The Nucleotide, Inhibitor, and Cation Binding Sites of P-type II ATPases


Corresponding author: G. Narahari Sastry,


P-type ATPases constitute a ubiquitous superfamily of cation transport enzymes, responsible for carrying out actions of paramount importance in biology such as ion transport and expulsion of toxic ions from cells. The harmonized toggling of gates in the extra- and intracellular domains explain the phenomenon of specific cation binding in selective physiological states. A quantitative understanding of the fundamental aspects of ion transport mechanism and regulation of P-type ATPases requires detailed knowledge of thermodynamical, structural, and functional properties. Computational studies have made significant contributions to our understanding of biological ion pumps. Various 3D structures of Ca2+-ATPase between E1 and E2 transition states have given a impetus to the theorists to work on the Na+K+- and H+K+-ATPase to address important questions about their function. The current review delineates the importance of cation, nucleotide, and inhibitor binding domains, with a focus on the therapeutic potential and biological relevance of the three P-type II ATPases. This will give an insight into the ion selectivity and their conduction across the transmembrane helices of P-type II ATPases, which may pave the way to a range of fundamental questions about the mechanism and aid in the efforts of structure- and analog-based drug design.

P-type ATPases are integral transmembrane (TM) proteins, which constitute a ubiquitous superfamily of ion transport enzymes. P-type ATPases transport specific cations across the cell membranes, and the direction of transport (inward, outward, or antiport) differs from one to the other. They have a range of about 6–12 transmembrane helices (TMH). Most of the P-type ATPases have 10 TMH. The high degree of similarity in sequence, function, and quaternary structure is responsible for the presence of similar transport mechanisms in P-type ATPases. Based on their phylogenetic analysis and function, the P-type ATPases are divided into five major subfamilies with unique substrates and subfamily-specific sequence motif (1,2) (Table 1). They have diverse functional roles such as regulation of action potentials in nervous tissues, secretion and reabsorption of solutes in the kidneys, secretion of acid in the stomach, absorption of nutrients in intestine, relaxation of muscles, Ca2+-dependent signal transduction, etc. The malfunctioning of pumps leads to many major diseases such as cardiovascular diseases, metabolic diseases, neurodegenerative diseases, cancer and are therefore considered promising drug targets in medical research (3,4).

Table 1.   The subfamily classification of the P-Type ATPases, the type of cations transported, and the number of transmembrane helices (1,2)
S. No.SubfamilyOrganismTransported speciesNo. of TM helices
  1. aThe type of metal ions transported is not yet unambiguously established.

1Type IAProkaryotesK+6 or 7
2Type IBProkaryotes and eukaryotesCu+, Ag+, Cd2+, Zn2+, Pb2+, Co2+8
3Type IIAProkaryotes and eukaryotesCa2+, Mn2+; SERCA10
4Type IIBProkaryotes and eukaryotesCa2+; PMCA10
5Type IICProkaryotes and eukaryotesNa+K+- and H+K+-ATPase10
6Type IIDEukaryotes (not in human)Na+a Ca2+a10
7Type IIIAProkaryotes and eukaryotes (plant and fungi)H+10
8Type IIIBProkaryotesMg2+10
9Type IVEukaryotesPhospholipidsa10 or 12
10Type VEukaryotesNo assigned specificity10 or 12

Mutations are observed to abolish the functioning completely or phenomenally alter the activity of ATPases. Thus, it is of wider interest to understand and explore the structural details of these processes and characterize the inhibitors that block the ion pump in one or the other state in order to comprehend the details of the mechanism of ion pumping (Table 2). P-type ATPases are macromolecules with ‘N’ (nucleotide binding), ‘P’ (phosphorylating), and ‘A’ (actuator) moving domains, and the transitions of the three domains are expected to be playing an important role in the flux of cations across the membrane against the electrochemical gradient(1–8). The dynamic relationship between the cation binding and their respective effect on the P-type ATPase is an area of profound scientific interest. Their structures and mechanism of ion translocation varies considerably since they are present in a diverse variety of biological membranes. The signature sequence of P-type ATPase family is a seven-amino-acid motif D-K-T-G-T-[LIVM]-[TIS] containing the important aspartate, which gets phosphorylated during the transport of cation. Figure 1 elucidates the ion transport mechanism and the role of each physiological state of Ca2+ ATPase (Sarco/endoplasmic reticulum calcium ATPase or SERCA) with a repertoire of crystal structures. Initially, two calcium ions bind to the calcium binding site (transmembrane domain) from the cytoplasmic side in E1 state. The cation binding site faces the cytoplasm in the E1 and the lumen or extracellular medium in E2. They have high affinity for the Ca2+ ions transferred from the cytoplasm in E1 state but low affinity in E2. Ca2+ binding is followed by the binding of ATP to the N-domain. The binding of both Ca2+ and ATP triggers a 30° conformational shift of the A-domain around the axis parallel to the membrane which is stabilized by non-covalent interactions between the interfaces of P- and A-domains. The γ-phosphate of ATP is then transferred to the aspartate (D351) of the signature sequence in the P-domain, forming the functionally important aspartyl phosphoanhydride intermediate (E1P-ADP). Active transport of cations takes place only after the formation of a covalent E1P-ADP intermediate. The phosphoryl transfer is followed by an 110° rotation of A-domain around the perpendicular axis to the membrane leading to close proximity between A-domain and P-domain. The TGES motif, located in a conformationally flexible loop of A-domain, is known to play a critical role in the formation of the E2P state and subsequent hydrolysis of the aspartyl phosphate. This autophosphorylation is the characteristic feature of P-type ATPases. The occluded Ca2+ ions in E1P-ADP state are released into the lumen or extracellular space during the transition to E2P state. At this juncture, the ADP sensitivity, that is, the ability of the phosphoenzyme to synthesize ATP from ADP, is lost. In the next step of cycle, as the Ca2+ released to the luminal side, the cation binding sites become very negatively charged. These negative charges of the cation binding sites are neutralized by the protonation of the acidic co-ordinating residues. The exchange of Ca2+ with H+ is favorable for the transmembrane domain to come closure. The TM domain closeness is coupled to the rearrangement of the A- and P-domains and responsible for the E2-P occluded state. The dephosphorylation of the aspartyl phosphate intermediate is assisted by the A-domain. Hydrolysis of phosphate and its release from the enzyme completes the cycle (5–7). The phosphorylation and dephosphorylation of the aspartate present in the P-domain of P-type ATPase follow the bimolecular nucleophilic substitution (SN2) reaction (8). Now, the cytoplasmic side of the TM domain opens to exchange the protons for two new Ca2+ ions.

Table 2.   The PDB ID, function, regulation, and their involvement in diseases of three known P-type II ATPases
 Ca2+-ATPase (α)Na+K+-ATPase (α, β, γ)H+K+-ATPase (α, β)
  1. aProton binding site correspond to that of H3O+ is shown.

  2. bResidues that are underlined represents back bone oxygen atom involved in cation binding.

  3. cTwo side-chain oxygens involved in cation binding.

PDB ID1su4, 2c9m, 1t5s, 1vfp, 3ba6, 1t5t, 1wpe, 2zbd, 2oa0, 2dqs, 2c8k, 2by4, 2c88, 2zbe, 2zbf, 2zbg, 3b9b, 3b9r, 1xp5, 2o9j, 1wpg, 3fgo, 2eas, 2eau, 1iwo, 2ear, 2c8l, 2eat, 2agv, 1kju, 2yfy, 3b9b, 3n5k, 3n8g, 3nal, 3nam, 3nan, 3fpb, 3fps, 3ar2, 3ar3, 3ar4, 3ar5, 3ar6, 3ar7, 3ar8, 3ar93n23, 3kdp, 3n2f, 2zxe, 3a3y, 3b8e, 1mo7, 1mo82xzb, 3ixz, 1iwc
FunctionsATP coupled exchange of Ca2+ from the cytosol to the SR lumen. Contributes to calcium sequestration involved in muscular excitation/contractionATP coupled exchange of Na+ and K+ ions across the plasma membrane. Electrical excitability of nerve and muscle tissues, secretion and reabsorption of solutes in the kidneys, isoform α4 might be required for the sperm motilityATP coupled exchange of H+ and K+ ions across the plasma membrane. Responsible for the acidification of the stomach
RegulationPhospholamban (PLN), Sarcolipin, Anti-apoptotic protein Bcl-2, OxytocinPhospholemman (FXYD1), γ-subunit (FXYD2), and other cell specific FXYD proteinsGastrin, Histamine
TargetsSkin disease, Hearing loss, Infertility, Cardiomyopathy, Brody disease, Darier–White disease, Segmental Darier, Acantholytic dyskeratotic epidermal nevi, Hopf disease, prostate cancer and infectious diseasesMcArdle disease, Natriuresis, Autosomal dominant polycystic kidney disease, Inflammatory bowel disease, Congestive heart diseaseGastric ulcers, Duodenal ulcers, Erosive esophagitis, GERD, Zollinger–Ellison syndrome
Cation binding sitesE1Ca2+ Site I: N768, E771, T799, D800, E908E1Na+ Site I: N783, E786, T814, D815, Q930E1H+ Site Ia: N792, E795, Q939
Ca2+ Site II: V304b, A305b, I307b, E309, N796, D800Na+ Site II: V329b, A330b, V332b, E334b, D811, D815H+ Site IIa: V338b, A339b, V341b, E820, D824
 Na+ Site III: Y778, G813b,T814, E961E2K+ Site I: K791, E795, E820, T823, D824, Q939
 E2K+ Site I: S782, E786, D811, D815, Q930K+ Site II: V338b, V341b, E343c, E795, E820
 K+ Site II: V329b, A330b, V332b, N783, E786, D811  
Figure 1.

 Different physiological states of SERCA ion antitransport (Ca2+/H+) mechanism. Active transport of cations is described in six steps: 1. Two Ca2+ bind to the transmembrane from the cytoplasmic side; 2. ATP binds to the N-domain; 3. Formation of a covalent aspartyl phosphoanhydride intermediate (γ-phosphate of the ATP transfer to the D351); 4. Ca2+/H+ exchange (Ca2+ transport to the lumen and acidic residues of cation binding sites get protonated) and ADP release; 5. Dephosphorylation of D351; 6. H+ is transported to the cytoplasmic side. Na+K+-ATPase and H+K+-ATPase are also expected to follow similar mechanism.

There are several crystal structures of different E1 to E2 conformations of the Ca2+-ATPase (9), one or two snapshots of E2 conformation of Na+K+-ATPase (10), and H+K+-ATPase (11) have been elucidated recently. Discrepancy in the solved and unsolved structure is attributed to the relative difficulty of studying membrane proteins. Since they are hydrophobic, they do not easily dissolve and are difficult to crystallize. Thus, the conventional methods of structure determination, solution NMR and X-ray crystallography, are not easily applied to membrane proteins. For this reason, theoretical and mathematical methods to predict transmembrane protein topologies offer an attractive alternative.

Till date, SERCA was used to build a homology model of H+K+-ATPase and Na+K+-ATPase to get the structural and functional details. Although the crystal structures of SERCA are significant for explanatory and predictive purposes, they cannot explain all features of other pumps; thus, a variety of different and complementary approaches are required. A wide variety of computational approaches, such as homology modeling, docking, quantum mechanics/molecular mechanics (QM/MM) molecular dynamics (MD) simulations, continuum electrostatic Poisson–Boltzmann theory, Poisson–Nernst–Planck electrodiffusion theory and kinetic rate models, etc. have been assisted to refine our understanding of P-type ATPases (12–14).

Noteworthy progress has been made in carefully constructing rational models of a number of intricate ion pumps with the help of homology modeling for the characterization of cation binding sites, which corroborated well with several experimental results and further substantiated the stoichiometry of the cations. The following sections describe the structural and functional aspects of the cation, nucleotide, and inhibitor binding sites of P-type II ATPase.

Cation Binding Sites

The binding sites for the biological cations in proteins have to be geometrically selective to facilitate the discrimination of cations, as they are the smallest entity in the body and have similar size of about Ca2+ (0.99 Å), Na+ (1.16 Å), K+ (1.35 Å), H+ (0.01 Å) radius, which is a most important factor in ionic-molecular interactions. The crystal structures of SERCA provide atomic models for two high-affinity Ca2+ binding sites located approximately in the center of the bilayer and between TM segments TM4, TM5, TM6, and TM8 in the E1 (2Ca2+) conformation and are about 5.7Å apart (Figure 2). According to Obara et al., (6) number of protons countertransport by Ca2+-ATPase is pH reliant and solely depends on the side (i.e., cytoplasmic or luminal) from which protons are introduced into the binding site. The mutational studies of E340A, T247K, and R822L have strikingly shown reduction in the Ca2+ binding transition (E2→ E1→ 2Ca2+·E1) and Ca2+ dissociation as well from site II back toward the cytosol but do not affect the apparent affinity for vanadate (inline image ). Further mutant D813L inhibited the Ca2+ binding transition, but not Ca2+ dissociation, and increased the apparent affinity for vanadate ion (15). The P312A mutant slows down the transition of 2Ca2+·E1P→E2P, assumed that this mutation stabilizes the Ca2+E1P conformation (16). Flipping or gating mechanism of E309 side chain plays a very important role to maintain the geometries of cation binding site in both E1 and E2 conformations. The E309Q or N796A mutant reduces the affinity of Ca2+, thereby limiting the binding of Ca2+ from site II but still retaining the high affinity of Ca2+ binding in site I (17). Side chain of E309, which points inward in the E2.BHQ, forms H-bonds with the carbonyl group of N796 and V304, but H-bond with the carbonyl of V304 requires protonation of E309. Therefore, the N796D mutation abolishes activity apparently by destroying site II. The side chain of N796 is stabilized by N101 (M2) through a water molecule. A significant observation was made with the flip of E309 peptide bond (the difference in ψ angle of E309 is 168°) from the β-strand conformation of E1 (2Ca2+) to the helix conformation in E2 (TG+BHQ) state. The side-chain amide of site II N796 protonates both the E771 and E309 to facilitate countertransport of H+; hence, this asparagine is critical and is changed to aspartate in Na+K+-ATPase and H+K+-ATPase that countertransport K+ instead of H+ (6). The continuum electrostatic calculations have shown that H-bond is formed by protonation of E58 and E908, which are likely to provide extra stability for the Ca2+ binding sites while reducing electrostatic repulsion between the clustered acidic residues. The four carboxyl groups of E771, D800, E309, and E908 clustered in the Ca2+ binding sites are found deleterious to structural integrity if left ionized. Thus, four residues are likely to be protonated in E2 (E771, D800, E309, and E908) and two in 2Ca2+·E1 (E309 and E908) (18). One significant finding is that K+ or Na+ binding sites are present on the surface of P-domain of SERCA. The binding site of K+ is lined up with three backbone carbonyl of residues L711, K712, and A714, and one side chain oxygen of E732 in the E1 conformation of Ca2+ ATPase. This K+ binding site is not conserved among P-type II ATPases but plays an important role in the function of SERCA. The mutational studies of E732Q and E732A have shown the stimulatory effect of K+ on dephosphorylation rate of SERCA (19).

Figure 2.

 The crystal structures of SERCA provide atomic models for two high-affinity Ca2+ binding sites located approximately in the center of the bilayer and between TM segments TM4, TM5, TM6, and TM8 in the E1 (2Ca2+) conformation Ca2+-ATPase.

The crystal structures of Na+K+-ATPase provide atomic models for three high-affinity Na+ (E1) and two K+ (E2) binding sites located approximately in the center of the bilayer and between TM segments TM4, TM5, TM6, TM8, and TM9 (Figure 3). It has been observed that two K+ are juxtaposed at a distance of 4.1 Å, and in contrast to SERCA, there is no bridging oxygen between these two metal centers. The third K+ binding site at cytoplasmic domain is responsible for the activation of dephosphorylation (20–22). Two water molecules are involved in H-bonding and van der Waals interaction with the P818 and T814 to maintain the octahedral geometry of cations in the cation binding sites. D334 (corresponding to the E309 of Ca2+-ATPase) serves as a gate for Na+ entry in the site I, and the additional proline residue in TM5 might assist the first Na+ to reach site III through site I. The octahedral co-ordination geometry of K+ in both sites are optimum, whereas the Na+ shows distorted octahedral geometry in all three sites. The mutational studies have shown that the size of side chain D334 plays a very important role in the occlusion of Na+ and K+ as D334QL mutations reduces the affinity of Na+ by two- to fourfold, whereas D334AE mutants are lethal (12). The β-subunit, a single spanning membrane protein of Na+K+-ATPase, affects the K+ binding and its transport by introducing structural change in the membrane domain. It also plays an important role in the targeting and stabilization of Na+K+-ATPase. Comparison of the structure of SERCA with β-subunit bound state of Na+K+-ATPase reveals countertransport of H+ instead of K+ in Ca2+-ATPase owing to the lack of β-subunit in Ca2+-ATPase. K+ binding depends on the unwinding of M7, which is stabilized by H-bonding β-subunit’s Y44 to carbonyl of G855 in Na+K+-ATPase, thus preventing its exposure to the hydrophobic core of the bilayer (20–22).

Figure 3.

 The crystal structures of E2 conformation of Na+K+-ATPase provides atomic models for two K+ binding sites located approximately in the center of the bilayer and between TM segments TM4, TM5, TM6.

The gastric proton pump H+K+-ATPase is considered to harbor two or three protons and two K+ binding cavities in the E1 and E2 conformations, respectively (23). The salt bridge between K791 (TM5) and E820 (TM6) is exclusively found in E2 conformation. Because of this salt bridge, there is space only for single K+ binding site, whereas it is absent in Na+K+ and Ca2+-ATPases as serine takes up the position of K791. The E820Q mutant disrupts the salt bridge and shows K+-independent ATPase activity, which means K+ does not exert any effect on the dephosphorylation of the phosphorylated intermediate. This indicates that E820Q mutant has lost K+ sensitivity and indirectly its E1 preference, which leads to the constitutive activation of H+K+-ATPase (24). According to our previous investigations, the transport of proton by the gastric proton pump probably requires the transported species to be H3O+ rather than H+ in E1 conformation (25). The single mutants D826A, I827A, L833G, and L833A have abolished the K+ activity and the phosphorylation capacity located on the same side of the α-helix of the TM6 segment. But the mutants L833V and L833M retained above activity, which suggested that the bulkiness or length of the side chain of L833 is important for the expression as well as the phosphorylation of the H+K+-ATPase (26). The N-terminal region of the β-subunit of H+K+-ATPase directly contacts with the phosphorylation domain of the α-subunit, probably through the conserved SYGQ sequence. This prevents the reverse reaction from E2P to E1P conformation and is responsible for generating large gradient across the membrane in vivo conditions (27). Therefore, H+K+-ATPase generates highest H+ gradient (>106 folds).

Role of Arenes in the Transmembrane Region

Many membrane proteins are involved in the transport of ions and small molecules, and these are likely to undergo significant conformational changes within their transmembrane section as the protein moves from one state to another. Albeit the time spent in one conformational state is not preferred over another, it can be envisioned that the location of the transmembrane aromatic residues is optimized to prevent the formation of an energetic well during transitions between states (28). The process of aromatic localization may be a general method of directing and stabilizing structural changes during conformational transitions within the transmembrane region of P-type ATPases (29,30). The E1 and E2 states have surprisingly showed contrasting views of the location of the aromatic residues in ATPases. Sequence analysis of gastric H+K+-ATPase, Na+K+-ATPase and Ca2+-ATPase has shown that the number of aromatic amino acid residues in the proton pump is higher compared to that in the calcium pump (Table 3). Moreover, the difference is mainly observed in the number of phenylalanine (F) and tyrosine (Y) residues, which interestingly have the benzene moiety in the aromatic side chain (Figure 4). Thus, it may not be irrational to speculate that the excess of F and Y residues in the proton pump compared to that in the calcium pump may be necessary for the hydronium ion and K+ transport aided by the arenes. Moreover, the Na+K+-ATPase also contains a large number of F and Y residues as it involves K+ transport, which is also expected to be assisted by arene rings. Aromatic residues are evenly distributed along the TMH and contributed as driving force to the dramatic shift toward the external and internal interfacial regions during cation transportation at the central cavity section in the pumps (30).

Table 3.   Number of aromatic amino acid residues present in the transmembrane (TM) domain of H+K+-ATPase (human), Ca2+-ATPase (SERCA1a), and Na+ K+-ATPase (human)
TM 12 (W, F)3 (Y, F, W)5 (W, F, F, F, W)
TM 22 (F, W)4 (F, Y, Y, F)3 (F, Y, Y)
TM 32 (F, W)5 (H, F, F, F, F)6 (H, F, H, F, F, F)
TM 41 (F)3 (F, F, Y)1 (F)
TM 53 (F, Y, F)2 (Y, Y)2 (Y, F)
TM 61 (W)2 (F, F)0
TM 75 (W, F, F, Y, Y)5 (Y, Y, F, F, F)2 (Y, F)
TM 803 (Y, F, F)4 (H, F, F, W)
TM 94 (W, H, F, Y)3 (F, F, Y)4 (F, F, F, Y)
TM 101 (W)5 (W, W, Y, F, Y)7 (W, W, F, F, Y, F, Y)
Figure 4.

 Proposal for the disposition of aromatic residues along the gastric proton pump. The spheres are the Cα carbons of the residues. The TM helices TM5 and TM6 are represented by blue ribbon.

Insight into the Nucleotide Binding Site

Inhibitors of the ATP binding are the essential chemical probe for clarifying molecular properties of the ATP binding site in the enzymes. Paxilline, an indole alkaloid mycotoxin from Penicillium paxilli, affects the high-affinity Ca2+ binding form of the ATPase (E1). At lower concentrations, paxilline inhibits the ATP-dependent acceleration of Ca2+ release from the phosphoenzyme and/or phosphoenzyme decay. At higher concentrations, paxilline competitively inhibits phosphoenzyme formation without affecting nucleotide binding (31). Ca2+ occlusion in the transmembrane domain is followed by ATP binding in the highly conserved nucleotide binding pocket of N-domain. Toyoshima et al. have suggested that the binding of Ca2+ gives mobility to the N-domain, but the absence of Ca2+ fixes its movement. The binding of ATP in the presence of Ca2+ results in a large conformational change to bring the γ-phosphate of ATP near E351 and permitting covalent transfer of γ-phosphate to one of the side chain carboxyl oxygen of E351, but the binding of ATP to the enzyme does not require the presence Ca2+ and Mg2+ (5) (Figure 5). Further Mueller et al. (30) have clearly shown that the binding of ATP in the N-domain exerts substantial changes in the structure and movement of P-domain rather than Ca2+, which has slight effect. The occupancy of high-affinity calcium binding site increases the amplitude of the N-domain motion by giving the flexibility to A-domain (32). F487, R489, R560, R687, D627, T625, and K684 are found to be very critical residue for ATP binding. It is also found that the AMPPCP or ATP mainly adopt antiglucoside/adenine conformation in solution whereas ADP displays significant conformational flexibility (7,33).

Figure 5.

 The Mg2+-activated ATP phosphoryl transferase activity of Ca2+-ATPase. (A) Details of the interaction of Mg2+–ACP in the Ca2+ ATPase phosphorylation site, showing co-ordination of Mg2+ (pink) by D351. (B) The transition state of phosphoryl transfer, as mimicked by the ADP–AlF3–D351 complex, with binding of two Mg2+ (pink), one of which is co-ordinated by γ-phosphate, as in (A), and the other one by β-phosphate. Red: N-domain; Yellow: P-domain.

Earlier studies have shown that ATP analog such as Co(NH3)4ATP is a competitive inhibitor with respect to MnATP for Na+K+-ATPase. Ptilomycalin A, from the marine sponge Ptilocaulis spiculifer, is the first competitive inhibitor of ATP except for the ATP analogs and furylacryloylphosphate which inhibits brain Na+K+-ATPase or SERCA at μm range. This inhibition becomes uncompetitive and anticompetitive with respect to Na+ and K+, respectively, in Na+K+-ATPase. However, these inhibitors are uncompetitive with Ca2+ in SERCA (34). Anthraquinone dyes, such as Cibacron Blue 3G-A, mimic the adenine nucleotides and bind nonspecifically in the ATP binding site of Na+K+-ATPase (35). Recent crystallized structure of Na+K+-ATPase has revealed the presence of only one ATP binding site in the N-domain, which has low and high binding affinities at the E2 and E1 conformations, respectively. The loop between TM helices 4–5 (H4-H5 loop) containing P and bulky N-domains has only one ATP binding pocket (20–22). The ATP binding site, found in the N-domain (R378–R589), is separated from the P-domain (L354–N377 and A590–L773), where the phosphorylation site (D369) is located. ATP binds to the E2(K) conformation with low affinity and greatly accelerates its transition to E1 conformation, which has high affinity for ATP. ATP binding site within the H4–H5 loop lined with the D446, F475, K480, N482, K501, G502, F548, C549, and C577 provides positive charge to the ATP binding pocket of Na+K+-ATPase. Although the R423 and D472 are found outside nucleotide binding pocket, they play an important role to maintain the shape of binding pocket through the H-bonding among them. D443, S477, and P489 are also other important residues for the proper positioning of nucleotide in the binding site. R544Q decreases the affinity of both ATP and ATP binding, but R544K mutant does not affect the ATP binding but decreases the affinity of ADP by a factor of 15. A stacking interaction between π-electron systems of aromatic ring of F475 and the adenine ring of ATP, which are parallel at a distance of about 3 Å, is important for the stabilization of ATP within the binding pocket (36,37). Comparison of the affinity of ATP and ADP for Na+K+-ATPase and their pH-dependent dissociation rate constants suggests that the higher affinity of ATP is mainly due to a slower dissociation rate of ATP (38). ATP accelerates the K+ transport without undergoing any hydrolysis in the E2 conformation of Na+K+-ATPase and toggle to the E1 conformation. Clarke et al. (39) have found that the conformational transition within a protein dimer E2:E2→E2:E1→E1:E1 occurs as a two-step reversible process in the absence of bound ATP, which shows the allosteric effect of ATP during the structural transition.

The homology modeling and electron crystallography structure of H+K+-ATPase showed that the ATP binding site of the gastric pump is structurally homologous to the other ATPases in which adenosine group lined up with the residues F491, K496, and K517. R560 is responsible for the tilted structure of ATP, which makes contacts with the ribose ring in both ATPase. The guanidine group of R249, which corresponds to K205 residue of SERCA comes within reach of γ-phosphate of the ATP. It was also demonstrated that proton pump can also utilize trinucleotides (NTPs) other than ATP in the catalytic reaction cycle of phosphoryl transfer and form phosphoenzyme intermediate (40).

Inhibitors and their Mechanism of Action

A range of compounds were found to display effective P-type ATPase inhibition (Figure 6), while most of the compounds were obtained in hit-and-trial ways. Inhibitors bind to the E2 conformation of the P-type ATPase and are responsible for the blockage of the ion pathway from the luminal or extracellular side (41). The lack of a good 3D structure for H+K+- and Na+K+-ATPase is the major bottleneck and precludes structure-based drug design (42).

Figure 6.

 Selected inhibitors of P-type II ATPase.

The plasma membrane Ca2+-ATPase (PMCA) inhibitors can be classified into three categories: phenolic, nonphenolic compounds, and calmodulin antagonists. Various flavonoids and alkylphenols of the former category of inhibitors have been reported to inhibit the PMCA and SERCA of Ca2+-ATPase. The bis-phenolic compounds might be of key importance in revising the function of PMCA and regulation of intracellular Ca2+, which may have potential cardiovascular activities (43).

Despite many SERCA inhibitors, 2,5-di-(tert-butyl)-1,4,benzohydroquinone (BHQ), cyclopiazonic acid (CPA) and thapsigargin (TG) are found to be highly promising lead molecules and stabilize E2 conformations of SERCA. 1,3-dibromo-2,4,6-tri (methylisothiouronium)benzene (Br2-TITU) stabilizes an E1-P conformation. BHQ, commonly used as an antioxidant, is a high-affinity inhibitor of SERCA having a binding site at the groove, surrounded by TM3, TM5, and TM7 distinct from TG (Figure 7). In fact, F256 on M3 is critical for binding of TG but has only small effects for BHQ. BHQ blocks the conformational changes in cytoplasmic gate, E309 in the Ca2+ ion binding site by occupying the space that would be necessary for the side chain to point outwards, hence halting the pump in an H+ occluded state (6). Cyclopiazonic acid, a toxic indole tetrameric acid can be obtained from Penicillium cyclopium, Aspergillus veriscolor, and Aspergillus flavus. Laursen et al. have explored significantly the role of divalent cation Mn2+ in the binding of cyclopiazonic acid to the Ca2+-ATPase receptor, which will be of high importance in the new lead identification. The binding pocket of CPA is lined by the TM1, TM2, TM3, and TM4 (44). Thapsigargin is a cell-permeable tumor-promoter naturally occurring sesquiterpene lactone, which provokes programmed cell death in both proliferative and quiescent cells including androgen-independent prostate cancer cells (45). It is a very potent and highly selective inhibitor of the endoplasmic reticulum isoform of Ca2+-ATPase but has little or no effect on the Ca2+-ATPase of hepatocyte or erythrocyte plasma membrane or of cardiac or skeletal muscle sarcoplasmic reticulum (46). Compared to the cardiac SR Ca2+-ATPase, ER Ca2+-ATPase has an extended C-terminal tail, which might provide a site for potential physiological regulation (47).

Figure 7.

 Inhibitor binding site of Ca2+-ATPase: Thapsigargin (TG) and 2,5-di-(tert-butyl)-1,4,benzohydroquinone (BHQ) are having distinct binding site at the groove, surrounded by M3, M5, and M7 for TG and M1, M2, M3, and M4 for BHQ.

The Na+K+-ATPase of different tissues, for example nerve, kidney, heart, etc. play a fundamental role in numerous physiological processes requiring a stringent control on its activity. Cardiac glycosides such as digoxin and digitoxin (digitalis) are important inhibitors for the treatment for congestive heart diseases (48). The Na+K+-ATPase is an important characterized target for cardiotonic steroids (10). Ouabain is one of the known cardiac steroid, which binds to the transmembrane domain (TM1–TM6) at the extracellular interface (Figure 8). The lactone ring of the ouabain adopts a chair conformation, which is buried within the transmembrane domain and stabilized mainly by π-π or stacking interaction with the F783 and F786 (TM5). Additionally, side chains of P118, D121, N122 (TM2), and T797 (TM6) and carbonyl group of G319, V322, and A323 (TM4) also take part in the stabilization of the lactone ring. The sugar moiety of ouabain is facing the extracellular side and interacts with the Q111, E312, R880, and D884 residues. The orientation of the ouabain binding site residues differ significantly between the high- and low-affinity complexes owing to the rearrangement of the TMH.

Figure 8.

 Inhibitor binding site of Na+K+-ATPase: Ouabain binds to the TM1–TM6 transmembrane domain at the extracellular side.

The positively charged aromatic guanidinium and isothiouronium derivatives, Br-TITU and Br2-TITU used to probe cation binding sites, act as high-affinity competitive blockers of K+ (Rb+) and Na+ at the cytoplasmic surface, thus activating dephosphorylation. Br-TITU or Br2-TITU at pH 9 causes irreversible inactivation of renal Na+K+-ATPase activity (49). Macrocyclic carbon suboxide (MCS) factors are potent inorganic inhibitors, inhibiting not only Na+K+-ATPase but also several P-type ATPases sharing many properties with putative endogenous digitalis-like factors. Macrocyclic carbon suboxide factors interact with the cytoplasmic side but not with the extracellular ouabain binding site in contrast to the cardiac-steroid-like inhibitors of the Na+K+-ATPase (50). Inhibitor 2-methoxy-3,8,9-trihydroxy coumestan (PCALC36) is a non-steroidal skeleton, which inhibits the Na+K+-ATPase by a mechanism of action different from the cardiac glycosides, forming a very stable complex with Na+K+-ATPase and serve as a structural paradigm to develop new inotropic drugs. Vanadate is an effective inhibitor in ionic conditions promoting the E2 conformation, while the inhibitory effect of PCALC36 is equal in ionic conditions favoring either the E1 or E2 conformation (51). One interesting finding is that palytoxin, which binds to the Na+K+-APTase from the extracellular face, transforms the pump into a cation channel by allowing the conformation in which both gates are opened that is not present in the native state of pump (52).

Proton pump inhibitors (PPIs), categorized as reversible and irreversible, are the two classes of drugs used to treat the gastric ulcer by selectively blocking the variant of ATPases (53). The reversible or noncovalent inhibitors such as 1,2,3-trimethyl-8-(pentafluorophenylmethoxy) imidazole [1,2-α] pyridinium iodide (TMPFPIP), a derivative of the imidazole [1,2-α]pyridine belonging to SCH28080 series, are K+-competitive inhibitors and bind with high affinity to the E2 conformation of the H+K+-ATPase, which has a low-affinity binding site for K+. The modeled E2 conformation proposed a single concerted ligand binding site for TMPFPIP occupying the vestibule created by TM5-6 loop (L809, P810, L811, and C813), TM6 (I814, I816, and F818), and TM8 (Y925, T929, and F932) (18,54). SCH28080 binds to the region of luminal surface consisting of TM4 (A335), TM5-6 loop (L809), TM6 (C813), and TM8 (Y922 and I940) (55). Mutant E822D showed eight times lower sensitivity to SCH 28080, whereas the E822A showed a higher sensitivity than the wild-type enzyme, suggesting that E822 is the site involved in binding with SCH 28080. C813, the common binding site for all PPIs, is exposed at the luminal surface of the pump (Figure 9). The additional cysteine, C822, which is bound by pantoprazole, is predicted to be buried in the transport domain of the gastric pump (18). Hinesol, a major component of the crude drug ‘So-jutsu’ (Atractylodis Lanceae Rhizoma) is an antigastric ulcer agent, which is more effective in selective inhibition of H+K+-ATPase E1 state, compared to Na+K+-ATPase, Mg2+-ATPase, Ca2+-ATPase, and H+-ATPase (56).

Figure 9.

 Inhibitor binding site of H+K+-ATPase: C813, the common binding site for all proton pump inhibitors, is exposed at the luminal surface of the pump.

The irreversible (covalent) class of H+K+-ATPase inhibitors comprises of substituted benzamidazoles such as lansoprazole, pantoprazole, rabeprazole and RO-18-5364 compete with omeprazole. Irreversible PPIs form a disulfide bond with C813, which is accessible from the luminal surface of E2 conformation and allows allocation of the binding site to a luminal vestibule adjacent to C813 enclosed by part of TM4 and TM5–TM6 loop. Omeprazole is an acid-accumulated and acid-activated prodrug that binds covalently to two cysteine residues at positions 813 (or 822) and 892, accessible from the acidic face of the pump, whereas lansoprazole binds to C321, C813 (or C822), and C892. Unlike other irreversible inhibitors that bind to cysteine 813 of TM5, pantoprazole binds to C822 deeper in the membrane domain of TM6 (57,58). In direct comparisons of lansoprazole, rabeprazole and pantoprazole versus omeprazole, symptom relief and healing rates are similar for gastric and duodenal ulcers and moderate-to-severe gastroesophageal reflux disease (GERD) (59).

SK&F97574 [3-Butryl-4-(2-methylamino)-8-(2-hydroxyethoxy) quinoline] is also a potent, reversible inhibitor of the gastric H+K+-ATPase. Duration of action of SK&F 97574 is longer than that of the histamine H2 receptor (H2RAs) antagonists such as cimetidine but shorter than that of covalent H+K+-inhibitors such as omeprazole (52,59). Consequent covalent inhibition of H+K+-ATPase blocks the final step of acid secretion; hence, the PPIs omeprazole, lansoprazole, and pantoprazole are more effective than H2RAs antagonists in controlling acid secretion. In humans, the half-life of the inhibitory effect on acid secretion is approximately 28 h for omeprazole and approximately 46 h for pantoprazole. An extension of PPI plasma half-lives is an obvious goal, possibly via exploitation of probable differences in the metabolism of the two enantiomers present in current PPI formulations: clinical data on the S-enantiomer of omeprazole (esomeprazole) suggest some improvement in acid control. An alternative is to generate a prodrug of a PPI; plasma levels of the PPI would thus depend on release of the active metabolite from the pro-drug, again extending drug half-life. Another area of active investigation is the development of acid-pump antagonists to inhibit acid secretion at its final step. Drugs that are widely used for treatment for acid-related diseases are either substituted pyridylmethylsulfinyl benzimidazole or imidazopyridine derivatives. They are all prodrugs that inhibit the acid-secreting gastric H+K+-ATPase by acid activation to reactive thiophiles that form disulfide bonds with one or more cysteine residues accessible from the exoplasmic surface of the enzyme. Proton pump inhibitors are metabolized extensively via the CYP450 enzyme system (60).

Summary and Outlook

Ion transports across membrane mediated by or against electrochemical gradient are processes of fundamental biophysical and biochemical importance. The recent advances in the structural characterization of SERCA pump in different physiological states triggered the enthusiasm to uncover the mechanistic details of ion antitransport process in P-type ATPases. The computational, mutational, X-ray or other structural characterization provides valuable insight in understanding the variation in cation, inhibitor, and nucleotide binding sites in different physiological states and their functional consequences. Inhibitors that block the ion pump in a defined state appear to have a huge therapeutic relevance.


M.C. thanks CSIR, New Delhi, for financial assistance. Financial support from Department of Science and Technology, New Delhi, in the form of Swarnajayanthi fellowship to G.N.S is acknowledged. The support of DAE-BRNS is also acknowledged.

End Box

Dr. G. Narahari Sastry*
Molecular Modeling Group, Indian Institute of Chemical Technology,
Hyderabad 500 607, India
Ph: +91-40-2719-1253
Fax: +91-40-27160512

Mukesh Chourasia
Molecular Modeling Group, Indian Institute of Chemical Technology,
Hyderabad 500 607, India