The extracellular fluid of animal cells has a salt concentration similar to that of seawater. However, cells must control their intracellular salt concentrations to facilitate specific processes, such as signal transduction and action potential propagation, and prevent unfavorable interactions with high concentrations of ions such as Ca²⁺. For instance, most animal cells contain a high concentration of K⁺ and a low concentration of Na⁺ relative to the external medium. These ionic gradients are generated by a specific transport system, an enzyme that is called the Na⁺–K⁺ pump or the Na⁺–K⁺ ATPase. The hydrolysis of ATP by the pump provides the energy needed for the active transport of Na⁺ out of the cell and K⁺ into the cell, generating the gradients. The pump is called the Na⁺–K⁺ ATPase because the hydrolysis of ATP takes place only when Na⁺ and K⁺ are present. This ATPase, like all such enzymes, requires Mg²⁺.

The change in free energy accompanying the transport of Na⁺ and K⁺ can be calculated. Suppose that the concentrations of Na⁺ outside and inside the cell are 143 and 14 mM, respectively, and the corresponding values for K⁺ are 4 and 157 mM. At a membrane potential of −50 mV and a temperature of 37°C, we can use the equation in the previous section to determine that the free-energy change in transporting 3 mol of Na⁺ out of the cell and 2 mol of K⁺ into the cell is 3(5.99) + 2(9.46) = +36.9 kJ mol⁻¹ (+8.8 kcal mol⁻¹). Under typical cellular conditions, the hydrolysis of a single ATP molecule per transport cycle provides sufficient free energy, about −50 kJ mol⁻¹ (−12 kcal mol⁻¹) to drive the uphill transport of these ions. The active transport of Na⁺ and K⁺ is of great physiological significance. Indeed, more than a third of the ATP consumed by a resting animal is used to pump these ions. The Na⁺–K⁺ gradient in animal cells controls cell volume, renders neurons and muscle cells electrically excitable, and drives the active transport of sugars and amino acids.

The purification of other ion pumps has revealed a large family of evolutionarily related ion pumps including proteins from bacteria, archaea, and all eukaryotes. Each of these pumps is specific for a particular ion or set of ions. Two are of particular interest: the sarcoplasmic reticulum Ca²⁺ ATPase (or SERCA) transports Ca²⁺ out of the cytoplasm and into the sarcoplasmic reticulum of muscle cells, and the gastric H⁺–K⁺ ATPase is the enzyme responsible for pumping sufficient protons into the stomach to lower the pH to 1.0. These enzymes and the hundreds of known homologs, including the Na⁺–K⁺ ATPase, are referred to as P-type ATPases because they form a key phosphorylated intermediate. In the formation of this intermediate, a phosphoryl group from ATP is linked to the side chain of a specific conserved aspartate residue in the ATPase to form phosphorylaspartate.

Membrane pumps function by mechanisms that are simple in principle but often complex in detail. Fundamentally, each pump protein can exist in two principal conformational states, one with ion-binding sites open to one side of the membrane and the other with ion-binding sites open to the other side (Figure 13.2). To pump ions in a single direction across a membrane, the free energy of ATP hydrolysis must be coupled to the interconversion between these conformational states.

371

We will consider the structural and mechanistic features of P-type ATPases by examining SERCA. The properties of this P-type ATPase have been established in great detail by relying on crystal structures of the pump in five different states. This enzyme, which constitutes 80% of the protein in the sarcoplasmic reticulum membrane, plays an important role in relaxation of contracted muscle. Muscle contraction is triggered by an abrupt rise in the cytoplasmic calcium ion level. Subsequent muscle relaxation depends on the rapid removal of Ca²⁺ from the cytoplasm into the sarcoplasmic reticulum, a specialized compartment for Ca²⁺ storage, by SERCA. This pump maintains a Ca²⁺ concentration of approximately 0.1 μM in the cytoplasm compared with 1.5 mM in the sarcoplasmic reticulum.

The first structure of SERCA to be determined had Ca²⁺ bound, but no nucleotides present (Figure 13.3). SERCA is a single 110-kDa polypeptide with a transmembrane domain consisting of 10 α helices. The transmembrane domain includes sites for binding two calcium ions. Each calcium ion is coordinated to seven oxygen atoms coming from a combination of side-chain glutamate, aspartate, threonine, and asparagine residues, backbone carbonyl groups, and water molecules. A large cytoplasmic headpiece constitutes nearly half the molecular weight of the protein and consists of three distinct domains, each with a distinct function. One domain (N) binds the ATP nucleotide, another (P) accepts the phosphoryl group on a conserved aspartate residue, and the third (A) serves as an actuator, linking changes in the N and P domains to the transmembrane part of the enzyme.

372

SERCA is a remarkably dynamic protein. For example, the structure of SERCA without bound Ca²⁺, but with a phosphorylaspartate analog present in the P domain, is shown in Figure 13.4. The N and P domains are now closed around the phosphorylaspartate analog, and the A domain has rotated substantially relative to its position in SERCA with Ca²⁺ bound and without the phosphoryl analog. Furthermore, the transmembrane part of the enzyme has rearranged significantly and the well-organized Ca²⁺-binding sites are disrupted. These sites are now accessible from the side of the membrane opposite the N, P, and A domains.

The structural results can be combined with other studies to construct a detailed mechanism for Ca²⁺ pumping by SERCA (Figure 13.5):

1. The catalytic cycle begins with the enzyme in its unphosphorylated state with two calcium ions bound. We will refer to the overall enzyme confomation in this state as E₁; with Ca²⁺ bound, it is E₁-(Ca²⁺)₂. In this conformation, SERCA can bind calcium ions only on the cytoplasmic side of the membrane. This conformation is shown in Figure 13.3.

2. In the E₁ conformation, the enzyme can bind ATP. The N, P, and A domains undergo substantial rearrangement as they close around the bound ATP, but there is no substantial conformational change in the transmembrane domain. The calcium ions are now trapped inside the enzyme.

3. The phosphoryl group is then transferred from ATP to Asp 351.

4. Upon ADP release, the enzyme again changes its overall conformation, including the membrane domain this time. This new conformation is referred to as E₂ or E₂-P in its phosphorylated form. The process of interconverting the E₁ and E₂ conformations is sometimes referred to as eversion.

373

In the E₂-P conformation, the Ca²⁺-binding sites become disrupted and the calcium ions are released to the side of the membrane opposite that at which they entered; ion transport has been achieved. This conformation is shown in Figure 13.4.

5. The phosphorylaspartate residue is hydrolyzed to release inorganic phosphate.

6. With the release of phosphate, the interactions stabilizing the E₂ conformation are lost, and the enzyme everts to the E₁ conformation.

The binding of two calcium ions from the cytoplasmic side of the membrane completes the cycle.

This mechanism likely applies to other P-type ATPases. For example, Na⁺–K⁺ ATPase is an α₂β₂ tetramer. Its α subunit is homologous to SERCA and includes a key aspartate residue analogous to Asp 351. The β subunit does not directly take part in ion transport. A mechanism analogous to that shown in Figure 13.5 applies, with three Na⁺ ions binding from the inside of the cell to the E₁ conformation and two K⁺ ions binding from outside the cell to the E₂ conformation.

Certain steroids derived from plants are potent inhibitors (K_i ≈ 10 nM) of the Na⁺–K⁺ pump. Digitoxigenin and ouabain are members of this class of inhibitors, which are known as cardiotonic steroids because of their strong effects on the heart (Figure 13.6). These compounds inhibit the dephosphorylation of the E₂-P form of the ATPase when applied on the extracellular face of the membrane.

Digitalis is a mixture of cardiotonic steroids derived from the dried leaf of the foxglove plant (Digitalis purpurea). The compound increases the force of contraction of heart muscle and is consequently a choice drug in the treatment of congestive heart failure. Inhibition of the Na⁺–K⁺ pump by digitalis leads to a higher level of Na⁺ inside the cell. The diminished Na⁺ gradient results in slower extrusion of Ca²⁺ by the sodium–calcium exchanger, an antiporter (Section 13.3). The subsequent increase in the intracellular level of Ca²⁺ enhances the ability of cardiac muscle to contract. It is interesting to note that digitalis was used effectively long before the discovery of the Na⁺–K⁺ ATPase. In 1785, William Withering, a British physician, heard tales of an elderly woman, known as “the old woman of Shropshire,” who cured people of “dropsy” (which today would be recognized as congestive heart failure) with an extract of foxglove. Withering conducted the first scientific study of the effects of foxglove on congestive heart failure and documented its effectiveness.

374

Analysis of the complete yeast genome revealed the presence of 16 proteins that clearly belong to the P-type ATPase family. More-detailed sequence analysis suggests that 2 of these proteins transport H⁺ ions, 2 transport Ca²⁺, 3 transport Na⁺, and 2 transport metals such as Cu²⁺. In addition, 5 members of this family appear to participate in the transport of phospholipids with amino acid head groups. These 5 proteins help maintain membrane asymmetry by transporting lipids such as phosphatidylserine from the outer to the inner leaflet of the bilayer membrane. Such enzymes have been termed “flippases.” Remarkably, the human genome encodes 70 P-type ATPases. All members of this protein family employ the same fundamental mechanism: the free energy of ATP hydrolysis drives membrane transport by means of conformational changes, which are induced by the addition and removal of a phosphoryl group at an analogous aspartate site in each protein.

Studies of human disease revealed another large and important family of active-transport proteins, with structures and mechanisms quite different from those of the P-type ATPase family. These pumps were identified from studies on tumor cells in culture that developed resistance to drugs that had been initially quite toxic to the cells. Remarkably, the development of resistance to one drug had made the cells less sensitive to a range of other compounds. This phenomenon is known as multidrug resistance. In a significant discovery, the onset of multidrug resistance was found to correlate with the expression and activity of a membrane protein with an apparent molecular mass of 170 kDa. This protein acts as an ATP-dependent pump that extrudes a wide range of small molecules from cells that express it. The protein is called the multidrug-resistance (MDR) protein or P-glycoprotein (“glyco” because it includes a carbohydrate moiety). Thus, when cells are exposed to a drug, the MDR pumps the drug out of the cell before the drug can exert its effects.

Analysis of the amino acid sequences of MDR and homologous proteins revealed a common architecture (Figure 13.7A). Each protein comprises four domains: two membrane-spanning domains and two ATP-binding domains. The ATP-binding domains of these proteins are called ATP-binding cassettes (ABCs) and are homologous to domains in a large family of transport proteins in bacteria and archaea. Transporters that include these domains are called ABC transporters. With 79 members, the ABC transporters are the largest single family identified in the E. coli genome. The human genome includes more than 150 ABC transporter genes.

The ABC proteins are members of the P-loop NTPase superfamily (Section 9.4). The three-dimensional structures of several members of the ABC transporter family have now been determined, including that of the bacterial lipid transporter MsbA. In contrast with the eukaryotic MDR protein, this protein is a dimer of 62-kDa chains: the amino-terminal half of each protein contains the membrane-spanning domain, and the carboxyl-terminal half contains the ATP-binding cassette (Figure 13.7B). Prokaryotic ABC proteins are often made up of multiple subunits, such as a dimer of identical chains, as above, or as a heterotetramer of two membrane-spanning domain subunits and two ATP-binding-cassette subunits. The consolidation of the enzymatic activities of several polypeptide chains in prokaryotes to a single chain in eukaryotes is a theme that we will see again. The two ATP-binding cassettes are in contact, but they do not interact strongly in the absence of bound ATP (Figure 13.8). On the basis of this structure and others, as well as on other experiments, a mechanism for active transport by these proteins has been developed (Figure 13.9):

375

1. The catalytic cycle begins with the transporter free of both ATP and substrate. While the distance between the ATP-binding cassettes in this form may vary with the individual transporter, the substrate binding region of the transporter faces inward.

2. Substrate enters the central cavity of the transporter from inside the cell. Substrate binding induces conformational changes in the ATP-binding cassettes that increase their affinity for ATP.

3. ATP binds to the ATP-binding cassettes, changing their conformations so that the two domains interact strongly with one another. The close interaction of the ABCs reorients the transmembrane helices such that the substrate binding site is now facing outside the cell (Figure 13.8, far right).

376

4. The outward facing conformation of the transporter has reduced affinity for the substrate, enabling the release of the substrate on the opposite face of the membrane.

5. The hydrolysis of ATP and the release of ADP and inorganic phosphate reset the transporter for another cycle.

Whereas eukaryotic ABC transporters generally act to export molecules from inside the cell, prokaryotic ABC transporters often act to import specific molecules from outside the cell. A specific binding protein acts in concert with the bacterial ABC transporter, delivering the substrate to the transporter and stimulating ATP hydrolysis inside the cell. These binding proteins are present in the periplasm, the compartment between the two membranes that surround some bacterial cells (Figure 12.34A).

Thus, ABC transporters use a substantially different mechanism from the P-type ATPases to couple the ATP hydrolysis reaction to conformational changes. Nonetheless, the net result is the same: the transporters are converted from one conformation capable of binding substrate from one side of the membrane to another that releases the substrate on the other side.