Voet D., Voet Ju.G. Biochemistry

Подождите немного. Документ загружается.

membrane without any accompanying leakage of small

molecules or ions.

b. Eukaryotes Express a Variety of

Glucose Transporters

The erythrocyte glucose transporter, known also as

GLUT1 (for glucose transporter 1) has a highly conserved

amino acid sequence (98% sequence identity between hu-

mans and rats), which suggests that all segments of this pro-

tein are functionally significant. GLUT1 is expressed in

most tissues, although in liver and muscle, tissues that are

highly active in glucose transport, it is present in only tiny

amounts. Three other glucose transporters, GLUT2,

GLUT3, and GLUT4, have been well characterized

(GLUT5 was originally thought to be a glucose transporter

but was later shown to be a fructose transporter).They are

40 to 65% identical to GLUT1 but have different tissue

distributions. For example,GLUT2 is prominent in pancre-

atic ␤ cells (which secrete insulin in response to increased

[glucose] in blood; Section 18-3F), liver (where its defects

result in symptoms resembling Type I glycogen storage dis-

ease; Section 18-4), and the intestine (which absorbs di-

etary glucose; Section 20-4A); GLUT3 is expressed in neu-

rons and the placenta, and GLUT4 occurs mainly in muscle

and fat cells. Note that the tissue distributions of these glu-

cose transporters correlate with the response of these tis-

sues to insulin: Liver is unresponsive to insulin (liver func-

tions, in part, to maintain the level of blood glucose; Section

18-3Fb), whereas muscle and fat cells take up glucose when

stimulated by insulin. Analysis of the human genome has

identified eight other members of the GLUT family,

GLUT6 through GLUT12 and HMIT (for H

⫹

-coupled

myo-inositol transporter), although they have yet to be

well characterized. All of them are members of the major

facilitator superfamily (MFS).

c. Cellular Glucose Uptake Is Regulated through

the Insulin-Sensitive Exocytosis/Endocytosis

of Glucose Transporters

Insulin stimulates fat and muscle cells to take up glucose.

Within 2 or 3 min after the administration of insulin to fat

cells, the J

max

for passive-mediated glucose transport into

these cells increases 20- to 30-fold, whereas the K

remains

constant. On withdrawal of the insulin, the rate of glucose

uptake returns to its basal level within 20 min to 2 h de-

pending on conditions. Neither the increase nor the de-

crease in the rate of glucose transport is affected by the

presence of protein synthesis inhibitors, so that these ob-

servations cannot be a consequence of the synthesis of new

glucose transporter or of a protein that inhibits it. How,

then, does insulin regulate glucose transport?

GLUT4 is the dominant glucose transporter in skeletal

muscle and adipose (fat) cells. In their basal state, these

cells store most of their GLUT4 in specialized GLUT4

storage vesicles. On insulin stimulation, these vesicles fuse

with the plasma membrane in a process known as exocyto-

sis (Fig. 20-11). The consequent increased number of

Section 20-2. Kinetics and Mechanisms of Transport 751

Figure 20-10 Alternating conformation model for glucose

transport. Such a system is also known as a “gated pore.” [After

Baldwin, S.A. and Lienhard, G.E., Trends Biochem. Sci. 6, 210

(1981).]

See the Animated Figures

Figure 20-11 Regulation of glucose uptake in muscle and fat

cells. Regulation is mediated by the insulin-stimulated exocytosis

(the opposite of endocytosis; Section 12-5Bc) of membranous

vesicles containing GLUT4 glucose transporters (left). On insulin

withdrawal, the process reverses itself through endocytosis

(right).

See the Animated Figures

Glucose

Binding

Dissociation

Recovery Transport

Glucose

GLUT4

Exocytosis

Endocytosis

Membranous

vesicle

Stimulation

by insulin

Plasma

membrane

JWCL281_c20_744-788.qxd 6/4/10 12:14 PM Page 751

cell-surface glucose transporters results in a proportional

increase in the cell’s glucose uptake rate. On insulin with-

drawal, the process is reversed through the endocytosis

of plasma membrane-embedded glucose transporters.

The deletion or mutation of GLUT4’s N-terminal eight

residues, particularly Phe 5, causes this transporter to ac-

cumulate in the plasma membrane. A Leu-Leu sequence

and an acidic motif near GLUT4’s C-terminus are likewise

essential for its sequestration by the cell’s endocytotic ma-

chinery. The way in which insulin controls this system,

which accounts for most of insulin’s effects on muscle and

fat cells, is imperfectly understood. However, it is clear

that this mechanism involves a tyrosine phosphorylation

cascade that is triggered by the binding of insulin to the in-

sulin receptor (Section 19-3Ac and Fig. 19-67) and includes

the activation of a class IA phosphoinositide 3-kinase

(PI3K; Section 19-4Da).

F. K

⫹

Channels: Ion Discrimination

Potassium ions diffuse from the cytoplasm (where [K

⫹

] ⬎

100 mM) to the extracellular space (where [K

⫹

] ⬍ 5 mM)

through transmembrane proteins known as K

⫹

channels, a

process that underlies numerous important biological

processes including maintenance of cellular osmotic bal-

ance, neurotransmission (Section 20-5), and signal transduc-

tion (Chapter 19). Although there is a large diversity of K

⫹

channels, even within single organisms, all of them have

similar sequences, exhibit comparable permeability charac-

teristics, and most importantly, are at least 10,000-fold

more permeable to K

⫹

than Na

⫹

. Since this high selectivity

(around the same as that of valinomycin; Section 20-2Cb)

implies energetically strong interactions between K

⫹

and the

protein, how can the K

⫹

channel maintain its observed nearly

diffusion-limited throughput rate of up to 10

ions per second

(a 10

-fold greater rate than that of valinomycin)?

a. The X-Ray Structure of KcsA Reveals the Basis

of K

⫹

Channel Selectivity

KcsA, the K

⫹

channel from Streptomyces lividans, is a

tetramer of identical 158-residue subunits. The X-ray

structure of its N-terminal 125-residue segment, deter-

mined by Roderick MacKinnon, reveals that each KcsA

subunit forms two nearly parallel transmembrane helices

that are inclined ⬃25° from the normal to the membrane

plane and which are connected by an ⬃20-residue pore

region (Fig. 20-12a). As is true of all known K

⫹

channels,

four such subunits associate to form a 4-fold rotationally

symmetric assembly surrounding a central pore. The four

inner (C-terminal) helices, which largely form the pore,

pack against each other near the cytoplasmic side of the

membrane much like the poles of an inverted teepee. The

four outer helices, which face the lipid bilayer, buttress

the inner helices but do not contact the adjacent outer he-

lices. The pore regions, which each consist of a so-called

turret, pore helix, and selectivity filter, occupy the open

extracellular end of the teepee, with the pore helices

fitting in between its poles. Several K

⫹

ions and ordered

water molecules are seen to occupy the central pore (Figs.

20-12b and 20-13a).

The 45-Å-long central pore has variable width: It starts

at its cytoplasmic side (Fig. 20-12b, bottom) as an ⬃6-Å-

diameter and 18-Å-long tunnel, the so-called internal

pore, whose entrance is lined with four anionic side chains

that presumably help exclude anions (red area at the bot-

tom of Fig. 20-12b).The internal pore then widens to form

a cavity ⬃10 Å in diameter. These regions of the central

pore are both wide enough so that a K

⫹

ion could move

through them in its hydrated state. However, the upper

part of the pore, the so-called selectivity filter, narrows to

3 Å, thereby forcing a transiting K

⫹

ion to shed its waters

of hydration.The walls of the internal pore and the cavity

are lined with hydrophobic groups that interact minimally

with diffusing ions (yellow area of the pore in Fig. 20-

12b). However, the selectivity filter (red area of the pore

at the top of Fig. 20-12b) is lined with closely spaced main

chain carbonyl oxygens of residues (Fig. 20-13a, top) that

are highly conserved in all K

⫹

channels (their so-called

signature sequence, TVGYG) and whose mutations dis-

rupt the ability of the channel to discriminate between K

⫹

and Na

⫹

ions.

What is the function of the cavity? Energy calculations

indicate that an ion moving through a narrow transmem-

brane pore must surmount an energy barrier that is maximal

at the center of the membrane. The existence of the cavity

reduces this electrostatic destabilization by surrounding the

ion with polarizable water molecules (Fig. 20-12c). In addi-

tion, the C-terminal ends of the four pore helices point di-

rectly at the center of the cavity, so that their helix dipoles

impose a negative electrostatic potential on the cavity that

lowers the electrostatic barrier facing a cation crossing a

lipid bilayer.

Remarkably, the K

⫹

ion occupying the cavity is liganded

by 8 ordered water molecules located at the corners of a

square antiprism (a cube with one face twisted by 45° with

respect to the opposite face) in which the K

⫹

ion is cen-

tered (Fig. 20-13a, bottom; K

⫹

in aqueous solution was

known to have such an inner hydration shell but it had

never before been visualized). The K

⫹

ion is precisely cen-

tered in the cavity but yet its liganding water molecules are

not in van der Waals contact with the walls of the cavity. In-

deed, there is room in the cavity for ⬃40 additional water

molecules although they are unseen in the X-ray structure

because they are disordered. This disorder arises because

the cavity is lined with hydrophobic groups (mainly the

side chains of Ile 100 and Phe 103; Fig. 20-13a) that interact

but weakly with water molecules, thus allowing them to in-

teract freely with the K

⫹

ion so as to form an outer hydra-

tion shell. What, then, holds the hydrated K

⫹

ion in place?

Apparently, it is very weak indirect hydrogen bonds involv-

ing such protein groups as the hydroxyl group of Thr 107

and possibly carbonyl O atoms from the pore and inner he-

lices. The absence of such an ordered hydration complex

when Na

⫹

rather than K

⫹

occupies the cavity is indicative

of a precise geometric match between the hydrated K

⫹

and

the cavity (the ionic radii of Na

⫹

and K

⫹

are 0.95 Å and

752 Chapter 20. Transport Through Membranes

JWCL281_c20_744-788.qxd 3/17/10 1:48 PM Page 752

1.33 Å, respectively).The cavity thereby provides a high ef-

fective K

⫹

concentration (⬃2M) at the center of the mem-

brane and positions the K

⫹

ion on the pore axis ready to

enter the selectivity filter.

How does the K

⫹

channel discriminate so acutely be-

tween K

⫹

and Na

⫹

ions? The main chain O atoms lining the

selectivity filter form a stack of rings (Fig. 20-13a, top) that

provide a series of closely spaced sites of appropriate di-

mensions for coordinating dehydrated K

⫹

ions but not the

smaller Na

⫹

ions. If the observed diameter of the selectiv-

ity filter is rigidly maintained, it would make the energy of

a dehydrated Na

⫹

in the selectivity filter considerably

Section 20-2. Kinetics and Mechanisms of Transport 753

Figure 20-12 X-ray structure of the KcsA K

⫹

channel.

(a) Ribbon diagram of the tetramer as viewed from within the

plane of the membrane with the cytoplasm below and the

extracellular region above. The protein’s 4-fold axis of rotation is

vertical and each of its identical subunits is differently colored.

(b) A cutaway diagram viewed similarly to Part a in which the

⫹

channel is represented by its solvent-accessible surface. The

surface is colored according to its physical properties, with

negatively charged areas red, uncharged areas white, positively

charged areas blue, and hydrophobic areas of the central pore

yellow. K

⫹

ions are represented by green spheres. (c) A schematic

diagram indicating how the K

⫹

channel stabilizes a cation in the

center of the membrane. The central pore’s 10-Å-diameter aqueous

cavity (which contains ⬃50 water molecules) stabilizes a K

⫹

ion

(green spheres) in the otherwise hydrophobic membrane interior.

In addition, the C-terminal ends of the pore helices (red) all point

toward the K

⫹

ion, thereby electrostatically stabilizing it via their

dipole moments (an ␣ helix has a strong dipole moment with its

negative end pointing toward the helix’s C-terminal end because

the bond dipoles of its component carbonyl and N¬H groups are

(a)

(b)

(c)

all parallel to the helix axis with their negative ends pointing

toward its C-terminal end; Fig. 8-11).This effect is magnified by the

low dielectric constant at the center of the membrane interior.

Electrostatic calculations indicate that the cavity is tuned to

maximally stabilize monovalent cations. [Courtesy of Roderick

MacKinnon, Rockefeller University. PDBid 1BL8.]

JWCL281_c20_744-788.qxd 3/17/10 1:48 PM Page 753

higher than that of hydrated Na

⫹

and thus account for the

⫹

channel’s high selectivity for K

⫹

ions. However, pro-

teins are not static structures. In fact, both X-ray evidence

and molecular dynamics simulations (Section 9-4a) indi-

cate that, at physiological temperatures, the atoms forming

the KcsA selectivity filter undergo thermal excursions av-

eraging ⬃1 Å, fluctuations sufficient to snugly cradle Na

⫹

ions with little energetic cost. Instead, as free energy calcu-

lations have demonstrated, it is the electrostatic interac-

tions of the carbonyl groups with the cation and with each

754 Chapter 20. Transport Through Membranes

(a)

(b)

(c)

Figure 20-13 Portions of the KcsA K

⫹

channel responsible for

its ion selectivity viewed similarly to Fig. 20-12. (a) The X-ray

structure of the residues forming the cavity (bottom) and

selectivity filter (top) but with the front and back subunits

omitted for clarity. Atoms are colored according to type, with

C yellow, N blue, O red, and K

⫹

ions represented by green spheres.

The water and protein O atoms that ligand the K

⫹

ions, including

those contributed by the front and back subunits, are represented

by red spheres.The coordination polyhedra formed by these

O atoms are outlined by thin white lines. (b and c) Two alternative

⫹

binding states of the selectivity filter, whose superposition is

presumed to be responsible for the electron density observed in

the X-ray structure of KcsA.Atoms are colored as in Part a.

Note that K

⫹

ions occupying the selectivity filter are interspersed

with water molecules and that the K

⫹

ion immediately above the

selectivity filter in Part b is farther above the protein than that in

Part c. Hence these ions maintain a constant spacing while

traversing the selectivity filter. [Part a based on an X-ray

structure by, and Parts b and c courtesy of, Roderick MacKinnon,

Rockefeller University. PDBid 1K4C.]

See Interactive

Exercise 14

JWCL281_c20_744-788.qxd 10/19/10 7:39 AM Page 754

other that confer specificity for binding K

⫹

ions. This is

consistent with the observation that no Na

⫹

-specific pro-

tein channels have evolved by refining the structure of a

KcsA-like channel.

Since the selectivity filter appears designed to specifi-

cally bind K

⫹

ions, how does it support such a high

throughput of these ions (up to 10

ions ⴢ s

⫺1

)? The struc-

ture in Fig. 20-13a shows what appear to be 4 K

⫹

ions in the

selectivity filter and two more just outside it on its extracel-

lular side. Such closely spaced positive ions would strongly

repel one another and hence represent a high energy situa-

tion. However, a variety of evidence suggests that this

structure is really a superposition of two sets of K

⫹

ions,

one with K

⫹

ions at the topmost position in Fig. 20-13a and

at positions 1 and 3 in the selectivity filter (Fig. 20-13b) and

the second with K

⫹

ions at the second position from the top

in Fig. 20-13a and at positions 2 and 4 in the selectivity fil-

ter (Fig. 20-13c; X-ray structures can show overlapping

atoms because they are averages of many unit cells).

Within the selectivity filter, the positions not occupied by

⫹

ions are instead occupied by water molecules that coor-

dinate the neighboring K

⫹

ions.

The electron density that is represented as the topmost

4 water molecules in Fig. 20-13a is highly elongated in the

vertical direction in this otherwise high-resolution (2.0 Å)

structure. Hence it is thought to actually arise from 8 water

molecules that ligand the topmost K

⫹

ion in Fig. 20-13b to

form an inner hydration shell similar to that of the K

⫹

the central cavity (Fig. 20-13a, bottom). Moreover, the four

water molecules liganding the topmost K

⫹

ion in Fig. 20-

13c also contribute to this electron density. This latter ring

of 4 waters provides half of the associated K

⫹

ion’s 8 lig-

anding O atoms. The others are contributed by the car-

bonyl O atoms of the 4 Gly 79 residues, which are properly

oriented to do so. It therefore appears that a dehydrated

⫹

ion transits the selectivity filter (moves to successive

positions in Figs. 20-13b,c) by exchanging the properly

spaced ligands extending from its walls and then exits into

the extracellular solution by exchanging protein ligands for

water molecules and hence again acquiring a hydration

shell.These ligands are spaced and oriented such that there

is little free energy change (estimated to be ⬍12 kJ ⴢ mol

⫺1

)

along the reaction coordinate via which a K

⫹

ion transits

the selectivity filter and enters the extracellular solution.

The rapid dehydration of the K

⫹

ion entering the selectiv-

ity channel from the cavity is, presumably, similarly man-

aged. The essentially level free energy landscape through-

out this process is, of course, conducive to the rapid transit

of K

⫹

ions through the ion channel and hence must be a

product of evolutionary fine-tuning. Energy calculations

indicate that mutual electrostatic repulsions between suc-

cessive K

⫹

ions, whose movements are concerted, balances

the attractive interactions holding these ions in the selec-

tivity filter and hence further facilitates their rapid transit.

G. Cl

ⴚ

Channels

⫺

channels, which occur in all cell types, permit the trans-

membrane movement of chloride ions along their concen-

tration gradient. In mammals, the extracellular Cl

⫺

concen-

tration is ⬃120 mM and the intracellular concentration is

⬃4 mM.

ClC channels form a large family of Cl

⫺

channels that

occur widely in all kingdoms of life.The X-ray structures of

ClC channels from two species of bacteria, determined by

Raimund Dutzler and MacKinnon, reveal, as biophysical

measurements had previously suggested, that ClC channels

are homodimers with each ⬃470-residue subunit forming

an anion-selective pore (Fig. 20-14). Each subunit consists

mainly of 18 mostly transmembrane ␣ helices that are re-

markably tilted with respect to the membrane plane and

have variable lengths compared to the transmembrane he-

lices in other integral proteins of known structures. The N-

and C-terminal halves of each subunit are related by a

pseudo-2-fold axis parallel to the plane of the membrane

and hence these two halves have opposite orientations in

Section 20-2. Kinetics and Mechanisms of Transport 755

Figure 20-14 X-ray structure of the ClC Cl

⫺

channel from

E. coli. Each subunit of the homodimer contains 18 ␣ helices of

variable lengths.The subunits are drawn in ribbon form with

one colored in rainbow order from its N-terminus (blue) to its

C-terminus (red) and the other pink.The two Cl

⫺

ions bound in

the selectivity filter of each subunit are represented by pale

(a)

(b)

green spheres. (a) View from within the membrane with the

extracellular surface above and the 2-fold axis relating the two

subunits vertical. (b) View from the extracellular side of the

membrane along the molecular 2-fold axis. [Based on an X-ray

structure by Raimund Dutzler and Roderick MacKinnon,

Rockefeller University. PDBid 1OTS.]

JWCL281_c20_744-788.qxd 3/17/10 1:48 PM Page 755

the membrane. This suggests that the ClC channel arose

through gene duplication although its two halves exhibit

only weak sequence similarity. Such antiparallel architec-

ture occurs in several types of transmembrane transport

proteins.

The ClC Cl

⫺

channel is located at the interface between

its N- and C-terminal halves. The specificity of the ClC

channel results from an electrostatic field established by

basic amino acids on the protein surface, which helps fun-

nel anions toward the pore, and by a selectivity filter

formed by the dipoles of several ␣ helices oriented with

their positively charged N-terminal ends pointing toward

the Cl

⫺

ions (opposite to their orientation in the KcsA

channel; Fig. 20-12c). This feature of the selectivity filter

helps attract Cl

⫺

ions, which are specifically coordinated by

main chain amide nitrogens and side chain hydroxyls from

Ser and Tyr residues. A positively charged residue such as

Lys or Arg, if it were present in the selectivity filter, would

probably bind a Cl

⫺

ion too tightly to facilitate its rapid

transit through the channel.

Unlike the KcsA channel, which has a central aqueous

cavity (Fig. 20-12c), the Cl

⫺

channel is hourglass-shaped,

with its narrowest part in the center of the membrane and

flanked by wider aqueous vestibules. A conserved Glu side

chain projects into the pore. This group would repel other

anions, suggesting that rapid Cl

⫺

flux requires a protein

conformational change in which the Glu side chain moves

aside. Another anion could push the Glu away, which ex-

plains why some Cl

⫺

channels appear to be activated by

⫺

ions; that is, they open in response to a certain concen-

tration of Cl

⫺

in the extracellular fluid.

H. Aquaporins

The observed rapid passage of water molecules across bio-

logical membranes had long been assumed to occur via

simple diffusion that was made possible by the small size

and high concentration of water molecules. However, cer-

tain cells, such as erythrocytes and those of the kidney, can

sustain particularly rapid rates of water transport, which

are reversibly inhibited by mercuric ion.This suggested the

existence of previously unrecognized protein pores that

conduct water through biological membranes. The first of

these proteins was discovered in 1992 by Peter Agre, who

named them aquaporins.

Aquaporins occur widely in all kingdoms of life. Plants

have up to 50 different aquaporins, which is indicative of

the importance of water transport to plant physiology. The

13 known mammalian aquaporins, AQP0 through AQP12,

are selectively expressed at high levels in tissues that rap-

idly transport water, such as kidneys, salivary glands, sweat

glands, and lacrimal glands (which produce tears). In fact,

kidneys alone employ seven different aquaporins, each

with specific locations and regulatory properties. There are

two subfamilies of aquaporins: those that permit only the

passage of water and those that also allow the passage of

small neutral molecules such as glycerol and urea and

hence are named aquaglyceroporins. Aquaporins permit

the passage of water molecules at extremely high rates (up

to ⬃3 ⫻ 10

per second) but, quite surprisingly, not protons

(really hydronium ions; H

⫹

), whose free passage would

discharge the cell’s membrane potential.

All known aquaporins are homotetramers, each of

whose subunits contain a water-transport channel (unlike

⫹

channels, whose transport channels lie along their 4-fold

axes; Section 20-2Fa). The X-ray structure of the most ex-

tensively studied aquaporin, bovine AQP1, reveals that

each of its 271-residue subunits consists mainly of six trans-

membrane ␣ helices plus two shorter helices that are com-

ponents of loops that extend only to the middle of the bi-

layer (Fig. 20-15a). Other aquaporins of known structure

have similar structures. The N- and C-terminal halves of

aquaporins are ⬃20% identical in sequence and related by

a pseudo-2-fold axis of symmetry that is parallel to the

plane of the membrane (Fig. 20-15a). Evidently, these seg-

ments arose through gene duplication. ClC channels have a

similar antiparallel architecture (Section 20-2G).

The helices in AQP1 surround an elongated hourglass-

shaped channel through the membrane (Fig. 20-16) that at

its narrowest point is ⬃2.8 Å wide, the diameter of a water

molecule. This region is formed by the side chains of the

highly conserved Phe 58, His 182, and Arg 197 (Fig. 20-15b,

lower right subunit) and hence is known as the ar/R con-

striction (ar for aromatic).The side chain of Cys 191, which

also forms part of the ar/R constriction, is the site of chan-

nel blockage by the binding of mercuric ion. For a water

molecule to pass through the ar/R constriction, it must

shed its shell of associated water molecules. This is facili-

tated by the side chains of His 182 and Arg 197. The water

molecules then continue in single file through the ⬃25-Å-

long and ⬃4-Å-wide portion of the channel, which is lined

with hydrophobic groups interspersed with several hydro-

gen bonding groups. The water molecules’ lack of interac-

tion with the hydrophobic walls of the channel facilitates

their rapid passage through the channel, whereas the hy-

drogen bonding groups reduce the energy barrier to water

transport. It is the balancing of these opposing factors that

is presumably responsible for aquaporin’s selective perme-

ability to water and its rapid transport rate.

If water were to pass through aquaporin as an uninter-

rupted chain of hydrogen-bonded molecules, then protons

would pass even more rapidly through the channel via pro-

ton jumping (Fig. 2-10; in order for more than one such se-

ries of proton jumps to occur, each water molecule in the

chain must reorient such that one of its protons forms a hy-

drogen bond to the next water molecule in the chain).

However, aquaporin interrupts this process by forming hy-

drogen bonds from the side chain NH

groups of the highly

conserved Asn 78 and Asn 194, to a water molecule that is

centrally located in the channel (Fig. 20-16). Consequently,

although this central water molecule can readily donate

hydrogen bonds to its neighboring water molecules in the

hydrogen bonded chain, it cannot accept one from them

nor reorient, thereby severing the “proton-conducting

wire.” Both of these Asn residues occur in the sequence

Asn-Pro-Ala (NPA), the signature sequence of aquaporins,

in which the Ala is located at the N-terminal end of each of

the half-spanning helices.

756 Chapter 20. Transport Through Membranes

JWCL281_c20_744-788.qxd 6/4/10 12:14 PM Page 756

Section 20-2. Kinetics and Mechanisms of Transport 757

Figure 20-15 X-ray structure of the aquaporin AQP1 from

bovine erythrocytes. (a) Ribbon diagram of an aquaporin

subunit colored in rainbow order from its N-terminus (blue) to

its C-terminus (red).The view is from within the membrane with

its extracellular surface above and along the subunit’s pseudo-2-

fold axis of symmetry. Note that the two helices closest to the

viewer (orange and blue-green) are both portions of loops that

extend only to the center of the bilayer.The four water

molecules that occupy the central portion of AQP1’s water-

transport channel are represented by red spheres. (b) View of the

aquaporin homotetramer from the extracellular surface along its

Figure 20-16 Schematic drawing of the water-conducting pore

of bovine aquaporin AQP1. The pore is viewed from within the

membrane with the extracellular surface above. The positions of

residues critical for preventing the passage of protons, other ions,

and small molecule solutes are indicated. [Courtesy of Peter

Agre, Johns Hopkins School of Medicine.]

(a)

(b)

4-fold axis.The subunit in the upper right is drawn in space-

filling form with C green, N blue, and O red; that in the upper left

is drawn in ribbon form colored in rainbow order from its

N-terminus (blue) to its C-terminus (red), that in the lower left is

represented by its solvent-accessible surface; and that in the

lower right displays the side chains forming the ar/R constriction

(those of Phe 58, His 182, Cys 191, and Arg 197) in stick form.

Each subunit forms a water-transport channel, which is most

clearly visible in the subunit drawn in space-filling form. [Based

on an X-ray structure by Bing Jap, University of California at

Berkeley. PDBid 1J4N.]

JWCL281_c20_744-788.qxd 6/4/10 12:15 PM Page 757

3 ATP-DRIVEN ACTIVE TRANSPORT

Mediated transport is categorized according to the stoi-

chiometry of the transport process (Fig. 20-17):

1. A uniport involves the movement of a single mole-

cule at a time. Maltoporin and GLUT1 are uniports.

2. A symport simultaneously transports two different

molecules in the same direction.

3. An antiport simultaneously transports two different

molecules in opposite directions.

The electrical character of ion transport is further spec-

ified as:

1. Electroneutral (electrically silent) if there is simulta-

neous charge neutralization, either by symport of oppo-

sitely charged ions or antiport of similarly charged ions.

Aquaporin is electroneutral.

2. Electrogenic if the transport process results in a

charge separation across the membrane. KcsA and ClC are

electrogenic.

Since the glucose concentration in blood plasma is gen-

erally higher than that in cells, GLUT1 normally transports

glucose into the erythrocyte, where it is metabolized via

glycolysis. Many substances, however, are available on one

side of a membrane in lower concentrations than are re-

quired on the other side of the membrane. Such substances

must be actively and selectively transported across the

membrane against their concentration gradients.

Active transport is an endergonic process that is often

coupled to the hydrolysis of ATP. How is this coupling ac-

complished? In endergonic biosynthetic reactions, it often

occurs through the direct phosphorylation of a substrate by

ATP; for example, the formation of UTP in the synthesis of

glycogen (Section 18-2B). Membrane transport, however,

is usually a physical rather than a chemical process; the

transported molecule is not chemically altered. Determin-

ing the mechanism by which the free energy of ATP hy-

drolysis is coupled to endergonic physical processes has

therefore been a challenging problem.

Three types of ATP hydrolyzing, transmembrane pro-

teins have been identified that actively transport cations:

1. P-type ATPases are located mostly in plasma mem-

branes and are so named because they are phosphorylated

by ATP during the transport process. P-type ATPases are

known that transport H

⫹

,Na

⫹

,Ca

2⫹

,Cu

2⫹

,Cd

2⫹

, and

2⫹

against their concentration gradients. They are distin-

guished from the other types of cation-translocating ATPases

by their inhibition by vanadate ( , a phosphate analog;

see Problem 8 in this chapter).

2. F-type ATPases (F

) function to translocate pro-

tons into mitochondria and bacterial cells, which in turn

powers ATP synthesis.They are discussed in Section 22-3C.

3. V-type ATPases are located in plant vacuolar mem-

branes and acidic vesicles, such as animal lysosomes, and

are homologous to the F-type ATPases.

Anions are transported by a fourth type of ATPase, the so-

called A-type ATPases. In this section, we discuss P-type

ATPases. We also examine a bacterial active transport

process, in which the molecules transported are concomi-

tantly phosphorylated, and the ABC transporters, which

transport a wide variety of substances across membranes.

In the next section, we study secondary active transport

systems, so called because they utilize the free energy of

electrochemical gradients generated by ion-pumping

ATPases to transport ions and neutral molecules against

their concentration gradients.

A. (Na

⫹

–K

⫹

)–ATPase of Plasma Membranes

One of the most thoroughly studied active transport sys-

tems is the (Na

⫹

–K

⫹

)–ATPase of plasma membranes. This

transmembrane protein, which was first isolated in 1957 by

Jens Skou, is often called the (Na

⫹

–K

⫹

) pump because it

pumps Na

⫹

out of and K

⫹

into the cell with the concomitant

hydrolysis of intracellular ATP. Unlike most P-type ATPases,

which are monomeric,(Na

⫹

–K

⫹

)–ATPases consist of ␣ and

␤ subunits. The ⬃1000-residue, nonglycosylated ␣ subunit

contains the enzyme’s ATP and ion binding sites. It is

highly conserved (98% identical among mammals) and

homologous to single-subunit P-type ATPases such as

the Ca

2⫹

-ATPase (Section 20-3B). The ⬃300-residue, gly-

cosylated ␤ subunit facilitates the correct insertion of the

␣ subunit into the plasma membrane and has been impli-

cated in K

⫹

transport.

The overall stoichiometry of the (Na

⫹

–K

⫹

)–ATPase re-

action is

3Na

⫹

(out) ⫹ 2K

⫹

(in) ⫹ ADP ⫹ P

3Na

⫹

(in) ⫹ 2K

⫹

(out) ⫹ ATP ⫹ H

O Δ

3⫺

758 Chapter 20. Transport Through Membranes

Figure 20-17 Uniport, symport, and antiport translocation

systems.

A (in)

B (in)

Symport

A (out)

B (out)

A (out)

B (out)

A (in)

B (in)

Antiport

A (in)

Uniport

A (out)

JWCL281_c20_744-788.qxd 3/17/10 1:48 PM Page 758

The (Na

⫹

–K

⫹

)–ATPase is therefore an electrogenic an-

tiport: Three positive charges exit the cell for every two

that enter. This extrusion of Na

⫹

enables animal cells to

control their water content osmotically; without function-

ing (Na

⫹

–K

⫹

) pumps, animal cells, which lack cell walls,

would swell and burst (recall that lipid bilayers are perme-

able to H

O; Section 12-2Ba). Moreover, the electrochemi-

cal potential gradient generated by the (Na

⫹

–K

⫹

) pump is

responsible for the electrical excitability of nerve cells

(Section 20-5Ba) and provides the free energy for the ac-

tive transport of glucose and amino acids into some cells

(Section 20-4A). In fact, all cells expend a large fraction of

the ATP they produce (typically 30% and up to 70% in

nerve cells) to maintain their required cytosolic Na

⫹

and K

⫹

concentrations.

a. ATP Phosphorylates an Essential Asp during the

Transport Process

The free energy of ATP hydrolysis powers the ender-

gonic transport of Na

⫹

and K

⫹

against an electrochemical

gradient. In coupling these two processes, a kinetic barrier

must somehow be erected against the “downhill” transport

of Na

⫹

and K

⫹

along their ion concentration gradients,

while simultaneously facilitating their “uphill” transport. In

addition, futile ATP hydrolysis must be prevented in the

absence of uphill transport. How the enzyme does so is by

no means well understood, although many of its mechanis-

tic aspects have been elucidated.

A key discovery was that the protein is phosphorylated

by ATP in the presence of Na

⫹

during the transport

process. The use of chemical trapping techniques demon-

strated that this phosphorylation occurs on an Asp residue

to form a highly reactive aspartyl phosphate intermediate.

For instance, sodium borohydride reduces acyl phosphates

to their corresponding alcohols. In the case of an aspartyl

phosphate residue, the alcohol is homoserine. By use of

[

H]NaBH

to reduce the phosphorylated enzyme, radioac-

tive homoserine was, in fact, isolated from the acid hy-

drolysate (Fig. 20-18). The phosphorylated residue, Asp

374, begins the highly conserved sequence DKTG that oc-

curs in the central region of the polypeptide chain.

b. The (Na

⫹

–K

⫹

)–ATPase Has Two Major

Conformational States

The observations that ATP phosphorylates the

(Na

⫹

–K

⫹

)–ATPase only in the presence of Na

⫹

, while the

aspartyl phosphate residue is only subject to hydrolysis in

the presence of K

⫹

, led to the realization that the enzyme

has two major conformational states, E1 and E2. These

states have different tertiary structures, different catalytic

activities, and different ligand specificities:

1. E1 has an inward-facing high-affinity Na

⫹

binding

site (K

⫽ 0.2 mM, well below the intracellular [Na

⫹

]) and

reacts with ATP to form the activated product E1

⬃

P only

when Na

⫹

is bound.

2. E2¬P has an outward-facing high-affinity K

⫹

bind-

ing site (K

⫽ 0.05M, well below the extracellular [K

⫹

])

and hydrolyzes to form P

⫹ E2 only when K

⫹

is bound.

c. An Ordered Sequential Kinetic Reaction

Mechanism Accounts for the Coupling of Active

Transport with ATP Hydrolysis

The (Na

⫹

–K

⫹

)–ATPase is thought to operate in accor-

dance with the following ordered sequential reaction

scheme (Fig. 20-19):

1. E1 ⴢ ATP, which acquired its ATP inside the cell,

binds 3Na

⫹

to yield the ternary complex E1 ⴢ ATP ⴢ 3Na

⫹

2. The ternary complex reacts to form the “high-energy”

aspartyl phosphate intermediate E1

⬃

P ⴢ 3Na

⫹

3. This “high-energy” intermediate relaxes to its “low-

energy” conformation, E2¬P ⴢ 3Na

⫹

, and releases its

bound Na

⫹

outside the cell; that is, Na

⫹

is transported

through the membrane.

4. E2¬P binds 2K

⫹

from outside the cell to form

E2¬P ⴢ 2K

⫹

5. The phosphate group is hydrolyzed,yielding E2 ⴢ 2K

⫹

6. E2 ⴢ 2K

⫹

changes conformation to E1, binds ATP,

and releases its 2K

⫹

inside the cell, thereby completing the

transport cycle.

The enzyme appears to have only one set of cation binding

sites, which apparently changes both its orientation and its

specificity during the course of the transport cycle.

Section 20-3. ATP-Driven Active Transport 759

Figure 20-18 Reaction of [

H]NaBH

with phosphorylated

(Na

⫹

–K

⫹

)–ATPase. The isolation of [

H]homoserine following

acid hydrolysis of the protein indicates that the original phospho-

rylated amino acid residue is Asp.

spartyl phosphate

residue

Homoserine

CH CH

OPO

NaB

acid hydrolysis

–

CH CH

OHC P

CH CH

OHC

COO

–

JWCL281_c20_744-788.qxd 6/4/10 1:20 PM Page 759

The obligatory order of the reaction requires that ATP

can be hydrolyzed only as Na

⫹

is transported “uphill.”

Conversely, Na

⫹

can be transported “downhill” only if ATP

is concomitantly synthesized. Consequently, although each

of the above reaction steps is, in fact, individually re-

versible, the cycle, as is diagrammed in Fig. 20-19, circulates

only in the clockwise direction under normal physiological

conditions; that is, ATP hydrolysis and ion transport are

coupled processes. Note that the vectorial (unidirectional)

nature of the reaction cycle results from the alternation of

the steps of the exergonic ATP hydrolysis reaction (Step 2,

Step 5, and ATP binding in Step 6) with the steps of the en-

dergonic ion transport process (Step 1, Steps 3 ⫹ 4, and K

⫹

release in Step 6).Thus, neither reaction can go to comple-

tion unless the other one also does.

d. Mutual Destabilization Accounts for the

Rate of Na

⫹

and K

⫹

Transport

The above ordered kinetic mechanism accounts only for

the coupling of active transport with ATP hydrolysis. In or-

der to maintain a reasonable rate of transport, the free ener-

gies of all its intermediates must be roughly equal. If some

intermediates were much more stable than others, the stable

intermediates would accumulate, thereby severely reducing

the overall transport rate. For example, in order for Na

⫹

be transported out of the cell, uphill, its binding must be

strong to E1 on the inside and weak to E2 on the outside.

Strong binding means greater stability and a potential bot-

tleneck.This difficulty is counteracted by the phosphoryla-

tion of E1 ⴢ 3Na

⫹

and its subsequent conformational

change to yield the low Na

⫹

affinity E2¬P (Steps 2 and 3,

Fig. 20-19). Likewise, the strong binding of K

⫹

to E2¬P on

the outside is attenuated by its dephosphorylation and con-

formational change to yield the low K

⫹

affinity E1 (Steps 5

and 6, Fig. 20-19). It is these mutual destabilizations that

permit Na

⫹

and K

⫹

to be transported at a rapid rate.

e. The X-Ray Structure of the (Na

⫹

–K

⫹

)–ATPase

Chikashi Toyoshima determined the X-ray structure of

shark (Na

⫹

–K

⫹

)–ATPase in complex with K

⫹

ions, an

MgF

2⫺

ion (a P

mimic), and a 74-residue subunit named

FXYD that functions as a tissue-specific regulator.This X-

ray structure (Fig.20-20) is that of the E2¬P ⴢ 2K

⫹

complex

(Fig. 20-19). The ␣ subunit of this ⬃160-Å-long protein

consists of a transmembrane domain (M) composed of 10

helices (␣M1–␣M10) of varied lengths and, from top to

bottom in Fig. 20-20, three well-separated cytoplasmic do-

mains: the nucleotide-binding domain (N), which binds

ATP; the actuator domain (A), so named because it partic-

ipates in the transmission of major conformational changes

(see below); and the phosphorylation domain (P), which

contains the protein’s phosphorylatable Asp residue. The

␤ subunit’s single transmembrane helix is tilted ⬃32° from

the normal to the plane of the membrane. The FXYD

subunit also has a single transmembrane helix but it is

nearly perpendicular to the plane of the membrane. The

MgF

2⫺

ion marks the ATPase’s catalytic site and is coordi-

nated by conserved residues from both its A and P do-

mains. Two K

⫹

ions are located ⬃4.1 Å apart in a common

binding cavity near the center of the ␣ subunit’s transmem-

brane domain that is formed, in large part, by the partial

unwinding of helices ␣M5 and ␣M7, and where they are

each liganded by several main chain carbonyl and side

chain oxygen atoms. The same cavity is implicated in bind-

ing the three Na

⫹

ions bound to the enzyme’s E1 form,with

two of these binding sites probably formed by the same

side chains that coordinate the K

⫹

ions and the third such

site formed, in part, by the side chains of the ␣ subunit’s

760 Chapter 20. Transport Through Membranes

Figure 20-19 Kinetic scheme for the active transport of Na

⫹

and K

⫹

by (Na

⫹

–K

⫹

)–ATPase.

Here (in) refers to the cytosol and (out) refers to the exterior of the cell.

binding

1 Formation of

“high-energy” aspartyl

phosphate intermediate

transport3

binding4

Phosphate

hydrolysis

transport and

ATP

binding

E1 • ATP E1 • ATP• 3Na

P • 3Na

ATP

Inside

3Na

(in)

Outside

3Na

(out)

ADP

(out)

P • 2K

E2 • 2K

JWCL281_c20_744-788.qxd 6/4/10 12:15 PM Page 760