Voet D., Voet Ju.G. Biochemistry

Подождите немного. Документ загружается.

small C-terminal helix. No channel leading from either side

of the membrane to this binding cavity is apparent.

f. Cardiac Glycosides Specifically Inhibit the

(Na

⫹

–K

⫹

)–ATPase

Study of the (Na

⫹

–K

⫹

)–ATPase has been greatly facili-

tated by the use of cardiac glycosides (also called car-

diotonic steroids), natural products that increase the inten-

sity of heart muscle contraction. Indeed, digitalis, an extract

of purple foxglove leaves (Fig. 20-21a), which contains a

mixture of cardiac glycosides including digitalin (Fig. 20-

21b), has been used to treat congestive heart failure for cen-

turies. The cardiac glycoside ouabain (pronounced wabane;

Fig. 20-21b), a product of the East African ouabio tree, has

been long used as an arrow poison. These two steroids,

which are still among the most commonly prescribed car-

diac drugs, inhibit the (Na

⫹

–K

⫹

)–ATPase by binding

strongly to an externally exposed portion of the enzyme

(the drugs are ineffective when injected inside cells) so as to

block Step 5 in Fig. 20-19.The resultant increase in intracel-

lular [Na

⫹

] stimulates the cardiac (Na

⫹

–Ca

2⫹

) antiport sys-

tem, which pumps Na

⫹

out of and Ca

2⫹

into the cell (Sec-

tion 22-1Bb). The increased cytosolic [Ca

2⫹

] boosts the

[Ca

2⫹

] in other cellular organelles, principally the sarcoplas-

mic reticulum (SR). Thus, the release of Ca

2⫹

to trigger

Section 20-3. ATP-Driven Active Transport 761

Figure 20-20 X-ray structure of shark (Na

⫹

–K

⫹

)–ATPase in

complex with K

⫹

and MgF

2⫺

ions. The protein is drawn in ribbon

form viewed parallel to the plane of the membrane (gray box) with

the cytosol above.The ␣ subunit is colored in rainbow order from

its N-terminus (blue) to its C-terminus (red), the ␤ subunit is

magenta, and the FXYD subunit is brown.The ␣ subunit’s bound

⫹

and MgF

2⫺

ions are drawn in space-filling form with K light

blue, Mg pink, and F light green. [Based on an X-ray structure by

Chikashi Toyoshima, University of Tokyo, Japan. PDBid 2ZXE.]

Ouabain

HOH

Digitalin

OCH

(b)

Figure 20-21 Cardiac glycosides. (a) The leaves of the purple

foxglove plant are the source of the heart muscle stimulant

digitalis. [iStockphoto.] (b) Digitalin, a major component of

digitalis, and ouabain, a cardiac glycoside isolated from the East

African ouabio tree, are among the most commonly prescribed

cardiac drugs.

JWCL281_c20_744-788.qxd 6/12/10 11:07 AM Page 761

muscle contraction (Section 35-3Cb) produces a larger than

normal increase in cytosolic [Ca

2⫹

], thereby intensifying the

force of cardiac muscle contraction. Ouabain, which was

once thought to be produced only by plants, is now known

to also be an animal hormone that is secreted by the adre-

nal cortex and functions to regulate cell [Na

⫹

] and overall

body salt and water balance.

762 Chapter 20. Transport Through Membranes

B. Ca

2⫹

–ATPase

Calcium ion often acts as a second messenger in a manner

similar to cAMP (Section 19-4). Transient increases in cy-

tosolic [Ca

2⫹

] trigger numerous cellular responses, including

muscle contraction (Section 35-3Ca), release of neurotrans-

mitters (Section 20-5Cb), and, as we have seen, glycogen

Figure 20-22 Mechanism of SERCA based on four X-ray

structures. (a) The X-ray structures of the SERCA complexes

indicated in lightface are models for the conformational states

indicated in boldface. These states roughly correspond to those in

the corners of Fig. 20-19. The structures are represented by their

ribbon diagrams embedded in their transparent surfaces (gray)

as viewed along the plane of the membrane with the cytosol

above. The A, N, and P domains are colored yellow, red, and blue,

the transmembrane helices M1–2, M3–4, M5–6, and M7–10 are

purple, green, tan, and gray, and bound Ca

2⫹

ions are represented

by gray spheres.ATP and its mimics are drawn in stick form

with C green, N blue, O red, and P orange. A conserved TGES

(a)

(b)

sequence is drawn in space-filling form in magenta. In the lower

right drawing, the position of the lipid bilayer is indicated by the

yellow rectangle. (b) Schematic diagram of the conformational

changes made by SERCA during its catalytic cycle.The protein

components are colored as in Part a except that helices M5–10

are all tan and Ca

2⫹

ions are represented by green spheres. In

addition, protons are represented by gray spheres, and the

residues that ligand Ca

2⫹

ions in the binding cavity are indicated

by red spheres. [Modified from drawings by Poul Nissen,

University of Aarhus, Denmark. PDBids 2C88, 3BA6, 3B9B, and

3B9R.]

JWCL281_c20_744-788.qxd 6/12/10 11:07 AM Page 762

phosphorylated [via a “high energy” bond, which is repre-

sented by a squiggle (

⬃

)] and in complex with the ADP

analog adenosine-5ⴕ-(␤-amino)diphosphate (AMPPN;

Fig. 22-23c).

2. The A domain rotates toward the phosphorylation

site, thereby contacting the P and N domains. This move-

ment pushes down on the M3–4 segment and pulls up on

the M1–2 segment, thereby separating the lower portions

of these segments from that of the M5–6 segment. This ex-

poses the Ca

2⫹

-binding cavity to the lumen of the SR and

forces apart the residues that had bound the Ca

2⫹

ions in

the cavity, resulting in the release of these ions into the lu-

men.A model for this E2¬P state is provided by the X-ray

structure of SERCA in its E2 state in which BeF

⫺

is cova-

lently linked to Asp 351 to form a mimic of a phosphate

group.

3. Two or three protons bind in the lumenally exposed

binding cavity. This, together with the exchange of the ADP

bound to the N domain for ATP results in movements of

transmembrane helices that again occludes the cavity from

the lumen. A model for the resulting E2¬P* ⴢ ATP state

[where the star (*) is indicative of a transition state] is pro-

vided by the X-ray structure of SERCA in its E2 state in

breakdown (Section 18-3Ce). Moreover, Ca

2⫹

is an impor-

tant activator of oxidative metabolism (Section 22-4).

The use of phosphate as a basic energy currency requires

cells to maintain a low internal [Ca

2⫹

] because, for example,

(PO

)

has a maximum aqueous solubility of 65 ␮M.

Thus, the [Ca

2⫹

] in the cytosol (⬃0.1 ␮M) is four orders

of magnitude less than it is in the extracellular spaces

(⬃1500 ␮M). This large concentration gradient is maintained

by the active transport of Ca

2⫹

across the plasma mem-

brane, the endoplasmic reticulum (ER; SR in muscle), and

the mitochondrial inner membrane. We discuss the mito-

chondrial system in Section 22-1Bb.The plasma membrane

and sarco(endo)plasmic reticulum each contain a P-type

2⫹

–ATPase (Ca

2⫹

pump) that actively pumps Ca

2⫹

out

of the cytosol at the expense of ATP hydrolysis. Their ki-

netic mechanisms match that of the (Na

⫹

–K

⫹

)–ATPase

(Fig. 20-19) except that two Ca

2⫹

ions replace the three Na

⫹

ions, two to three H

⫹

ions replace the three K

⫹

ions, and

(out) refers to the outside of the cell for plasma membrane

2⫹

–ATPase or to the lumen of the sarco(endo)plasmic

reticulum for the Ca

2⫹

–ATPase of that membrane.

a. X-Ray Structures of the Ca

⫹

–ATPase

Suggest Its Mechanism

The X-ray structures of the 994-residue Ca

2⫹

–ATPase

from rabbit muscle sarcoplasmic reticulum [also known as

SERCA for sarco(endo)plasmic reticulum Ca

2⫹

-ATPase]

in its complexes with a variety of analogs of ATP and its

components have been determined,the first of which were re-

ported by Toyoshima. SERCA, which constitutes 90% of SR

membrane protein, is a 140-Å-long monomeric protein that

closely resembles the ␣ subunit of the (Na

⫹

–K

⫹

)–ATPase

(Fig. 20-20). It contains a 10-helix transmembrane domain

(M), an actuator domain (A), a nucleotide-binding do-

main (N) to which the ATP binds, and a phosphorylation

domain (P) that contains the phosphorylatable Asp 351.

The two Ca

2⫹

ions are bound in a cavity, similar to that in

the (Na

⫹

–K

⫹

)–ATPase, which is formed, in large part, by

the disruption of helices M4 and M6 in this region.

Four of these structures, all determined by Poul Nissen,

collectively suggest a mechanism of action of P-type ATPases

resulting from a vectorial series of conformational changes

driven by ATP binding, hydrolysis, and release (Figs. 20-22):

1. Let us start with the X-ray structure of SERCA in

its E2 conformation in complex with the inhibitor thapsi-

gargin (TG; Fig. 22-23a), which stabilizes SERCA’s E2

state, and the nonhydrolyzable ATP analog adenosine-5ⴕ-

(␤,␥-methylene)triphosphate (AMPPCP; Fig. 22-23b).

This structure provides a model for the E2 ⴢ ATP state

(Fig. 20-22, upper left). The exchange of 2 to 3 protons for

2 cytosolic Ca

2⫹

ions causes a rotation of the A domain

that occludes the binding cavity.This activates the ATPase

catalytic site to phosphorylate Asp 351 yielding ADP,

which releases the N domain from the P domain due

to the cleavage of the ATP’s ␤,␥-phosphodiester bond

that had previously linked these two domains. A model

for the resulting Ca

E1⬃P ⴢ ADP state is provided by the

X-ray structure of SERCA in its E1 state with Asp 351

Section 20-3. ATP-Driven Active Transport 763

Figure 20-23 Molecular formulas of several SERCA

inhibitors. (a) Thapsigargin (TG), a product of the plant Thapsia

garganica, (b) AMPPCP, and (c) AMPPN.

Thapsigargin (TG)

(a)

OH OH

–

Adenosine-5-(β,γ-methylene)triphosphate

(AMPPCP)

–

(b)

Adenosine-5-(β-amino)diphosphate

(AMPPN)

OH OH

–

(c)

JWCL281_c20_744-788.qxd 7/1/10 7:18 AM Page 763

C. (H

⫹

–K

⫹

)–ATPase of Gastric Mucosa

Parietal cells of the mammalian gastric mucosa secrete HCl

at a concentration of 0.15M (pH 0.8). Since the cytosolic

pH of these cells is 7.4, this represents a pH difference of

6.6 units, the largest known in eukaryotic cells.The secreted

protons are derived from the intracellular hydration of

by carbonic anhydrase:

The secretion of H

⫹

involves the participation of an

⫹

–K

⫹

)–ATPase, an electroneutral antiport with struc-

ture and properties similar to that of Ca

2⫹

–ATPase. Like

other P-type ATPases, it is phosphorylated during the

transport process. In this case, however, the K

⫹

, which en-

ters the cell as H

⫹

is pumped out, is subsequently external-

ized by its electroneutral cotransport with Cl

⫺

. HCl is

therefore the overall transported product.

a. Cimetidine and Omeprazole Prevent Gastric

Ulcers and Heartburn

For many years, effective treatment of peptic ulcers,

which was a frequently fatal condition caused by the attack

of stomach acid on the gastric mucosa, often required sur-

gical removal of the affected portions of the stomach or

sometimes the entire stomach. The discovery, by James

Black, of cimetidine,

which inhibits stomach acid secretion, all but eliminated

the need for this dangerous and debilitating surgery. The

⫹

–K

⫹

)–ATPase of the gastric mucosa is activated by his-

tamine stimulation of the H

-receptor in a process medi-

ated by cAMP. Cimetidine (trade name Tagamet), which

became available in 1976, was the first of several widely

used drugs that competitively inhibit the binding of hista-

mine to this receptor. These drugs likewise alleviate the

symptoms of gastroesophageal reflux disease (GERD,

commonly known as heartburn), which is caused by the re-

guritation of stomach acid into the esophagus, a widespread

and painful condition that, when chronic, can damage the

esophagus and cause esophageal cancer. Heartburn can

also be relieved by antacids but the effects of cimetidine

and its analogs are longer lasting (6–10 hours vs 1–2 hours

for antacids, although the former require ⬃30 minutes to

take effect) and can be taken before meals to prevent

heartburn from occurring. A drawback of most of these

substances is that they inhibit several cytochrome P450s

and thereby interfere with the metabolism of numerous

widely used drugs (Section 15-4Bc).

Cimetidine

Histamine

⫹ H

O Δ HCO

⫺

⫹ H

⫹

764 Chapter 20. Transport Through Membranes

complex with AMPPCP and AlF

⫺

, which is covalently

linked to Asp 351 to form a trigonal bipyramidal mimic of

the hydrolytic transition state (Fig. 16-6b).

4. The dephosphorylation of Asp 351, as stimulated by

the previous binding of ATP, motivates conformational

changes that opens a channel to the cytosol that permits

the exchange of protons with Ca

2⫹

ions, thereby complet-

ing the catalytic cycle.

This mechanism explains how P-type ATPases transport

cations against their steep electrochemical gradients.

b. Calmodulin Regulates the Plasma

Membrane Ca

2⫹

Pump

For a cell to maintain its proper physiological state, it

must regulate the activities of its ion pumps precisely. The

regulation of the Ca

2⫹

pump in the plasma membrane is

controlled by the level of Ca

2⫹

through the mediation of

calmodulin (CaM).This ubiquitous eukaryotic Ca

2⫹

-binding

protein participates in numerous cellular regulatory

processes including, as we have seen, the control of glyco-

gen metabolism (Section 18-3Ce).

2⫹

–Calmodulin activates the Ca

2⫹

–ATPase of plasma

membranes. The activation, as deduced from the study of the

isolated ATPase, results in a decrease in its K

for Ca

2⫹

from

20 to 0.5 ␮M.Ca

2⫹

–CaM activates the Ca

2⫹

pump by binding

to an inhibitory polypeptide segment of the pump in a man-

ner similar to the way in which Ca

2⫹

–CaM activates its tar-

get protein kinases (Section 18-3Cf). Evidence supporting

this mechanism comes from proteolytically excising the

2⫹

pump’s CaM-binding polypeptide, yielding a truncated

pump that is active even in the absence of CaM. Synthetic

peptides corresponding to this CaM-binding segment not

only bind Ca

2⫹

–CaM but inhibit the truncated pump by in-

creasing its K

for Ca

2⫹

and decreasing its V

max

. This sug-

gests that, in the absence of Ca

2⫹

–CaM, the CaM-binding

segment of the pump interacts with the rest of the protein so

as to inhibit its activity. When the Ca

2⫹

concentration in-

creases, Ca

2⫹

–CaM forms and binds to the CaM-binding

segment of the pump in a way that causes it to dissociate

from the rest of the pump, thereby relieving the inhibition.

Now we can see how Ca

2⫹

regulates its own cytoplasmic

concentration: At Ca

2⫹

levels below calmodulin’s ⬃1 ␮M

dissociation constant for Ca

2⫹

, the Ca

2⫹

–ATPase is relatively

inactive due to autoinhibition by its CaM-binding segment. If,

however, the [Ca

2⫹

] rises to this level, Ca

2⫹

binds to calmod-

ulin, which, in turn, binds to the CaM-binding segment so as

to relieve the inhibition, thereby activating the Ca

2⫹

pump:

(CaM* indicates activated calmodulin). This interaction de-

creases the pump’s K

for Ca

2⫹

to below the ambient [Ca

2⫹

thereby causing Ca

2⫹

to be pumped out of the cytosol.When

the [Ca

2⫹

] decreases sufficiently, Ca

2⫹

dissociates from

calmodulin and this series of events reverses itself,thereby in-

activating the pump.The entire system is therefore analogous

to a basement sump pump that is automatically activated by

a float when the water reaches a preset level.

2⫹

–CaM* ⴢ pump(active)

2⫹

⫹ CaM Δ Ca

2⫹

–CaM* ⫹ pump(inactive) Δ

JWCL281_c20_744-788.qxd 7/1/10 7:19 AM Page 764

energy” phosphoryl donor for ATP synthesis in the pyruvate

kinase reaction of glycolysis; Section 17-2J). The PTS simul-

taneously transports and phosphorylates sugars. Since the cell

membrane is impermeable to sugar phosphates, once they

enter the cell, they remain there. Some of the PTS-transported

sugars are listed in Table 20-2.

The PTS system involves two soluble cytoplasmic pro-

teins, Enzyme I (EI) and HPr (for histidine-containing

phosphocarrier protein), which participate in the transport

of all sugars (Fig. 20-24). In addition, for each sugar the sys-

tem transports, there is a specific transmembrane transport

protein EII, which consists of at least three functional com-

ponents: two that are cytoplasmic, EIIA and EIIB, and a

transmembrane channel, EIIC. These three components as-

sociate differently in different EII’s. In E. coli, for example,

EIIA, EIIB, and EIIC are separate subunits in cellobiose-

specific EII; EIIB and EIIC are covalently linked and EIIA

is separate in glucose-specific EII; and all three components

are present on a single peptide in mannitol-specific EII.

Glucose transport, which resembles that of other sugars,

involves the transfer of a phosphoryl group from PEP to

glucose with net inversion of configuration about the phos-

phorus atom.Since each phosphoryl transfer involves inver-

sion (Section 16-2A), an odd number of transfers must be

Omeprazole (trade names Prilosec and Nexium),

directly inhibits the (H

⫹

–K

⫹

)–ATPase by forming an

adduct with the side chain of its Cys 831. Since its introduc-

tion in 1989, omeprazole has greatly supplanted the use of

cimetidine and its analogs due to the irreversible nature of

its inhibition (which reduces acid secretion by up to 99%)

and its lack of drug–drug interactions. Hence,omeprazole is

presently one of the most frequently used drugs worldwide.

D. Group Translocation

Group translocation is a variation of ATP-driven active

transport that most bacteria use to import certain sugars. It

is required for many bacterial processes, both useful and

harmful (to humans), such as those that produce cheese,soy

sauce, and dental cavities. It differs from active transport in

that the molecules transported are simultaneously modified

chemically. The most extensively studied example of group

translocation is the phosphoenolpyruvate-dependent phos-

photransferase system (PTS) of E. coli discovered by Saul

Roseman in 1964. Phosphoenolpyruvate (PEP) is the phos-

phoryl donor for this system (recall that PEP is the “high-

OCH

Omeprazole

Section 20-3. ATP-Driven Active Transport 765

Table 20-2 Some of the Sugars Transported by the E.coli

PEP-Dependent Phosphotransferase

System (PTS)

Glucose Galactitol

Fructose Mannitol

Mannose Sorbitol

N-Acetylglucosamine Xylitol

PEP

EI phos-

phorylation

by PEP

1. Phosphoryl

transfer

to HPr

2. Phosphoryl

transfer

to EIIA

HPr

~ P

Glucose-6-P

ATP

cAMP + PP

EIIBC

glc

~ P

EIIBC

glc

Glucose

Glycerol + ATP

Glycerol-3-phosphate

+ ADP

Inside Outside

Phosphoryl

transfer

to EIIBC

Phosphoryl

transfer

to EIIBC

Phosphoryl transfer

to glucose

EI ~ PPyruvate

IIA

glc

IIA

glc

~ P

glycerol

kinase

adenylate

cyclase

lactose

permease

Lactose + H

Figure 20-24 Transport of glucose by the PEP-dependent

phosphotransferase system (PTS). HPr and EI are cytoplasmic proteins

common to all sugars transported. EIIA

glc

and EIIBC

glc

are proteins

specific for glucose. EIIA

glc

inhibits non-PTS transport proteins such as

the lactose permease (Section 20-4B) and enzymes such as glycerol

kinase. Adenylate cyclase is activated by the presence of EIIA

glc

⬃P

(or possibly inhibited by the presence of EIIA

glc

JWCL281_c20_744-788.qxd 6/4/10 12:16 PM Page 765

involved. Four phosphorylated protein intermediates have

been identified, indicative of five phosphoryl transfers:

The transport process occurs as follows (Fig. 20-24):

1. PEP phosphorylates EI at N3 (N

) of His 189 to form

a reactive phosphohistidine adduct.

2. The phosphoryl group is transferred to N1 (N

␦

) of

His 15 on HPr. His is apparently a favored phosphoryl

group acceptor in phosphoryl-transfer reactions. It also

participates in the phosphoglycerate mutase reaction of

glycolysis (Section 17-2Ha).

3. HPr

⬃

P continues the phosphoryl-transfer chain by

phosphorylating EIIA

glc

at N3 of His 90.

4. The fourth phosphoryl transfer is to Cys 421 of EIIB

glc

5. The phosphoryl group is finally transferred from

EIIB

glc

to glucose, which, in the process, is transported

across the membrane by EIIC

glc

. Glucose is released into

the cytoplasm only after it has been phosphorylated to

glucose-6-phosphate (G6P).

Thus the transport of glucose is driven by its indirect, exergonic

phosphorylation by PEP. The PTS is an energy-efficient sys-

tem since only one ATP equivalent is required to both trans-

port and phosphorylate glucose. When the active transport

and phosphorylation steps occur separately, as they do in

many cells, two ATPs are hydrolyzed per glucose processed.

a. Bacterial Sugar Transport Is

Genetically Regulated

The PTS is more complex than the other transport sys-

tems we have encountered, probably because it is part of a

complicated regulatory system governing sugar transport.

When any of the sugars transported by the PTS is abun-

dant, the active transport of sugars, which enter the cell via

other transport systems, is inhibited.This inhibition, called

catabolite repression, is mediated through the cAMP con-

centration (Section 31-3C). cAMP activates the transcrip-

tion of genes that encode various sugar transport proteins,

including lactose permease (Section 20-4B). The presence

of glucose results in a decrease in [cAMP], which, in turn,

represses the synthesis of these other sugar transport pro-

teins. Direct inhibition of the sugar transport proteins

themselves, as well as of certain enzymes, also occurs.

The mechanism for control of [cAMP] is thought to reside

in EIIA

glc

,which is transiently phosphorylated in Step 3 of the

PTS transport process (Fig.20-24).When glucose is plentiful,

this enzyme is present mostly in its dephospho form since

Phosphohistidine residue

–

PEP S EI S HPr S EIIA

glc

S EIIBC

glc

S glucose

EIIA

glc

⬃

P rapidly transfers its phosphoryl group through

EIIBC

glc

to glucose. Under these conditions, adenylate cy-

clase is inactive, although whether dephospho EIIA

glc

in-

hibits this enzyme or EIIA

glc

⬃

P activates it is unclear.

However, dephospho EIIA

glc

binds to and inhibits many

non-PTS transporters and enzymes that participate in the

metabolism of sugars other than glucose (the metabolite of

choice for many bacteria), including lactose permease and

glycerol kinase (Section 17-5A). In the absence of glucose,

EIIA

glc

is converted to EIIA

glc

⬃

P, thereby relieving the inhi-

bition of non-PTS transporters. In addition, adenylate cy-

clase is activated to produce cAMP, which, in turn, induces

the increased production of some of the non-PTS trans-

porters and enzymes that EIIA

glc

inhibits. This is a form of

energy conservation for the cell.Why synthesize the proteins

required for the transport and metabolism of all sugars

when the metabolism of only one sugar at a time will do?

b. The X-Ray Structure of EIIA

glc

Complex with Glycerol Kinase

The X-ray structures of EIIA

glc

, both alone and in com-

plex with one of its regulatory targets, glycerol kinase,

which were determined by James Remington and Rose-

man, have revealed how EIIA

glc

inhibits at least some of its

targets and why EIIA

glc

⬃

P does not do so. EIIA

glc

con-

tains two His residues, His 75 and His 90, that are required

for phosphoryl transfer, although only His 90 is necessary

for EIIA

glc

to accept a phosphate from HPr. The X-ray

structure of E. coli EIIA

glc

alone reveals that these two His

residues lie in close proximity (their N3 atoms are 3.3 Å

apart) in a depression on the surface of the protein that is

surrounded by a remarkable ⬃18-Å-diameter hydropho-

bic ring consisting of 11 Phe, Val, and Ile side chains.

The X-ray structure of EIIA

glc

in complex with glycerol

kinase (Fig. 20-25) confirms that this hydrophobic gasket is

indeed the site of interaction between the two proteins and

reveals how the phosphorylation of His 90 disrupts this inter-

action.The two active site His residues, which are completely

buried within the hydrophobic interaction surface, coordi-

nate a previously unanticipated Zn

2⫹

ion, which is addition-

ally coordinated to Glu 478 of glycerol kinase and a water

molecule. The phosphorylation of EIIA

glc

His 90 to yield EI-

glc

⬃

P no doubt disrupts this intermolecular interaction,

thereby releasing glycerol kinase and reversing its inhibition.

E. ABC Transporters

The ABC transporters, which widely occur in all kingdoms of

life, are named for their highly conserved, ⬃100-residue ATP-

binding cassette. They form a large superfamily of transmem-

brane proteins that collectively transport a wide variety of sub-

stances across membranes, including ions, sugars, peptides,

lipids, metabolites, and numerous toxins and drugs. For exam-

ple, the acquired resistance of cancer cells to chemotherapeu-

tic agents is often due to the selection of cells that overexpress

the ABC transporter named multidrug resistance (MDR)

transporter (also called P-glycoprotein), which pumps a wide

assortment of amphiphilic substances—including many

drugs—out of the cell. Similar proteins in bacteria are, in

many cases, responsible for their antibiotic resistance. The

766 Chapter 20. Transport Through Membranes

JWCL281_c20_744-788.qxd 6/4/10 12:16 PM Page 766

(b)

(a)

human genome encodes 48 ABC transporters and their de-

fects are responsible for several inherited diseases includ-

ing cystic fibrosis (Section 12-4Bf),Tangier disease (Section

12-5Cd), and adrenoleukodystrophy (Section 25-2Fa).

ABC transporters consist minimally of four modules: two

highly conserved cytosolic nucleotide-binding domains, and

two transmembrane domains that typically contain six trans-

membrane helices each. In bacteria, the four domains are

contained on two or four separate polypeptides. In eukary-

otes, a single polypeptide includes all four domains. Bacterial

ABC transporters mediate the import as well as the export of

a variety of compounds, whereas their eukaryotic counter-

parts function only as exporters that transport molecules out

of the cell or into intracellular compartments such as the ER.

The ABC transporter of gram-negative bacteria named

MsbA, a close homolog of the MDR transporter, functions

to transport lipopolysaccharides (Section 11-3Bc) and their

glycolipid component,lipid A, from the cytoplasmic leaflet of

the plasma membrane to its periplasmic leaflet; that is, MsbA

is a flippase (Section 12-4Aa). X-ray structures of Msb,deter-

mined by Geoffrey Chang, reveal that its two identical

582-residue subunits are extensively intertwined (Fig. 20-26).

In the nucleotide-free (apo) structure (Fig. 20-26a), each

subunit contributes four of its six transmembrane helices

Section 20-3. ATP-Driven Active Transport 767

Figure 20-25 X-ray structure of E. coli EIIA

glc

(yellow, a

168-residue monomer) in complex with one of its regulatory

targets, glycerol kinase (blue, a tetramer of identical 501-residue

subunits). The two proteins associate, in part, by tetrahedrally

coordinating a Zn

2⫹

ion via the side chains of His 75 and His 90 of

EIIA

glc

, a carboxylate oxygen from Glu 478 of glycerol kinase, and

a water molecule. These groups are shown in ball-and-stick form

with C gray, N blue, O red, and Zn

2⫹

white. The Zn

2⫹

-mediated

interaction between EIIA

glc

and glycerol kinase inactivates

glycerol kinase, presumably through an induced-fit mechanism.

The phosphorylation of EIIA

glc

at His 90 disrupts this interaction,

thereby reversing the inhibition of glycerol kinase. [Courtesy of

James Remington, University of Oregon. PDBid 1GLA.]

Figure 20-26 X-ray structures of the ABC transporter MsbA.

(a) E. coli MsbA in the absence of bound nucleotide (Open

Apo). (b) S. typhimurium MsbA in complex with AMPPNP

(Nucleotide Bound). On the left, the structures are drawn in

ribbon form viewed along the plane of the membrane with the

periplasmic space above. One subunit in each homodimer is

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

(red) and the other subunit is gray. The AMPPNP in Part b is

drawn in stick form in blue. Transmembrane helices (TM1–6),

extracellular loops (EL1–3), and intracellular helices (IH1–2) are

labeled. On the upper right, the structures are drawn as

semitransparent surface diagrams with one subunit cyan and the

other gray, and showing the embedded transmembrane helices

(1–6) of the cyan subunit.The horizontal lines delineate the lipid

bilayer. On the lower right, the structures are represented by

schematic diagrams with one subunit cyan and the other white

and with the positions of the transmembrane helices indicated.

[Modified from drawings by Geoffrey Chang,The Scripps

Research Institute, La Jolla, California. PDBids 3B5W and 3B60.]

JWCL281_c20_744-788.qxd 7/1/10 7:19 AM Page 767

(TM1–3 and 6) to one leg of the ⌳-shaped homodimer and

two (TM4–5) to the other leg. The nucleotide-binding do-

mains (NBDs; globular portions at the end of each leg) are

⬃50 Å apart. However, in its complex with AMPPNP (Fig.

20-26b), the two legs of the ⌳ have swung together by

⬃60° to bring the NBDs into contact, with TM3 and 6 hav-

ing swung away from TM1–2 to contact TM4–5.This opens

a space between the subunits on the periplasmic side of

the membrane, which presumably permits a lipopolysac-

charide that had bound to MsbA from the cytosolic leaflet

of the plasma membrane to be released into its periplas-

mic leaflet. The unidirectional transport of substrate may

be facilitated by the different sizes of the openings, and

possibly, by a reduced affinity for substrate by MsbA’s

ATP-bound form.

4 ION GRADIENT–DRIVEN

ACTIVE TRANSPORT

Systems such as the (Na

⫹

–K

⫹

)–ATPase discussed above

utilize the free energy of ATP hydrolysis to generate elec-

trochemical potential gradients across membranes. Con-

versely, the free energy stored in an electrochemical poten-

tial gradient may be harnessed to power various endergonic

physiological processes. Indeed, ATP synthesis by mito-

chondria and chloroplasts is powered by the dissipation of

proton gradients generated through electron transport and

photosynthesis (Sections 22-3C and 24-2D). In this section

we discuss active transport processes that are driven by the

dissipation of ion gradients. We consider three examples:

intestinal uptake of glucose by the Na

⫹

–glucose symport,

uptake of lactose by E. coli lactose permease, and the mito-

chondrial ATP–ADP translocator.

A. Na

⫹

–Glucose Symport

Nutritionally derived glucose is actively concentrated in

brush border cells of the intestinal epithelium by an Na

⫹

dependent symport (Fig. 20-27). It is transported from

these cells to the circulatory system via GLUT2 (Section

20-2Eb), a passive-mediated glucose uniport located on the

capillary side of the cell. Note that although the immediate

energy source for glucose transport from the intestine is the

⫹

gradient, it is really the free energy of ATP hydrolysis

that powers this process through the maintenance of the Na

⫹

gradient by the (Na

⫹

–K

⫹

)–ATPase. Nevertheless, since glu-

cose enhances Na

⫹

resorption, which in turn osmotically

enhances water resorption, glucose (possibly as sucrose), in

addition to salt and water, should be fed to individuals suf-

fering from severe salt and water losses resulting from diar-

rhea (drinking only water or a salt solution is ineffective

since they are rapidly excreted from the gastrointestinal

tract). It is estimated that the introduction of this simple

and inexpensive treatment,called oral rehydration therapy,

has decreased the annual number of human deaths from

severe diarrhea, mostly in children in less-developed coun-

tries, from 4.6 to 1.6 million.

a. Active and Passive Glucose Transporters Exhibit

Differential Drug Susceptibilities

The two glucose transport systems are inhibited by dif-

ferent drugs:

1. Phlorizin inhibits Na

⫹

-dependent glucose transport.

768 Chapter 20. Transport Through Membranes

ATP

Glucose uniport,

GLUT2

glucose symport

ADP + P

(Na

–K

)–ATPase

Glucose

Microvilli

Intestinal

lumen

To capillaries

Brush border cell

(a) Small intestine

(b) Villus

Intestinal folds

Intestinal

mucosa

Lumen

Capillaries

Columnar epithelial cells

(brush border)

Villi

Figure 20-27 Glucose transport in the intestinal epithelium.

The brushlike villi lining the small intestine greatly increase its

surface area (a), thereby facilitating the absorption of nutrients.

The brush border cells from which the villi are formed (b)

actively concentrate glucose from the intestinal lumen in

symport with Na

⫹

(c), a process that is driven by the

(Na

⫹

–K

⫹

)–ATPase, which is located on the capillary side of the

cell and functions to maintain a low internal [Na

⫹

].The glucose is

exported to the bloodstream via GLUT2, a separate passive-

mediated uniport system (Section 20-2Eb).

JWCL281_c20_744-788.qxd 7/1/10 7:30 AM Page 768

2. Cytochalasin B inhibits Na

⫹

-independent glucose

transport.

Phlorizin binds only to the external surface of the Na

⫹

dependent glucose transporter,whereas cytochalasin B binds

to the cytoplasmic surface of the Na

⫹

-independent glucose

transporter. This further indicates that these proteins are

asymmetrically inserted into membranes. The use of these

inhibitors permits the actions of the two glucose trans-

porters to be studied separately in intact cells.

Kinetic studies indicate that the Na

⫹

–glucose symport

binds its substrates, Na

⫹

and glucose, in random order (Fig.

20-28),although binding of Na

⫹

increases the affinity of the

transporter for glucose to such an extent that the upper

pathway is heavily favored. Only when both substrates are

bound, however, does the protein change its conformation

to expose the binding sites to the inside of the cell.This re-

quirement for concomitant Na

⫹

and glucose transport pre-

vents the wasteful dissipation of the Na

⫹

gradient.

D-Glucose-β

Phlorizin

Cytochalasin B

B. Lactose Permease

Gram-negative bacteria such as E. coli contain several ac-

tive transport systems for concentrating sugars.We have al-

ready discussed the PTS system.Another extensively stud-

ied system, lactose permease (also known as galactoside

permease), utilizes the proton gradient across the bacterial

cell membrane to cotransport H

⫹

and lactose (Fig. 20-29).

The proton gradient is metabolically generated through

oxidative metabolism in a manner similar to that in mito-

chondria (Section 22-3B). The electrochemical potential

gradient created by both these systems is used mainly to

drive the synthesis of ATP.

Section 20-4. Ion Gradient–Driven Active Transport 769

Figure 20-28 The Na

⫹

–glucose symport system represented

as a Random Bi Bi kinetic mechanism. Binding of Na

⫹

increases

the affinity of the transporter for glucose to such an extent that

the upper pathway is heavily favored.T

and T

, respectively,

••

Glc

Glc Glucose

Glc

•

••

Glc

•

Glc

•

Glc

Recovery

InsideOutside

Transport

Release

+ Lactose

E-1E-2

Binding E-2

• H

Lactose

Oxidative

metabolism

E-2 • H

• Lactose E-1 • H

• Lactose

Figure 20-29 Kinetic mechanism of lactose permease in

E. coli. H

⫹

binds first to E-2 outside the cell, followed by lactose.

They are released in random order from E-1 inside the cell. E-2

must bind both lactose and H

⫹

in order to change conformation

to E-1, thereby cotransporting these substances into the cell. E-1

changes conformation to E-2 when neither lactose nor H

⫹

bound, thus completing the transport cycle.

represent the transport protein with its binding sites exposed to

the outer and inner surfaces of the membrane. [After Crane,

R.K. and Dorando, F.C., in Martonosi, A.N. (Ed.), Membranes

and Transport, Vol. 2, p. 154, Plenum Press (1982).]

JWCL281_c20_744-788.qxd 6/4/10 2:09 PM Page 769

How do we know that lactose transport requires the

presence of a proton gradient? Ronald Kaback has estab-

lished the requirement for this gradient through the fol-

lowing observations:

1. The rate of lactose transport into bacteria is in-

creased enormously by the addition of

D-lactate, an energy

source for transmembrane proton gradient generation.

Conversely, inhibitors of oxidative metabolism, such as

cyanide, block both the formation of the proton gradient

and lactose transport.

2. 2,4-Dinitrophenol, a proton ionophore that dissi-

pates transmembrane proton gradients (Section 22-3D),

inhibits lactose transport into both intact bacteria and

membrane vesicles.

3. The fluorescence of dansylaminoethylthiogalactoside,

a competitive inhibitor of lactose transport, is sensitive to

the polarity of its environment and thus changes when it

binds to lactose permease. Fluorescence measurements in-

dicate that it does not bind to membrane vesicles that con-

tain lactose permease in the absence of a transmembrane

proton gradient.

a. Lactose Permease Has Two Major

Conformational States

Lactose permease is a 417-residue monomer that, like

the mammalian glucose transporters (Section 20-2Eb), to

which it is distantly related, consists mainly of 12 trans-

membrane helices with its N- and C-termini in the cyto-

plasm. Like the (Na

⫹

–K

⫹

)–ATPase, it has two major con-

formational states (Fig. 20-29):

1. E-1, which has a low-affinity lactose binding site fac-

ing the interior of the cell.

2. E-2, which has a high-affinity lactose binding site fac-

ing the exterior of the cell.

E-1 and E-2 can interconvert only when their H

⫹

and lac-

tose binding sites are either both filled or both empty. This

prevents not only dissipation of the H

⫹

gradient without

cotransport of lactose into the cell, but also transport of

OH H

HOH

OH H

HOH

OH H

HOH

Lactose

Dansylaminoethylthiogalactoside

lactose out of the cell without cotransport of H

⫹

against its

concentration gradient.

The X-ray structure of lactose permease in complex

with a tight-binding lactose analog, determined by Kaback

and So Iwata, reveals that this protein consists of two

structurally similar and 2-fold pseudosymmetrically re-

lated domains containing six transmembrane helices each

(Fig. 20-30a).A large internal hydrophilic cavity is open to

770 Chapter 20. Transport Through Membranes

Figure 20-30 X-ray structure of lactose permease from E. coli.

(a) Ribbon diagram as viewed from the membrane with the

cytoplasmic side up.The protein’s 12 transmembrane helices are

colored in rainbow order with the N-terminus purple and the

C-terminus pink.The bound lactose analog is represented by

black spheres. (b) Surface model viewed as in Part a but with the

two helices closest to the viewer in Part a removed to reveal the

lactose-binding cavity. The surface is colored according to its

electrostatic potential with positively charged areas blue,

negatively charged areas red, and neutral areas white. [Courtesy

of H. Ronald Kaback, UCLA. PDBid 1PV7.]

(a)

(b)

JWCL281_c20_744-788.qxd 6/4/10 1:37 PM Page 770