Magnesium enhances aurintricarboxylic acid’s inhibitory action on the plasma membrane Ca2+-ATPase

Effect of ATA on the PMCA activity

Figure 1b shows PMCA Ca²⁺-ATPase activity as a function of increasing ATA concentrations. The continuous line corresponds to the fitting of Eq. (1) to the experimental data,

where v₀ and v_∞ correspond to the PMCA Ca²⁺-ATPase activity in the absence of ATA and when its concentration tends to infinity, K_i is the ATA concentration at which half the maximum effect is observed, and n is the Hill coefficient. The values obtained for these parameters were 0.073 ± 0.003 µM; v_∞ = 3.0 ± 1.6% and n = 1.8 ± 0.1.

The PII-ATPases can hydrolyze other substrates in the absence of the transported ion, indicating that this phosphatase activity is not linked to ion transport. PMCA hydrolyzes p-nitrophenyl phosphate (pNPP), and this phosphatase activity is inhibited by Ca²⁺, suggesting its association with the E2 conformation¹⁹. Figure 1c shows the phosphatase activity of PMCA as a function of the ATA concentration. The continuous line represents the fitting of Eq. (1) to the experimental data, but in this case, v₀ and v_∞ refer to the PMCA phosphatase activity. The values obtained for these parameters were: K_i = 0.097 ± 0.004 µM; v_∞ = 2.6 ± 0.6%, and n = 2.5 ± 0.2. These results indicate that ATA inhibits both the Ca²⁺-ATPase and phosphatase activities of PMCA isolated from human erythrocytes with high apparent affinity.

Effect of calcium, magnesium, and ATP on inhibition of PMCA by ATA

Figure 2a shows the Ca²⁺-ATPase activity of PMCA as a function of free Ca²⁺ in the presence of different ATA concentrations. In all cases, the experimental data were described by Eq. (5). Table 1 shows the parameter obtained in the absence and presence of 0.15 µM ATA. When ATA concentrations increased, K_Ca (Fig. 2d) and V_max decreased (Fig. 2g), and the K_Ca/V_max ratio was not constant (inset in Fig. 2g). These results suggest that in PMCA, ATA behaves as a mixed-type inhibitor with respect to Ca²⁺. In this mechanism, the inhibitor binds with different affinity to the free enzyme or the enzyme–substrate complex, meaning that the presence of the substrate modifies the affinity for the inhibitor and vice versa²⁰. This agrees with the fact that ATA also inhibits the phosphatase activity of PMCA with apparent high affinity (Fig. 1c).

Like other members of the P-ATPase family, Mg²⁺ is an essential cofactor for PMCA. On one hand, the true substrate is ATP·Mg, and on the other, PMCA possesses a binding site for Mg²⁺. Figure 2b shows the PMCA Ca²⁺-ATPase activity as a function of free Mg²⁺ in the presence of different ATA concentrations. The continuous lines correspond to the fitting of Eq. (5) to the experimental data. In the presence of increasing ATA concentrations, K_Mg (Fig. 2e) and V_max (Fig. 2h) decreased (see also Table 1), and the K_Mg /V_max ratio remained approximately constant (inset in Fig. 2h). These results suggest that ATA behaves as an uncompetitive inhibitor with respect to Mg²⁺. In this mechanism, the inhibitor binds to the enzyme–substrate complex, meaning an increase in the apparent affinity for the substrate is observed²⁰.

Figure 2c shows the PMCA Ca²⁺-ATPase activity as a function of ATP (ATP·Mg) in the presence of different ATA concentrations. The experimental data were described by Eq. (5) and K_ATP represents an overall value between the PMCA apparent affinity by ATP at the catalytic site (high affinity) and the regulatory site (low affinity)²¹. K_ATP (Fig. 2f) remained approximately constant up to 0.1 µM ATA and then increased, while V_max (Fig. 2i) decreased throughout the range evaluated (see also Table 1). Thus, the K_ATP/V_max ratio as a function of ATA concentration was not constant (inset in Fig. 2i). These results suggest that in PMCA, ATA behaves as a mixed-type inhibitor with respect to ATP, i.e., ATA does not bind to the ATP-binding site but affects the affinity of PMCA for this nucleotide.

Interaction of ATA with magnesium

The interaction of ATA with metals has been described early and its employ to detect aluminum in water in a standard assay²². Besides, ATA can form complexes with alkaline earth metals such as Sr²⁺, Ba²⁺, Ca²⁺, and Mg²⁺ (Fig. 3a)²³. Thus, the cooperative behavior observed in the inhibition of Ca²⁺-ATPase activity (Fig. 1b) could be because the molecule that binds to PMCA is ATA·Mg (or ATA·Ca²⁺), and this complex forms in low amounts at low ATA concentrations.

Interaction of ATA and Mg²⁺ evaluated by fluorescence. (a) Structure of Triaurinecarboxylic Acid coordinated with Mg²⁺ (ATA·Mg). (b) Excitation spectra in the presence of increasing ATA concentrations in the reconstitution medium of PMCA. Inset. Fluorescence at 306 nm as a function of ATA concentration (from top to bottom: 0.4; 0.8; 1.6; 3.1; 6.3; 12.5; 25 and 50 µM ATA). (c) ATA fluorescence at 306 nm as a function of the Mg²⁺ concentration. The λ_em was 425 nm and the fluorescence in the absence of Mg²⁺ (1 mM EDTA) was normalized to 1. The discontinuous line represents the fitting of Eq. (2) to the experimental data and the values obtained for b and F_∞ parameter were 24.1 ± 2.1 µM and 0.860 ± 0.002, respectively. Inset. Excitation spectra of ATA in the presence of increasing Mg²⁺ concentrations (from top to bottom: 0; 20; 50; 100 and, 1000 µM Mg²⁺). Panel (b) and inset in (c) show representative experiments, Panel (c) corresponds to the results obtained in three independent experiments.

Figure 3b shows the fluorescence excitation spectrum of ATA in the reaction medium in which the Ca²⁺-ATPase activity of PMCA was determined, but in the absence of ATP and protein. The excitation maximum was observed at 306 nm, and the fluorescence intensity increased linearly as a function of the ATA concentration (inset in Fig. 3b). These results agree with those previously described for aqueous solutions of ATA at pH 7.4⁷ and suggest that the molecule is a monomer at the evaluated concentrations.

Figure 3c shows the fluorescence at 306 nm of 10 µM ATA as a function of the Mg²⁺ concentrations. The addition of Mg²⁺ decreased the ATA fluorescence with a saturable behavior without significant changes in other characteristics of the spectrum (inset in Fig. 3c). The experimental data were well described by Eq. (2),

where F₀ and F_∞ correspond to the fluorescence in the absence of Mg²⁺ and when its concentration tends to infinity, respectively, and b corresponds to the added Mg²⁺ concentration at which half the maximum effect is achieved. The values obtained for b and F_∞ were 24.1 ± 2.1 µM and 0.860 ± 0.002, respectively. A similar result was obtained in the absence of Mg²⁺ and increasing concentrations of Ca²⁺ (data not shown). These results suggest that in the conditions in which Ca²⁺-ATPase activity was determined (Fig. 1b), most of the ATA would be found as ATA·Mg complex and, consequently, it would be the form that inhibits PMCA. In agreement with this, the phosphatase activity of PMCA, which is determined in the absence of Ca²⁺ and in the presence of Mg²⁺, was inhibited by ATA (Fig. 1c).

Binding of ATA to PMCA assessed by fluorescence

ATA fluorescence in the presence of PMCA

To evaluate the binding of ATA to PMCA, we studied the changes in fluorescence of both molecules in different experimental conditions.

Figure 4a shows the emission spectrum of ATA in the absence and presence of PMCA plus Ca²⁺ and Mg²⁺. Under these conditions, the conformational equilibrium of PMCA is shifted towards the E1Ca state. The ATA fluorescence increased in the presence of PMCA, and the λ_em maximum shifted slightly towards blue (from 425 to 429 nm) indicating that the ATA is in a more hydrophobic environment¹¹. The fluorescence of ATA bound to PMCA increases as a function of ATA concentration with a saturable behavior (Fig. 4b), whereas in the absence of PMCA (free ATA) fluorescence increases linearly (inset in Fig. 4b). The continuous line corresponds to the fit of Eq. (3) to the experimental data,

where Kʹ_ATA corresponds to the ATA concentration at which half the maximum effect is achieved. The value obtained for Kʹ_ATA was 0.16 ± 0.02 µM. These results suggest that, under equilibrium conditions, only one molecule of ATA binds to PMCA with high affinity.

Binding of ATA to PMCA assayed by fluorescence. (a) Fluorescence emission spectrum of 1.5 μM ATA in the absence (green line) and presence (red line) of PMCA. The λ_ex was 306 nm and the fluorescence intensity at 429 nm in the presence of PMCA was normalized to 1. (b) Relative fluorescence of ATA bound to PMCA at 429 nm as a function of the ATA concentration. The continuous line represents the fitting of Eq. (3) to the experimental data and the value obtained for Kʹ_ATA was 0.16 ± 0.02 µM. Inset. Relative fluorescence of ATA at 429 nm as a function of ATA concentration in the absence of PMCA. The continuous line represents the fitting of a linear equation to the experimental data. (c) Fluorescence emission of PMCA in the absence (dashed black line) and in the presence of increasing concentrations of ATA. The λ_ex was 290 nm and the fluorescence intensity at 332 nm in the absence of ATA was normalized to 1. Inset. Fluorescence emission spectra of PMCA (red line) and ATA (dashed gray line), and fluorescence excitation spectrum of ATA (green line). (d) Intrinsic fluorescence of PMCA at 332 nm as a function of ATA concentration. The continuous line represents the fitting of Eq. (2) to the experimental data and the values obtained for F_min and Kʹʹ_ATA were 0.69 ± 0.09 and 0.19 ± 0.02 µM, respectively. Inset. Stern–Volmer plot, where F₀ and F are the fluorescence at 332 nm in the absence and presence of different concentrations of ATA (quencher), respectively. The dashed line is illustrative to facilitate visualization of the behavior of the experimental data. Panels (a–c) show representative experiments, Panel (d) shows the data from three independent experiments.

Intrinsic fluorescence of PMCA in the presence of ATA

The intrinsic fluorescence of PMCA is due to the presence of aromatic amino acids (Tyr, Phe, and Trp). Among these, Trp have the highest quantum yield, are very sensitive to environmental changes, and are the most susceptible to energy transfer (quenching) by charged residues or other ligands²⁴. Figure 4c (inset) shows the excitation and emission spectra of PMCA and ATA. The emission spectrum of PMCA (red line) overlaps with the excitation spectrum of ATA (green line), indicating that the energy transfer from PMCA to ATA is possible. Furthermore, the emission spectrum of ATA (dashed gray line) does not overlap with that of PMCA, ensuring that the fluorescence at 332 nm arises solely from the intrinsic fluorescence of the protein. When increasing concentrations of ATA were added to PMCA and the samples were excited at 290 nm, a decrease in PMCA fluorescence concomitant with an increase in ATA emission was observed (Fig. 4c). Under these conditions, the main contribution of PMCA fluorescence is due to the Trp residues²⁵.

Figure 4d shows that the PMCA fluorescence at 332 nm as a function of the ATA concentration decreased and then remained constant. Equation (4) was fitted to the experimental data,

where F₀ and F_min correspond to the intrinsic fluorescence of PMCA in the absence of ATA and when its concentrations tend to infinite, respectively, and Kʹʹ_ATA is the ATA concentration at which half the maximum effect is achieved. The values obtained for F_min and Kʹʹ_ATA were 0.69 ± 0.09 and 0.19 ± 0.02 µM, respectively. The value of Kʹʹ_ATA was not significantly different from Kʹ_ATA (from Eq. (3)) indicating that it is the same phenomenon observed from the changes that occur in PMCA (Fig. 4d) or in the ligand (Fig. 4b) when the PMCA·ATA complex is formed. Note, if we consider the approximation that most of ATA is in the free form ([ATA]_free ⁓ [ATA]_TOTAL), the value of Kʹʹ_ATA (and Kʹ_ATA) corresponds to the value of the apparent dissociation constant (K_D) for the PMCA·ATA complex.

The representation of F₀/F as a function of the quencher concentration is known as the Stern–Volmer plot and provides information about the interaction between a fluorophore and a quencher. Here, F₀ and F represent the fluorescence in the absence and presence of a given quencher concentration ([Q]), respectively. The Stern–Volmer relationship assumes interaction between a single fluorophore and a quencher, proposing a linear relationship between F₀/F and [Q]. Consequently, if PMCA had a single Trp residue (fluorophore) near the ATA-binding site, an increase in ATA concentration (quencher) would lead to a proportional decrease in the intrinsic fluorescence of the protein, resulting in a linear Stern–Volmer relationship. However, when multiple fluorophore molecules exist, such as in proteins with several Trp, the quencher may have varying degrees of accessibility to these residues, causing deviation from linearity in the Stern–Volmer plot with a negative curvature²⁴. In the presence of saturating ATA concentrations, the intrinsic fluorescence of PMCA decreased by 30%, indicating that only a reduced group of Trp is close to the ATA-binding site and is accessible to quenching. The remaining intrinsic fluorescence (F_min) is attributed to the residues distant from the ATA-binding site and thus, not accessible to quenching. The Stern–Volmer plot for PMCA and ATA deviates from linearity with a negative curvature (inset in Fig. 4d), suggesting the presence of more than one Trp near the ATA-binding site, each with different accessibility to quenching by ATA.

Effect of ATA on the nucleotide-binding pocket of PMCA

It has been proposed that ATA binds to the nucleotide-binding site of enzymes and inhibit their ATPase activity⁶. However, our results show that in PMCA, ATA behaves as a mixed-type inhibitor with respect to ATP, i.e. ATA does not bind to the ATP-binding site but its binding decreases the affinity of PMCA for this substrate. Thus, to evaluate whether ATA binds the ATP-binding site of PMCA in equilibrium conditions, we assayed the fluorescence of the PMCA∙ATA complex in the absence of Ca²⁺ (E2 state) to prevent the pump from cycling and in the presence of increasing ATP concentrations.

If ATP displaces ATA, the fluorescence of the PMCA∙ATA complex should decrease until values close to free ATA. On the contrary, the addition of ATP (ATP·Mg) produced an increase in the fluorescence, although without significant changes in the emission spectrum (Fig. 5a). Figure 5b shows that the fluorescence of the PMCA∙ATA complex at 429 nm increased with a saturable behavior (dotted line) as a function of ATP concentration, indicating that the binding of ATP to PMCA·ATA complex produces changes in the ATA environment. In the absence of PMCA, the addition of ATP produced a slight and linear increase in fluorescence (inset in Fig. 5b), consistent with its poor quantum yield under these experimental conditions²⁶. Half of the maximum effect on the fluorescence was observed at 2.8 ± 0.1 mM ATP (Eq. 6), this value is two orders of magnitude greater than the dissociation constant for ATP in the E2 state of PMCA previously reported²¹. These results indicate that ATP binds to the PMCA∙ATA complex with a lower affinity than that to the free enzyme.

Effect of ATA on the nucleotide-binding pocket of PMCA. (a) Relative fluorescence of ATA bound to PMCA (PMCA·ATA complex) in the absence (dashed black line) and presence of increasing concentrations of ATP. (b) Relative fluorescence of ATA bound to PMCA (429 nm) as a function of ATP concentration. The dashed line corresponds to the fitting of Eq. (6) to the experimental data (see “Methods” section for details). Inset. Relative fluorescence of ATA (429 nm) as a function of ATP concentration in the absence of PMCA. In A and B, the fluorescence of ATA bound to PMCA at 429 nm in the absence of ATP was normalized to 1. (c) Relative fluorescence of ATA bound to PMCA (429 nm) in E2·ATP state. The continuous line represents the fitting of Eq. (3) to the experimental data and Kʹ_ATA was 0.9 ± 0.1 µM. (d) Relative fluorescence of Eo bound to PMCA (PMCA·Eo complex) in the absence (red line) and presence of increasing ATA concentrations. The gray line shows the fluorescence of Eo in the absence of PMCA. The fluorescence of Eo bound to PMCA at 543 nm in the absence of ATA was normalized to 1. (e) Relative fluorescence of Eo bound to PMCA (543 nm) as a function of ATA concentration. Inset. Emission spectra of ATA (red line) and the Eo (dashed red line), and excitation spectrum of Eo (green line). (f) Relative fluorescence of ATA bound to the PMCA·Eo complex (429 nm) as a function of ATA concentration. The dashed line corresponds to the fitting of Eq. (6) to the experimental data. In Panels (c,f), the fluorescence at 2 µM ATA was normalized to 1. Panels (a–d) show representative experiments, Panels (e,f) show the data from three independent experiments.

In another set of experiments, PMCA was incubated in the presence of 1 mM ATP and in the absence of Ca²⁺ (E2·ATP state). Then, increasing concentrations of ATA were added, and the fluorescence of the PMCA·ATA complex was determined (Fig. 5c). In the absence of ATA, the fluorescence of the PMCA·ATP complex was close to zero, and it increased as the concentration of ATA increased, indicating that ATA binds to the PMCA·ATP complex. The value of Kʹ_ATA obtained from the fitting of Eq. (3) to the experimental data was 0.8 ± 0.1 µM, a value significantly higher than that obtained in the absence of ATP (Fig. 4b). These results show that ATP binds to the PMCA·ATA complex and induces a change in the environment of ATA, affecting its fluorescence (Fig. 4b). Since ATP has a low quantum yield, we used the fluorescent probe eosin to evaluate whether ATA induces changes in the ATP-binding site environment.

Eosin (Eo) is a fluorescent probe that binds with high affinity (0.1 µM) to the ATP-binding site of PMCA inhibiting the Ca²⁺-ATPase activity by a competitive mechanism²⁷. Eosin fluorescence is sensitive to changes in the hydrophobicity of its environment; thus, it can sense conformational changes that affect the nucleotide-binding pocket in the PMCA·Eo complex. When Eo binds to PMCA, its fluorescence increases and the emission maximum shifts from 539 to 543 nm. Consequently, the emission and excitation spectra of ATA and Eo do not overlap (inset in Fig. 5e), allowing us to specifically evaluate the fluorescence of ATA or Eo in the PMCA·Eo·ATA complex. Furthermore, the effect of ATA on the nucleotide-binding pocket can be studied in the presence of Ca²⁺ (E1Ca). To evaluate whether ATA displaces Eo from the ATP-binding site of PMCA, the PMCA·Eo complex was formed, and increasing concentrations of ATA were added. Then, the fluorescence at 543 nm (PMCA·Eo complex) and at 423 nm (PMCA·ATA complex) was determined in each sample.

Figure 5d shows the emission spectrum of Eo in the absence (gray line) and the presence of PMCA and Ca²⁺ (PMCA·Eo complex, red line). The addition of ATA to the PMCA·Eo complex produced a slight increase in the fluorescence and a slight redshift of the emission maximum, suggesting that the environment of Eo in the nucleotide-binding pocket changes. The fluorescence of the PMCA·Eo complex at 543 nm as a function of ATA concentration remained constant (Fig. 5e), indicating that ATA does not displace Eo from the nucleotide-binding pocket. In the same samples, the fluorescence of the PMCA·ATA complex increased with a saturable behavior (Fig. 5f), suggesting that ATA binds to the PMCA·Eo complex. Half of the maximum effect on the fluorescence was observed at 0.72 ± 0.01 µM ATA (Eq. 6), a value higher than that obtained in the absence of Eo (Fig. 4b). Taken together, these results indicate that ATA can bind to the PMCA·ATP complex (and to PMCA·Eo) but with a lower affinity than to the free enzyme. Consistent with this, ATP also binds to the PMCA·ATA complex with lower affinity than to free PMCA.

The ATA-binding site assayed by flexible molecular docking

To find a possible binding site of ATA in PMCA, we performed flexible molecular docking studies. From this point forward, we will denote the immediate environment surrounding the ATP-binding site as nucleotide-binding pocket.

The ATP-binding site

Because PMCA structure has not been resolved in the presence of ATP, we first performed molecular docking of ATP (ATP·Mg) on PMCA1. Figure 6a shows the predicted ATP-binding site of PMCA1 in the absence of Ca²⁺ (E2). As residues interacting with ATP are conserved in P-ATPases, we compared the results obtained by molecular docking in PMCA1 (first position =  − 9.3 kcal/mol) with residues identified in the E2·ATP structure of SERCA1^28,29 (detailed in parentheses in the text). All residues identified in PMCA1 are conserved in PMCA4 (see legend of Fig. 6a). In PMCA, the γ-phosphate interacts with Lys773 (Lys684³⁰) and Lys476 through salt bridges, and with Thr708 (Thr 625), Gly709 (Gly626), and Thr477 (Thr353) through hydrogen bonds. Lys476 and Thr477 belong to the 475DKTG motif, which contains the phosphorylatable Asp475, and Thr708 and Gly709 belong to the 708TGDN motif³¹. Nearby, Asp475 (Asp351) and Asp797 (Asp703) interact through the Mg²⁺ ion²⁹, Thr799 (Val705), and Asp801 stabilize the nucleotide through hydrogen bonds. Residues Asp797, Thr799, and Asp801 belong to the 795TGDVND motif (701TGDXND), which together with the 708TGDN motif, is involved in the coordination of Mg²⁺ upon the binding of ATP³². This suggests that the predicted site may be the ATP-binding site in PMCA.

The ATA-binding site. (a) Propose ATP-binding site of PMCA. All residues identified in PMCA1 are conserved in PMCA4 as follow (PMCA1 → PMCA4): Asp475–Lys476–Thr477 → Asp465–Lys466–Thr467; Thr708–Gly709 → Thr696–Gly697; Lys773 → Lys661; Asp797–Trh799–Asp801 → Asp785–Trh787–Asp789. (b) Propose ATA·Mg-binding site on the PMCA1 structure. All residues identified in PMCA1 are conserved in PMCA4 as follow (PMCA1 → PMCA4): Asp475–Thr477 → Asp465–Thr467; Val707–Thr708–Gly709 → Val695–Thr696–Gly697; SER768–Pro770 → SER756–Pro758. (c) Trp residues of PMCA near ATA·Mg: Trp759 (P-domain) at 20.8 Å, Trp844 (P-domain) at 29.9 Å and Trp664 (N-domain) at 34.5 Å. These residues correspond to Trp 747, Trp 832 and Trp652 in PMCA4. (d) Relative fluorescence of ATA bound to E2, E1Ca, E1(Ca)P and E2P (E2·BeF) states of PMCA. The fluorescence of ATA bound to E2 state was normalized to 1. (e) Binding of ATA·Mg to PMCA·ATP complex. (f) Binding of PMCA·ATP to ATA·Mg complex. Color code: The oxygen, carbon, phosphorus and hydrogen atoms are colored red, blue, orange and white, respectively. The magnesium ion is represented as a green sphere. The ATA molecule is shown in blue light (b) or pink (c,e,f). In (d), the values are the mean ± SE of three independent experiments. *P < 0.01 with respect to E2.

The ATA-binding site

In the molecular docking of ATA uncomplexed with Mg (ATA(−)) all predicted conformations overlap with the ATP-binding site. In the most probable conformation (− 8.3 kcal/mol), all residues that interact with ATA(−) coincide with those proposed for the interaction with ATP (Supplementary Fig. S1). Therefore, the proposed ATA(−) binding site on PMCA does not match experimental results showing that this ligand does not bind to the ATP-binding site. In the molecular docking of ATA·Mg, all predicted conformations are possible and are close to the ATP-binding site (Supplementary Fig. S2). Figure 6b shows the ATA·Mg-binding site in the most likely conformation (− 9.6 kcal/mol). ATA·Mg interacts through Mg²⁺ ions with Asp475 and Val707 and, establishes hydrogen bonds with Thr477, Thr708, and Gly709. Thus, ATA·Mg would interact with residues of the 708TGDN and 475DKTG motifs, involved in the interaction with Mg²⁺ and the γ-phosphate of ATP at the phosphorylation site (Asp475)³². Furthermore, ATA·Mg interacts with Ser768 through a hydrogen bond and is stabilized by hydrophobic interactions with Pro770. These latter residues are in the nucleotide-binding pocket but are not involved in the interaction with ATP (Fig. 6a). Unlike ATA(−), ATA·Mg does not interact with Lys773 and Lys476, two key residues for the interaction with ATP³¹. A similar result has been described in the serine/threonine phosphatase Stp1⁴, where ATA binds into the active pocket inhibiting the phosphatase activity by a mixed-type mechanism. On the other hand, three of the ten Trp residues of PMCA are located less than 35 Å apart from ATA-Mg: Trp759 (P-domain) at 20.8 Å, Trp844 (P-domain) at 29.9 Å and Trp664 (N-domain) at 34.5 Å (Fig. 6c). The other Trp are located at the interface with the transmembrane domain, at a distance greater than 35 Å from the ATA·Mg. This agrees with the fact that only a small fraction of the Trp of PMCA is differentially accessible to quenching by ATA (Fig. 4d). All residues mentioned for the ATA·Mg interaction in PMCA1 are conserved in PMCA4 (see legend of Fig. 6b).

The nucleotide-binding pocket of PMCA is less accessible to water in the E1Ca and E2 states than in the EP analogous states^14,27. These changes in the hydrophobicity affect the fluorescence of probes that bind in the nucleotide-binding pocket. The quantum yield of Eo is higher in hydrophobic environments; thus, its fluorescence is higher when it is bound to the E1Ca and E2 states than to the E2P and E1(Ca)P states. Since the molecular docking results propose that ATA·Mg binds in the nucleotide-binding pocket, we evaluated the fluorescence of ATA bound to PMCA in the states analogous to E2P and E1(Ca)P. To this end, we phosphorylated PMCA from ATP in the presence of Ca²⁺ and lanthanum to stabilize the E1(Ca)P state, or stabilized the E2P-analogous form in the absence of Ca²⁺ and the presence of BeF₃⁻ (see “Methods” section for details). Then, we determined the fluorescence of the PMCA⋅ATA complex. Similar to that occurs with Eo^14,27, the fluorescence of ATA bound to PMCA in the E1Ca and E2 states was similar, while when the pump was stabilized in the states analogous to E2P and E1(Ca)P, the fluorescence decreased (Fig. 6d). Consistent with the site proposed by flexible molecular docking (Fig. 6b), these results show that ATA would be sensitive to changes in the nucleotide-binding pocket of PMCA.

The binding of ATP and ATA to PMCA

While the flexible molecular docking results indicate a partial overlap between the ATA·Mg-binding site and the ATP-binding site, experimental evidence suggests that this interaction does not follow a competitive inhibition mechanism, i.e. ATA and ATP bind simultaneously to PMCA. To test this, we assayed whether ATA·Mg and ATP could bind both to the PMCA nucleotide-binding pocket. Figure 6e shows the molecular docking of ATA·Mg to the PMCA·ATP complex in the most probable conformation (Fig. 6a). The ten predicted conformations are feasible and are near the ATP-binding site, but the binding energy of ATA·Mg increases from − 9.8 to − 8.7 kcal/mol. In the presence of ATP, ATA·Mg maintains its interaction with Thr477 and interacts with Glu605 through the Mg²⁺ ion. It is further stabilized by hydrogen bonds with Lys476 and Asn711 (part of the 708TGDN motif) and by a salt bridge with Arg582. Notably, Arg582 is homologous to Arg489 in SERCA and stabilizes the γ-phosphate of the ATP analog adenosine 5′-[α,β-methylene]diphosphate (AMPPCP) in the E2-AMPPCP state³¹. Figure 6f shows the molecular docking of ATP to the PMCA·ATA·Mg complex in the most probable conformation (Fig. 6b). Only four predicted conformations are feasible and the binding energy of ATA·Mg increases from − 9.3 to − 8.4 kcal/mol in the most probable position. ATP binds near the ATA·Mg, where it is stabilized by hydrogen bonds with Asn711, Asn713, Thr714, and Glu605, among other residues. Although ATA·Mg is not coordinated by residues of the 795TGDXND motif, the γ-phosphate of ATP loses interaction with these residues and is coordinated by others in the nucleotide-binding pocket. Furthermore, in the PMCA·ATA·Mg complex, ATP would lose interaction with Lys773 and the 475DKTG motif. It is important to note that in these experiments the molecular docking of ATA·Mg or ATP on the PMCA-ATP or PMCA·ATA·Mg complex is conducted, respectively. Consequently, the ligand already bound within the complex doesn’t exhibit free movement, as would be the case, given the reversible binding nature of both molecules to PMCA. Thus, the primary objective of these experiments is to assess the compatibility of both ligands within the nucleotide-binding pocket. However, the interaction with different residues must be analyzed with prudence.

These findings suggest that while the ATA·Mg binding site partially overlaps with the ATP binding site, both ligands can accommodate within the nucleotide-binding pocket, albeit with lower binding affinity compared to free PMCA.