorbance of samples was corrected by subtracting the absorbance of buffer alone after addition of the same concentration of the particular salt. Chloride salts of alkaline-earth metals (Mg2+, Ca2+, Sr2+ and Ba2+) and transition metals (Mn2+, Co2+, Ni2+, Cu2+ and Zn2+) were used for these experiments as a source of cations. Co2+ was used only for WT protein. These transition metals were chosen for the present study because they belong to period 4 of the periodic table and are of comparable ionic radii as Ca2+. Mn2+ (having higher affinity for CASQ2 than Ca2+) was used to achieve molecular compaction at 0.2 mM, then Ca2+ was added and turbidity was measured as described above. All turbidity measurements were repeated at least three times.
Limited proteolysis by trypsin
Purified WT and mutants of CASQ2 were subjected to proteolysis by Tos-Phe-CH2Cl (tosylphenylalanylchloromethane; ‘TPCK’)-treated trypsin (New England Biolabs). The reactions were performed in the absence and presence (1 and 2 mM) of alkaline earth metal cations (Mg2+, Ca2+ and Sr2+). The reactions were also performed in the absence and presence (0.5 and 1 mM) of transition metal cations (Ni2+, Zn2+ and Mn2+). Proteins were incubated in the reaction buffer recommended by the supplier, 50 mM Tris/HCl (pH 8.0), for 20 min at 25 °C in a reaction volume of 200 μl. Then, the indicated cation was added from a stock solution (~20 times concentrated solution) of its chloride salt and again incubated for 20 min. Then the indicated protease was added at a protease/CASQ2 ratio of 1:100 (w/w). The digestion was limited to 30 min, and the samples were quickly mixed with Laemmli sample buffer (Bio-Rad) and heated at 99 °C for 2 min. Equal amounts of protein from each sample were loaded and analysed by SDS/PAGE (12% gel). The gels were stained with Coomassie Brilliant Blue R250 staining solution (Bio-Rad), destained and imaged. Each of the trypsinolysis experiments was repeated at least three times.
Sequence comparison and computational studies
Homologous CASQ sequences were collected by BLAST and analysed by multiple alignments using similar approaches as described previously by us . The accession numbers of the sequences used are provided in the Supplementary Experimental section at http://www.BiochemJ.org/bj/435/bj4350391add.htm. The illustrations showing the conservation of sequences around the CPVT mutations were generated using Weblogo 3.0 . To examine further the effect of the mutations on the 3D (3-dimensional) structure and surface charge, we used MD/FEP (molecular dynamics/free energy perturbation) simulations. For this, the initial 3D co-ordinates were take from the X-ray crystal structure of CASQ2 (chain A; PDB ID 1SJI) and was improved and corrected using the Schrödinger Suite as described previously . The corrected structure was mutated (R33Q, L167H, D307H and P308L) using the builder module of Maestro interface (Schrödinger). All of these structures (both WT and mutated) were subjected to MD/FEP simulations. The MD/FEP studies were carried out employing the DESMOND MD package (Desmond Molecular Dynamics System, version 2.X; D.E. Shaw Research). The complete method for the system building, energy minimization and MD/FEP simulation are described in detail in the Supplementary Experimental section.
CASQ2 maintains a relatively higher affinity for Ca2+ over other alkaline earth metals
It is presently unknown whether Ca2+ binding to CASQ2 is due to its location inside the SR rich in Ca2+ or due to its preference to Ca2+ over other ions. In the present study, we chose to explore ion selectivity to Ca2+ by comparing the effect of various metal groups on CASQ2 folding and aggregation. Previous studies have indicated that CASQ2 is highly structured in the presence of 300 mM KCl , whereas at low salt (25 mM NaCl or 100 mM KCl) it is less structured [23,26]. This prompted us to investigate cation-induced folding of CASQ2. We first investigated the effect of alkaline earth divalent (Mg2+, Ca2+ and Sr2+) and alkali monovalent (Na+ and K+) cations on CASQ2 conformation using CD spectroscopy. The CD spectrum obtained for WT CASQ2 under low cationic strength (20 mM NaCl) showed that it is less compact (Figure 1A), as reported previously by us [23,29]. On the other hand the CD spectrum at high cationic strength (300 mM NaCl) of the buffer showed high ellipticity with two negative minima around 222 and 208 nm (Figure 1B), as reported by Kim et al. . This indicates that the CASQ2 protein has undergone a significant conformational transition from a molten globule structure into a compact molecule. The CD spectra obtained with 300 mM NaCl and KCl were overlapping, and CD ellipticity remained unaltered with a corresponding increase in the concentration of Na+ or K+. Next, we examined the effect of divalent cations. Of the three alkaline earth metals tested (Mg2+, Ca2+ and Sr2+), Ca2+ brought about the conformational transition at the lowest concentration (300 μM). However, a similar transition required higher concentrations of Mg2+ and Sr2+ (1.0 and 1.2 mM respectively). These results suggest that CASQ2 maintains a relatively higher affinity for Ca2+ over other alkaline earth metals at the monomer level (Figure 1). These studies were repeated with the CPVT mutants R33Q, L167H, D307H and P308L, which are located in three different structural domains of CASQ2 (see Figure 4B). The CD spectra of the mutants R33Q, D307H and P308L both at low (20 mM Na/KCl) and at high (300 mM Na/KCl, and 1.0 Mg2+, 300 μM Ca2+ and 1.2 mM Sr2+) cationic strength overlapped with that of the WT protein, indicating that the structural transition is not affected. On the other hand, the L167H mutant failed to undergo the conformational transition in the presence of both alkali and alkaline earth metals.
Cation-induced conformational transition of WT CASQ2 and its CPVT mutants
Ca2+ is the only alkaline earth metal able to induce aggregation of WT CASQ2
Although molecular compactness can be brought about by all of the alkali and alkaline earth cations, as observed by CD spectroscopy, whether aggregation can be initiated by these cations was investigated further by turbidimetry and trypsinolysis. As shown in Figure 2, studies with alkaline earth metals (Mg2+, Ca2+, Sr2+ and Ba2+) revealed that WT CASQ2 only undergoes aggregation (as indicated by becoming significantly turbid) with Ca2+, and not by Mg2+, Sr2+ or Ba2+. Aggregation of WT CASQ2 cannot be induced by monovalent cations (Na+ and K+) (Supplementary Figure S1 at http://www.BiochemJ.org/bj/435/bj4350391add.htm). The domain I mutant (R33Q) had significantly compromised ability to become turbid under low Ca2+ concentrations, but gained turbidity under higher concentrations (~3.5–6.0 mM). The domain II mutation (L167H) completely inhibited the gain in turbidity induced by Ca2+. The R33Q and L167H mutants did not respond to any other alkaline earth metals (Figure 2) and did not gain turbidity. The domain III mutants D307H and P308L showed altered Ca2+-induced behaviour as described below. The CPVT mutants did not gain turbidity in the presence of monovalent cations (Supplementary Figure S1). We investigated further the effect of alkaline earth metals on CASQ2 by limited proteolysis; the WT protein gained resistance to trypsin in the presence of Ca2+ and became more resistant with increases in [Ca2+]. In contrast, there was no pronounced increase in resistance to trypsin digestion with progressive increases in [Mg2+] or [Sr2+]. On the other hand, both R33Q and L167H mutants, which have impaired aggregation, remained highly sensitive to trypsin in the presence/absence of different concentrations of [Mg2+], [Ca2+] or [Sr2+] as shown in Figures 2(E)–2(G).
Alkaline earth metals induce the aggregation of WT and mutant CASQ2
Domain III mutations D307H and P308L lead to a loss of selectivity towards Ca2+
The domain III mutations D307H and P308L showed altered molecular behaviour in comparison with WT CASQ2 in the presence of alkaline earth metals. We found that these mutants showed a gradual increase in turbidity up to ~1.5 mM Ca2+, but they failed to respond to increasing [Ca2+] up to 4 mM, whereas at high [Ca2+] above 4.0 mM the turbidity reached the maximum achievable level (Figure 2). The loss of Ca2+ sensitivity in mutants D307H and P308L (between 1.5 and 4 mM CaCl2) prompted us to investigate their behaviour towards other alkaline earth metals. We found that, in the presence of Mg2+, the mutants D307H and P308L became turbid with increasing [Mg2+], whereas the WT and other mutants failed to respond (Figure 2). The maximum absorbance observed in the presence of Mg2+ was lower compared with Ca2+. The other two alkaline earth metals Sr2+ and Ba2+ (with larger ionic radii) did not bring about turbidity with either mutant. We examined further the molecular properties of these mutants in the presence of [Mg2+], [Ca2+] or [Sr2+] using tryptic digestion. To our surprise, both D307H and P308L became progressively resistant to trypsin with an increase in [Mg2+] in addition to [Ca2+]; however, this resistance was not observed with [Sr2+] (Figure 2). The EC50 values (the concentration required to achieve 50% of maximum achievable turbidity) are shown in Table 1. These results showing a loss of ion selectivity due to both D307H and P308L mutations suggested that they might alter the surface charge distribution. This point was investigated further using computational dynamics.
|Parameter||EC50 (mM)||CT (mM)||EC50 (mM)||CT (mM)||EC50 (mM)||CT (mM)||EC50 (mM)||CT (mM)||EC50||CT (mM)|
|Parameter||EC50 (mM)||CT (mM)||EC50 (mM)||CT (mM)||EC50 (mM)||CT (mM)||EC50 (mM)||CT (mM)||EC50||CT (mM)|
Transition metal ions are capable of promoting the aggregation of WT CASQ2
Our experiments with alkali and alkaline earth cations indicated that Ca2+ is the most favourable metal to support aggregation of WT CASQ2, indicating that ionic size is a key factor. We wanted to investigate whether co-ordination bonding ability is also a determinant of the CASQ2–cation interaction. Therefore we chose to use transition metals which have a high co-ordination bonding ability to study the molecular behaviour of CASQ2. Turbidimetric measurements showed that all of the transition metals (Mn2+, Co2+, Ni2+, Cu2+ and Zn2+) can induce aggregation of WT CASQ2 (Figure 3). The behaviour of the WT protein in the presence of various cations was replotted together to highlight the different turbidity curves obtained in the presence of different cations (Supplementary Figure S2 at http://www.BiochemJ.org/bj/435/bj4350391add.htm), which might be due either to polymerization and/or aggregation. Moreover, all of the five transition metals brought about the conformational transition as observed by CD spectroscopy (Supplementary Figure S3 at http://www.BiochemJ.org/bj/435/bj4350391add.htm). Out of these transition metals used, Ni2+, Zn2+ and Mn2+ were selected for further investigation using limited proteolysis because they are present in significant amounts in the human body (Supplementary Table S1 at http://www.BiochemJ.org/bj/435/bj4350391add.htm). For this study, proteolysis was performed at 0.5 and 1.0 mM concentrations of the transition metal ions. At 0.5 mM MnCl2, CASQ2 became significantly protected; however, at this concentration, neither Ni2+ nor Zn2+ conferred protection against trypsin digestion of WT CASQ2. On the other hand, at 1.0 mM of both Mn2+ and Ni2+, WT CASQ2 became resistant to trypsinization, but at 1.0 mM Zn2+ the protein still remained comparatively susceptible with a lesser quantity of the native band (Figure 3).
Transition metals are capable of promoting aggregation of CASQ2
Transition metals can promote aggregation of R33Q and L167H
We found that, in the presence of transition metals (Mn2+, Ni2+, Cu2+ and Zn2+), the domain I and domain II mutants (R33Q and L167H) can undergo aggregation (Figure 3). For R33Q, Mn2+ generated the highest absorbance at the lowest concentration and, for L167H, it was Cu2+. The two domain III mutants (D307H and P308L) showed the least sensitivity to Cu2+ and did not respond to the intermediate concentrations of Ni2+ (between 1.5 and 4 mM). Mn2+ was the only transition metal capable of promoting aggregation of all of the mutants and WT CASQ2. As R33Q and L167H mutants could undergo aggregation, their conformation was compared further using limited proteolysis with trypsin. Interestingly, both mutants gained resistance to trypsinolysis progressively with increases in metal concentration (Figure 3), indicating that gradual aggregation can be achieved. The R33Q and L167H mutants failed to show resistance in the presence of Ca2+, but became resistant in the presence of Zn2+ and Mn2+. Next, we examined whether the mutant L167H could undergo the key conformational transition in the presence of transition metals using CD spectroscopy. The L167H mutant, which did not respond to alkali and alkaline earth metals, could undergo conformational transition when induced by transition metals, similar to WT CASQ2 (Supplementary Figure S4 at http://www.BiochemJ.org/bj/435/bj4350391add.htm). The concentration of transition metals needed to bring about the conformational transition for each of the mutants as well as the EC50 values are shown in Table 1. We first used Mn2+ to promote monomer compaction of L167H and then studied its ability to form aggregates in the presence of Ca2+. The turbidity increased progressively with the addition of Ca2+ and the turbidity curve of L167H resembled that of WT CASQ2 and the Ca2+ concentration required to bring about the saturation of turbidity was similar. The results suggests that, once monomer compaction is achieved with Mn2+, the L167H mutant can undergo aggregation in a Ca2+-dependent manner (Figure 4A).
CPVT mutations in CASQ2 disrupt evolutionarily conserved residues
Despite significant progress in CASQ2 structure–function research, it is still unknown whether CASQ2 has a preferential selectivity towards Ca2+ or if the CASQ2 interaction with Ca2+ is merely a consequence of its location in the lumen of the SR. Therefore, in the present study, we examined the basis of ion selectivity of CASQ2 by exposing CASQ2 to alkali, alkaline earth and transition metals. We also investigated how CPVT mutations occurring in different domains of CASQ2 affect cation-induced folding and aggregation/polymerization. The major findings of the present study are that (i) native CASQ2 undergoes critical molecular compaction before the formation of dimers and aggregates of CASQ2; (ii) native CASQ2 shows preferential ionic selectivity to Ca2+ over other metal ions; (iii) ion-selectivity depends on cationic properties of individual metal ions but not abundance; and, lastly, (iv) CPVT mutations occurring in different domains can cause a distinct alteration in the overall function of the CASQ2 molecule and could affect ion selectivity.
One of the most interesting observations made in the present study is that CASQ2 remains in a loosely packed molten globular structure at a low cationic strength of the medium. The molten globule is an intermediate state during the protein folding pathway in which the folding has already reached a critical point (with hydrophobic collapse), but many minor contacts and/or close residue–residue interactions present in the native state are yet to form [34–36]. With an increase in the cationic strength of the medium, the protein first undergoes a key conformational transition to form a compact monomer. Many different cations can promote this conformational transition (at different ionic strengths), and this transition process is intricately associated with cation interaction with CASQ2. Although the above structural transition can be brought about by many cations, only specific divalent cations can support aggregation of the protein. Our present results also agree with the published findings that alkali metals, such as Na+ and K+, can cause molecular compaction, but they cannot support CASQ2 aggregation [37–39]. The results show that, among the alkaline earth metals, only Ca2+ can promote CASQ2 to aggregate into higher order structures. Studies by Wei et al.  showed that CASQ2 remains either as a monomer or dimer at [Ca2+] up to 1 mM. We found that 0.3 mM [Ca2+] is required for monomer compaction and CASQ2 aggregation occurs at [Ca2+] higher than 1 mM. These results are in agreement with published findings showing that CASQ2 aggregates only when [Ca2+] exceeds 1 mM [20,23,26]. We have shown previously that polymeric CASQ2 returns to a monomeric/dimeric form when [Ca2+] is lowered, suggesting that the polymerization state is reversible . In addition, our studies suggest that both the ionic size and co-ordinate bonding ability are the determinants of CASQ2 aggregation. These results are in agreement with published findings and prove that cation binding is intertwined with CASQ folding and aggregation/polymerization [20,21,39,41].
In the present study, we investigated the molecular behaviour of CPVT mutations located in all of the three structural domains of CASQ2. The location of each mutation in the molecule and charge distribution around the mutation location are shown in Figure 4(B). Notable among the mutants is L167H because it fails to undergo the key structural transition (molecular compaction), which is in agreement with published results [26,28]. We analysed further the structural conservation of the amino acids present around Leu167. Multiple sequence alignment of vertebrate CASQ (CASQ1 and CASQ2) and invertebrate CASQ showed that the β-7 sheet (KLIGF) containing Leu167 is highly conserved. Moreover, Leu167 has been found to be replaced by arginine or valine residues in only seven out of 33 species (Figure 4C). Neither Ca2+ nor any of the other alkaline earth metals can bring about the compaction of the L167H mutant and therefore cannot support its aggregation. To gain further insight into how introduction of a histidine residue affects the structure of the protein leading to a loss of Ca2+ sensitivity we performed molecular dynamics studies. We found that Leu167 contained in a β-sheet (KLIGF) is sandwiched between the two α-helices (α-5 and α-6) and stabilizes the local structure of domain II in two ways (Figure 5A). First, Leu167, through its backbone, establishes hydrogen bond contact with Phe216 of the neighbouring β-sheet (β-9) so that the β-sheets become arranged in an almost planar orientation, forming the core of domain II. Leu167 also establishes hydrogen bond contact through its side chain and Phe182 of the α-6 helix (Figure 5A). Secondly, the hydrophobic side chain of Leu167 also establishes weak van der Waals interactions with the surrounding residues Phe181, Phe195, Phe216, Val214 and Val241 (Figure 5C). The replacement of Leu167 with a histidine residue introduces an imidazole ring into the hydrophobic region that not only disturbs the hydrogen bond network, but also introduces steric hindrance, which is non-conducive to the formation of the van der Waals interactions and leads to an alteration in the 3D geometry as shown in Figures 5(B) and 5(D) (also see Supplementary Figure S5 at http://www.BiochemJ.org/bj/435/bj4350391add.htm). Therefore it becomes energetically unfavourable (Table 1) and cannot undergo the key molecular compaction. It is noteworthy to point out that L167H is the only human CPVT mutation reported to exert an arrhythmic phenotype in a compound heterozygous condition and interacts poorly with RyR2 in vivo [17,27].
Computational studies showing structural alterations caused by CPVT mutations
Another intriguing finding of the present study was that the two neighbouring mutants D307H and P308L generate similar turbidity curves and proteolytic profiles in the presence of all of the metals studied. Notable is their biphasic turbidity curve in Ca2+ and sensitivity towards Mg2+. Our sequence conservation analysis showed that they disrupt a very highly conserved loop, which is composed of negatively charged residues (Figure 4D). Employing molecular dynamics we found that the introduction of a histidine residue in the D307H mutation displaced the surface charge of the neighbouring residues (Glu275, Ser277, Asp309 and Asp310) as shown in Figure 5(F). On the other hand P308L introduces a hydrophobic side chain that repels the surrounding hydrophobic residues (Phe273, Leu326, Phe311 and Trp318), which leads to displacement of surface charge of residues Glu275, Ser277, Asp309 and Asp310 (Figure 5G). Therefore it appears that when the domain III charge distribution pattern is disturbed it could greatly modify CASQ2 behaviour to metal ions, including loss of ion selectivity towards Ca2+, which could have important consequences for muscle function. It has been shown that, in addition to Ca2+, Mg2+ found inside the SR can also regulate RyR2 opening and Ca2+ release [10,11] and there has been growing evidence that the relative ratio of free Mg2+ and Ca2+ on both sides of the SR membrane is important in regulating RyR2-mediated Ca2+ release [10,42–44]. Moreover, studies have shown a link between Mg2+ imbalance and cardiac arrhythmia [13,45,46].
It is not surprising that we were able to observe aggregation of mutants (R33Q and L167H) in the presence of transition metals that failed to respond Ca2+. Our results suggests that transition metals (Ni2+, Mn2+, Cu2+, Co2+ and Zn2+), which can form a higher number of co-ordinate bonds (due to incomplete ultimate and penultimate orbitals), can promote molecular compaction as well as aggregation. Moreover, they can vary the bond angles and length and, due to this flexibility, they can interact with CASQ2 in an induced-fit manner. Mn2+ is an especially unique element having a half-filled outer orbital and its ionic radius can vary depending on its valence state. Because of these properties it can form the highest number of co-ordinate bonds and promote a CASQ2 conformation which is energetically favourable to undergo aggregation. However the precise role of these transition metals in cardiac excitation–contraction coupling is presently unknown, although results suggests that Zn2+ is present inside the cardiac SR and Mn2+ is present in the cardiomyocytes . In conclusion, our present results demonstrate that the preferential binding of CASQ2 to Ca2+ is based on the cationic properties of Ca2+ but not abundance. Furthermore the findings obtained in the present study show that CPVT mutations in CASQ2 affect cation-induced folding and compaction at the monomer level and subsequent aggregation thus compromising CASQ2 function.
We thank Dr Brandon J. Biesiadecki and Dr Jonathan P. Davis (The Ohio State University College of Medicine, Columbus, OH, U.S.A.) for critical reading of the manuscript prior to submission and for helpful suggestions. We thank Joseph Ostler and Leslie A. Rowland for editorial assistance and input prior to submission. N.C.B. thanks Dr Ashish Arora and Professor Rita Bernhardt for mentoring in protein biochemistry.
Naresh Bal and Muthu Periasamy conceived the idea and designed the experiments with help from Nivedita Jena, Ashoke Sharon and Sandor Gyorke. Naresh Bal, Nivedita Jena, Danesh Sopariwala, Tuniki Balaraju, Sana Shaikh, Chandralata Bal and Ashoke Sharon planned and performed the experiments. Naresh Bal, Muthu Periasamy, Nivedita Jena and Ashoke Sharon analysed the data. Naresh Bal and Muthu Periasamy wrote the manuscript with help from Nivedita Jena, Ashoke Sharon and Sandor Gyorke. Studies involving biophysical, biochemical and bioinformatics were performed by Naresh Bal, Nivedita Jena, Danesh Sopariwala and Sana Shaikh. Studies involving computational modelling were performed by Tuniki Balaraju, Chandralata Bal and Ashoke Sharon.
This work was supported, in part, by the National Institutes of Health [grant number R01 HL64014 (to M. P.)]. N.C.B. was supported by a postdoctoral fellowships from the American Physiological Society (physiological genomics) and the American Heart Association. T.B., C.B. and A.S. were supported by the Department of Science and Technology, Government of India.
Present address, The Ohio State University College of Pharmacy, Columbus, OH 43210, U.S.A.
These authors contributed equally to the study.