Medical Physiology, 3rd Edition

Solute Transport Across Cell Membranes

In passive, noncoupled transport across a permeable membrane, a solute moves down its electrochemical gradient

We are all familiar with the way that water can flow from one side of a dike to another, provided the water levels on the two sides of the dike are different and the water has an open pathway (a breach in the dike) to move from one side to the other. In much the same way, a substance can passively move across a membrane that separates two compartments when there is both a favorable driving force and an open pathway through which the driving force can exert its effect.

When a pathway exists for transfer of a substance across a membrane, the membrane is said to be permeable to that substance. The driving force that determines the passive transport of solutes across a membrane is the electrochemical gradient or electrochemical potential energy difference acting on the solute between the two compartments. This electrochemical potential energy difference includes a contribution from the concentration gradient of the solute—the chemical potential energy difference—and, for charged solutes (e.g., Na+, Cl), a contribution from any difference in voltage that exists between the two compartments—the electrical potential energy difference.

This concept of how force and pathway determine passive movement of solutes is most easily illustrated by the example of passive, noncoupled transport. Noncoupled transport of a substance X means that movement of X across the membrane is not directly coupled to the movement of any other solute or to any chemical reaction (e.g., the hydrolysis of ATP). What, then, are the driving forces for the net movement of X? Clearly, if the concentration of X is higher in the outside compartment ([X]o) than in the inside compartment ([X]i), and assuming no voltage difference, the concentration gradient will act as the driving force to bring about the net movement of X across the membrane from outside to inside (Fig. 5-2). If [X] is the same on both sides but there is a voltage difference across the membrane—that is, the electrical potential energy on the outside (ψo) is not the same as on the inside (ψi)—this voltage difference will also drive the net movement of X, provided X is charged. The concentration gradient for X and the voltage difference across the membrane are the two determinants of the electrochemical potential energy difference for X between the two compartments. Because the movement of X by such a noncoupled mechanism is not directly coupled to the movement of other solutes or to any chemical reactions, the electrochemical gradient for X is the only driving force that contributes to the transport of X. Thus, the transport of X by a noncoupled, passive mechanism must always proceed “downhill,” in the direction down the electrochemical potential energy difference for X.


FIGURE 5-2 Uncoupled transport of a solute across a cell membrane. The net passive movement of a solute (X) depends on both the difference in concentration between the inside of the cell ([X]i) and the outside of the cell ([X]o) and the difference in voltage between the inside of the cell (ψi) and the outside of the cell (ψo).

Regardless of how X moves passively through the membrane—whether X moves through lipid or through a membrane protein—the direction of the overall driving force acting on X determines the direction of net transport. In the example in Figure 5-2, the overall driving force favors net transport from outside to inside (influx). However, X may still move from inside to outside (efflux). Movement of X across the membrane in one direction or the other is known as unidirectional flux. The algebraic sum of the two unidirectional fluxes is the net flux, or the net transport rate. Net transport occurs only when the unidirectional fluxes are unequal. In Figure 5-2, the overall driving force makes unidirectional influx greater than unidirectional efflux, which results in net influx.

When no net driving force is acting on X, we say that X is at equilibrium across the membrane and there is no net transport of X across the membrane. However, even when X is in equilibrium, there may be and usually are equal and opposite movements of X across the membrane. Net transport takes place only when the net driving force acting on X is displaced from the equilibrium point, and transport proceeds in the direction that would bring X back to equilibrium.

Equilibrium is actually a special case of a steady state. In a steady state, by definition, the conditions related to X do not change with time. Thus, a transport system is in a steady state when both the driving forces acting on it and the rate of transport are constant with time. Equilibrium is the particular steady state in which there is no net driving force and thus no net transport.

How can a steady state persist when X is not in equilibrium? Returning to the dike analogy, the downhill flow of water can be constant only if some device, such as a pump, keeps the water levels constant on both sides of the dike. A cell can maintain a nonequilibrium steady state for X only when some device, such as a mechanism for actively transporting X, can compensate for the passive movement of X and prevent the intracellular and extracellular concentrations of X from changing with time. This combination of a pump and a leak maintains both the concentrations of X and the passive flux of X.

At equilibrium, the chemical and electrical potential energy differences across the membrane are equal but opposite

As noted in the preceding section, the driving force for the passive, uncoupled transport of a solute is the electrochemical potential energy difference for that solute across the membrane that separates the inside (i) from the outside (o). We define the electrochemical potential energy difference as follows: imageN5-3




Electrochemical Potential Energy Difference for an Ion Across a Cell Membrane

Contributed by Peter Aronson, Emile Boulpaep, Walter Boron

The chemical potential energy, or partial molar Gibbs free energy, µX, of an uncharged solute X is


(NE 5-2)

where [X] is the concentration (more precisely, the chemical activity) of the solute, R is the gas constant (R = 8.314 joules/[K ⋅ mole]), and T is the temperature in kelvins (K = 273.16 + °C). Thus, µX has the units of energy per mole of X (joule/mole). Note that “potential” in the often-used term chemical potential is shorthand for “chemical potential energy.” In the case of a cell, we must consider the chemical potential energy both on the inside (µX,i) and on the outside (µX,o):


(NE 5-3)

Thus, if µX,o > µX,i (i.e., if [X]o > [X]i), then X will spontaneously move from the outside to the inside. On the other hand, if µX,o < µX,i, then X will spontaneously move from the inside to the outside.

We can define the chemical potential energy difference (ΔµX) as


(NE 5-4)

If solute X is charged, we must also consider the difference in partial molar free energy (ΔµX,elec) due to the voltage difference across the cell membrane. If the voltage inside the cell is ψi and the voltage outside the cell is ψo, then this voltage difference is (ψi − ψo), which is also known as the membrane voltage (Vm). This electrical portion of the partial molar free energy change is the electrical work (joules/mole) needed to move the charge, which is on X, across the membrane and into the cell. According to the laws of physics, the electrical work per mole is the product of the voltage difference and the amount of charge/mole moved. Thus, we must multiply the voltage difference (joules/coulomb) by the Faraday constant, F (coulombs/mole), and the valence of the ion X, zX (unitary charges/ion):


(NE 5-5)

The total free energy change (ΔũX) required to move X into the cell is simply the sum of the chemical and electrical terms:


(NE 5-6)

Equation NE 5-6 is the same as Equation 5-6 on page 106 in the main text.

where zX is the valence of X, T is absolute temperature, R is the gas constant, and F is the Faraday constant. The first term on the right-hand side of Equation 5-6, the difference in chemical potential energy, describes the energy change (joules per mole) as X moves across the membrane if we disregard the charge—if any—on X. The second term, the difference in electrical potential energy, describes the energy change as a mole of charged particles (each with a valence of zX) moves across the membrane. The difference (ψi − ψo) is the voltage difference across the membrane (Vm), also known as the membrane potential.

By definition, X is at equilibrium when the electrochemical potential energy difference for X across the membrane is zero:



Thus, ΔũX is the net driving force (units: joules/mole). When ΔũX is not zero, X is not in equilibrium and will obviously tend either to enter the cell or to leave the cell, provided a pathway exists for X to cross the membrane.

It is worthwhile to consider two special cases of the equilibrium state (see Equation 5-7). First, when either the chemical or the electrical term in Equation 5-6 is zero, the other must also be zero. For example, when X is uncharged (zX = 0), as in the case of glucose, equilibrium can occur only when [X] is equal on the two sides of the membrane. Alternatively, when X is charged, as in the case of Na+, but the voltage difference (i.e., Vm) is zero, equilibrium likewise can occur only when [X] is equal on the two sides of the membrane. Second, when neither the chemical nor the electrical term in Equation 5-6 is zero, equilibrium can occur only when the two terms are equal but of opposite sign. Thus, if we set ΔũX in Equation 5-6 to zero, as necessary for a state of equilibrium,



This relationship is the Nernst equation, which describes the conditions when an ion is in equilibrium across a membrane. Given values for [X]i and [X]o, X can be in equilibrium only when the voltage difference across the membrane equals the equilibrium potential (EX), also known as the Nernst potential. Stated somewhat differently, EX is the value that the membrane voltage would have to have for X to be in equilibrium. imageN5-4 If we express the logarithm to the base 10, then for the special case in which the temperature is 29.5°C,




Difference Between Vm and EX

Contributed by Emile Boulpaep, Walter Boron

Equation 5-6 on page 106 in the text (shown here as Equation NE 5-7) states that—for ion X—the electrochemical potential energy difference across the cell membrane is


(NE 5-7)

Here, each of the three major terms enclosed by horizontal braces has the dimension of energy per mole (e.g., joules/mole or kcal/mole). If we divide Equation NE 5-7 through by zXF we obtain:


(NE 5-8)

Each of the three major terms in Equation NE 5-8 now has the dimension of voltage. In other words, in dividing an energy term (units: joules/mole) by zXF (units: coulombs/mole), we are left with joules/coulomb, which is the definition of a volt. The first term on the right side of Equation NE 5-7 is nothing more than the negative of the Nernst potential that we introduced in Equation 5-8. The second term on the right side of Equation NE 5-8 is, of course, membrane potential. When Vm = EX, the ion is in equilibrium. Otherwise, the difference (Vm − EX) is the net electrochemical driving force—expressed in units of volts—that acts on ion X as the ion crosses the membrane. On page 151 of the text, we use this force to derive an expression in Equation 6-15 for the electrical current carried by ion X as the ion crosses the membrane:


(NE 5-9)

Equation NE 5-9 is written in such a way that an inward current (i.e., the movement of a positively charged species into the cell or of a negatively charged species out of the cell) is negative.

Equation NE 5-9 allows us to predict the direction that ion X will passively move (if indeed it can move at all) across the membrane. Of course, if Vm = EX, there will be no net movement of the ion at all. If Vmis more negative than EX, then the membrane voltage is too negative for X to be in equilibrium. As a result, if X is positive, the cation will tend to passively enter the cell. For example, Na+ generally tends to enter cells passively because Vm (e.g., −80 mV) is generally more negative than ENa (e.g., +67 mV in Fig. 6-10). If X is negative, the anion will tend to passively exit the cell. For example, Cl generally tends to exit cells passively because, in most cells, Vm (−60 mV) is generally more negative than ECl (e.g., −47 mV).

The opposite is true, of course, if Vm is more positive than EX.

At normal body temperature (37°C), the coefficient is ~61.5 mV instead of 60 mV. At 20°C, it is ~58.1 mV. imageN5-5


60mV per 10-fold Concentration Change

Contributed by Emile Boulpaep, Walter Boron

We start with Equation 5-8 (shown here as Equation NE 5-10):


(NE 5-10)

R is 8.314 joules/(K ⋅ mole), F is 96,484 coulombs/mole, and T is the temperature in kelvins (K = 273.16 + °C). In order to convert the natural logarithm to the logarithm in base 10, we must multiply the “ln” term by ln(10), which is ~2.303. For the term 2.303 RT/F to be exactly 60 mV, the temperature must be 29.5°C (302.66 K):


(NE 5-11)

Indeed, this is Equation 5-9.

To illustrate the use of Equation 5-9 we compute EX for a monovalent cation, such as K+. If [K+]i is 100 mM and [K+]o is 10 mM, a 10-fold concentration gradient, then



Thus, a 10-fold gradient of a monovalent ion such as K+ is equivalent, as a driving force, to a voltage difference of 60 mV. For a divalent ion such as Ca2+, a 10-fold concentration gradient can be balanced as a driving force by a voltage difference of 60 mV/2, or only 30 mV.

(Vm − EX) is the net electrochemical driving force acting on an ion

When dealing with an ion (X), it is more convenient to think about the net driving force in voltage (units: millivolts) rather than electrochemical potential energy difference (units: joules per mole). If we divide all terms in Equation 5-6 by the product of valence and the Faraday constant (zXF), we obtain



Because the energy terms previously expressed as joules per mole were divided by coulombs per mole (i.e., zXF)—all three energy terms enclosed in braces are now in units of joules per coulomb or volts. The term on the left is the net electrochemical driving force acting on ion X. The first term on the right, as defined in Equation 5-8, is the negative of the Nernst equilibrium potential (−EX). The second term on the right is the membrane voltage (Vm). Thus, a convenient equation expressing the net driving force is



In Table 5-3, we use this equation—along with the values in Table 5-2 for extracellular (i.e., interstitial) and intracellular concentrations and a typical Vm of −60 mV—to compute the net driving force of Na+, K+, Ca2+, Climage, and H+. When the net driving force is negative, cations will enter the cell and anions will exit. Stated differently, when Vm is more negative than EX (i.e., the cell is too negative for X to be in equilibrium), a cation will tend to enter the cell and an anion will tend to exit.


Net Electrochemical Driving Forces Acting on Ions in a Typical Cell*




X = −(RT/zXF) ln ([X]i/[X]o)

m − EX)

Na+ 145 mM

15 mM

−60 mV

+61 mV

−121 mV

K+ 4.5 mM

120 mM

−60 mV

−88 mV

+28 mV

Ca2+ 1.2 mM

10−7 M

−60 mV

+125 mV

−185 mV

Cl 116 mM

20 mM

−60 mV

−47 mV

−13 mV

image 25 mM

16 mM

−60 mV

−12 mV

−48 mV

H+ 40 nM
pH 7.4

63 nM

−60 mV

−12 mV

−48 mV

*Calculated at 37°C using −RT/zxF = −26.71 mV.

In simple diffusion, the flux of an uncharged substance through membrane lipid is directly proportional to its concentration difference

The difference in electrochemical potential energy of a solute X across the membrane is a useful parameter because it allows us to predict whether X is in equilibrium across the cell membrane (i.e., Is ΔũX = 0?) or, if not, whether X would tend to passively move into the cell or out of the cell. As long as the movement of X is not coupled to the movement of another substance or to some biochemical reaction, the only factor that determines the direction of net transport is the driving force ΔũX = 0. The ability to predict the movement of X is independent of any detailed knowledge of the actual transport pathway mediating its passive transport. In other words, we can understand the overall energetics of X transport without knowing anything about the transport mechanism itself, other than knowing that it is passive.

So far, we have discussed only the direction of net transport, not the rate. How will the rate of X transport vary if we vary the driving force ΔũX? Unlike determining the direction, determining the rate—that is, the kinetics—of transport requires knowing the peculiarities of the actual mechanism that mediates passive X transport.

Most transport systems are so complicated that a straightforward relationship between transport rate and ΔũX may not exist. Here we examine the simplest case, which is simple diffusion. How fast does an uncharged, hydrophobic solute move through a lipid bilayer? Gases (e.g., CO2), a few endogenous compounds (e.g., steroid hormones), and many drugs (e.g., anesthetics) are both uncharged and hydrophobic. Imagine that such a solute is present on both sides of the membrane but at a higher concentration on the outside (see Fig. 5-2). Because X has no electrical charge and because [X]o is greater than [X]i, the net movement of X will be into the cell. How fast X moves is described by its flux (JX); namely, the number of moles of X crossing a unit area of membrane (typically 1 cm2) per unit time (typically 1 second). Thus, JXhas the units of moles/(square centimeter ⋅ second). The better that X can dissolve in the membrane lipid (i.e., the higher the lipid-water partition coefficient of X), the more easily X will be able to traverse the membrane-lipid barrier. The flux of X will also be greater if X moves more readily once it is in the membrane (i.e., a higher diffusion coefficient) and if the distance that it must traverse is short (i.e., a smaller membrane thickness). We can combine these three factors into a single parameter called the permeability coefficient of X (PX).imageN5-6 Finally, the flux of X will be greater as the difference in [X] between the two sides of the membrane increases (a large gradient).


Definition of the Permeability Coefficient

Contributed by Emile Boulpaep, Walter Boron

The permeability coefficient (P) is


(NE 5-12)

D is the diffusion coefficient (in cm2/s), β is the partition coefficient (concentration in the lipid divided by the concentration in the bulk aqueous phase; a dimensionless number), and a is the thickness of the membrane (in cm). Thus, the units of the permeability coefficient are cm/s.


(NE 5-13)

These concepts governing the simple diffusion of an electrically neutral substance were quantified by Adolf Fick in the 1800s and applied by others to the special case of a cell membrane. They are embodied in the following equation, which is a simplified version of Fick's law:



As already illustrated in Figure 5-2, we can separate the net flux of X into a unidirectional influx (image) and a unidirectional efflux (image). The net flux of X into the cell is simply the difference between the unidirectional fluxes:



Thus, unidirectional influx is proportional to the outside concentration, unidirectional efflux is proportional to the inside concentration, and net flux is proportional to the concentration difference (not the ratio [X]o/[X]i, but the difference [X]o − [X]i). In all cases, the proportionality constant is PX.

A description of the kinetic behavior of a transport system (see Equation 5-14)—that is, how fast things move—cannot violate the laws of energetics, or thermodynamics (see Equation 5-6)—that is, the direction in which things move to restore equilibrium. For example, the laws of thermodynamics (see Equation 5-6) predict that when the concentration gradient for a neutral substance is zero (i.e., when [X]o/[X]i = 1), the system is in equilibrium and therefore the net flux must be zero. The law of simple diffusion (see Equation 5-14), which is a kinetic description, also predicts that when the concentration gradient for a neutral substance is zero (i.e., [X]o − [X]i = 0), the flux is zero.

Some substances cross the membrane passively through intrinsic membrane proteins that can form pores, channels, or carriers

Because most ions and hydrophilic solutes of biological interest partition poorly into the lipid bilayer, simple passive diffusion of these solutes through the lipid portion of the membrane is negligible. Noncoupled transport across the plasma membrane generally requires specialized pathways that allow particular substances to cross the lipid bilayer. In all known cases, such pathways are formed from integral membrane proteins. Three types of protein pathways through the membrane are recognized:

1. The membrane protein forms a pore that is always open (Fig. 5-3A). Physiological examples are the porins in the outer membranes of mitochondria, cytotoxic pore-forming proteins such as the perforin released by lymphocytes, and perhaps the aquaporin water channels. A physical equivalent is a straight, open tube. If you look though this tube, you always see light coming through from the opposite side.


FIGURE 5-3 Three types of passive, noncoupled transport of a solute (X) through integral membrane proteins.

2. The membrane protein forms a channel that is alternately open and closed because it is equipped with a movable barrier or gate (see Fig. 5-3B). Physiological examples include virtually all ion channels, such as the ones that allow Na+, Cl, K+, and Ca2+ to cross the membrane. The process of opening and closing of the barrier is referred to as gating. Thus, a channel is a gated pore, and a pore is a nongated channel. A physical equivalent is a tube with a shutter near one end. As you look through this tube, you see the light flickering as the shutter opens and closes.

3. The membrane protein forms a carrier surrounding a conduit that never offers a continuous transmembrane path because it is equipped with at least two gates that are never open at the same time (see Fig. 5-3C). Between the two gates is a compartment that can contain one or more binding sites for the solute. If the two gates are both closed, one (or more) of the transiting particles is trapped, or occluded, in that compartment. Physiological examples include carriers that move single solutes through the membrane by a process known as facilitated diffusion, which is discussed in the next section. A physical equivalent is a tube with shutters at both ends. As you look through this tube, you never see any light passing through because both shutters are never open simultaneously.

Water-filled pores can allow molecules, some as large as 45 kDa, to cross membranes passively

Some membrane proteins form pores that provide an aqueous transmembrane conduit that is always open (see Fig. 5-3A). Among the large-size pores are the porins (Fig. 5-4) found in the outer membranes of gram-negative bacteria and mitochondria. Mitochondrial porin allows solutes as large as 5 kDa to diffuse passively from the cytosol into the mitochondria's intermembrane space.


FIGURE 5-4 Structure of the PhoE porin of Escherichia coli. A, Top view of a porin trimer that shows the backbones of the polypeptide chains. Each of the three identical monomers, which are shown in different colors, contains 330 amino acids. The center of each monomer is a pore. B, Side view of a porin trimer. The extracellular surface is shown at the top. Each monomer consists of a β barrel with 16 antiparallel β sheets (i.e., adjacent polypeptide strands are oriented in opposite directions) surrounding a large cavity that at its narrowest point has an oval cross section (minimum and maximum internal diameters, 0.7 × 1.1 nm). The images are based on high-resolution electron microscopy at a resolution of 3.5 Å (0.35 nm). (From Jap BK, Walian PJ: Structure and functional mechanisms of porins. Physiol Rev 76:1073–1088, 1996.)

One mechanism by which cytotoxic T lymphocytes kill their target cells is the release of monomers of a pore-forming protein known as perforin. Perforin monomers polymerize within the target cell membrane and assemble like staves of a barrel to form large, doughnut-like channels with an internal diameter of 16 nm. The passive flow of ions, water, and other small molecules through these pores kills the target cell. A similar pore plays a crucial role in the defense against bacterial infections. The binding of antibodies to an invading bacterium (“classic” pathway), or simply the presence of native polysaccharides on bacteria (“alternative” pathway), triggers a cascade of reactions known as the complement cascade. This cascade culminates in the formation of a doughnut-like structure with an internal diameter of 10 nm. This pore is made up of monomers of C9, the final component of the complement cascade.

The nuclear pore complex (NPC), which regulates traffic into and out of the nucleus (see p. 21), is remarkably large. The NPC is made up of at least 30 different proteins and has a molecular mass of 108 Da and an outer diameter of ~100 nm. It can transport huge molecules (approaching 106 Da) in a complicated process that involves ATP hydrolysis. In addition to this active component of transport, the NPC also has a passive component. Contained within the massive NPC is a simple aqueous pore with an internal diameter of ~9 nm that allows molecules <45 kDa to move between the cytoplasm and nucleus but almost completely restricts the movement of globular proteins that are larger than ~60 kDa.

The plasma membranes of many types of cells have proteins that form channels just large enough to allow water molecules to pass through. The first water channel to be studied was aquaporin 1 (AQP1), a 28-kDa protein. AQP1 belongs to a larger family of aquaporins (AQPs) that has representatives in organisms as diverse as bacteria, plants, and animals. In mammals, the various AQPs have different tissue distributions, different mechanisms of regulation, and varying abilities to transport small neutral molecules other than water. In the lipid bilayer, AQP1 (Fig. 5-5) exists as tetramers. Each monomer consists of six membrane-spanning helices as well as two shorter helices that dip into the plane of the membrane. These structures form a permeation pathway for the single-file diffusion of water. For his discovery of the aquaporins, Peter Agre shared the 2003 Nobel Prize in Chemistry. imageN5-7


FIGURE 5-5 Structure of the human AQP1 water channel. A, Top view of an AQP tetramer, showing transmembrane helices H1–H6 and shorter helices HE and HB. Each of the four identical monomers is made up of 269 amino acids and has a pore at its center. B, Side view of AQP. The extracellular surface is shown at the top. The images are based on high-resolution electron microscopy at a resolution of 3.8 Å (0.38 nm). (From Murata K, Mitsuoka K, Hirai T, et al: Structural determinants of water permeation through aquaporin-1. Nature 407:599–605, 2000. © 2000 Macmillan Magazines Ltd.)


Peter Agre

For more information about Peter Agre and the work that led to his Nobel Prize, visit (accessed October 2014).

Gated channels, which alternately open and close, allow ions to cross the membrane passively

Gated ion channels, like the AQPs just discussed, consist of one or more polypeptide subunits with α-helical membrane-spanning segments. These channels have several functional components (see Fig. 5-3B). The first is a gate that determines whether the channel is open or closed, with each state reflecting a different conformation of the membrane protein. Second, the channel generally has one or more sensors that can respond to one of several different types of signals: (1) changes in membrane voltage, (2) second-messenger systems that act at the cytoplasmic face of the membrane protein, or (3) ligands, such as neurohumoral agonists, that bind to the extracellular face of the membrane protein. These signals regulate transitions between the open and closed states. A third functional component is a selectivity filter, which determines the classes of ions (e.g., anions or cations) or the particular ions (e.g., Na+, K+, Ca2+) that have access to the channel pore. The fourth component is the actual open channel pore (see Fig. 5-3B). Each time that a channel assumes the open conformation, it provides a continuous pathway between the two sides of the membrane so that ions can flow through it passively by diffusion until the channel closes again. During each channel opening, many ions flow through the channel pore, usually a sufficient number to be detected as a small current by sensitive patch-clamp techniques (see p. 154).

Na+ Channels

Because the electrochemical driving force for Na+ (Vm − ENa) is always strongly negative (see Table 5-3), a large, inwardly directed net driving force or gradient favors the passive movement of Na+ into virtually every cell of the body. Therefore, an open Na+ channel will act as a conduit for the passive entry of Na+. One physiological use for channel-mediated Na+ entry is the transmission of information. Thus, voltage-gated Na+ channels are responsible for generating the action potential (e.g., “nerve impulse”) in many excitable cells. Another physiological use of Na+ channels can be found in epithelial cells such as those in certain segments of the renal tubule and intestine. In this case, the ENaC Na+ channels are largely restricted to the apical surface of the cell, where they allow Na+ to enter the epithelial cell from the renal tubule lumen or intestinal lumen. This passive influx is a key step in the movement of Na+ across the entire epithelium, from lumen to blood.

K+ Channels

The electrochemical driving force for K+ (Vm − EK) is usually fairly close to zero or somewhat positive (see Table 5-3), so K+ is either at equilibrium or tends to move out of the cell. In virtually all cells, K+ channels play a major role in generating a resting membrane voltage that is inside negative. Other kinds of K+ channels play a key role in excitable cells, where these channels help terminate action potentials.

Ca2+ Channels

The electrochemical driving force for Ca2+ (Vm − ECa) is always strongly negative (see Table 5-3), so Ca2+ tends to move into the cell. When Ca2+ channels are open, Ca2+ rapidly enters the cell down a steep electrochemical gradient. This inward movement of Ca2+ plays a vital role in transmembrane signaling for both excitable and nonexcitable cells as well as in generating action potentials in some excitable cells.

Proton Channels

The plasma membranes of many cell types contain Hv1 H+ channels. Under normal conditions, the electrochemical driving force for H+ generally tends to move H+ into cells if Hv1 channels are open (see Table 5-3). However, Hv1 channels tend to be closed under normal conditions and activate only when the membrane depolarizes or the cytoplasm acidifies—that is, when the driving force favors the outward movement of H+. Hv1 channels may therefore help mediate H+ extrusion from the cell during states of strong membrane depolarization (e.g., during an action potential) or severe intracellular acidification.

Anion Channels

Most cells contain one or more types of anion-selective channels through which the passive, noncoupled transport of Cl—and, to a lesser extent, image—can take place. The electrochemical driving force for Cl(Vm − ECl) in most cells is modestly negative (see Table 5-3), so Cl tends to move out of these cells. In certain epithelial cells with Cl channels on their basolateral membranes, the passive movement of Cl through these channels plays a role in the transepithelial movement of Cl from lumen to blood.

Some carriers facilitate the passive diffusion of small solutes such as glucose

Carrier-mediated transport systems transfer a broad range of ions and organic solutes across the plasma membrane. Each carrier protein has a specific affinity for binding one or a small number of solutes and transporting them across the bilayer. The simplest passive carrier-mediated transporter is one that mediates facilitated diffusion. Below, we will introduce cotransporters (which carry two or more solutes in the same direction) and exchangers (which move them in opposite directions).

All carriers that do not either hydrolyze ATP or couple to an electron transport chain are members of the solute carrier (SLC) superfamily,imageN5-8 which is organized according to the homology of the deduced amino-acid sequences (Table 5-4). Each of the 52 SLC families contains up to 53 genes that encode proteins that share a relatively high amino-acid sequence identity (20% to 25%). Moreover, each gene may encode multiple variants (see Fig 4-19). Members of an SLC family may differ in molecular mechanism (i.e., facilitated diffusion, cotransport, exchange), kinetic properties (e.g., solute specificity and affinity), regulation (e.g., phosphorylation), sites of membrane targeting (e.g., plasma membrane versus intracellular organelles), tissues in which they are expressed (e.g., kidney versus brain), or developmental stage at which they are expressed.


The SLC Superfamily of Solute Carriers imageN5-8




SLC1 (7)*

Glutamate transporters

EAAT1, 2, 3, 4, 5
ASCT1, 2

SLC2 (14)

Facilitated transport of hexoses

GLUT1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 14; HMIT

SLC3 (2)

Heavy subunits of heterodimeric amino-acid transporters (with SLC7)

rBAT system
4F2hc system

SLC4 (10)

image exchangers and cotransporters

AE1, 2, 3 (Cl-HCO3 exchangers)
NBCe1, 2 (electrogenic Na/HCO3 cotransporters)
NBCn1, 2 (electroneutral Na/HCO3 cotransporters)
NDCBE (Na+-driven Cl-HCO3 exchanger)

SLC5 (12)

Na/glucose cotransporters

SGLT1, 2, 3, 4, 5 (glucose);
NIS (iodide); SMIT1, 2(myoinositol)
SMCT1, 2 (monocarboxylates)
SMVT (biotin)
CHT (choline)

SLC6 (21)

Na+- and Cl-coupled cotransport of “neurotransmitters”

B0AT1 (Na+-coupled amino acid)
GAT1, 2, 3; GBT1 (Na+- and Cl-coupled GABA)
ATB0+ (Na+- and Cl-coupled amino acids)
NET (norepinephrine transporters)
SERT (serotonin)
DAT (dopamine)

SLC7 (14)

Transporter subunits of heterodimeric amino-acid transporters (with SLC3)

LAT1, 2; y+LAT1, 2
CAT1, 2, 3, 4

SLC8 (3)

Na-Ca exchangers

NCX1, 2, 3

SLC9 (13)

Na-H exchangers

NHE1, 2, 3, 4, 5, 6, 7, 8, 9
NHA1, 2

SLC10 (7)

Na/bile-salt cotransporters


SLC11 (2)

H+-driven metal-ion cotransporters


SLC12 (9)

Cation-coupled Cl cotransporters

NKCC1, 2 (Na/K/Cl cotransporter)
NCC (Na/Cl cotransporter)
KCC1, 2, 3, 4 (K/Cl cotransporter)

SLC13 (5)

Na+-coupled sulfate and carboxylate cotransporters

NaDC1, 3 (mono-, di-, and tricarboxylates)
NaS1, 2 (sulfate)

SLC14 (2)

Facilitated transport of urea

UT-A1, 2, 3, 4, 5, 6; UT-B1, 2

SLC15 (4)

H+-driven oligopeptide cotransporters

PepT1, 2
PhT1, 2

SLC16 (14)

Monocarboxylate transporters

MCT1, 2, 3, 4 (H+-coupled monocarboxylate cotransporter)
MCT8, 10 (facilitated diffusion of aromatic amino acids)

SLC17 (9)

Type I Na/phosphate cotransporters and vesicular Glu transporters

NPT1, 3, 4
VGLUT1, 2, 3

SLC18 (4)

Vesicular monoamine transporters

VMAT1, 2 (H+-amine exchanger)
VAChT (H+-acetylcholine exchanger)

SLC19 (3)

Folate/thiamine transporters

ThTr1, 2

SLC20 (2)

Type III Na/phosphate cotransporters

PiT-1, 2

SLC21 (11) or SLCO

Organic anion and cation transporters

OATP1, 2, 3, 4, 5, 6 (bile salts, thyroid hormones, prostaglandins)

SLC22 (23)

Organic cation, anion, zwitterion transporters

OCT1, 2, 3 (facilitated diffusion of organic cations)
OCTN1 (organic cation-H exchanger)
OCTN2 (Na/organic cation cotransporter)
OAT1, 2, 3, 4, 5, 7, 10 (exchange or facilitated diffusion of organic anions)
URAT1 (urate exchanger)

SLC23 (4)

Na/ascorbic acid transporters

SVCT1, 2, 3, 4

SLC24 (5)

Na+/(Ca2+-K+) exchanger

NCKX1, 2, 3, 4, 5

SLC25 (53)

Mitochondrial carriers

ANC1, 2, 3, 4 (ANT1, 2, 3, 4)

SLC26 (11)

Multifunctional anion exchangers

DRA (Cl-HCO3 exchanger)
Pendrin (exchanges image, Cl or I)
Prestin (exchanges Cl, formate, oxalate; a motor in cochlear hair cells)
CFEX (exchanges Climage, oxalate, formate)

SLC27 (6)

Fatty-acid transporters

FATP1, 2, 3, 4, 5, 6

SLC28 (3)

Na/nucleoside transporters

CNT1, 2, 3

SLC29 (4)

Facilitative nucleoside transporters

ENT1, 2, 3, 4

SLC30 (10)

Zinc efflux transporter

ZnT1, 2, 3, 4, 5, 6, 7, 8, 9, 10

SLC31 (2)

Copper importer

CTR1, 2

SLC32 (1)

Vesicular inhibitory amino-acid transporter

VIAAT (exchange of H+ for GABA or glycine)

SLC33 (1)

Acetyl–coenzyme A transporter


SLC34 (3)

Type II Na/phosphate cotransporters

NaPi-IIa, b, c

SLC35 (30)

Nucleoside-sugar transporter


SLC36 (4)

H+-coupled amino-acid cotransporters

PAT1, 2, 3, 4

SLC37 (4)

Sugar-phosphate/phosphate exchanger

SPX1, 2, 3, 4

SLC38 (11)

Na+-driven neutral amino acids (system A and N)

SNAT1, 2, 4 (system A, cotransports amino acids with Na+)
SNAT3, 5 (system N, cotransports amino acids with Na+ in exchange for H+)

SLC39 (14)

Metal-ion transporters

ZIP1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14 (uptake of Zn2+)

SLC40 (1)

Basolateral Fe2+ transporter

Ferroportin (FPN1)

SLC41 (3)

MgtE-like magnesium transporter


SLC42 (3)

image, CO2 channels


SLC43 (3)

Na+-independent, system L–like amino-acid transporter

LAT3, 4

SLC44 (5)

Choline-like transporter

CTL1, 2, 3, 4, 5

SLC45 (4)

Putative sugar transporter


SLC46 (3)

Folate transporter


SLC47 (2)

Multidrug and toxin extrusion transporter

MATE1, 2

SLC48 (1)

Heme transporter

HRG-1 (heme responsive gene 1)

SLC49 (4)

FLVCR-related transporter


SLC50 (1)

Sugar efflux transporter


SLC51 (2)

Steroid-derived molecule transporters


SLC52 (3)

Riboflavin transporter

RFVT1, 2, 3

*In parentheses is the number of human genes in the family. Members of individual families generally are named SLCxAy (e.g., SLC4A4), where x is a number identifying the family and y is a number identifying the gene within that family.

GABA, gamma-aminobutyric acid.


The SLC Superfamily of Solute Carriers

Contributed by Emile Boulpaep, Walter Boron

The SLC superfamily was the subject of a series of reviews—one per family member—in 2013. The reference below is the introduction to the series.


Hediger MA, Clémençon B, Burrier RE, Bruford EA. The ABCs of membrane transporters in health and disease (SLC series): Introduction. Mol Aspects Med. 2013;34:95–107.

Carrier-mediated transport systems behave according to a general kinetic scheme for facilitated diffusion that is outlined in Figure 5-3C. This model illustrates how, in a cycle of six steps, a carrier can passively move a solute X into the cell.

This mechanism can mediate only the downhill, or passive, transport of X. Therefore, it mediates a type of diffusion, called facilitated diffusion. When [X] is equal on the two sides of the membrane, no net transport will take place, although equal and opposite unidirectional fluxes of X may still occur.

In a cell membrane, a fixed number of carriers is available to transport X. Furthermore, each carrier has a limited speed with which it can cycle through the steps illustrated in Figure 5-3C. Thus, if the extracellular X concentration is gradually increased, for example, the influx of X will eventually reach a maximal value once all the carriers have become loaded with X. This situation is very different from the one that exists with simple diffusion—that is, the movement of a solute through the lipid phase of the membrane. Influx by simple diffusion increases linearly with increases in [X]o, with no maximal rate of transport. As an example, if X is initially absent on both sides of the membrane and we gradually increase [X] on one side, the net flux of X (JX) is described by a straight line that passes through the origin (Fig. 5-6A). However, with carrier-mediated transport, JX reaches a maximum (Jmax) when [X] is high enough to occupy all the carriers in the membrane (see Fig. 5-6B). Thus, the relationship describing carrier-mediated transport follows the same Michaelis-Menten kinetics as do enzymes:



This equation describes how the velocity of an enzymatic reaction (V) depends on the substrate concentration ([S]), the Michaelis constant (Km), and the maximal velocity (Vmax). The comparable equation for carrier-mediated transport is identical, except that fluxes replace reaction velocities:



Thus, Km is the solute concentration at which JX is half of the maximal flux (Jmax). The lower the Km, the higher the apparent affinity of the transporter for the solute.


FIGURE 5-6 Dependence of transport rates on solute concentration. A, The net flux of the solute X through the cell membrane is JXB, The maximal flux of X (Jmax) occurs when the carriers are saturated. The flux is half of its maximal value (imageJmax) when the concentration of X is equal to the Km.

Historically, the name carrier suggested that carrier-mediated transport occurs as the solute binds to a miniature ferryboat that shuttles back and forth across the membrane. Small polypeptides that act as shuttling carriers exist in nature, as exemplified by the antibiotic valinomycin. Such “ion carriers,” or ionophores, bind to an ion on one side of the membrane, diffuse across the lipid phase of the membrane, and release the ion on the opposite side of the membrane. Valinomycin is a K+ ionophore that certain bacteria produce to achieve a selective advantage over their neighbors. However, none of the known carrier-mediated transport pathways in animal cell membranes are ferries.

Examples of membrane proteins that mediate facilitated diffusion are the GLUT glucose transporters (Fig. 5-7), members of the SLC2 family (see Table 5-4). The GLUTs have 12 membrane-spanning segments as well as multiple hydrophilic polypeptide loops facing either the ECF or the ICF. They could not possibly act as a ferryboat shuttling back and forth across the membrane. Instead, some of the membrane-spanning segments of carrier-mediated transport proteins most likely form a permeation pathway through the lipid bilayer, as illustrated by the amphipathic membrane-spanning segments 7, 8, and 11 in Figure 5-7. These membrane-spanning segments, as well as other portions of the protein, probably also act as the gates and solute-binding sites that allow transport to proceed in the manner outlined in Figure 5-3C.


FIGURE 5-7 Structure of the GLUT family of glucose transporters. The 12 membrane-spanning segments are connected to each other by intracellular and extracellular loops.

The SLC2 family includes 14 transporters (GLUTs). Whereas GLUT1 is constitutively expressed on the cell surface, GLUT4 in the basal state is predominantly present in the membranes of intracellular vesicles, which represent a storage pool for the transporters. Because a solute such as glucose permeates the lipid bilayer so poorly, its uptake by the cell depends strictly on the activity of a carrier-mediated transport system for glucose. Insulin increases the rate of carrier-mediated glucose transport into certain cells by recruiting the GLUT4 isoform to the plasma membrane from the storage pool (see p. 1047).

Two other examples of transporters that mediate facilitated diffusion are the urea transporter (UT) family, which are members of the SLC14 family (see Table 5-4), and the organic cation transporter (OCT) family, which are members of the SLC22 family. Because OCT moves an electrical charge (i.e., carries current), it is said to be electrogenic.

The physical structures of pores, channels, and carriers are quite similar

Pores, ion channels, and carriers all have multiple transmembrane segments surrounding a solute permeation pathway. Moreover, some channels also contain binding sites within their permeation pathways, so transport is saturable with respect to ion concentration. However, pores, channels, and carriers are fundamentally distinct kinetically (Table 5-5). Pores, such as the porins, are thought to be continuously open and allow vast numbers of particles to cross the membrane. No evidence suggests that pores have conformational states. Channels undergo conformational transitions between closed and open states. When they are open, they are open to both intracellular and extracellular solutions simultaneously. Thus, while the channel is open, it allows multiple ions, perhaps millions, to cross the membrane per open event. Because the length of time that a particular channel remains open varies from one open event to the next, the number of ions flowing through that channel per open event is not fixed. Carriers have a permeation pathway that is virtually never open simultaneously to both intracellular and extracellular solutions. Whereas the fundamental event for a channel is opening, the fundamental event for a carrier is a complete cycle of conformational changes. Because the binding sites in a carrier are limited, each cycle of a carrier can transport only one or a small, fixed number of solute particles. Thus, the number of particles per second that can move across the membrane is generally several orders of magnitude lower for a single carrier than for a single channel.


Comparison of Properties of Pores, Channels, and Carriers






Water channel (AQP1)

Shaker K+ channel

Glucose transporter (GLUT1)

Conduit through membrane

Always open

Intermittently open

Never open

Unitary event

None (continuously open)


Cycle of conformational changes

Particles translocated per “event”

6 × 104*


Particles translocated per second

Up to 2 × 109

106–108 when open


*Assuming a 100-pS (picosiemens) channel, a driving force of 100 mV, and an opening time of 1 ms.

We have seen how carriers can mediate facilitated diffusion of glucose, which is a passive or downhill process. However, carriers can also mediate coupled modes of transport. The remainder of this section is devoted to these carriers, which act as pumps, cotransporters, and exchangers.

The Na-K pump, the most important primary active transporter in animal cells, uses the energy of ATP to extrude Na+ and take up K+

Active transport is a process that can transfer a solute uphill across a membrane—that is, against its electrochemical potential energy difference. In primary active transport, the driving force needed to cause net transfer of a solute against its electrochemical gradient comes from the favorable energy change that is associated with an exergonic chemical reaction, such as ATP hydrolysis. In secondary active transport, the driving force is provided by coupling the uphill movement of that solute to the downhill movement of one or more other solutes for which a favorable electrochemical potential energy difference exists. A physical example is to use a motor-driven winch to lift a large weight into the air (primary active transport) and then to transfer this large weight to a seesaw, on the other end of which is a lighter child. The potential energy stored in the elevated weight will then lift the child (secondary active transport). For transporters, it is commonly the favorable inwardly directed Na+ electrochemical gradient, which itself is set up by a primary active transporter, that drives the secondary active transport of another solute. In this and the next section, we discuss primary active transporters, which are also referred to as pumps. The pumps discussed here are all energized by ATP hydrolysis and hence are ATPases.

As a prototypic example of a primary active transporter, consider the nearly ubiquitous Na-K pump (or Na,K-ATPase, NKA). This substance was the first enzyme recognized to be an ion pump, a discovery for which Jens Skou shared the 1997 Nobel Prize in Chemistry. imageN5-9 The Na-K pump is located in the plasma membrane and has both α and β subunits (Fig. 5-8A). The α subunit, which has 10 transmembrane segments, is the catalytic subunit that mediates active transport. The β subunit, which has one transmembrane segment, is essential for proper assembly and membrane targeting of the Na-K pump. Four α isoforms and two β isoforms have been described. These isoforms have different tissue and developmental patterns of expression as well as different kinetic properties.


FIGURE 5-8 Model of the sodium pump. A, Schematic representation of the α and β subunits of the pump. B, The protein cycles through at least eight identifiable stages as it moves 3 Na+ ions out of the cell and 2 K+ ions into the cell. Pi, phosphate.


Jens Skou

For more information about Jens Skou and the work that led to his Nobel Prize, visit (accessed October 2014).

With each forward cycle, the pump couples the extrusion of three Na+ ions and the uptake of two K+ ions to the intracellular hydrolysis of one ATP molecule. By themselves, the transport steps of the Na-K pump are energetically uphill; that is, if the pump were not an ATPase, the transporter would run in reverse, with Na+ leaking into the cell and K+ leaking out. Indeed, under extreme experimental conditions, the Na-K pump can be reversed and forced to synthesize ATP! However, under physiological conditions, hydrolysis of one ATP molecule releases so much free energy—relative to the aggregate free energy needed to fuel the uphill movement of three Na+ and two K+ ions—that the pump is poised far from its equilibrium and brings about the net active exchange of Na+ for K+ in the desired directions.

Although animal cells may have other pumps in their plasma membranes, the Na-K pump is the only primary active transport process for Na+. The Na-K pump is also the most important primary active transport mechanism for K+. In cells throughout the body, the Na-K pump is responsible for maintaining a low [Na+]i and a high [K+]i relative to ECF. In most epithelial cells, the Na-K pump is restricted to the basolateral side of the cell.

The Na-K pump exists in two major conformational states: E1, in which the binding sites for the ions face the inside of the cell; and E2, in which the binding sites face the outside. The Na-K pump is a member of a large superfamily of pumps known as E1-E2 ATPases or P-type ATPases. It is the ordered cycling between these two states that underlies the action of the pump. Figure 5-8B is a simplified model showing the eight stages of this catalytic cycle of the α subunit:

Stage 1: ATP-bound E1  ATP state. The cycle starts with the ATP-bound E1 conformation, just after the pump has released its bound K+ to the ICF. The Na+-binding sites face the ICF and have high affinities for Na+.

Stage 2: Na+-bound E1  ATP 3Na+ state. Three intracellular Na+ ions bind.

Stage 3: Occluded E1-P (3Na+) state. The ATP previously bound to the pump phosphorylates the pump at an aspartate residue. Simultaneously, ADP leaves. This phosphorylation triggers a minor conformational change in which the E1 form of the pump now occludes the three bound Na+ ions within the permeation pathway. In this state, the Na+-binding sites are inaccessible to both the ICF and ECF.

Stage 4: Deoccluded E2-P 3Na+ state. A major conformational change shifts the pump from the E1 to the E2 conformation and has two effects. First, the pump becomes deoccluded, so that the Na+-binding sites now communicate with the extracellular solution. Second, the Na+ affinities of these binding sites decrease.

Stage 5: Empty E2-P state. The three bound Na+ ions dissociate into the external solution, and the protein undergoes a minor conformational change to the empty E2-P form, which has high affinity for binding of extracellular K+. However, the pore still communicates with the extracellular solution.

Stage 6: K+-bound E2-P 2K+ state. Two K+ ions bind to the pump.

Stage 7: Occluded E2  (2K+) state. Hydrolysis of the acylphosphate bond, which links the phosphate group to the aspartate residue, releases the inorganic phosphate into the intracellular solution and causes a minor conformational change. In this E2 ⋅ (2K+) state, the pump occludes the two bound K+ ions within the permeation pathway so that the K+-binding sites are inaccessible to both the ECF and ICF.

Stage 8: Deoccluded E1  ATP 2K+ state. Binding of intracellular ATP causes a major conformational change that shifts the pump from the E2 back to the E1 state. This conformational change has two effects. First, the pump becomes deoccluded, so that the K+-binding sites now communicate with the intracellular solution. Second, the K+ affinities of these binding sites decrease.

Stage 1: ATP-bound E1  ATP state. Dissociation of the two bound K+ ions into the intracellular solution returns the pump to its original E1 ⋅ ATP state, ready to begin another cycle.

Because each cycle of hydrolysis of one ATP molecule is coupled to the extrusion of three Na+ ions from the cell and the uptake of two K+ ions, the stoichiometry of the pump is three Na+ to two K+, and each cycle of the pump is associated with the net extrusion of one positive charge from the cell. Thus, the Na-K pump is electrogenic.

Just as glucose flux through the GLUT1 transporter is a saturable function of [glucose], the rate of active transport by the Na-K pump is a saturable function of [Na+]i and [K+]o. The reason is that the number of pumps is finite and each must bind three Na+ ions and two K+ ions. The transport rate is also a saturable function of [ATP]i and therefore depends on the metabolic state of the cell. In cells with high Na-K pump rates, such as renal proximal tubules, a third or more of cellular energy metabolism is devoted to supplying ATP to the Na-K pump.

A hallmark of the Na-K pump is that it is blocked by a class of compounds known as cardiac glycosides, examples of which are ouabain and digoxin; digoxin is widely used for a variety of cardiac conditions. These compounds have a high affinity for the extracellular side of the E2-P state of the pump, which also has a high affinity for extracellular K+. Thus, the binding of extracellular K+ competitively antagonizes the binding of cardiac glycosides. An important clinical correlate is that hypokalemia (a low [K+] in blood plasma) potentiates digitalis toxicity in patients.

Besides the Na-K pump, other P-type ATPases include the H-K and Ca pumps

The family of P-type ATPases—all of which share significant sequence similarity with the α subunit of the Na-K pump—includes several subfamilies.

H-K Pump

Other than the Na-K pump, relatively few primary active transporters are located on the plasma membranes of animal cells. In the parietal cells of the gastric gland, an H-K pump (HKA) extrudes H+ across the apical membrane into the gland lumen. Similar pumps are present in the kidney and intestines. The H-K pump mediates the active extrusion of H+ and the uptake of K+, all fueled by ATP hydrolysis, probably in the ratio of two H+ ions, two K+ ions, and one ATP molecule. Like the Na-K pump, the H-K pump is composed of α and β subunits, each with multiple isoforms. The α subunit of the H-K pump also undergoes phosphorylation through E1 and E2 intermediates during its catalytic cycle (see Fig. 5-8B) and, like the α subunit of the Na-K pump, is a member of the P2C subfamily of P-type ATPases. The Na-K and H-K pumps are the only two P-type ATPases with known β subunits, all of which share significant sequence similarity.

Ca Pumps

Most, if not all, cells have a primary active transporter at the plasma membrane that extrudes Ca2+ from the cell. These pumps are abbreviated (for plasma-membrane Ca-ATPase), and at least four PMCA isoforms appear in the P2B subfamily of P-type ATPases. These pumps exchange one H+ for one Ca2+ for each molecule of ATP that is hydrolyzed.

Ca pumps (or Ca-ATPases) also exist on the membrane surrounding such intracellular organelles as the sarcoplasmic reticulum (SR) in muscle cells and the endoplasmic reticulum (ER) in other cells, where they play a role in the active sequestration of Ca2+ into intracellular stores. The SERCAs (for sarcoplasmic and endoplasmic reticulum calcium ATPases) appear to transport two H+ and two Ca2+ ions for each molecule of ATP hydrolyzed. imageN5-10 The three known SERCAs, which are in the P2A subfamily of P-type ATPases, are expressed in different muscle types (see Table 9-1).


Crystal Structure of SERCA1

Contributed by Emile Boulpaep, Walter Boron

SERCA is a P-type ATPase, as is the Na-K pump. In 2000, Toyoshima and colleagues determined the x-ray crystal structure of SERCA with two Ca2+ ions bound. This was the first crystal structure identified for any P-type pump or ATPase. In 2002, Toyoshima and Nomurai determined the x-ray crystal structure again, but with no Ca2+ bound. In their 2004 paper, Toyoshima and Mizutani crystallized the protein with a bound nonhydrolyzable ATP analog AMP-PNP (adenylyl-imidodiphosphate) and one Mg2+ (under physiological conditions, ATP—the energy source—binds as a complex with Mg2+), as well as two Ca2+ ions (the transported species) occluded within a channel in the protein.


Toyoshima C, Mizutani T. Crystal structure of the calcium pump with a bound ATP analogue. Nature. 2004;430:529–535.

Toyoshima C, Nakasako M, Nomura H, Ogawa H. Crystal structure of the calcium pump of sarcoplasmic reticulum at 2.6Å resolution. Nature. 2000;405:647–655.

Toyoshima C, Nomurai H. Structural changes in the calcium pump accompanying the dissociation of calcium. Nature. 2002;418:605–611.

Toyoshima C, Nomura H, Sugita Y. Structural basis of ion pumping by Ca2+-ATPase of sarcoplasmic reticulum. FEBS Lett. 2003;555:106–110.

Other Pumps

Among the other P-type ATPases is the copper pump ATP7B. This member of the P1B subfamily of P-type ATPases is mutated in Wilson disease (see Box 46-5).

The F-type and the V-type ATPases transport H+

F-type or FoF1 ATPases

The ATP synthase of the inner membrane of mitochondria, also known as an F-type or FoF1 ATPase, catalyzes the final step in the ATP synthesis pathway. imageN5-11


ATP Synthase

A Pump in Reverse

Contributed by Emile Boulpaep, Walter Boron

The apparent paradox of how the same “pump” protein can act both as an ATPase and an ATP synthase can be resolved if we recognize that the pump can either hydrolyze ATP and use the energy to pump H+out of the mitochondrion or—in the physiological direction—use the energy of the inwardly directed H+ gradient to synthesize ATP.

The FoF1 ATPase of mitochondria (Fig. 5-9A) looks a little like a lollipop held in your hand. The hand-like Fo portion is embedded in the membrane and serves as the pathway for H+ transport. The Fo portion has at least three different subunits (a, b, and c), for an overall stoichiometry of ab2c10–12. The lollipop-like F1 portion is outside the plane of the membrane and points into the mitochondrial matrix. The “stick” consists of a γ subunit, with an attached ε subunit. The “candy” portion of F1, which has the ATPase activity, consists of three alternating pairs of α and β subunits as well as an attached δ subunit. Thus, the overall stoichiometry of F1 is α3β3γδε. The entire FoF1 complex has a molecular mass of ~500 kDa.


FIGURE 5-9 The FoF1 ATPase and its role as the ATP synthase in the mitochondrial synthesis of ATP. A, A cartoon of the FoF1 ATPase. The pump has two functional units, Fo (which historically stood for oligomycin-sensitive factor) and Fl (which historically stood for factor 1). Fo is the transmembrane portion that contains the ion channel through which the H+ passes. The F1 is the ATPase. In one complete cycle, the downhill movement of H+ ions causes the c subunits of Fo and the axle formed by the subunits of F1 to rotate 360 degrees in three 120-degree steps, which causes the α and β subunits to sequentially synthesize and release 3 ATP molecules, for a synthase stoichiometry of ~3 H+ per ATP. However, the mitochondrion uses ~1 additional H+ to import inorganic phosphate (Pi) and to exchange cytosolic ADP for mitochondrial ATP. Thus, a total of ~4 H+ would be needed per ATP. B, Complexes I, III, and IV of the respiratory chain use the energy of NADH to pump H+ out of the mitochondrial matrix; the consensus is 10 H+ ions per NADH molecule. The resulting H+ gradient causes the mitochondrial FoF1 ATPase to run as an ATP synthase. Thus, the mitochondrion synthesizes (10 H+/NADH) × (1 ATP/4 H+) = 2.5 ATP/NADH. Similarly, the consensus is that complexes III and IV use the energy of 1 FADH2 molecule to pump 6 H+ ions out of the mitochondrial matrix (not shown). Thus, the mitochondrion synthesizes (6 H+/FADH2) × (1 ATP/4 H+) = 1.5 ATP/FADH2.

A fascinating property of the FoF1 ATPase is that parts of it rotate. We can think of the hand, stick, and candy portions of the FoF1 ATPase as having three distinct functions. (1) The hand (the c proteins of Fo) acts as a turbine that rotates in the plane of the membrane, driven by the H+ ions that flow through the turbine—down the H+ electrochemical gradient—into the mitochondrion. (2) The stick is an axle (γ and ε subunits of F1) that rotates with the turbine. (3) The candy (the α and β subunits of F1) is a stationary chemical factory—energized by the rotating axle—that synthesizes one ATP molecule for each 120-degree turn of the turbine/axle complex. In addition, the a and b subunits of Fo, and possibly the δ subunit of F1, form a stator that holds the candy in place while the turbine/axle complex turns. Paul Boyer and John Walker shared part of the 1997 Nobel Prize in Chemistry for elucidating this “rotary catalysis” mechanism. imageN5-12


Paul Boyer and John Walker

For more information about Paul Boyer and John Walker and the work that led to their Nobel Prize, visit (accessed October 2014).

Under physiological conditions, the mitochondrial FoF1 ATPase runs as an ATP synthase (i.e., “backward” for an H pump)—the final step in oxidative phosphorylation—because of a large, inwardly directed H+ gradient across the inner mitochondrial membrane (see Fig. 5-9B). The citric acid cycle captures energy as electrons and transfers these electrons to reduced nicotinamide adenine dinucleotide (NADH) imageN5-13 and reduced flavin adenine dinucleotide (FADH2). NADH and FADH2 transfer their high-energy electrons to the electron transport chain, which consists of four major complexes on the inner membrane of the mitochondrion (see Fig. 5-9B). As this “respiratory chain” transfers the electrons from one electron carrier to another, the electrons gradually lose energy until they finally combine with 2 H+ and image O2 to form H2O. Along the way, three of the four major complexes of the respiratory chain (I, III, IV) pump H+ across the inner membrane into the intermembrane space (i.e., the space between the inner and outer mitochondrial membranes). These “pumps” are not ATPases. The net result is that electron transport has established a large out-to-in H+ gradient across the mitochondrial inner membrane.



Contributed by Alisha Bouzaher

NADH and NAD+ are, respectively, the reduced and oxidized forms of nicotinamide adenine dinucleotide (NAD) and their close analogs are NADPH and NADP+, the reduced and oxidized forms of nicotinamide adenine dinucleotide phosphate (NADP). The coenzymes NADH and NADPH each consist of two nucleotides joined at their phosphate groups by a phosphoanhydride bond. NADPH is structurally distinguishable from NADH by the additional phosphate group residing on the ribose ring of the nucleotide, which allows enzymes to preferentially interact with either molecule.

Total concentrations of NAD+/NADH (10−5M) are higher in the cell by approximately 10-fold compared to NADP+/NADPH (10−6M). Ratios of the oxidized and reduced forms of these coenzymes offer perspective into the metabolic activity of the cell. The high NAD+/NADH ratio favors the transfer of a hydride from a substrate to NAD+ to form NADH, the reduced form of the molecule and oxidizing agent. Therefore, NAD+ is highly prevalent within catabolic reaction pathways where reducing equivalents (carbohydrate, fats, and proteins) transfer protons and electrons to NAD+. NADH acts as an energy carrier, transferring electrons from one reaction to another. Conversely, the NADP+/NADPH ratio is low, favoring the transfer of a hydride to a substrate oxidizing NADPH to NADP+. Thus, NADPH is utilized as a reducing agent within anabolic reactions, particularly the biosynthesis of fatty acids.


Nelson DL, Cox MM. Lehninger Principles of Biochemistry. 6th ed. AH Freeman: New York; 2012.

Wikipedia. s.v. Nicotinamide adenine dinucleotide.

The FoF1 ATPase—which is complex V in the respiratory chain—can now use this large electrochemical potential energy difference for H+. The H+ ions then flow backward (i.e., down their electrochemical gradient) into the mitochondrion through the FoF1 ATPase, which generates ATP in the matrix space of the mitochondrion from ADP and inorganic phosphate. The entire process by which electron transport generates an H+ gradient and the FoF1 ATPase harnesses this H+ gradient to synthesize ATP is known as the chemiosmotic hypothesis. Peter Mitchell, who proposed this hypothesis in 1961, received the Nobel Prize in Chemistry for his work in 1978. imageN5-14


Peter Mitchell

For more information about Peter Mitchell and the work that led to his Nobel Prize, visit (accessed October 2014).

The precise stoichiometry is unknown but may be one ATP molecule synthesized for every three H+ ions flowing downhill into the mitochondrion (one H+ for each pair of αβ subunits of F1). imageN5-15 If the H+gradient across the mitochondrial inner membrane reverses, the FoF1 ATPase will actually function as an ATPase and use the energy of ATP hydrolysis to pump H+ out of the mitochondrion. Similar FoF1 ATPases are also present in bacteria and chloroplasts.


ATP Molecules Synthesized per NADH

Contributed by Emile Boulpaep, Walter Boron

Glycolysis and the citric acid cycle generate the reducing equivalents NADH and FADH2, and then the inner membrane of the mitochondria converts the energy of these reducing equivalents to ATP in two steps. First, the electron transport chain uses the energy from NADH and FADH2 to pump H+ from the mitochondrial matrix into the intermembrane space between the mitochondrial inner and outer membranes, converting O2 to H2O in the final step. Second, the ATP synthase uses the energy stored in the H+ gradient to generate ATP from ATP plus inorganic phosphate.

Generation of the H+ Gradient

For each NADH consumed in the inner matrix of the mitochondrion, it appears that complex I and complex III of the electron transport chain (see Fig. 5-9) each pump 4 H+ from the matrix, across the inner membrane, and into the intermembrane space, and complex IV pumps out an additional 2 H+. Thus, for each NADH, the consensus is that the mitochondrion pumps 10 H+. The FADH2 from succinate feeds into the electron transport chain at complex II (which is actually succinate dehydrogenase), bypassing complex I. Thus, for each FADH2, the consensus is that the mitochondrion pumps 6 H+.

Generation of ATP

Regardless of whether the reducing equivalents come from NADH or FADH2, the result of their being processed by the electron transport chain is a steep electrochemical H+ gradient across the inner membrane. The FoF1 ATPase, also located in the inner membrane, uses the energy in this inwardly directed H+ gradient to synthesize ATP from ADP and inorganic phosphate. In other words, the FoF1 ATPase usually functions as an ATP synthase. As outlined in the text on page 118 and in Figure 5-9, the ATP synthase consists of (1) a group of 10 to 12 c subunits that forms an H+ channel, which rotates like a turbine in the plane of the membrane as protons pass through the channel; (2) a shaft (consisting of the γ and ε subunits), which rotates with the turbine and extends into the mitochondrial matrix; the shaft also projects deeply into (3) a stationary globular structure that consists of three pairs of αβ subunits. In addition, one a subunit, two b subunits, and a δ subunit hold the complex together.

At any one time, the β subunit of one αβ pair is empty (β-empty), the β subunit of another binds ADP + Pi (β-ADP), and the β subunit of the third binds ATP (β-ATP). (Note: Pi is inorganic phosphate.) It is believed that each time three H+ pass through the ATP synthase, the turbine (10 to 12 c subunits) and the shaft (γ and ε subunits) rotate together by 120 degrees. This rotation brings the tip of the γ subunit into contact with a new pair of αβ subunits in the stationary F1 portion of the ATP synthase, causing this β subunit to shift from the β-ATP to the β-empty conformation—and release a just-synthesized ATP. Simultaneously, the previously empty β subunit shifts from the β-empty to the β-ADP conformation (ready to create a new ATP), and the β subunit that previously was binding ADP + Pi creates a new ATP by shifting to the β-ATP conformation. Thus, each time a trio of protons passes through the ATP synthase into the mitochondrial inner matrix, the shaft rotates by 120 degrees and completes the synthesis of one ATP molecule. A complete 360-degree rotation of the shaft (which would require nine protons) would generate three new ATP molecules, with each αβ pair passing through each of the three possible conformations. It is important to note that the above H+/ATP stoichiometry is inferred from the structure of the ATP synthase (three pairs of αβ subunits) as well as a wealth of biochemical experiments.

Ancillary Transport Processes

Even though the ATP synthase per se appears to have a stoichiometry of one ATP for every three protons, we have not addressed the overall process of ATP synthesis. As we shall see, the mitochondrion must accomplish two additional tasks in order to complete the process.

First, the mitochondrion must transport one Pi into its inner matrix for each ATP to be synthesized. This uptake of Pi appears to be accomplished by an image cotransporter. In other words, the mitochondrion must take up one H+ (previously pumped out by the electron transport chain) to energize the uptake of each Pi.

Second, the mitochondrion must import one ADP3− molecule for each ATP to be synthesized. In addition, the mitochondrion must export the newly synthesized ATP4− molecule. Both jobs are accomplished by the same ADP-ATP exchanger, which is energized in part by the electrical gradient (Δψ) that the electron transport chain generates along with the chemical H+ gradient as it pumps H+ out across the mitochondrial inner membrane.

Thus, the consensus is that the mitochondrion needs to import four H+ to synthesize one ATP molecule:

1. Three H+ to energize a 120-degree turn of the ATP synthase and thus convert one ADP + one Pi to one ATP.

2. One H+ to take up the Pi.

Overall ATP/NADH Stoichiometry

In summary, it appears that, for each NADH processed, the electron transport chain extrudes 10 protons. Furthermore, it appears that the ATP synthase and ancillary transporters can generate one ATP molecule from the inward movement of four protons. Thus, the overall ATP/NADH stoichiometry would be 2.5 ATP molecules for each NADH molecule. Because the processing of FADH2 results in the extrusion of only six protons, the overall ATP/FADH2 stoichiometry would be 1.5 ATP molecules for each FADH2 molecule. Keep in mind that these figures—which we use in examples throughout the text after the first three printings of the text—are the best current estimates; more about this below.

NADH Shuttle Mechanisms

An additional consideration is the access of NADH to complex I in the respiratory chain. In animal cells, the NADH must approach complex I from the matrix side of the mitochondrial inner membrane. This is not a problem for NADH generated inside the mitochondrial matrix by pyruvate dehydrogenase or the citric acid cycle (see Fig. 58-11). However, NADH generated by glycolysis (see Fig. 58-6A) cannot directly cross the mitochondrial inner membrane. As a result, animal cells use two complicated shuttle mechanisms to move the NADH reducing equivalents indirectly across the mitochondrial inner membrane.

Cells throughout most of the body—but not skeletal muscle or brain—use the malate-aspartate shuttle to bring NADH equivalents across the mitochondrial inner membrane into the mitochondrial matrix. The process, which is described in comprehensive biochemistry texts, involves six steps: (1) Malate dehydrogenase uses NADH and H+ to convert oxaloacetate to malate in the intermembrane space, regenerating NAD+ in the process. (2) The malate–α-ketoglutarate exchanger in the mitochondrial inner membrane imports the malate into the matrix. (3) Malate dehydrogenase in the mitochondrial matrix uses NAD+ to convert the newly imported malate back to oxaloacetate, regenerating NADH+ and H+ in the process. (4) Aspartate aminotransferase in the matrix converts the oxaloacetate and a glutamate to α-ketoglutarate and aspartate. (5) The aforementioned malate–α-ketoglutarate exchanger recycles the α-ketoglutarate back to the intermembrane space, and the glutamate-aspartate exchanger does the same for aspartate. (6) In the intermembrane space, aspartate aminotransferase converts the newly exited aspartate and α-ketoglutarate to glutamate and oxaloacetate, completing the cycle.

The net effect is to shuttle NADH indirectly into the matrix, where it can approach complex I. Using NADH shuttled in this way, the electron transport chain can pump 10 protons from the matrix into the intermembrane space, a number that can produce 2.5 ATP molecules.

Skeletal muscle and brain use a very different two-step method to process the NADH that is produced by glycolysis, the glycerol-3-phosphate shuttle: (1) The cytosolic enzyme glycerol-3-phosphate dehydrogenase uses NADH and H+ to convert dihydroxyacetone phosphate to glycerol-3-phosphate, regenerating NAD+ in the process. (2) The enzyme glycerol-3-phosphate dehydrogenase on the outer surface of the mitochondrial inner membrane converts glycerol-3-phosphate dehydrogenase in the intermembrane space to dihydroxyacetone phosphate, thereby regenerating the latter and releasing it into the intermembrane space. In the process, the dehydrogenase converts FAD to FADH2. Because this FADH2 enters the electron transport chain at complex III, it can fuel the extrusion of only six protons across the mitochondrial inner membrane. Thus, using reducing equivalents shuttled this way, the electron transport chain has an ATP/NADH stoichiometry of only 1.5 ATP molecules per NADH molecule.

Note that after the third printing of the book, we have used the values of 2.5 or 1.5 ATP per NADH to indicate that the energy yield depends on the shuttle system. In the brain, of course, we can simply refer to the value of 1.5 (see Fig. 11-10).

Our Confidence in the Cited Figures for the ATP/NADH Stoichiometry

Although the figures of 2.5 or 1.5 ATP molecules per NADH are the consensus, they should not be considered absolute at this time. Although investigators have invested considerable effort in attempting to determine the stoichiometry experimentally, the task is a daunting one for several reasons. First, some of the protons extruded by the electron transport chain can leak back into the mitochondrial matrix by pathways other than the ATP synthase. In fact, brown fat cells make use of this bypass to generate heat (see Fig. 57-6). Second, the proton traffic is computed from measured pH changes and values of buffering power that are sometimes difficult to know with certainty. Third, the precise stoichiometry of the image cotransporter depends on the precise pH values on either side of the mitochondrial inner membrane because the reaction image has a pK of ~6.8. Fourth, the intricacies of the ATP synthase have yet to be fully worked out (i.e., the stoichiometry may not always be precisely one ATP synthesized for every three protons entering through the ATP synthase per se).

V-type H Pump

The membranes surrounding such intracellular organelles as lysosomes, endosomes, secretory vesicles, storage vesicles, and the Golgi apparatus contain a so-called vacuolar-type (or V-type) H-ATPase that pumps H+ from the cytoplasm to the interior of the organelles. The low pH generated inside these organelles is important for sorting proteins, dissociating ligands from receptors, optimizing the activity of acid hydrolases, and accumulating neurotransmitters in vesicles. The apical membranes of certain renal tubule cells as well as the plasma membranes of certain other cells also have V-type H pumps that extrude H+from the cell. These V-type H pumps, unlike the gastric H-K pump, are independent of K+. Instead, the V-type H pump is similar to the hand-held lollipop–like structure of the F-type ATPase, with which it shares a significant—although low—level of amino-acid homology. For example, the hand of the V-type pump has only six subunits, but each is twice as large as a c subunit in the F-type ATPase.

ATP-binding cassette transporters can act as pumps, channels, or regulators

The so-called ABC proteins all have a motif in their amino-acid sequence that is an ATP-binding cassette (ABC). In humans, this family includes at least 49 members in seven subfamilies named ABCA through ABCG (Table 5-6). Some are pumps that presumably hydrolyze ATP to provide energy for solute transport. Some may hydrolyze ATP, but they do not couple the liberated energy to perform active transport. In other cases, ATP regulates ABC proteins that function as ion channels or regulators of ion channels or transporters.


ABC Transporters




ABCA (12)


ABCA1 (cholesterol transporter)

ABCB (11)


ABCB1 (MDR1 or P-glycoprotein 1)
ABCB4 (MDR2/3)
ABCB11 (bile salt export pump, BSEP)

ABCC (13)


ABCC1 (multidrug resistance–associated protein 1, MRP1)
ABCC8 (sulfonylurea receptor, SUR1)

ABCD (4)


ABCD1 (ALD; mediates uptake of fatty acids by peroxisomes)

ABCE (1)


ABCE1 (RNASELI; blocks ribonuclease L)

ABCF (3)


ABCF1 (lacks transmembrane domains)

ABCG (5)


ABCG2 (breast cancer resistance protein, BCRP; transports sulfated steroids, uric acid, xenobiotics)
ABCG5/ABCG8 (heterodimer of “half” ABCs that transport cholesterol)

*In parentheses is the number of human genes in the family.

The nomenclature of the MDRs is especially confusing because different and conflicting numbering systems have been used for different species; we use the human numbering system.

ABCA Subfamily

ABC1 (ABCA1) is an important transporter for mediating the efflux of phospholipids and cholesterol from macrophages and certain other cells.

MDR Subfamily

The multidrug-resistance transporters (MDRs) are ATPases and primary active transporters. The MDR proteins are tandem repeats of two structures, each of which has six membrane-spanning segments and a nucleotide-binding domain (NBD) that binds ATP. MDR1, also called P-glycoprotein, extrudes cationic metabolites and drugs across the cell membrane. The substrates of MDR1 appear to have little in common structurally, except that they are hydrophobic. A wide variety of cells express MDRs, including those of the liver, kidney, and gastrointestinal tract. MDR1 plays an important and clinically antagonistic role in cancer patients in that it pumps a wide range of anticancer drugs out of cancer cells, thereby rendering cells resistant to these drugs.

MRP/CFTR Subfamily

Another member of the ABC superfamily that is of physiological interest is the cystic fibrosis transmembrane conductance regulator (CFTR), which is mutated in the hereditary disease cystic fibrosis (Box 43-1). CFTR is a 170-kDa glycoprotein that is present at the apical membrane of many epithelial cells. CFTR functions as a low-conductance Cl channel as well as a regulator of other ion channels.

Like MDR1, CFTR has two membrane-spanning domains (MSD1 and MSD2), each composed of six membrane-spanning segments (Fig. 5-10). Also like MDR1, CFTR has two NBDs (NBD1 and NBD2). Unlike MDR1, however, a large cytoplasmic regulatory (R) domain separates the two halves of CFTR. The regulatory domain contains multiple potential protein kinase A and protein kinase C phosphorylation sites. Phosphorylation of these sites, under the influence of neurohumoral agents that control fluid and electrolyte secretion, promotes activation of CFTR. The binding of ATP to the NBDs also controls channel opening and closing. Thus, ATP regulates the CFTR Cl channel by two types of mechanisms: protein phosphorylation and interaction with the NBDs. imageN5-16


FIGURE 5-10 Cystic fibrosis transmembrane conductance regulator (CFTR). The CFTR Cl channel has two membrane-spanning domains (MSD1 and MSD2). A large cytoplasmic regulatory (R) domain separates the two halves of the molecule, each of which has an ATP-binding domain (NBD1 and NBD2). The most common mutation in cystic fibrosis is the deletion of the phenylalanine at position 508 (ΔF508) in the NBD1 domain. (Model modified from Riordan JR, Rommens JM, Kerem B, et al: Identification of the cystic fibrosis gene: Cloning and complementary DNA. Science 245:1066–1073, 1989.)


Regulation of the CFTR Channel by ATP

Contributed by Emile Boulpaep, Walter Boron

CFTR is phosphorylated by protein kinase A (PKA) at several sites within its R domain (eFig. 5-1). Modest phosphorylation causes a conformational change in the R domain that makes NBD1 accessible to ATP. Additional phosphorylation also makes NBD2 accessible to ATP. When ATP binds to NBD1 and is subsequently hydrolyzed, the channel opens, but it then rapidly closes once the ADP and phosphate dissociate (“flickery opening”). However, if a second ATP binds to NBD2, the channel is stabilized in its open state (“long opening”). ATP hydrolysis at NBD2 terminates the long opening and is thus necessary for CFTR to return to its closed state. Dephosphorylation of the R domain by protein phosphatases returns CFTR to its resting state. The control of CFTR by ATP hydrolysis is reminiscent of the control of G-protein activity by GTP hydrolysis (see p. 53).

The R domain of CFTR can also be phosphorylated by PKC. PKC enhances the stimulatory effect of PKA on CFTR Cl transport, but alone it appears to have little direct effect on CFTR function.


EFIGURE 5-1 A widely accepted model of how ATP regulates CFTR both by phosphorylation and ATP hydrolysis. The channel is closed in the three channel states in the top row (nothing bound to the NBDs). The channel is in a “flickery” open state as it makes the transition from the top row to the middle row (NBD1 occupied). Finally, the channel is in a stable or “long” open state in the third row (NBD1 and NBD2 occupied). Pi, inorganic phosphate. (Data from Gadsby DC, Dousmanis AG, Nairn AC: ATP hydrolysis cycles the gating of CFTR Cl channels. Acta Physiol Scand Suppl 643:247–256, 1998.)

Cotransporters, one class of secondary active transporters, are generally driven by the energy of the inwardly directed Na+ gradient

Like pumps or primary active transporters, secondary active transporters can move a solute uphill (against its electrochemical gradient). However, unlike the pumps, which fuel the process by hydrolyzing ATP, the secondary active transporters fuel it by coupling the uphill movement of one or more solutes to the downhill movement of other solutes. The two major classes of secondary active transporters are cotransporters (or symporters) and exchangers (or antiporters). Cotransporters are intrinsic membrane proteins that move the “driving” solute (the one whose gradient provides the energy) and the “driven” solutes (which move uphill) in the same direction.

Na/Glucose Cotransporter

The Na/glucose cotransporter (SGLT) is located at the apical membrane of the cells that line the proximal tubule and small intestine (Fig. 5-11A). The SGLTs, which belong to the SLC5 family (see Table 5-4), consist of a single subunit with 14 membrane-spanning segments. SGLT2 moves one Na+ ion with each glucose molecule (i.e., 1 : 1 stoichiometry of Na+ to glucose), whereas the SGLT1 isoform moves two Na+ions with each glucose molecule. Drugs that inhibit SGLT2 are useful in treating type 2 diabetes in adults because these agents reduce glucose reabsorption in the kidney, thereby lowering plasma [glucose]. imageN5-17


FIGURE 5-11 Representative cotransporters.


Clinical Use of SGLT2 Inhibitors

Contributed by Emile Boulpaep

SGLT2 inhibitors are novel “glucuretics” that have been approved for the treatment of type 2 diabetes. The low-affinity, high-capacity cotransporter SGLT2 reabsorbs the bulk of the filtered glucose in the S1/S2 segments of the proximal convoluted tubule, whereas the low-affinity, high-capacity cotransporter SGLT1 in the S3 segment reabsorbs the remainder. Thus, specific SGLT2 inhibition causes the bulk of filtered glucose to be excreted in the urine, whereas about 10% of filtered glucose is still reabsorbed by SGLT1, so that [glucose]plasma is prevented from falling below normal. However, in clinical practice, SGLT2 inhibitors inhibit only 30% to 50% of renal glucose reabsorption.

Phlorizin, the 2′-glucoside of phloretin, is a natural compound in the bark of fruit trees and a competitive inhibitor of both SGLT1 and SGLT2. Before the discovery of insulin, phlorizin was used in the treatment of diabetes, although the compound is poorly absorbed by the gastrointestinal tract and not stable. The U.S. Food and Drug Administration approved two specific SGLT2 inhibitors for oral use in the treatment of type 2 diabetes in adults: canagliflozin and dapagliflozin (Farxiga®). Urinary tract infections and genital fungal infections are common adverse effects of this treatment due to chronically high glucose concentration in the urine.


Abdul-Ghani MA, DeFronzo RA. Lowering plasma glucose concentration by inhibiting renal sodium-glucose co-transport. J Intern Med. 2014;276(4):352–363.

For an Na/glucose cotransporter with 1 : 1 stoichiometry, the overall driving force is the sum of the electrochemical potential energy difference for Na+ and the chemical potential energy difference for glucose. Thus, the highly favorable, inwardly directed Na+ electrochemical gradient can drive the uphill accumulation of glucose from the lumen of the kidney tubule or gut into the cell. Figure 5-12 shows how the Na+gradient drives glucose accumulation into membrane vesicles derived from the brush border of renal proximal tubules. Equilibrium is achieved when the electrochemical potential energy difference for Na+ in one direction is balanced by the chemical potential energy difference for glucose in the opposite direction:




FIGURE 5-12 Na+-driven glucose uptake into brush-border membrane vesicles. imageN5-22


Using Membrane Vesicles to Study Glucose Transport

Contributed by Gerhard Giebisch, Erich Windhager, Emile Boulpaep, Walter Boron

We describe the membrane-vesicle technique in imageN33-5.Figure 5-12 illustrates the use of this technique to explore how the Na+ gradient affects glucose uptake. The vesicles are made from brush-border membrane vesicles (i.e., made from the apical membrane of the proximal tubule). In the absence of Na+ in the experimental medium, glucose enters renal brush-border membrane vesicles slowly until reaching an equilibrium value (green curve in the central graph of Fig. 5-12). At this point, internal and external glucose concentrations are identical. The slow increase in intravesicular [glucose] occurs by diffusion in the absence of Na+. In contrast, adding Na+ to the external medium establishes a steep inwardly directed Na gradient, which dramatically accelerates glucose uptake (red curve in the central graph of Fig. 5-12). The result is a transient “overshoot” during which glucose accumulates above the equilibrium level. Thus, in the presence of Na+, the vesicle clearly transports glucose uphill. Similar gradients of other cations, such as K+, have no effect on glucose movement beyond that expected from diffusion alone.

A negative cell voltage can also drive Na/glucose cotransport, even when there is no Na+ gradient. In experiments in which the internal and external Na+ concentrations are the same, making the inside of the vesicles electrically negative accelerates glucose uptake (not shown in Fig. 5-12).

In vesicle experiments performed on vesicles made from the basolateral membrane, the overshoot in intravesicular [glucose] does not occur, even in the presence of an inward Na+ gradient. Thus, the Na/glucose cotransporter is restricted to the apical membrane.

We can express ΔũNa in terms of the Na+ concentrations and membrane voltage and can express Δµglucose in terms of the glucose concentrations. imageN5-18 If we substitute these expressions into Equation 5-17, we derive the following relationship for the maximal glucose concentration gradient that can be generated by a given electrochemical potential energy difference for Na+:




Maximal Glucose Gradient Achievable by SGLT1 and SGLT2

Contributed by Emile Boulpaep, Walter Boron


As noted on page 121 (see Equation 5-17; shown here as Equation NE 5-14), SGLT2—the Na/glucose cotransporter with a 1 : 1 stoichiometry of Na+ to glucose—is in equilibrium when


(NE 5-14)

ΔũNa is the electrochemical energy difference across the cell membrane for Na+, and Δµglucose is the chemical energy difference for glucose (because glucose has no charge, the electrical energy difference for glucose across the membrane is zero). Starting from the definition of electrochemical energy difference in Equation 5-6, we can express ΔũNa in terms of the Na+ concentrations and membrane potential:


(NE 5-15)

Similarly, we can express Δµglucose in terms of the glucose concentrations:


(NE 5-16)

If we substitute these last two expressions into Equation NE 5-14, we obtain the following equation, which describes the relationships between the Na+ and glucose concentrations when SGLT2 is in equilibrium:


(NE 5-17)


As noted in the text in Equation 5-19 (shown here as Equation NE 5-18), SGLT1—the Na/glucose cotransporter with a 2 : 1 stoichiometry of Na+ to glucose—is in equilibrium when


(NE 5-18)

If we substitute Equations NE 5-15 and NE 5-16 into Equation NE 5-18, we obtain the following equation, which describes the relationships between the Na+ and glucose concentrations when SGLT1 is in equilibrium:


(NE 5-19)

We can use Equation NE 5-17 (for SGLT2) and Equation 5-6 (for SGLT1) to compute the maximum achievable glucose gradients. Simply insert the values for [Na+]i, [Na+]o, and Vm as discussed on page 121.

In an epithelial cell that has a 10-fold Na+ concentration gradient and a 60-mV inside-negative voltage across the apical membrane, the Na+ electrochemical gradient can generate a 10 × 101, or 100-fold, glucose concentration gradient across the plasma membrane. In other words, the 10 : 1 Na+ concentration gradient buys a 10-fold glucose gradient, and the Vm of −60 mV buys another 10-fold. However, the leakage of glucose out of the cell by other pathways at the basolateral membrane prevents the Na/glucose cotransporter from coming to equilibrium.

An Na/glucose cotransporter with 2 : 1 stoichiometry is capable of generating an even larger concentration gradient for glucose across the plasma membrane. Such a cotransporter would be in equilibrium when



The maximal glucose gradient is



In the same epithelial cell with a 10-fold Na+ concentration gradient and a Vm of −60 mV, the Na+ electrochemical gradient can generate a glucose concentration gradient of 102 × 102, or 10,000-fold! In other words, the 10 : 1 Na+ concentration gradient—when squared for two Na+ ions—buys a 100-fold glucose gradient, and the −60 mV membrane voltage—when multiplied by two for the effective charge on two Na+ions—buys another 100-fold.

Because the cotransporter protein has specific sites for binding Na+ and glucose and because the number of transporters is fixed, the rate of transport by SGLT is a saturable function of the glucose and Na+concentrations.

Na+-Driven Cotransporters for Organic Solutes

Functionally similar, but structurally distinct from one another, are a variety of Na+ cotransporters in the proximal tubule and small intestine. Na+-driven amino-acid transporters (see Fig. 5-11B) belong to both the SLC6 and SLC38 families (see Table 5-4). SLC13 includes Na+-coupled cotransporters for monocarboxylates, dicarboxylates, and tricarboxylates; SLC5 includes Na+-coupled cotransporters for monocarboxylates.

Na/HCO3 Cotransporters

The Na/HCO3 cotransporters (NBCs) belong to the SLC4 family, imageN5-19 and are a key group of acid-base transporters. In the basolateral membranes of certain epithelial cells, the electrogenic NBCs (NBCe1/e2, e for electrogenic) appear to operate with an image stoichiometry of 1 : 3 (see Fig. 5-11D) and—for typical ion and voltage gradients—mediate electrogenic image efflux. Here, these NBCs mediate image absorption into the blood. In most other cells, these same two transporters operate with a stoichiometry of 1 : 2—probably because of the absence of a key protein partner—and mediate the electrogenic image influx (see Fig. 5-11E). Finally, the electroneutral NBCs (NBCn1/n2, n for electroneutral) operate with an image stoichiometry of 1 : 1 (see Fig. 5-11F) and also mediate image influx. In these last two cases, the Na+ electrochemical gradient drives the uphill accumulation of image, which is important for epithelial image secretion and for the regulation of intracellular pH (pHi) to relatively alkaline values.


image Transporters in the SLC4 Family

Contributed by Emile Boulpaep, Walter Boron

So far, investigators have identified ten human genes in the SLC4 family of solute carriers. These genes include three that encode Cl-HCO3 exchangers (the so-called anion exchangers, or AE1, AE2, and AE3), and five that encode Na+-coupled image transporters. In addition, one gene encodes an Na/borate cotransporter, and another gene encodes a protein—termed AE4—of controversial function. For a discussion of these subjects, consult the review by Romero and colleagues.

The five Na+-coupled image transporters include the two electrogenic Na/HCO3 cotransporters (NBCe1 and NBCe2), two electroneutral Na/HCO3 cotransporters (NBCn1 and NBCn2), and a single Na+-driven Cl-HCO3 exchanger (NDCBE).

NBCe1 and NBCe2 appear to be able to transport Na+ and image in an image stoichiometry of either 1 : 3 (as in the renal proximal tubule; see Fig. 39-4A) or 1 : 2 (as in most other cells, including the pancreatic duct; Fig. 43-6). Preliminary evidence from the laboratory of Boron suggests that NBCe1 and NBCe2—at least when operating with a stoichiometry of 1 : 2—in fact transport image rather than image. Of course, image arises from image in the reaction image + H+, which has a pK of ~10.3.

NDCBE appears to move 1 Na+ and 2 image ions into the cell in exchange for 1 Cl (which moves out of the cell). Extensive kinetic data are consistent with the hypothesis that the NDCBE from the squid axon in fact transports either Na+ plus image (in exchange for Cl) or perhaps the image ion pair (in exchange for Cl). The image forms rapidly, and reversibly, from Na+ and imageimage.

Work on NBCn2 (Parker et al, 2008) is consistent with the hypothesis that this electroneutral Na/HCO3 cotransporter—which appears to transport 1 Na+ and 1 image into the cell—may in fact mediate the uptake of 1 Na+ and 1 image in exchange for 1 intracellular image. The net effect of exchanging 1 extracellular image for 1 intracellular image would be the uptake of 1 image. In fact, it is intriguing to speculate that all members of the SLC4 family are in fact exchangers, and that the proteins that appear to mediate cotransport in fact mediate an exchange that “nets out” as apparent cotransport.

In addition to the SLC4 family, several members of the SLC26 family can also carry image, although these SLC26 proteins tend to be less selective in the anions that they transport. For a review of this gene family, consult Alper and Sharma.


Alper SL, Sharma AK. The SLC26 gene family of anion transporters and channels. Mol Aspects Med. 2013;34:494–515.

Boron WF. Intracellular-pH-regulating mechanism of the squid axon: Relation between the external Na+ and image dependences. J Gen Physiol. 1985;85:325–345.

Boron WF, De Weer P. Intracellular pH transients in squid giant axons caused by CO2, NH3, and metabolic inhibitors. J Gen Physiol. 1976;67:91–112.

Boron WF, Knakal RC. Intracellular pH-regulating mechanism of the squid axon. Dependence on extracellular pH. J Gen Physiol. 1992;99:817–837.

Boron WF, Knakal RC. Intracellular pH-regulating mechanism of the squid axon: Interaction between DNDS and extracellular Na+ and imageJ Gen Physiol. 1989;93:123–150.

Boron WF, Russell JM. Stoichiometry and ion dependencies of the intracellular-pH-regulating mechanism in squid giant axons. J Gen Physiol. 1983;81:373–399.

Hogan EM, Cohen MA, Boron WF. K+- and image-dependent acid–base transport in squid giant axons: Base efflux. J Gen Physiol. 1995;106:821–844.

Hogan EM, Cohen MA, Boron WF. K+- and image-dependent acid–base transport in squid giant axons: Base influx. J Gen Physiol. 1995;106:845–862.

Parker MD, Musa-Aziz R, Rojas JD, et al. Characterization of human SLC4A10 as an electroneutral Na/HCO3 cotransporter (NBCn2) with Cl self-exchange activity. J Biol Chem. 2008;283:12777–12788.

Romero MF, Fulton CM, Boron WF. The SLC4 family of image transporters. Pflugers Arch. 2004;447:495–509.

Na+-Driven Cotransporters for Other Inorganic Anions

Important examples of other Na+-driven cotransporters include the Na/I cotransporter (SLC5A5), the sulfate cotransporter (SLC13 family; see Table 5-4), and the inorganic phosphate cotransporter (NaPi; see Fig. 5-11C)—which are members of the SLC17, SLC20, and SLC34 families.

Na/K/Cl Cotransporter

The three types of cation-coupled Cl cotransporters all belong to the SLC12 family. The first is the Na/K/Cl cotransporter (NKCC), which harnesses the energy of the inwardly directed Na+ electrochemical gradient to drive the accumulation of Cl and K+ (see Fig. 5-11G). One variant of this cotransporter, NKCC1 (SLC12A2), is present in a wide variety of nonepithelial cells (see p. 131), as well as in the basolateral membranes of some epithelial cells. Another variant of this cotransporter, NKCC2 (SLC12A1), is present on the apical membrane of cells lining the thick ascending limb of the loop of Henle in the kidney (see p. 757). A characteristic of the NKCCs is that they are inhibited by furosemide and bumetanide, which are called loop diuretics because they increase urine flow by inhibiting transport at the loop of Henle. Because of its sensitivity to bumetanide, NKCC is sometimes called the bumetanide-sensitive cotransporter (BSC).

Na/Cl Cotransporter

The second type of cation-coupled Cl cotransporter is found in the apical membrane of the early distal tubule of the kidney (see p. 758). This K+-independent Na/Cl cotransporter (NCC or SLC12A3) is blocked by thiazide diuretics rather than by loop diuretics (see Fig. 5-11H). For this reason, NCC has also been called the thiazide-sensitive cotransporter (TSC).

K/Cl Cotransporter

The third type of cation-coupled Cl cotransporter is the Na+-independent K/Cl cotransporter (KCC) family (KCC1 to KCC4 or SLC12A4 to SLC12A7). Because the Na-K pump causes K+ to accumulate inside the cell, the K+ electrochemical gradient is outwardly directed across the plasma membrane (see Fig. 5-11I). In addition, pathways such as the Na/K/Cl cotransporter and the Cl-HCO3 anion exchanger (see below) bring Cl into the cell, so that in most cells the Cl electrochemical gradient is also outwardly directed. Thus, the net driving force acting on the KCC favors the exit of K+ and Cl from the cell (see pp. 131–132).

H+-Driven Cotransporters

Although the majority of known cotransporters in animal cells are driven by the inward movement of Na+, some are instead driven by the downhill, inward movement of H+. The H/oligopeptide cotransporter PepT1 and related proteins are members of the SLC15 family. PepT1 is electrogenic and responsible for the uptake of small peptides (see Fig. 5-11J) from the lumen into the cells of the renal proximal tubule (see pp. 777–778) and small intestine (see p. 923). The H+-driven amino-acid cotransporters (e.g., PAT1) are members of the SLC36 family. The monocarboxylate cotransporters, such as MCT1, are members of the SLC16 family. They mediate the electroneutral, H+-coupled flux of lactate, pyruvate, or other monocarboxylates across the cell membranes of most tissues in the body (see Fig. 5-11K). In the case of lactate, MCT1 can operate in either the net inward or net outward direction, depending on the lactate and H+ gradients across the cell membrane. MCT1 probably moves lactate out of cells that produce lactate by glycolysis but into cells that consume lactate. The divalent metal-ion cotransporter (DMT1), a member of the SLC11 family, couples the influx of H+ to the influx of ferrous iron (Fe2+) as well as to the influx of a variety of other divalent metals, some of which (Cd2+, Pb2+) are toxic to cells (see Fig. 5-11L). DMT1 is expressed at high levels in the kidney and proximal portions of the small intestine.

Exchangers, another class of secondary active transporters, exchange ions for one another

The other major class of secondary active transporters is the exchangers, or antiporters. Exchangers are intrinsic membrane proteins that move one or more “driving” solutes in one direction and one or more “driven” solutes in the opposite direction. In general, these transporters exchange cations for cations or anions for anions.

Na-Ca Exchanger

The nearly ubiquitous Na-Ca exchangers (NCXs) belong to the SLC8 family (see Table 5-4). They most likely mediate the exchange of three Na+ ions per Ca2+ ion (Fig. 5-13A). NCX is electrogenic and moves net positive charge in the same direction as Na+. Under most circumstances, the inwardly directed Na+ electrochemical gradient across the plasma membrane drives the uphill extrusion of Ca2+ from the cell. Thus, in concert with the plasma-membrane Ca pump, this transport system helps maintain the steep, inwardly directed electrochemical potential energy difference for Ca2+ that is normally present across the plasma membrane of all cells.


FIGURE 5-13 Representative exchangers.

NCX uses the inwardly directed Na+ electrochemical gradient to drive the secondary active efflux of Ca2+. With a presumed stoichiometry of three Na+ per Ca2+, the effectiveness of the Na+ electrochemical gradient as a driving force is magnified; thus, NCX is at equilibrium when the Ca2+ electrochemical gradient is balanced by three times the Na+ electrochemical gradient:






In a cell with a 10-fold Na+ concentration gradient and a Vm of −60 mV, the electrochemical potential energy difference for Na+ can buy a Ca2+ concentration gradient of 103 × 101, or 10,000-fold, which is the Ca2+gradient across most cell membranes. Thus, the effect of the 10-fold inward Na+ concentration gradient is cubed and can account for a 103-fold Ca2+ concentration gradient across the plasma membrane. In addition, the stoichiometry of three Na+ per Ca2+ produces a net inflow of one positive charge per transport cycle. Thus, the 60-mV inside-negative Vm acts as the equivalent driving force to another 10-fold concentration gradient.

Na-H Exchanger

The Na-H exchangers (NHEs), which belong to the SLC9 family (see Table 5-4), mediate the 1 : 1 exchange of extracellular Na+ for intracellular H+ across the plasma membrane (see Fig. 5-13B). One or more of the nine imageN5-20 known NHEs are present on the plasma membrane of almost every cell in the body. Through operation of NHEs, the inwardly directed Na+ electrochemical gradient drives the uphill extrusion of H+ from the cell and raises pHi. The ubiquitous NHE1, which is present in nonepithelial cells as well as on the basolateral membranes of epithelia, plays a major role in pHi regulation (see p. 127) and cell volume (see p. 131). NHE3 is present at the apical membranes of several epithelia, where it plays a major role in acid secretion (see p. 827) and Na+ absorption.


The NHEs

Contributed by Emile Boulpaep, Walter Boron

Consult the review by Donowitz et al for an overview of the NHE family of Na-H exchangers (also known as the SLC9 family of “solute-linked carriers”).


Donowitz M, Ming Tse C, Fuster D. SLC9/NHE gene family, a plasma membrane and organellar family of Na+/H+ exchangers. Mol Aspects Med. 2013;34:236–251.

Another cation exchanger—the multidrug and toxin extrusion transporter (MATE; SLC47 family)—mediates the uptake of H+ across the apical membrane of renal proximal tubule cells and hepatocytes in exchange for organic cations, which MATE secretes into the lumen.

Na+-Driven Cl-HCO3 Exchanger

A second Na+-coupled exchanger that is important for pHi regulation is the Na+-driven Cl-HCO3 exchanger (NDCBE), a member of the SLC4 family (see Table 5-4). This electroneutral transporter couples the movement of one Na+ ion and the equivalent of two image ions imageN5-19 in one direction to the movement of one Cl ion in the opposite direction (see Fig. 5-13C). NDCBE uses the inwardly directed Na+electrochemical gradient to drive the uphill entry of image into the cell. Thus, like the Na-H exchangers, NDCBE helps keep pHi relatively alkaline. NDCBE may also participate in Na+ reabsorption (i.e., movement from lumen to blood) across certain tubule cells in the kidney.

Cl-HCO3 Exchanger

A third group of exchangers that are involved in acid-base transport are the Cl-HCO3 exchangers (see Fig. 5-13D) that function independently of Na+. These may be members of either the SLC4 or the SLC26 families. Virtually all cells in the body express one of the three electroneutral SLC4 or Cl-HCO3 exchangers, also known as anion exchangers (AE1 to AE3). AE1 is important for transporting image into the red blood cell in the lung and out of the red blood cell in peripheral tissues (see p. 656). In other cells, where the inwardly directed Cl gradient almost always drives image out of the cell, AE2 and AE3 play important roles in pHi regulation (see p. 127) by tending to acidify the cell. Moreover, the uptake of Cl often plays a role in the regulation of cell volume (see p. 130).

Several members of the SLC26 family can function as Cl-HCO3 exchangers and thereby play important roles in epithelial Cl and image transport. Because the stoichiometry need not be 1 : 1, SLC26 transport can be electrogenic. As described next, even SLC26 proteins that exchange Cl for image also transport a wide variety of other anions.

Other Anion Exchangers

A characteristic of the SLC26 family is their multifunctionality. For example, SLC26A6—present in the apical membranes of renal proximal tubule cells—can mediate Cl-formate exchange and Cl-oxalate exchange (see Fig. 5-13E). These activities appear to be important for the secondary active uptake of Cl and for secretion of oxalate. Pendrin not only mediates Cl-HCO3 exchange but may also transport I, which may be important in the thyroid gland (see p. 1006).

Anion exchangers other than those in the SLC4 and SLC26 families also play important roles. The organic anion–transporting polypeptides (OATPs) are members of the SLC21 family. In the liver, OATPs mediate the uptake of bile acids, bilirubin, and the test substrate bromsulphthalein. Another member of the SLC21 family is the prostaglandin transporter (PGT), which mediates the uptake of prostanoids (e.g., prostaglandins E2 and F and thromboxane B2).

The organic anion transporters (OATs) are members of the diverse SLC22 family. The OATs—by exchange or facilitated diffusion—mediate the uptake of endogenous organic anions (see Fig. 5-13F), as well as drugs, including penicillin and the test substrate p-aminohippurate. URAT1, another SLC22 member, is an exchanger that mediates urate transport in the renal proximal tubule. Surprisingly, the OCT transporters that mediate the facilitated diffusion of organic cations (see p. 115) are also members of SLC22.