Voet D., Voet Ju.G. Biochemistry

Подождите немного. Документ загружается.

permits a cognate aminoacyl–tRNA to enter the peptidyl

transferase center. The irreversible GTPase reaction must

precede this proofreading step because otherwise the dis-

sociation of a noncognate tRNA (the release of its anti-

codon from the codon) would simply be the reverse of the

initial binding step, that is, it would be part of the initial se-

lection step rather than proofreading. GTP hydrolysis

therefore provides the second context necessary for proof-

reading; it is the entropic price the system must pay for accu-

rate tRNA selection.

F. Chain Termination

Polypeptide synthesis under the direction of synthetic

mRNAs such as poly(U) terminates with a peptidyl–tRNA

in association with the ribosome. However, the translation

of natural mRNAs, which contain the Stop codons UAA,

UGA, or UAG, results in the release of free polypeptides.

Accurate termination is essential, not only because it pre-

vents the wasteful synthesis of nonfunctional polypeptides,

but also because prematurely terminated polypeptides

may be toxic.

a. Prokaryotic Termination

In E.coli, chain termination has several stages (Fig. 32-60):

1. The termination codons, the only codons that nor-

mally have no corresponding tRNAs, are recognized by

class I release factors (Table 32-9): RF-1 recognizes UAA

and UAG, whereas the 39% identical RF-2 recognizes

UAA and UGA. Swapping a conserved PXT tripeptide in

RF-1 with a conserved SPF tripeptide in RF-2 interchanges

their Stop codon specificities, which suggests that these

tripeptides mimic anticodons.

2. On binding to their corresponding Stop codon, RF-1

and RF-2 induce the transfer of the peptidyl group from

Section 32-3. Ribosomes and Polypeptide Synthesis 1391

P site

3′

A site

Empty

mRNA

Nascent polypeptide

5′

UAA

3′5′

UAA

RF-1

RF-3

RF-1

3′5′

UAA

RF-1

GDP

GTP

RF-3

RF-1

GDP

Peptidyl–tRNA

COO

–

Uncharged

tRNA

Polypeptide

GDP

GTP

EF-G

RRF

3′5′

UAA

3′5′

UAA

GTP

RF-3

GDP

RF-3

GTP

EF-G

RRF

3′

5′

GDP

EF-G

50S subunit

30S subunit

Figure 32-60 Termination pathway in E. coli ribosomes. RF-1

recognizes the Stop codons UAA and UAG, whereas RF-2 (not

shown) recognizes UAA and UGA. Eukaryotic termination

follows an analogous pathway but requires only a single class I

release factor, eRF1, that recognizes all three Stop codons.

JWCL281_c32_1338-1428.qxd 8/4/10 4:45 PM Page 1391

tRNA to water rather than to an aminoacyl–tRNA, thereby

releasing the completed polypeptide (Fig 32-61). This occurs

with an error rate of 10

5

without proofreading.The class I

release factors act at the ribosomal A site as is indicated by

the observations that they compete with suppressor tRNAs

for termination codons and that they cannot bind to the ri-

bosome simultaneously with EF-G.A GGQ tripeptide that

is universally conserved in all class I release factors is im-

plicated in catalyzing the hydrolysis of the peptidyl–tRNA

ester linkage (see below).

3. Once the newly synthesized polypeptide has been re-

leased from the ribosome, the class II release factor RF-3,

in its complex with GDP, binds to the ribosome at the same

site as do EF-Tu and EF-G. In fact, the X-ray structure of

RF-3 ⋅ GDP resembles that of EF-Tu ⋅ GMPPNP (Fig. 32-

47). Free RF-3 has a greater affinity for GDP than GTP but

on binding to the ribosome–RF-1/2 complex, it exchanges

its bound GDP for GTP. The resulting change in the con-

formation of RF-3, as seen in cryo-EM studies, causes it to

bind more tightly to the ribosome and expel the RF-1/2.

RF-3 is not required for cell viability although it is neces-

sary for maximum growth rate; RF-3 only accelerates the

dissociation of RF-1/2 from the ribosome by ⬃5-fold.

4. The interaction of RF-3 ⋅ GTP with the ribosome

stimulates it to hydrolyze its bound GTP, much as occurs

with EF-Tu ⋅ GTP and EF-Tu ⋅ GTP. The resulting RF-3 ⋅

GDP then dissociates from the ribosome. Subsequently, ri-

bosomal recycling factor (RRF) binds in the ribosomal A

site followed by EF-G  GTP. RRF, which was discovered

by Akira Kaji, is essential for cell viability.

5. EF-G hydrolyzes its bound GTP, which causes RRF

to be translocated to the P site and the tRNAs previously

in the P and E sites (the latter not shown in Fig.32-60) to be

released. Finally, the small and large ribosomal subunits

separate, a process that is facilitated by the binding of IF-3

(Section 32-3Cc), and RRF, EF-G  GDP, and mRNA are

released. The ribosomal subunits can then participate in a

new round of initiation (Fig. 32-43).

b. Eukaryotic Termination

Chain termination in eukaryotes resembles that in

prokaryotes, but it has only one class I release factor, eRF1,

that recognizes all three Stop codons. It is unrelated in se-

quence to RF-1 and RF-2. However, the eukaryotic class II

release factor, eRF3, resembles RF-3 in both sequence and

function. Nevertheless, eRF3 is essential for eukaryotic cell

viability.

c. The Ribosome Binds RF-1 and RF-2 in a

Conformation That Catalyzes Peptide Release

The X-ray structures of the T. thermophilus ribosome

with RF-1, an mRNA containing a UAA Stop codon, and

deacylated tRNAs in its P and E sites was determined by

Noller (Fig. 32-62a), and the closely similar structures con-

taining the tRNAs, RF-2, and an mRNA with a UAA or a

UGA Stop codon, were respectively determined by Noller

and Ramakrishnan. These are all product complexes since

they lack peptidyl groups on their P-site tRNAs. The struc-

turally similar RF-1 (Fig. 32-62b) and RF-2 each consist of

four domains with domains 2 and 4 occupying the ribo-

some’s decoding center (DC), where they contact the

mRNA’s Stop codon, and with domain 3 occupying the

peptidyl transferase center (PTC), where it interacts with

the ribose residue of the P-site tRNA’s A76 to which a pep-

tidyl group would be linked in a substrate complex. The

deletion of domain 1 does not affect peptide release activ-

ity but is required for the RF-3–facilitated dissociation of

RF-1/2 from the ribosome (see below).

1392 Chapter 32. Translation

Figure 32-61 Ribosome-catalyzed hydrolysis of peptidyl–tRNA to form a polypeptide and free tRNA.

O OH

OPO

–

Adenine

CHR

n–1

–

CHR

n–1

Peptidyl–tRNA

tRNA

HO OH

OPO

–

Adenine

tRNA

...

Polypeptide

JWCL281_c32_1338-1428.qxd 9/7/10 2:23 PM Page 1392

The binding of RF-1 or RF-2 in the DC causes A530 and

A1492 of the 16S RNA to flip out from their resting state

(Fig. 32-59a) as occurs with the binding of a tRNA to its

cognate codon (Fig. 32-59b). However,A1493 does not flip

out because in doing so it would clash with domain 2 of ei-

ther release factor. Instead, it stacks on A1913 of the 23S

RNA. The Stop codons are recognized by hydrogen bond-

ing and van der Waals interactions with the similarly lo-

cated PXT and SPF tripeptides on domain 2 of RF-1 and

RF-2. However, the observation that mutations in RF-2

distant from its SQF motif result in altered specificity sug-

gests that Stop codon recognition arises from a subtle bal-

ance of binding energy and conformational changes as we

have seen to be the case for codon recognition by tRNA

(Section 32-3Ea).

In the PTC, the GGQ tripeptide on domain 3 of both

RF-1 and RF-2 contacts the 3¿-terminal ribose residue

(A76) of the deacylated P-site tRNA. Both Gly residues

adopt backbone conformations that are forbidden for

other amino acid residues, which accounts for the observa-

tions that the mutation of either residue results in an up to

-fold reduction in the rate of peptide release. The main

chain NH group of the Gln residue is hydrogen bonded to

the 3¿-OH group of A76, which, it is hypothesized, positions

it to also hydrogen bond to and thereby stabilize the tran-

siton state oxyanion in the hydrolysis reaction. In agree-

ment with this hypothesis, the mutation of the Gln residue

to Pro, which lacks a main chain NH group, abolishes the

hydrolysis reaction. The side chain of the Gln residue is

pointed away from the ribose (top of Fig. 32-62b) where, it

is proposed, it helps position a water molecule for an in-

line nucleophilic attack on the scissile ester bond. In addi-

tion, the observation that peptide release is nearly abol-

ished by the removal of the 2¿-OH from the P-site tRNA’s

3¿ terminal residue suggests that peptide release utilizes a

substrate-assisted proton shuttle mechanism similar to that

of peptide bond formation (Section 32-3Di). Finally, the

binding of RF-1/2 shifts U2585 so as to expose the other-

wise protected scissile ester bond to nucleophilic attack

(Section 32-3Di). Nevertheless, the formulation of a defin-

itive mechanism for peptide release must await the X-ray

structure of a ribosome in complex with both a release fac-

tor and a peptidyl–tRNA, that is, a substrate complex.

How does the binding of a release factor to a Stop

codon in the DC induce the ⬃75-Å distant PTC to hy-

drolyze the scissile ester bond? In the X-ray structures of

RF-1 or RF-2 alone, their PXT/SPF and GGQ motifs are

only ⬃23 Å apart due to a change in conformation of their

switch–loop segments (Fig. 32-62b) relative to that in the

ribosomal complexes. The conformation of the

Section 32-3. Ribosomes and Polypeptide Synthesis 1393

Figure 32-62 X-ray structure of the T. thermophilus ribosome

in complex with RF-1, a UAA Stop codon–containing mRNA,

and deacylated tRNAs in its P and E sites. (a) The overall

structure with proteins shown in ribbon form and all RNAs

shown in ladder form but the mRNA, which is drawn in worm

form in green.The ribosome is semitransparent with its 23S RNA

gray, its 5S RNA light blue, its 16S RNA cyan, the 50S subunit’s

proteins magenta, and the 30S subunit’s proteins purple. The

tRNAs occupying the P and E sites are orange and red, and the

RF-1, which in part occupies the ribosomal A site, is yellow. (b)

Close-up of the interactions between the P-site tRNA, mRNA,

(b)

and RF-1.The ribosome, tRNA, and mRNA are drawn as in Part

a and the RF-1 is colored with its domains 1, 2, 3, and 4 green,

yellow, blue, and magenta, respectively.The so-called

switch–loop, which connects domains 3 and 4 and undergoes a

major conformational rearrangement between the free and

ribosome-bound RF-1, is orange.The PVT tripeptide implicated

in Stop codon recognition and the GGQ tripeptide implicated in

catalyzing the ester hydrolysis reaction are drawn in stick form in

red. [Courtesy of Harry Noller, University of California at Santa

Cruz. PDBids 3D5A and 3D5B.]

(a)

JWCL281_c32_1338-1428.qxd 9/7/10 2:23 PM Page 1393

switch–loop observed in the ribosomal complexes is only

possible when a Stop codon is recognized. This is because

the flipping out of A1493, which only occurs when a tRNA

binds its cognate codon in the DC (Fig. 32-59), or the fail-

ure of A1913 to stack on A1493, alters the binding pocket

for the ribosomally bound switch–loop. Noller has there-

fore proposed that the binding of a Stop codon by its cor-

responding release factor and the consequent rearrange-

ment of both its switch–loop and the DC cooperatively

permit the GGQ motif to bind to the PTC in a way that

catalyzes peptide release.

d. RRF Binds in the Ribosomal A Site

The X-ray structure of T. thermophilus ribosomal recy-

cling factor (RRF), determined by Yoshikazu Nakamura,

reveals it to be a two-domain structure that resembles

tRNA in its overall shape (Fig. 32-63). The comparison of

this structure with those of several other bacterial RRFs in-

dicates that the two linkers connecting the RRF domains

are flexible such that domain II can rotate about the axis of

the three-helix bundle forming domain I.

The X-ray structure of the T. thermophilus ribosome in

complex with RRF in its A site, the anticodon stem–loop

(ASL) of tRNA

Phe

in its P site, tRNA

Met

in its E site, and an

mRNA with a UAG Stop codon in the A site was deter-

mined by Ramakrishnan (Fig. 32-64). Domain I of the RRF

spans the A and P sites of the 50S ribosome, a position in

which the tip of its domain I would clash with a tRNA in

the P site (and which rationalizes why a ribosomal complex

of RRF and a tRNA in the P site has not been crystallized).

This suggests that RRF binding forces a tRNA bound in

the P site into the P/E hybrid binding state (Section 32-

3Dl). Previous structural studies suggested that RRF bind-

ing induced changes in the bridges connecting the small

and large subunits (Section 32-3Ae). However, no such

changes are observed in the above structure. Perhaps they

occur in the P/E hybrid state.

e. GTP Hydrolysis Speeds Up Ribosomal Processes

What is the role of the GTP hydrolysis reactions medi-

ated by the various ribosomally associated G proteins (IF-

2, EF-Tu, EF-G, and RF-3 in bacteria)? Translation occurs

in the absence of GTP, albeit slowly, so that the free energy

of the peptidyl transferase reaction is sufficient to drive the

entire translational process. Moreover, none of the GTP

1394 Chapter 32. Translation

Figure 32-63 X-ray structure of T. thermophilus RRF. This

monomeric protein is drawn in ribbon form colored in rainbow

order from its N-terminus (blue) to its C-terminus (red). [Based

on an X-ray structure by Yoshikazu Nakamura,The University of

Tokyo, Japan. PDBid 1EH1.]

Figure 32-64 X-ray structure of the T. thermophilus ribosome

in complex with RRF in its A site, the anticodon stem–loop

(ASL) of tRNA

Phe

in its P site, tRNA

Met

in its E site, and an

mRNA with a UAG Stop codon in the A site. The ribosomal

RNAs are shown as semitransparent surface diagrams with 23S

RNA light blue, 5S RNA blue-green, and 16S RNA yellow.The

ribosomal proteins are drawn in ribbon form with 50S subunit

proteins blue and 30S subunit proteins tan.The RRF, tRNAs, and

mRNA are represented by their surface diagrams with domains I

and II of RRF red and blue, the mRNA magenta, the P-site ASL

purple, and the E-site tRNA

Met

gray. [Courtesy of Venki

Ramakrishnan, MRC Laboratory of Molecular Biology, Cam-

bridge, U.K. PDBids 2V46 and 2V47.]

JWCL281_c32_1338-1428.qxd 8/19/10 10:06 PM Page 1394

hydrolysis reactions yields a “high-energy” covalent inter-

mediate as does, say, ATP hydrolysis in numerous biosyn-

thetic reactions. Instead, the ribosomal binding of a

G-protein induces it to hydrolyze its bound GTP to GDP

resulting a conformational change that causes the ribo-

some to carry out a particular process (

binding for IF-2, accommodation for EF-Tu, translocation

for EF-G, and RF-1/2 release for RF-3) and release the re-

sulting G-protein ⋅ GDP complex. The high rate and irre-

versibility of the GTP hydrolysis reaction ensures that the

various complex ribosomal processes to which it is coupled,

initiation, elongation, and termination, will themselves be

fast and irreversible. In essence, G-protein  GTP com-

plexes act as Maxwell’s demons to trap the ribosome in

functionally productive conformations. Hence, as we saw to

be the case for ribosomal proofreading (Section 32-3Eb),

the ribosome utilizes the free energy of GTP hydrolysis to

gain a more ordered (lower entropy) state rather than a

higher energy state as often occurs in ATP-dependent

processes.

G. Protein Synthesis Inhibitors: Antibiotics

Antibiotics are bacterially, fungally, or synthetically pro-

duced substances that inhibit the growth of microorganisms.

Antibiotics are known to inhibit a variety of essential

biological processes, including DNA replication (e.g.,

ciprofloxacin; Section 29-3Cd), transcription (e.g., rif-

amycin B; Section 31-2Bb), and bacterial cell wall synthesis

(e.g., penicillin; Section 11-3Bb). However, the majority of

known antibiotics, including a great variety of medically

useful substances, block translation. This situation is pre-

sumably a consequence of the translational machinery’s

enormous complexity, which makes it vulnerable to disrup-

tion in many ways. Antibiotics have also been useful in an-

fMet–tRNA

Met

alyzing ribosomal mechanisms because, as we have seen for

puromycin (Section 32-3Df), the blockade of a specific

function often permits its biochemical dissection into its

component steps.Table 32-10 and Fig.32-65 present several

medically significant and/or biochemically useful transla-

tional inhibitors. We study the mechanisms of a few of the

best characterized of them below.

a. Streptomycin

Streptomycin, which was discovered in 1944 by Selman

Waksman, is a medically important member of a family of

antibiotics known as aminoglycosides that inhibit prokary-

otic ribosomes in a variety of ways. At low concentrations,

streptomycin induces the ribosome to characteristically

misread mRNA: One pyrimidine may be mistaken for the

other in the first and second codon positions and either

pyrimidine may be mistaken for adenine in the first posi-

tion. This inhibits the growth of susceptible cells but does

not kill them. At higher concentrations, however, strepto-

mycin prevents proper chain initiation and thereby causes

cell death.

Certain streptomycin-resistant mutants (str

) have ri-

bosomes with an altered protein S12 compared with

streptomycin-sensitive bacteria (str

). Intriguingly, a

change in base C912 of 16S rRNA (which lies in its central

domain; Fig. 32-27a) also confers streptomycin resistance.

(Some mutant bacteria are not only resistant to strepto-

mycin but dependent on it; they require it for growth.) In

partial diploid bacteria that are heterozygous for strepto-

mycin resistance (str

/str

), streptomycin sensitivity is

dominant. This puzzling observation is explained by the

finding that, in the presence of streptomycin, str

ribo-

somes remain bound to initiation sites, thereby excluding

str

ribosomes from these sites. Moreover, the mRNAs

in these blocked complexes are degraded after a few

Section 32-3. Ribosomes and Polypeptide Synthesis 1395

Table 32-10 Some Ribosomal Inhibitors

Inhibitor Action

Chloramphenicol Inhibits peptidyl transferase on the prokaryotic large subunit

Cycloheximide Inhibits peptidyl transferase on the eukaryotic large subunit

Erythromycin Inhibits translocation by the prokaryotic large subunit

Fusidic acid Inhibits elongation in prokaryotes by binding to EF-G  GDP in a way that prevents its dissociation

from the large subunit

Paromomycin Increases the ribosomal error rate

Puromycin An aminoacyl–tRNA analog that causes premature chain termination in prokaryotes and eukaryotes

Streptomycin Causes mRNA misreading and inhibits chain initiation in prokaryotes

Tetracycline Inhibits the binding of aminoacyl–tRNAs to the prokaryotic small subunit

Diphtheria toxin Catalytically inactivates eEF2 by ADP-ribosylation

Ricin/abrin/-sarcin Ricin and abrin are poisonous plant glycosidases that catalytically inactivate the eukaryotic large subunit

by hydrolytically depurinating a specific highly conserved A residue of the 28S RNA, which is located on

the so-called sarcin–ricin loop that forms a critical part of the ribosomal factor–binding center; -sarcin

is a fungal protein that cleaves a specific phosphodiester bond in the sarcin–ricin loop

JWCL281_c32_1338-1428.qxd 8/19/10 10:06 PM Page 1395

1396 Chapter 32. Translation

C C

NH CHCl

Chloramphenicol

CHOH

Cycloheximide

N(CH

)

HOH

OCH

Erythromycin

COOH

OCCH

Fusidic acid

OH OH

N(CH

)

Tetracycline

HOH

OH H

HNH

OH H

HNH

Paromomycin

CNH

HHO

HN HH

Streptomycin

CHO H

OH O

NH C

Figure 32-65 Selection of antibiotics that act as translational inhibitors.

JWCL281_c32_1338-1428.qxd 8/19/10 10:06 PM Page 1396

minutes, which allows the str

ribosomes to bind to newly

synthesized mRNAs as well.

b. Chloramphenicol

Chloramphenicol, the first of the “broad-spectrum” an-

tibiotics, inhibits the peptidyl transferase activity on the

large subunit of prokaryotic ribosomes. However, its clini-

cal uses are limited to only severe infections because of its

toxic side effects, which are caused, at least in part, by the

chloramphenicol sensitivity of mitochondrial ribosomes.

The 23S RNA is implicated in chloramphenicol binding by

the observation that some of its mutants are chloram-

phenicol resistant. Indeed, X-ray studies indicate that

chloramphenicol binds in the large subunit’s polypeptide

exit tunnel in the vicinity of the A site. This explains why

chloramphenicol competes for binding with the 3¿ end of

aminoacyl–tRNAs as well as with puromycin (whose ribo-

somal binding site overlaps that of chloramphenicol) but

not with peptidyl–tRNAs. These observations suggest that

chloramphenicol inhibits peptidyl transfer by interfering

with the interactions of ribosomes with A site–bound

aminoacyl–tRNAs.

c. Paromomycin

Paromomycin, a clinically useful aminoglycoside antibi-

otic, increases the ribosomal error rate. The X-ray structure

of the 30S subunit in complex with paromomycin (Fig. 32-66)

reveals that it binds to the interior of the RNA loop in

which the bases of A1492 and A1493 are normally stacked

(Fig. 32-59a).This causes these bases to flip out of the loop

and assume a conformation resembling that in the

codon–anticodon–30S subunit complex (Fig. 32-59b). In-

deed, this codon–anticodon–30S subunit complex is not

significantly disturbed by the binding of paromomycin. As

we have seen in Section 32-3Ea, the 30S subunit employs

A1492 and A1493 to ascertain whether the first two

codon–anticodon base pairs are Watson–Crick base pairs,

that is, whether the incoming tRNA is cognate to the codon

in the A site. Noncognate tRNAs normally have insuffi-

cient codon–anticodon binding energy to flip A1492 and

A1493 out of the loop and consequently are rejected by the

ribosome. However, the binding of paromomycin to the

30S subunit pays the energetic price of these base flips.This

facilitates the ribosomal acceptance (stabilizes the binding)

of near-cognate aminoacyl–tRNAs and hence the erro-

neous incorporation of their amino acid residues into the

polypeptide being synthesized.

d. Tetracycline

Tetracycline and its derivatives are broad-spectrum

antibiotics that bind to the small subunit of prokaryotic ri-

bosomes, where they inhibit aminoacyl–tRNA binding. An

X-ray structure of tetracycline in complex with the 30S

subunit reveals that tetracycline mainly binds in a crevice

comprised of only the 3¿ major domain of 16S RNA (Fig.

32-27a) and which is located in the neck of the 30S subunit

just above its A site.This permits the initial screening of the

aminoacyl–tRNA to proceed but physically blocks its ac-

commodation into the peptidyl transferase (A/A) site after

EF-Tu–catalyzed GTP hydrolysis has occurred, resulting in

the release of the tRNA. Hence, in addition to preventing

protein synthesis, tetracycline binding causes the unpro-

ductive hydrolysis of GTP which, since this occurs every

time a cognate aminoacyl–tRNA binds to the ribosome,

poses an enormous energetic drain on the cell. The nu-

cleotides forming the tetracycline binding site are poorly

conserved in eukaryotic ribosomes, thereby accounting for

tetracycline’s bacterial specificity.

Tetracycline also blocks the stringent response (Section

31-3I) by inhibiting (p)ppGpp synthesis. This indicates that

deacylated tRNA must bind to the A site in order to acti-

vate stringent factor.

Tetracycline-resistant bacterial strains have become

quite common, thereby precipitating a serious clinical

problem. Resistance is often conferred by a decrease in

bacterial cell membrane permeability to tetracycline rather

than any alteration of ribosomal components.

e. Diphtheria Toxin

Diphtheria is a disease resulting from bacterial infection

by Corynebacterium diphtheriae that harbor the bacterio-

phage corynephage . Diphtheria was a leading cause of

childhood death until the late 1920s when immunization be-

came prevalent. Although the bacterial infection is usually

confined to the upper respiratory tract, the bacteria secrete

a phage-encoded protein, known as diphtheria toxin (DT),

which is responsible for the disease’s lethal effects. Diphthe-

ria toxin specifically inactivates the eukaryotic elongation

factor eEF2, thereby inhibiting eukaryotic protein synthesis.

Section 32-3. Ribosomes and Polypeptide Synthesis 1397

Figure 32-66 X-ray structure of the 30S ribosome in complex

with the antibiotic paromomycin. The view and coloring are the

same as those in Fig. 32-59 with the paromomycin (PAR) drawn

in stick form in yellow-green. [Courtesy of Venki Ramakrishnan,

MRC Laboratory of Molecular Biology, Cambridge, U.K. PDBid

1IBK.]

JWCL281_c32_1338-1428.qxd 8/4/10 4:45 PM Page 1397

The pathogenic effects of diphtheria are prevented, as

was discovered in the 1880s, by immunization with toxoid

(formaldehyde-inactivated toxin). Individuals who have

contracted diphtheria are treated with antitoxin from

horse serum, which binds to and thereby inactivates DT, as

well as with antibiotics to combat the bacterial infection.

DT is a member of the family of bacterial toxins that in-

cludes cholera toxin (CT) and pertussis toxin (PT; Section

19-2Ce). It is a monomeric 535-residue protein that is readily

cleaved past its Arg residues 190, 192, and 193 by trypsin

and trypsinlike enzymes. This hydrolysis occurs around the

time diphtheria toxin encounters its target cell, yielding

two fragments, A and B, which, nevertheless, remain linked

by a disulfide bond. The B fragment’s C-terminal domain

binds to a specific receptor on the plasma membrane of

susceptible cells, thereby inducing DT’s uptake into the en-

dosome (Fig. 12-91) via receptor-mediated endocytosis

(Section 12-5Bc; free fragment A is devoid of toxic activ-

ity).The endosome’s low pH of 5 triggers a conformational

change in the B fragment’s N-terminal domain, which then

inserts into the endosomal membrane so as to facilitate the

entry of the A fragment into the cytoplasm. The disulfide

bond linking the A and B subunits is then cleaved by the

cytoplasm’s reducing environment.

Within the cytosol, the A fragment catalyzes the ADP-

ribosylation of eEF2 by NAD



thereby inactivating this elongation factor. Since the A

fragment acts catalytically, one molecule is sufficient to

ADP-ribosylate all of a cell’s eEF2s, which halts protein

synthesis and kills the cell. Only a few micrograms of diph-

theria toxin are therefore sufficient to kill an unimmu-

nized individual.

Diphtheria toxin specifically ADP-ribosylates a modi-

fied His residue on eEF2 known as diphthamide:

NH CH

OH OH

ADP-Ribosylated diphthamide

OADP

ADP-ribosyl group

His residue

N(CH

)

diphtheria toxin

(inactive)

(active)

+ NAD

eEF2

DP-ribosyl-eEF2 + Nicotinamide + H

Diphthamide occurs only in eEF2 (not even in its bacter-

ial counterpart, EF-G), which accounts for the specificity

of diphtheria toxin in exclusively modifying eEF2 (recall

that CT ADP-ribosylates a specific Arg residue on G

s

and PT ADP-ribosylates a specific Cys residue on G

;

Section 19-2C). Since diphthamide occurs in all eukary-

otic eEF2s, it probably is essential to eEF2 activity. Yet,

certain mutant cultured animal cells, which have unim-

paired capacity to synthesize proteins, lack the enzymes

that post-translationally modify His to diphthamide

(although mutating the diphthamide His to Asp, Lys, or

Arg inactivates translation). Perhaps the diphthamide

residue has a control function.

4 CONTROL OF EUKARYOTIC

TRANSLATION

The rates of ribosomal initiation on prokaryotic mRNAs

differ by factors of up to 100, a variation that is largely a

consequence of their different Shine–Dalgarno se-

quences. Moreover, the genes forming an operon are of-

ten expressed in decreasing molar amounts from the

operon’s 5¿ end to its 3¿ end. For example, the proteins

specified by the E. coli lac operon (Section 31-1Ab), -

galactosidase, galactose permease, and thiogalactoside

transacetylase, are produced in molar ratios of 10:5:2.

Such polarity may arise when the initiation codon of a

gene that lacks a Shine–Dalgarno sequence is very near

the Stop codon of its upstream gene, a situation that oc-

curs most often when the Stop codon overlaps the initia-

tion codon as in the sequence AUGA. The translation of

the upstream gene will then be required for the transla-

tion of the downstream gene, a phenomenon termed

translational coupling. The polarity arises because a ribo-

some often dissociates from the mRNA on encountering

the upstream gene’s Stop codon.Alternatively, an mRNA

may fold in a way that masks an internal Shine–Dalgarno

sequence, for example, by the base pairing of a segment

adjacent to the Shine–Dalgarno sequence to a down-

stream element of the preceding gene. Such Shine–Dal-

garno sequences only become available when a ribosome

that is translating the preceding gene disrupts the folded

structure.

Genetic expression in prokaryotes is largely transcrip-

tionally controlled (Section 31-3). This is apparently be-

cause prokaryotic mRNAs have lifetimes of only a few

minutes and, hence, it is a more efficient use of resources to

control their transcription. Nevertheless, the expression of

certain prokaryotic genes is translationally controlled,

most notably those encoding the ribosomal proteins

(which comprise 10% of cellular proteins), which must be

produced in equimolar amounts. The production of riboso-

mal proteins is controlled, in part, through a process in

which a ribosomal protein binds to the mRNA of the

operon encoding it in the vicinity of a translational start

site located near the mRNA’s 5¿ end so as to inhibit its

translational initiation. However, each such protein binds

1398 Chapter 32. Translation

JWCL281_c32_1338-1428.qxd 9/7/10 2:23 PM Page 1398

more tightly to an rRNA in forming the ribosome. Conse-

quently, only when there is an excess of that protein will it

inhibit its own translation as well as those of other proteins

encoded by its operon.

Eukaryotic cells, whose mRNAs have lifetimes of hours

or days, respond to many of their needs through transla-

tional control. In this section, we examine how eukaryotic

translation is regulated via the phosphorylation/dephos-

phorylation of eIF2 and eIF4E. We then consider transla-

tional control by mRNA masking and cytoplasmic

polyadenylation and end by discussing the uses of anti-

sense oligonucleotides.

A. Regulation of eIF2

Four important pathways for the regulation of translation

in eukaryotes involve the phosphorylation of the con-

served Ser 51 on the  subunit of eIF2 (eIF2; recall that

eIF2 is an  trimer that conducts to the 40S

ribosomal subunit, and the resulting complex scans the

bound mRNA for the initiating AUG codon to form the

48S preinitiation complex; Section 32-3Cd). The so-called

eIF2 kinases that do so share a conserved kinase domain

but have unique regulatory domains.

a. Heme Availability Controls Globin Translation

Reticulocytes synthesize protein, almost exclusively he-

moglobin, at an exceedingly high rate and are therefore a

favorite subject for the study of eukaryotic translation. He-

moglobin synthesis in fresh reticulocyte lysates proceeds

normally for several minutes but then abruptly stops be-

cause of the inhibition of translational initiation and the

consequent polysome disaggregation. This process is pre-

vented by the addition of heme [a mitochondrial product

(Section 26-4A) that this in vitro system cannot synthe-

size], thereby indicating that globin synthesis is regulated by

heme availability. The inhibition of globin translational ini-

tiation is also reversed by the addition of the eukaryotic

initiation factor eIF2 and by high levels of GTP.

In the absence of heme, reticulocyte lysates accumulate

an eIF2 kinase named heme-regulated inhibitor [HRI;

also called heme-controlled repressor (HCR)]. HRI is a

homodimer whose 629-residue subunits each contain two

heme-binding sites. When heme is plentiful, both of these

sites are occupied and the protein, which is autophospho-

rylated at several Ser and Thr residues, is inactive. How-

ever, when heme is scarce, one of these sites loses its

bound heme, thereby activating HRI to autophosphory-

late itself at several additional sites and to phosphorylate

Ser 51 of eIF2.

Phosphorylated eIF2 can participate in the ribosomal

initiation process in much the same way as unphosphory-

lated eIF2. This puzzling observation was clarified by the

discovery that GDP does not dissociate from phosphory-

lated eIF2 at the completion of the initiation process as it

normally does through a process facilitated by eIF2B act-

ing as a GEF (Fig. 32-44). This is because phosphorylated

eIF2 forms a much tighter complex with eIF2B than does

unphosphorylated eIF2. This sequesters eIF2B (Fig. 32-

Met–tRNA

Met

67), which is present in lesser amounts than eIF2, thereby

preventing the regeneration of the eIF2  GTP required

for translational initiation.The presence of heme reverses

this process by inhibiting HRI, whereon the phosphory-

lated eIF2 molecules are reactivated through the action

of eIF2 phosphatase, which is unaffected by heme. The

reticulocyte thereby coordinates its synthesis of globin

and heme.

b. Interferons Protect against Viral Infection

Interferons are cytokines that are secreted by virus-infected

vertebrate cells. On binding to surface receptors of other cells,

interferons convert them to an antiviral state,which inhibits the

replication of a wide variety of RNA and DNA viruses. In-

deed, the discovery of interferons in the 1950s arose from the

observation that virus-infected individuals are resistant to in-

fection by a second type of virus.

There are three families of interferons: type  or leuko-

cyte interferon (165 residues; leukocytes are white blood

cells), the related type  or fibroblast interferon (166

residues; fibroblasts are connective tissue cells), and type 

or lymphocyte interferon (146 residues; lymphocytes are

immune system cells). Interferon synthesis is induced by the

double-stranded RNA (dsRNA) that is generated during in-

fection by both DNA and RNA viruses, as well as by the syn-

thetic dsRNA poly(I)  poly(C). Interferons are effective an-

tiviral agents in concentrations as low as 3  10

14

M, which

makes them among the most potent biological substances

known. Moreover, they have far wider specificities than an-

tibodies raised against a particular virus. They have there-

fore elicited great medical interest, particularly since some

cancers are virally induced (Section 19-3B). Indeed, they

are in clinical use against certain tumors and viral infec-

tions.These treatments are made possible by the production

of large quantities of these otherwise quite scarce proteins

through recombinant DNA techniques (Section 5-5G).

Interferons prevent viral proliferation largely by inhibit-

ing protein synthesis in infected cells (lymphocyte interferon

Section 32-4. Control of Eukaryotic Translation 1399

Figure 32-67 Model for heme-controlled protein synthesis in

reticulocytes.

ATP ADP

HRI

(active)

Inhibited

by heme

pro-HRI

(inactive)

eIF2 phosphatase

eIF2 eIF2

eIF-2B

eIF2

eIF2B

P P

JWCL281_c32_1338-1428.qxd 8/4/10 4:45 PM Page 1399

ATP ADP

Double-stranded

RNA-activated

protein kinase (PKR)

eIF2 Phosphatase

eIF2 eIF2

eIF2B

eIF2

eIF2B

dsRNA

Interferon-induced

2,5A synthetase

dsRNA

RNase L

(inactive)

ATP

2,5 A

(2',5')-phospho-

diesterase

ATP + AMP

RNase L

(active)

mRNA Nucleotides

(a)

(b)

Inhibition of Translation

mRNA Degradation

P P

also modulates the immune response). They do so in two

independent ways (Fig. 32-68):

1. Interferons induce the production of an eIF2 ki-

nase, double-stranded RNA–activated protein kinase

[PKR; also known as double-stranded RNA–activated in-

hibitor (DAI); 551 residues], which on binding dsRNA,

dimerizes and autophosphorylates itself. This activates

PKR to phosphorylate eIF2 at its Ser 51, thereby inhibit-

ing ribosomal initiation and hence the proliferation of

viruses in virus-infected cells. The importance of PKR to

cellular antiviral defense is indicated by the observation

that many viruses express inhibitors of PKR.

2. Interferons also induce the synthesis of (2,5)-

oligoadenylate synthetase (2,5A synthetase). In the pres-

ence of dsRNA, this enzyme catalyzes the synthesis from

ATP of the unusual oligonucleotide pppA(2p5A)

where

n  1 to 10. This compound, 2,5-A, activates a preexisting

endonuclease, RNase L, to degrade mRNA, thereby inhibit-

ing protein synthesis. 2,5-A is itself rapidly degraded by an

enzyme named (2,5)-phosphodiesterase so that it must be

continually synthesized to maintain its effect.

The independence of the 2,5-A and PKR systems is

demonstrated by the observation that the effect of 2,5-A

on protein synthesis is reversed by added mRNA but not

by added eIF2. [Recall that RNA interference (RNAi; Sec-

tion 31-4At) constitutes an alternative dsRNA-based an-

tiviral defense.]

c. PERK Prevents the Buildup of Unfolded Proteins

in the ER

PKR-like endoplasmic reticulum kinase (PERK), a

1087-residue transmembrane protein, resides in the endo-

plasmic reticulum (ER) membrane of all multicellular eu-

karyotes. It is repressed by its binding to the ER-resident

chaperone BiP (Section 12-4Bf).When the ER contains an

excessive amount of unfolded proteins (caused by various

forms of stress such as high temperatures), BiP dissociates

from PERK, thereby activating PERK to phosphorylate

eIF2 at its Ser 51 and hence inhibit translation. Thus

PERK functions to protect the cell from the irreversible

damage caused by the accumulation of unfolded proteins

in the ER.

Wolcott–Rallison syndrome is a genetic disease charac-

terized mainly by insulin-dependent (type I) diabetes that

develops in early infancy (type I diabetes usually first ap-

pears in childhood; Section 27-4B). It is caused by muta-

tions in the catalytic domain of PERK. This results in the

death of pancreatic  cells, in which PERK is particularly

abundant. Multiple systemic disorders subsequently occur

including osteoporosis (reduction in the quantity of bone)

and growth retardation.

d. GCN2 Regulates Amino Acid Biosynthesis

GCN2 (1590 residues),the sole eIF2 kinase in yeast, is a

transcriptional activator of the gene encoding GCN4, a tran-

scriptional activator of numerous yeast genes, many of which

encode enzymes that participate in amino acid biosynthetic

pathways. The C-terminal domain of GCN2, which resem-

bles histidyl–tRNA synthetase (HisRS), preferentially binds

uncharged tRNAs (whose presence is indicative of an insuf-

ficient supply of amino acids). The binding of an uncharged

tRNA to this HisRS-like domain activates the adjacent

eIF2 kinase domain and thereby inhibits translational initi-

ation, although at only a modest level.

Despite this inhibition of yeast protein synthesis, acti-

vated GCN2 induces the expression of GCN4. This seem-

ingly paradoxical property of GCN2, as Alan Hinnebusch

explained, arises from the fact that GCN4 mRNA contains

four short so-called upstream open reading frames

1400 Chapter 32. Translation

Figure 32-68 The action of interferon. In interferon-treated

cells, the presence of dsRNA, which normally results from a viral

infection, causes (a) the inhibition of translational initiation and

(b) the degradation of mRNA, thereby blocking translation and

preventing virus replication.

JWCL281_c32_1338-1428.qxd 8/19/10 10:06 PM Page 1400