Voet D., Voet Ju.G. Biochemistry

Подождите немного. Документ загружается.

by bisulfite treatment (Section 30-7b), which reestablishes

the “missing” base pair as U ⴢ A. Similarly, Sec–tRNA

Sec

which is also not bound by EF-Tu (but rather by SELB;

Section 32-2De), has 8 bp in its acceptor stem vs 7 bp in

those of other elongator tRNAs. However, initiator tRNAs

from several sources have fully base paired acceptor stems,

and the U1 ⴢ A72 base pair of tRNA

Gln

is opened up on

binding to GlnRS (Section 32-2Cc).

c. EF-Tu Undergoes a Major Conformational Change

on Hydrolyzing GTP

Morten Kjeldgaard and Nyborg determined the X-ray

structures of T. aquaticus EF-Tu (405 residues) in com-

plex with GMPPNP and the 70% identical E. coli EF-Tu

(393 residues) in complex with GDP (Fig. 32-48).The con-

formation of EF-Tu in its complex with only GMPPNP

closely resembles that in its ternary complex with

Phe–tRNA

Phe

and GMPPNP (Fig. 32-47). However, com-

parison of the GMPPNP and GDP complexes indicates

that, on hydrolyzing its bound GTP, EF-Tu undergoes a

major structural reorganization. Its greatest local confor-

mational changes occur in the Switch I and Switch II re-

gions of domain 1, which in all G-proteins signal the state

of the bound nucleotide to interacting partners (Section

19-2Cb; here domains 2 and 3): Switch I converts from a ␤

hairpin to a short ␣ helix and the ␣ helix of Switch II shifts

toward the C-terminus by 4 residues. As a consequence,

this latter helix reorients by 42°, which results in domain 1

rigidly changing its orientation with respect to domains 2

and 3 by a dramatic 91° rotation.The tRNA binding site is

thereby eliminated.

d. EF-Ts Disrupts the Binding of GDP to EF-Tu

EF-Tu has a 100-fold higher affinity for GDP than GTP.

Hence, replacement of the EF-Tu–bound GDP by GTP

must be facilitated by the interaction of EF-Tu with EF-Ts

(Fig. 32-46, top). The X-ray structure of the EF-Tu ⴢ EF-Ts

complex, determined by Stephen Cusack and Reuben

Section 32-3. Ribosomes and Polypeptide Synthesis 1381

Figure 32-47 X-ray structure of the ternary complex of yeast

Phe–tRNA

Phe

, Thermus aquaticus EF-Tu, and GMPPNP. The

EF-Tu is drawn in ribbon form colored in rainbow order from its

N-terminus (blue) to its C-terminus (red).The tRNA is shown in

stick form colored according to atom type with C green, N blue,

O red, and P orange and with orange rods linking successive P

atoms.The tRNA’s appended aminoacyl–Phe residue and the

GMPPNP that is bound to the EF-Tu are drawn in space-filling

form with C atoms cyan and yellow, respectively. Two bound

2⫹

ions are represented by magenta spheres. [Based on an

X-ray structure by Jens Nyborg, University of Aarhus, Århus,

Denmark. PDBid 1TTT.]

Figure 32-48 Comparison of the X-ray structures of EF-Tu in

its complexes with GDP and GMPPNP. The protein is

represented by its C

␣

backbone with domain 1, its GTP-binding

domain, purple in the GDP complex and red in the GMPPNP

complex. Domain 2 and domain 3, which have the same

orientation in both complexes, are green and cyan.The bound

GDP and GMPPNP are shown in stick form with C yellow, N

blue, O red, and P green. [Courtesy of Morten Kjeldgaard and

Jens Nyborg, University of Aarhus, Århus, Denmark. PDBid

1EFT.]

See Interactive Exercise 45

JWCL281_c32_1338-1428.qxd 10/19/10 8:24 AM Page 1381

Leberman, reveals that the EF-Tu has a conformation re-

sembling that of its GDP complex (Fig. 32-49) but with

its domains 2 and 3 swung away from domain 1 by ⬃18°.

EF-Ts is an elongated molecule that binds along the right side

of EF-Tu as shown in Fig. 32-49, where it contacts EF-Tu’s

domains 1 and 3. The intrusive interactions of EF-Ts side

chains with the GDP binding pocket on EF-Tu domain 1

disrupts the Mg

2

ion binding site. This reduces the affinity

of EF-Tu for GDP, thereby facilitating its exchange for

GTP (after EF-Ts has dissociated), which has a 10-fold

higher concentration in the cell than does GDP (the GEF-

containing segment of Sos similarly interferes with Mg

2

binding and hence guanine nucleotide binding by Ras; Sec-

tion 19-3Cf). EF-Tu’s subsequent binding of a charged

elongator tRNA increases its affinity for GTP.

e. Transpeptidation

In the transpeptidation stage of the elongation cycle

(Fig. 32-46), the peptide bond is formed through the nucle-

ophilic displacement of the P-site tRNA by the amino

group of the 3¿-linked aminoacyl–tRNA in the A site (Fig

32-39). The nascent polypeptide chain is thereby length-

ened at its C-terminus by one residue and transferred to

the A-site tRNA. The reaction occurs without the need of

activating cofactors such as ATP because the ester linkage

between the nascent polypeptide and the P-site tRNA is a

“high-energy” bond. The peptidyl transferase center that

catalyzes peptide bond formation is located entirely on the

large subunit as is demonstrated by the observation that in

high concentrations of organic solvents such as ethanol, the

large subunit alone catalyzes peptide bond formation. The

organic solvent apparently distorts the large subunit in a

way that mimics the effect of small subunit binding.

f. Puromycin Is an Aminoacyl–tRNA Analog

The ribosomal elongation cycle was originally character-

ized through the use of the antibiotic puromycin (Fig.32-50).

This product of Streptomyces alboniger, which resembles the

3¿ end of Tyr–tRNA, causes the premature termination of

polypeptide chain synthesis. Puromycin, in competition with

the mRNA-specified aminoacyl–tRNA but without the

need of elongation factors, binds to the ribosomal A site

which, in turn, catalyzes a normal transpeptidation reaction

to form peptidyl–puromycin. Yet, the ribosome cannot

catalyze the transpeptidation reaction in the next elonga-

tion cycle because puromycin’s “amino acid residue” is

linked to its “tRNA” via an amide rather than an ester bond.

Polypeptide synthesis is therefore aborted and the pep-

tidyl–puromycin is released.

In the absence of the elongation factor EF-G (see be-

low), an active ribosome cannot bind puromycin because

its A site is at least partially occupied by a peptidyl–tRNA.

A newly initiated ribosome, however, violates this rule; it

1382 Chapter 32. Translation

Figure 32-50 Puromycin. This antibiotic (left)

resembles the 3¿-terminus of tyrosyl–tRNA (right).

Figure 32-49 X-ray structure of the E. coli EF-Tu  EF-Ts

complex. Domains 1, 2, and 3 of EF-Tu are magenta, green, and

cyan, respectively, and EF-Ts is orange. [Based on an X-ray

structure by Stephen Cusack and Reuben Leberman, EMBL,

Grenoble Cedex, France. PDBid 1EFU.]

Puromycin

tRNA

O OH

Tyrosyl–tRNA

–

JWCL281_c32_1338-1428.qxd 8/4/10 4:45 PM Page 1382

...

–

O OH

NH CH

O OH

OCH

(a) (b)

tRNA (P site) tRNA (A site)

NH CH

NH OH

n–1

Peptidyl transferase

tetrahedral intermediate

CCdA-p-Puro

catalyzes fMet–puromycin formation. These observations

demonstrated the functional existence of the ribosomal P

and A sites and established that binds directly

to the P site, whereas other aminoacyl–tRNAs must first en-

ter the A site.

g. Transpeptidation: The Ribosome Is a Ribozyme

What is the nature of the peptidyl transferase center,

that is, does it consist of RNA, protein, or both? Since all

proteins, including those associated with ribosomes, are ri-

bosomally synthesized, the primordial ribosome must have

preceded the primordial proteins and hence consisted

entirely of RNA. Despite this (in hindsight) obvious evolu-

tionary argument, the idea that rRNA functions catalyti-

cally was not seriously entertained until after it had been

discovered that RNA can, in fact, act as a catalyst (Section

31-4Ae). Several other observations further indicate that

the ribosome is a ribozyme:

1. The absence from the 50S subunit of any one of its

proteins but L2, L3, and L4 does not abolish its peptidyl

transferase function.

2. rRNAs are more highly conserved throughout evolu-

tion than are ribosomal proteins.

3. Most mutations that confer resistance to antibiotics

that inhibit protein synthesis occur in genes encoding

rRNAs rather than ribosomal proteins.

Nevertheless, the unambiguous demonstration that rRNA

functions catalytically in polypeptide synthesis proved to be

fMet–tRNA

Met

surprisingly elusive. Noller succeeded in showing that the T.

thermophilus large ribosomal subunit from which ⬃95% of

the protein had been removed by treatment with SDS and

proteinase K followed by phenol extraction (which dena-

tures proteins; Section 6-6A) maintained 80% of its pep-

tidyl transferase activity in a model reaction. Moreover, this

activity was abolished by RNase treatment. However, since

the remaining protein was due to several intact ribosomal

proteins (which are presumably sequestered within the 23S

RNA), it could be argued that these proteins are essential

for ribosomal catalytic function,a reasonable expectation in

light of the 3.5 billion years over which ribosomal proteins

and RNAs have coevolved.

Steitz and Moore unequivocally determined the nature

of the peptidyl transferase center through its identification

in the X-ray structure of the 50S subunit. Peptide bond for-

mation presumably resembles the reverse of peptide bond

hydrolysis such as that catalyzed by serine proteases (Sec-

tion 15-3C). The ribosomal reaction’s tetrahedral interme-

diate (Fig. 32-51a) is mimicked by a compound synthesized

by Michael Yarus that consists of the trinucleotide CCdA

linked to puromycin via a phosphoramidite group (Fig. 32-

51b).This compound, which is named CCdA-p-Puro, binds

tightly to the ribosome so as to inhibit its peptidyl trans-

ferse activity. The X-ray structure of the 50S subunit in

complex with CCdA-p-Puro reveals that the inhibitor

binds to domain V of the 23S RNA (Fig. 32-27b) at the en-

trance to the ⬃100-Å-long polypeptide exit tunnel that

runs through to the back of the subunit (Figs. 32-30 and 32-

34c). There, the inhibitor is completely enveloped in RNA

Section 32-3. Ribosomes and Polypeptide Synthesis 1383

Figure 32-51 Ribosomal tetrahedral intermediate and its

analog. (a) The chemical structure of the tetrahedral intermediate

(red C) in ribosomally mediated peptide bond formation in

which the A-site aminoacyl residue is Tyr. (b) CCdA-p-Puro, the

transition state analog of the tetrahedral intermediate in Part a

produced by linking the 3¿-OH group of CCdA to the amino

group of puromycin’s O-methyltyrosine residue via a phosphoryl

group.

JWCL281_c32_1338-1428.qxd 8/4/10 4:45 PM Page 1383

with no protein side chain approaching closer than ⬃18 Å to

the inhibitor’s phosphoramidite group and with the nearest

2

ion 8.5 Å away—both too far away to participate in

catalysis. Moreover, all the nucleotides that contact the

CCdA-p-Puro are 95% conserved among all three king-

doms of life. Clearly, the ribosomal transpeptidase reaction

is catalyzed by RNA.

Despite the foregoing, the X-ray structure of the T.

thermophilus ribosome in complex with tRNA and

mRNA (Fig. 32-34) reveals that the N-terminal tail of L27

interacts with the phosphate group of residue C75 of the

A-site tRNA. Moreover, L16 interacts with the elbow re-

gion of the A-site tRNA via interactions between con-

served Arg residues and phosphate and 2¿-OH groups in

the elbow region of the tRNA. These observations con-

firm previous mutational studies indicating that L16 and

L27 help stabilize the binding of the A-site tRNA. How-

ever, some organisms lack L27, which indicates that its

binding to A-site tRNA is not part of an evolutionarily

conserved mechanism.

h. Peptide Bond Formation Does Not Occur via

Acid–Base Catalysis

The ribosomal peptidyl transferase reaction occurs

⬃10

-fold faster than the uncatalyzed reaction. How does

the ribosome catalyze this reaction? Peptide bond forma-

tion is naively expected to proceed via the nucleophilic at-

tack of the amino group on the carbonyl group of an ester

to form a tetrahedral intermediate that collapses to an

amide and an alcohol (Fig. 32-39). However, in the physio-

logical pH range, the attacking amino group is predomi-

nantly in its ammonium form (RNH



), and hence lacks

the lone pair necessary to undertake a nucleophilic attack.

This suggests that the peptidyl transferase reaction is cat-

alyzed in part by a general base that abstracts a proton

from the ammonium group to generate the required free

amino group (RNH

Inspection of the peptidyl transferase center in H.

marismortui reveals that the only basic group within 5 Å of

the inferred position of the attacking amino group is atom

N3 of the invariant rRNA base A2486 (A2451 in E. coli). It

is ⬃3 Å from and hence hydrogen bonded to the attacking

amino group (Fig.32-52).This further suggests that the pro-

tonated A2486-N3 electrostatically stabilizes the oxyanion

of the tetrahedral reaction intermediate and then donates

the proton to the leaving group of the P-site tRNA to yield

a 3¿-OH group (general acid catalysis). However, in order

for A2486-N3 to act as a general base in abstracting the

proton from an ammonium group (whose pK is ⬃10), it

must have a pK of at least 7 (recall that proton transfers be-

tween hydrogen-bonded groups occur at physiologically

significant rates only when the pK of the proton donor is

no more than 2 or 3 pH units greater than that of the pro-

ton acceptor; Section 15-3Dd).Yet, the pK of N3 in AMP is

3.5. Moreover, several lines of evidence indicate that

A2486 does not function as an acid–base catalyst including

(1) the model displayed in Fig. 32-52 indicates that the

tetrahedral intermediate’s oxyanion would point away

from and hence could not be stabilized by protonated

A2486-N3; and (2) mutagenic replacement of A2486 by

any other base does not greatly alter the ribosome’s cat-

alytic rate.

i. The Ribosome Is an Entropy Trap

If the peptidyl transfer reaction does not involve

acid–base catalysis, what is the origin of the ribosome’s

catalytic power? Marina Rodnina and Richard Wolfenden

noted that the uncatalyzed reaction of esters with amines

to form amides occurs quite facilely in aqueous solution.

They therefore measured the rates of both uncatalyzed

peptide bond formation by model compounds and pep-

tidyl transfer by the ribosome at several different temper-

atures. This provided values of and , the re-

action’s change in the enthalpy and entropy of activation

by the ribosome relative to the uncatalyzed reaction. Here

¢¢H

‡

cat

 T ¢¢S

‡

cat

 ¢¢G

‡

cat

 ¢G

‡

cat

(uncat)  ¢G

‡

(cat),

¢¢S

‡

cat

¢¢H

‡

cat

1384 Chapter 32. Translation

Figure 32-52 Model of the substrate complex of the 50S

ribosomal subunit. Atoms are colored according to type with the

A-site substrate C and P purple, P-site substrate C and P green,

23S rRNA C and P orange, N blue, and O red.The attacking

amino group of the A-site aminoacyl residue is held in position

for nucleophilic attack (cyan arrow) on the carbonyl C of the

P-site aminoacyl ester through hydrogen bonds (dashed black

lines) to A2486-N3 and the 2¿-O of the P-site A76. [Courtesy of

Peter Moore and Thomas Steitz, Yale University.]

where is the change in the reaction’s free energy ¢¢G

‡

cat

of activation by the ribosome, G

‡

(cat), relative to that of

the uncatalyzed reaction, G

‡

(uncat) (Section 14-1C).

The measured value of is 19 kJ  mol

1

, a quan-

tity that would be positive, not negative, if the ribosomal

reaction had a significant component of chemical catalysis

such as acid–base catalysis and/or the formation of new

hydrogen bonds. In contrast, the value of is 52 kJ 

mol

1

, which indicates that the Michaelis complex in the ri-

bosomal reaction is significantly more ordered relative to

the transition state than is the uncatalyzed reaction. This

value of largely accounts for the observed ⬃10

fold rate enhancement of the ribosomal reaction relative to

T ¢¢S

‡

cat

T ¢¢S

‡

cat

¢¢H

‡

cat

JWCL281_c32_1338-1428.qxd 8/4/10 4:45 PM Page 1384

the uncatalyzed reaction (the rate enhancement by the ri-

bosome is given by ; Section 14-1Cd). Evidently, the

ribosome enhances the rate of peptide bond formation by

properly positioning and orienting its substrates and/or ex-

cluding water from the preorganized electrostatic environ-

ment of the active site (a form of reactant ordering) rather

than by chemical catalysis.

The X-ray structures of the large ribosomal subunit in

complex with aminoacyl–tRNA and peptidyl–tRNA indi-

cate that the ribosome uses an induced fit mechanism, as

occurs in enzymes such as hexokinase (Section 17-2Aa).

Conformational changes in the 23S rRNA, presumably

triggered by proper binding of the aminoacyl–tRNA in the

A site, orient the ester group of the peptidyl–tRNA for nu-

cleophilic attack. The hydrogen bond between the 2¿-OH

of the P-site A76 and the attacking amino group (Fig. 32-

52) is crucial in doing so. In fact, replacing this 2¿-OH group

with H or F reduces the reaction rate by at least 10

. This

suggests that the peptidyl transferase reaction occurs via

the substrate-assisted proton shuttle mechanism dia-

grammed in Fig. 32-53. Although the peptidyl transferase

reaction is relatively sluggish compared to the reactions

catalyzed by many protein enzymes, it is sufficiently fast to

keep up with the other ribosomal processes (which collec-

tively link together 10–20 residues/s).Apparently, the ribo-

some’s peptidyl transferase function is a molecular fossil

from the RNA world.

In the absence of a tRNA in the A site, the ester bond

linking the peptidyl group to the P-site tRNA is shielded by

U2585 of the 23S RNA from nucleophilic attack by water,

which would otherwise release the peptidyl group from the

ribosome.

¢¢G

‡

cat

>RT

j. Translocation: The Ribosome Moves to the

Next Codon

In the translocation stage of the elongation cycle, the now

uncharged P-site tRNA (at first but subsequently an

elongator tRNA) is transferred to the E site (not shown in

Fig. 32-46), its former occupant having been previously ex-

pelled (see below). Simultaneously, the peptidyl–tRNA in the

A site, together with its bound mRNA, is moved to the P site.

This prepares the ribosome for the next elongation cycle.

The maintenance of the peptidyl–tRNA’s codon–anticodon

association is no longer necessary for amino acid specifica-

tion. Rather, it acts as a place-keeper that permits the ribo-

some to precisely step off the three nucleotides along the

mRNA required to preserve the reading frame. Indeed, the

observation that frameshift suppressor tRNAs induce a

four-nucleotide translocation (Section 32-2Ea) indicates

that mRNA movement is directly coupled to tRNA move-

ment.An Mg

2

-stabilized kink in the mRNA between the A

and P codons apparently helps prevent slippage.

k. EF-G Structurally Mimics the

EF-Ts  tRNA Complex

The translocation process requires the participation of

elongation factor EF-G (also called EF2), which binds to

the ribosome together with GTP and is only released on

hydrolysis of the GTP to GDP  P

(Fig. 32-46). EF-G re-

lease is a prerequisite for beginning the next elongation cy-

cle because EF-G and EF-Tu bind to the same site of the ri-

bosome and hence their binding is mutually exclusive.

The X-ray structure of T. thermophilus EF-G  GMPPNP,

determined by Anders Liljas and Derek Logan, reveals a

tadpole-shaped monomeric protein that consists of five

tRNA

Met

Section 32-3. Ribosomes and Polypeptide Synthesis 1385

Figure 32-53 The mechanism of ribosome-catalyzed peptidyl

transfer. The nucleophilic attack of the -amino group of the

aminoacyl–tRNA (red) on the carbonyl C of the peptidyl–tRNA

(blue) occurs in concert with a proton shuttle that involves the

O3¿ and 2¿-OH of the P-site A76 together with the -amino

A siteP siteA siteP siteA siteP site

OOH

n+1

tRNA

OOH

tRNA

n+1

tRNA

n+1

tRNA

group of the aminoacyl–tRNA.The reaction proceeds through a

transition state (center; enclosed by square brackets) that contains

a six-membered ring of partially bonded atoms and which

collapses to the reaction products drawn on the right.

JWCL281_c32_1338-1428.qxd 9/7/10 2:23 PM Page 1385

domains (Fig. 32-54).The first two domains closely resemble

those in EF-Tu  GMPPNP rather than those in EF-Tu 

GDP (Fig. 32-48). This, it is argued, is because the two elon-

gation factors have reciprocal functions with EF-Tu  GTP

facilitating the conversion of the ribosome from its post- to

its pre-translocational state and EF-G  GTP promoting the

reverse transition. This idea is supported by the intriguing

observation that the Phe–tRNA

Phe

 EF-Tu  GMPPNP and

EF-G  GMPPNP complexes are almost identical in appear-

ance: EF-G’s three C-terminal domains (yellow through red

in Fig. 32-54), which have no counterparts in EF-Tu, closely

resemble the EF-Tu–bound tRNA in shape, a remarkable

case of macromolecular mimicry. Indeed, EF-G occupies the

same ribosomal site as does EF-Tu (Fig. 32-55).

EF-G is unusual among G-proteins in that it has no cor-

responding GEF. However, its N-terminal guanine nu-

cleotide–binding domain contains a unique  helical insert

(green in Fig. 32-54) that contacts the domain’s conserved

core at sites analogous to those in EF-Tu that interact with

EF-Ts. This suggests that this subdomain acts as an inter-

nal GEF.

1386 Chapter 32. Translation

Figure 32-55 X-ray structure of the T. thermophilus ribosome

in its complexes with (a) EF-G and (b) EF-Tu. The ribosome is

drawn in ribbon form with its 23S RNA orange, its 50S subunit

proteins brown, its 16S RNA cyan, and its 30S proteins blue. Its

bound tRNAs, mRNA, EF-G, and EF-Tu are represented by

their surface diagrams with E-site tRNA yellow, P-site tRNA

green,A-site tRNA magenta, mRNA black, and both EF-G and

EF-Tu red. In Part b, the A-site tRNA bound to EF-Tu is in the

A/T conformation (see below).The positions of the decoding

center (DC) and the peptidyl transferase center (PTC) are

indicated. [Courtesy of Martin Schmeing and Venki

Ramakrishnan, MRC Laboratory of Molecular Biology,

Cambridge, U.K. PDBids 2WRI, 2WRJ, 2WRN, and 2WRO.]

Figure 32-54 X-ray structure of EF-G from T. thermophilus in

complex with GMPPNP. The protein is drawn in ribbon form

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

C-terminus (red).The GMPPNP is drawn in space-filling form

colored according to atom type with C yellow, N blue, O red, and

P orange. An Mg

2

ion that is bound to the GMPPNP is

represented by a magenta sphere. Portions of the structure are

not visible. Note the remarkable resemblance in shape between

this structure and that of Phe–tRNA

Phe

 EF-Tu  GMPPNP (Fig.

32-47). [Based on an X-ray structure by Anders Liljas and Derek

Logan, Lund University, Lund, Sweden. PDBid 2BV3.]

(a)

(b)

JWCL281_c32_1338-1428.qxd 9/7/10 2:23 PM Page 1386

Transpeptidation

Accommodation

= tRNA = amino acid

residue

Site:

50S

30S

Binding state:

Post-translocational

state

E/E P/P

EPA

A/AP/P

EP A

A/AP/P

EPA

A/PP/E

P/P A/T

aa-tRNA

•

EF-Tu

•

GTP

EF-Tu

•

tRNA

EF-G

•

GTP

EF-G

•

GDP+ P

EF-G

EF-Tu

•

GDP + P

Pre-translocational

state

GTP

l. Translocation Occurs via Intermediate States

Chemical footprinting studies (Section 31-2Ab) by

Noller revealed that certain bases in the 16S rRNA are

protected by tRNAs bound in the ribosomal A and P sites

and that certain bases in the 23S rRNA are protected by

tRNAs in the A, P, and E sites.Almost all of these protected

bases are absolutely conserved in evolution and many of

them have been implicated in ribosomal function through

biochemical or genetic studies.

Variations in chemical footprinting patterns during the

elongation cycle together with the more recently determined

X-ray and cryo-EM structures indicate that the translocation

of tRNA occurs in several discrete steps (Fig. 32-56):

1. Let us start with the ribosome in its post-transloca-

tional state: a deacylated tRNA bound to the E subsites of

both the 30S and 50S subunits (the E/E binding state), a

peptidyl–tRNA bound in the P subsites of both subunits

Section 32-3. Ribosomes and Polypeptide Synthesis 1387

Figure 32-56 Ribosomal binding states in the elongation cycle.

Note how this scheme elaborates the classic elongation cycle

diagrammed in Fig. 32-46. The drawings are accompanied by

17-Å-resolution cryo-EM–based images of the E. coli 70S

ribosome in the corresponding binding states in which the 30S

subunit is transparent yellow, the 50S subunit is transparent blue,

and the tRNAs and elongation factors are colored as in the

drawing they accompany. [Cryo-EM images courtesy of Knud

Nierhaus, Max-Planck-Institut für Molekulare Genetik, Berlin,

Germany, and Joachim Frank,Wadsworth Center, State

University of New York at Albany.]

JWCL281_c32_1338-1428.qxd 8/19/10 10:06 PM Page 1387

(the P/P state), and the A site empty. An aminoacyl–tRNA

(aa–tRNA) in ternary complex with EF-Tu and GTP binds

to the A site accompanied by the release of the E-site

tRNA (but see below). This yields a complex in which the

incoming aa–tRNA is bound in the 30S subunit’s A subsite

via a codon–anticodon interaction (recall that the mRNA

is bound to the 30S subunit) but with the EF-Tu preventing

the entry of the tRNA’s aminoacyl end into the 50S

subunit’s A subsite, an arrangement termed the A/T state

(T for EF-Tu).

2. EF-Tu hydrolyzes its bound GTP to GDP  P

and is

released from the ribosome. This permits the aa–tRNA to

fully bind to the A site (the A/A state), a process called ac-

commodation in which the 3¿ end of the tRNA swings

around by nearly 70 Å.

3. The peptidyl transferase reaction occurs, yielding the

pre-translocational state.

4. The acceptor end of the new peptidyl–tRNA shifts

from the A subsite of the 50S subunit to its P-subsite, while

the tRNA’s anticodon end remains associated with the A

subsite of the 30S subunit (yielding the A/P hybrid binding

state). The acceptor end of the newly deacylated tRNA si-

multaneously moves from the P subsite to the E subsite of

the 50S subunit while its anticodon end remains associated

with the P subsite of the 30S subunit (the P/E state).

5. The ribosomal binding of the EF-G  GTP complex

and the subsequent GTP hydrolysis impel the anticodon

ends of these tRNAs, together with their bound mRNA, to

move relative to the small ribosomal subunit such that the

peptidyl–tRNA assumes the P/P state and the deacylated

tRNA assumes the E/E state (the post-translocational

state), thereby completing the elongation cycle.

The binding of tRNA to the A and E sites, as Nierhaus

has shown, exhibits negative allosteric cooperativity. In the

pre-translocational state, the E site binds the newly deacy-

lated tRNA with high affinity (the E site is sterically unable

to bind an aminoacyl–tRNA), whereas the empty A site

has low affinity for aminoacyl–tRNA. However, in the

post-translocational state, the ribosome has undergone a

conformational change that converts the A site to a high-

affinity state and the E site to a low-affinity state, which

consequently releases the deacylated tRNA when

aa–tRNA  EF-Tu  GTP binds to the A site.Thus, the E site

is not simply a passive holding site for spent tRNAs but

performs an essential function in the translation process.

The GTP hydrolysis by the elongation factors EF-Tu and

EF-G as well as the peptidyl transferase reaction appar-

ently function to reduce the activation barriers between

these conformational states.The unidirectional A S P S E

flow of tRNAs through the ribosome is thereby facilitated.

Certain aspects of the foregoing mechanism are not

fully resolved. For example, X-ray studies of the 70S ribo-

some in complex with three tRNAs (e.g., Fig. 32-34) sug-

gest that the E-site tRNA is not released from the ribo-

some until Step 2 of Fig. 32-56. However, Nierhaus and

Frank argue that these complexes were crystallized in the

presence of unphysiologically high tRNA concentrations.

Whatever the case, it is clear that the changes in binding

states result in large-scale tRNA movements, in some in-

stances 50 Å. Moreover, cryo-EM studies indicate that on

binding EF-G  GDP(CH

)P (like GMPPNP but with a

group rather than an NH group bridging its  and 

phosphates), the 30S subunit rotates with respect to the

50S subunit by 6° clockwise when viewed from the 30S sub-

unit’s solvent side, which results in a maximum displace-

ment of ⬃19 Å at the periphery of the ribosome.This rota-

tion is accompanied by many smaller conformational

changes in both subunits, particularly in the regions about

the entrance and exit to the mRNA channel. Clearly, we

are far from fully understanding how the ribosome works

at the molecular level.

m. The Eukaryotic Elongation Cycle Resembles That

of Prokaryotes

The eukaryotic elongation cycle closely resembles that of

prokaryotes. In eukaryotes, the functions of EF-Tu and EF-

Ts are respectively assumed by the eukaryotic elongation

factors eEF1A and eEF1B, with yeast eEF1B consisting of

two subunits: eEF1B, which catalyzes nucleotide ex-

change, and eEF1B, which has unknown function (in

higher eukaryotes, eEF1B contains a third subunit,

eEF1B, that possesses a nucleotide exchange activity sim-

ilar to that of eEF1B). Likewise, eEF2 functions in a man-

ner analogous to EF-G. However, the corresponding eu-

karyotic and prokaryotic elongation factors are not

interchangeable.

The X-ray structure of yeast eEF1A  eEF1B (Fig. 32-57),

determined by Kjeldgaard and Nyborg, reveals that

eEF1A structurally resembles the homologous EF-Tu (Fig.

32-49), whereas eEF1B exhibits no resemblance to EF-

Ts, either in sequence or in structure. Nevertheless,

eEF1B functionally interacts with eEF1A much as EF-Ts

interacts with EF-Tu: Both GEFs associate with their cor-

responding G-protein so as to disrupt the Mg

2

binding site

associated with its bound guanine nucleotide.

E. Translational Accuracy

The genetic code is normally expressed with remarkable fi-

delity. We have already seen that transcription and tRNA

aminoacylation both proceed with high accuracy (Sections

31-2Ec and 32-2Ce).The accuracy of ribosomal mRNA de-

coding was estimated from the rate of misincorporation of

[

S]Cys into highly purified flagellin, an E. coli protein

(Section 35-3I) that normally lacks Cys. These measure-

ments indicated that the mistranslation rate is ⬃10

4

errors

per codon. This rate is greatly increased in the presence of

streptomycin, an antibiotic that increases the rate of ribo-

somal misreading (Section 32-3Ga). From the types of

reading errors that streptomycin is known to induce, it was

deduced that the mistranslation arose almost entirely from

the confusion of the Arg codons CGU and CGC for the

Cys codons UGU and UGC.The above error rate is there-

fore largely caused by mistakes in ribosomal decoding.

An aminoacyl–tRNA is selected by the ribosome only

according to its anticodon. Yet the binding energy loss

1388 Chapter 32. Translation

JWCL281_c32_1338-1428.qxd 8/19/10 10:06 PM Page 1388

arising from a single base mismatch in a codon–anticodon in-

teraction is estimated to be ⬃12 kJ  mol

1

, which, accord-

ing to Eq. [32.1], cannot account for a ribosomal decoding

accuracy of less than ⬃10

2

errors per codon. Moreover,

the base pairing interaction between the UUU codon for

Phe and the GAA anticodon of tRNA

Phe

would be naively

expected to be less stable than the incorrect pairing be-

tween the UGC codon for Ser and the GCG anticodon of

tRNA

Arg

. This is because both interactions have one G  U

base pair and the former correct interaction’s remaining

two A  U base pairs are weaker than the latter incorrect

interaction’s remaining two G  C base pairs. Evidently, the

ribosome has some sort of proofreading mechanism that

increases its overall decoding accuracy.

a. The Ribosome Monitors the Formation of a

Correct Codon–Anticodon Complex

As we have seen (Figs. 32-55b and 32-56), the amino-

acyl–tRNA  EF-Tu  GTP ternary complex initially binds to

the ribosome with the tRNA in the A/T binding state.The

tRNA only assumes the fully bound A/A state (accommo-

dation) after the GTP has been hydrolyzed and the EF-Tu 

GDP complex has been released from the ribosome.These

two states presumably permit the ribosome to double-

check (proofread) the codon–anticodon complex that the

mRNA makes with the incoming tRNA.

The X-ray structure of the T. thermophilus 30S subunit

in complex with a U

hexanucleotide mRNA and a 17-nt

RNA consisting of the tRNA

Phe

anticodon stem–loop

(Fig. 32-11, although its nucleotides are unmodified), de-

termined by Ramakrishnan, revealed how an mRNA-

specified tRNA initially binds to the ribosome. The

codon–anticodon association is stabilized by its interac-

tions with three universally conserved ribosomal bases,

A1492, A1493, and G530 (Fig. 32-58):

1. The first codon–anticodon base pair, that between

mRNA U1 and tRNA A36, is stabilized by the binding of

the rRNA A1493 base in the base pair’s minor groove (Fig.

32-58a).

2. The second codon–anticodon base pair, that between

U2 and A35, is bolstered by A1492 and G530, which both

bind in this base pair’s minor groove (Fig. 32-58b).

3. The third codon–anticodon base pair (the wobble

pair; Section 32-2Db), that between U3 and G34, is rein-

forced through minor groove binding by G530 (Fig. 32-58c).

Section 32-3. Ribosomes and Polypeptide Synthesis 1389

Figure 32-58 Codon–anticodon interactions in the ribosome.

The (a) first, (b) second, and (c) third codon–anticodon base

pairs as seen in the X-ray structure of the T. thermophilus 30S

subunit in complex with U

(a model mRNA) and the 17-nt

anticodon stem–loop of tRNA

Phe

(whose anticodon is GAA).

The structures are drawn in ball-and-stick form embedded in

Figure 32-57 X-ray structure of yeast eEF1A  eEF1B␣.

Domains 1, 2, and 3 of eEF1A are magenta, green, and cyan,

respectively, and eEF1B is orange. The complex is oriented so

as to emphasize the structural resemblance between eEF1A and

the similarly colored EF-Tu in its complex with EF-Ts (Fig.

32-49). Note the lack of resemblance between eEF1B and

EF-Ts. [Based on an X-ray structure by Morten Kjeldgaard and

Jens Nyborg, University of Aarhus, Århus, Denmark. PDBid 1F60.]

their semitransparent van der Waals surfaces. Codons are purple,

anticodons are yellow, and rRNA is brown or gray with non-C

atoms colored according to type (N blue, O red, and P green).

Protein C atoms are gray and Mg

2

ions are represented by

magenta spheres. [Courtesy of Venki Ramakrishnan, MRC

Laboratory of Molecular Biology, Cambridge, U.K. PDBid 1IBM.]

(a) (b) (c)

JWCL281_c32_1338-1428.qxd 8/19/10 10:06 PM Page 1389

This latter interaction appears to be less stringent than

those in the first and second codon–anticodon positions,

which is consistent with the need for the third codon–

anticodon pairing to tolerate non-Watson–Crick base pairs

(Section 32-2D).

Comparison of this structure with that of the 30S subunit

alone reveals that the foregoing rRNA nucleotides un-

dergo conformational changes on the formation of a

codon–anticodon complex (Fig. 32-59). In the absence of

tRNA, the bases of A1492 and A1493 stack in the interior

of an RNA loop but flip out of this loop to form the

codon–anticodon complex, whereas the G530 base

switches from the syn to the anti conformation (Section

29-2Aa).These interactions enable the ribosome to mon-

itor whether an incoming tRNA is cognate to the codon

in the A site; a non-Watson–Crick base pair could not

bind these ribosomal bases in the same way. Indeed, any

mutation of A1492 or A1493 is lethal because pyrim-

idines in these positions could not reach far enough to in-

teract with the codon–anticodon complex or G530 and

because a G in either position would be unable to form

the required hydrogen bonds and its N2 would be sub-

jected to steric collisions. An incorrect codon–anticodon

provides insufficient free energy to bind the tRNA to the

ribosome and it therefore dissociates from it, still in its

ternary complex with EF-Tu and GTP.

GTP Hydrolysis by EF-Tu Is a Thermodynamic

Prerequisite to Ribosomal Proofreading

A proofreading step must be entirely independent of the

initial selection step. Only then can the overall probability

of error be equal to the product of the probabilities of er-

ror of the individual selection steps. We have seen that

DNA polymerases and aminoacyl–tRNA synthetases

maintain the independence of their two selection steps by

carrying them out at separate active sites (Sections 30-2Ac

and 32-2Ce). Yet the ribosome only recognizes the incom-

ing aminoacyl–tRNA according to its anticodon’s comple-

mentarity to the codon in the A site. Consequently, the ri-

bosome must somehow examine this codon–anticodon

interaction in two separate ways.

The formation of a correct codon–anticodon complex

induces EF-Tu to hydrolyze its bound GTP, although how

this occurs is unclear (note that EF-Tu’s GTPase domain is

bound in the 50S subunit which, together with the observa-

tion that GTP hydrolysis requires an intact tRNA, suggests

that the hydrolysis signal is at least in part transmitted

through the tRNA). The resulting conformational change

in EF-Tu (Fig. 32-48) swings its bound tRNA into the A/A

state (accommodation), a process that moves the 3¿ end of

the tRNA by nearly 70 Å. This, it is hypothesized, subjects

the codon–anticodon interaction to a strain that only a cor-

rect pairing can withstand. The codon–anticodon interac-

tion is thereby subjected to a second screening that only

1390 Chapter 32. Translation

Figure 32-59 Ribosomal decoding site. The X-ray structures

of T. thermophilus 30S subunit (a) alone and (b) in its complex

with U

and the 17-nt anticodon stem–loop of tRNA

Phe

.The

RNAs are drawn as ribbons with their nucleotides in paddle

form with tRNA gold, A-site mRNA purple, rRNA gray, and

(a) (b)

nucleotides that undergo conformational changes red. Protein

S12 is tan and Mg

2

ions are represented by red spheres. Compare

Part b with Fig. 32-58. [Courtesy of Venki Ramakrishnan, MRC

Laboratory of Molecular Biology, Cambridge, U.K. PDBids

(a) 1FJF and (b) 1IBM.]

JWCL281_c32_1338-1428.qxd 8/19/10 10:06 PM Page 1390