Báo cáo khoa học: Staphylococcus aureus elongation factor G – structure and analysis of a target for fusidic acid pdf

Staphylococcus aureus elongation factor G – structure and
analysis of a target for fusidic acid
Yang Chen, Ravi Kiran Koripella, Suparna Sanyal and Maria Selmer
Department of Cell and Molecular Biology, Uppsala University, Sweden
Introduction
Protein synthesis, translation of mRNA into protein, is
performed on the ribosome. To synthesize a protein,
the ribosome goes through the phases of initiation,
elongation, termination, and recycling, each phase
being assisted by a number of protein translation fac-
tors [1,2]. Some of these factors, in prokaryotes initia-
tion factor 2, elongation factor Tu (EF-Tu), elongation
factor G (EF-G), and release factor 3, are GTPases,
which drive the process forwards using GTP as the
energy source. Among these, only EF-G participates
in two distinct steps of the translation cycle: elonga-
tion and ribosome recycling. During elongation, after
formation of each new peptide bond, EF-G binds to
the ribosome and, under GTP hydrolysis, catalyses
translocation, the concerted movement of mRNA,
together with A-site and P-site tRNAs, to expose a
new A-site codon [3,4]. Recycling takes place when the
translating ribosome has reached a stop codon and
released the nascent peptide. At this point, EF-G and
ribosome recycling factor bind to the post-termination
complex to catalyse the disassembly of the complex [5–
7]. EF-G has a low intrinsic activity in GTP hydrolysis
that is stimulated by the interaction with the ribosome
[8,9]. The currently prevalent model states that EF-G
Keywords
antibiotic resistance; crystallography;

elongation factor G (EF-G); fusidic acid;
switch region
Correspondence
M. Selmer, Department of Cell and
Molecular Biology, Uppsala University,
BMC, Box 596, 751 24 Uppsala, Sweden
Fax: +46 18 536971
Tel: +46 18 4714177
E-mail:
Database
The atomic coordinates and observed
structure factors are available in the Protein
Data Bank database under the accession
number 2XEX
(Received 22 April 2010, revised 28 June
2010, accepted 14 July 2010)
doi:10.1111/j.1742-4658.2010.07780.x
Fusidic acid (FA) is a bacteriostatic antibiotic that locks elongation factor G
(EF-G) on the ribosome in a post-translocational state. It is used clinically
against Gram-positive bacteria such as pathogenic strains of Staphylo-
coccus aureus, but no structural information has been available for EF-G
from these species. We have solved the apo crystal structure of EF-G from
S. aureus to 1.9 A
˚
resolution. This structure shows a dramatically different
overall conformation from previous structures of EF-G, although the indi-
vidual domains are highly similar. Between the different structures of free
or ribosome-bound EF-G, domains III–V move relative to domains I–II,
resulting in a displacement of the tip of domain IV relative to domain G.
In S. aureus EF-G, this displacement is about 25 A

˚
relative to structures of
Thermus thermophilus EF-G in a direction perpendicular to that in previous
observations. Part of the switch I region (residues 46–56) is ordered in
a helix, and has a distinct conformation as compared with structures of
EF-Tu in the GDP and GTP states. Also, the switch II region shows a new
conformation, which, as in other structures of free EF-G, is incompatible
with FA binding. We have analysed and discussed all known fusA-based
fusidic acid resistance mutations in the light of the new structure of EF-G
from S. aureus, and a recent structure of T. thermophilus EF-G in complex
with the 70S ribosome with fusidic acid [Gao YG et al. (2009) Science 326,
694–699]. The mutations can be classified as affecting FA binding, EF-G–
ribosome interactions, EF-G conformation, and EF-G stability.
Abbreviations
EF-G, elongation factor G; EF-Tu, elongation factor Tu; EM, electron microscopy; FA, fusidic acid; PDB, Protein Data Bank.
FEBS Journal 277 (2010) 3789–3803 ª 2010 The Authors Journal compilation ª 2010 FEBS 3789
binds to the ribosome in GTP form, hydrolyses GTP,
releases inorganic phosphate and, through a conforma-
tional change, drives tRNA translocation [10] or ribo-
some recycling [7]. However, there are recent
indications that EF-G may act differently in transloca-
tion and ribosome recycling [11].
The crystal structure of EF-G from Thermus thermo-
philus was first solved in 1994 in complex with GDP
[12] as well as in apo form [13]. Since then, structures of
several mutants of EF-G from the same bacterium have
been solved [14–16]. EF-G forms an extended structure
consisting of five domains (Fig. 1A). The domain G
(domain I) and domain II form a globular structure
that is conserved in all other ribosomal GTPases. The

exception is the additional subdomain G¢, which
is inserted in domain G, and exists only in release factor
3 and EF-G. As in other GTPases, domain G
contains a conserved P-loop, which coordinates the
a-phosphate and b-phosphate, and two so-called switch
regions, which coordinate the c-phosphate and change
conformation between a tense GTP state and a relaxed
GDP state [17].
Ribosomal complexes that have been stalled by the
locking of EF-G to the ribosome with either a nonhy-
drolysable GTP analogue [18] or the antibiotic fusidic
acid (FA) and GDP [19–21] have been visualized with
cryo-electron microscopy (EM) and single-particle
reconstructions. Recently, a 3.6 A
˚
crystal structure of
EF-G bound to the T. thermophilus ribosome showed
the FA-binding site for the first time, and revealed the
detailed interactions of EF-G with the ribosome [22]
(summarized in Fig. 1B). As compared with this FA-
stabilized, ribosome-bound conformation, crystal struc-
tures of T. thermophilus EF-G display different confor-
mations, where domains III–V have rotated relative to
domains I–II, resulting in the position of the tip of
domain IV differing by 20 A
˚
(Fig. 2; discussed further
below). It appears that ribosome binding is the main
trigger of the conformational change in EF-G, as in
solution it can accommodate GDP or GTP without

forcing any major changes in its global conformation
[15,23,24].
FA is a clinically used steroid antibiotic that locks
EF-G on the ribosome after GTP hydrolysis and trans-
location [25]. FA binds to a pocket between domains
I, II and III of EF-G, and seems to lock EF-G in a
conformation intermediate between the GTP-bound
and GDP-bound forms [22]. Staphylococcus aureus is
one of the major clinical targets for FA treatment.
However, very few studies have been performed using
EF-G from this species. In this study, we have solved
the apo crystal structure of S. aureus EF-G to 1.9 A
˚
resolution, allowing us to examine the generality of
conclusions drawn from the T. thermophilus EF-G
structures and to pinpoint the role of amino acids
that are mutated in isolated FA-resistant strains of
S. aureus [26,27].
Results and Discussion
Structure solution
S. aureus EF-G was crystallized in a mixture of poly-
ethylene glycol 3350 and NaCl in Tris ⁄ HCl buffer at
pH 8.7. The crystals grew in space group P2
1
and
diffracted to 1.9 A
˚
resolution (Table 1). There are two
molecules in the asymmetric unit, forming a noncrys-
tallographic two-fold symmetry. b-Sheets from domain

V of molecules A and B form an extended b-sheet,
and helix A
V
packs in an antiparallel fashion to the
equivalent helix in molecule B. Residues 2–38, 64–441
and 445–692 in molecule A and residues 2–41, 46–56,
65–441 and 447–692 in molecule B were ordered and
could be built into the electron density maps. The
ordered part includes domain III, which is disordered
in structures of wild-type T. thermophilus EF-G
[12,13]. In molecule B, part of the switch I region
could be built. The entire switch I region is disordered
in all previous EF-G structures from T. thermophilus
[12–16], including the EF-G–70S complex structure
with GDP and FA [22], and has been suggested to be
ordered only in the ribosome-bound GTP state, as
Table 1. Summary of crystallographic data and refinement.
Data collection statistics
Resolution (A
˚
)
a
46.2–1.9 (2.0–1.9)
R
meas
[48] (%) 6.1 (55.5)
I ⁄ r
I
16.5 (3.1)
Completeness (%) 92.7 (78.5)

Redundancy 3.71 (3.56)
Refinement statistics
Resolution (A
˚
) 46.2–1.9
Number of unique reflections ⁄ test set 110 987 ⁄ 4777
R
work
⁄ R
free
(%) 18.7 ⁄ 22.4
Molecules per asymmetrical unit 2
Number of atoms
Protein 10 352
Water 509
Ions 4
Average B-factor (A
˚
2
) 22.4
Rmsd from ideality
Bond lengths (A
˚
) 0.020
Bond angles (°) 1.53
Ramachandran statistics
Residues in most favoured regions (%) 96.75
Residues in additional allowed regions (%) 2.87
Residues in disallowed regions (%) 0.38
a

Values in parentheses represent the highest-resolution bin.
Crystal structure of Staphylococcus aureus EF-G Y. Chen et al.
3790 FEBS Journal 277 (2010) 3789–3803 ª 2010 The Authors Journal compilation ª 2010 FEBS
indicated by recent proteolytic cleavage experiments
[28]. It is also disordered in structures of the eukary-
otic equivalent, eEF2 [29]. In the present structure, res-
idues 46–56 form a helix that was only clearly visible
in difference Fourier maps after refinement of the rest
of the structure. The density for residues 42–45 was
too weak to allow interpretation, but this region would
need to be in an extended conformation to bridge
16.5 A
˚
between the a carbons of residues 41 and 46.
In molecule A, the density for the entire switch I
region is too weak for interpretation, indicating higher
flexibility.
Attempts to soak S. aureus EF-G crystals with GDP
as well as nonhydrolysable GTP analogues resulted in
partial occupancy of GDP in the nucleotide-binding
site. Therefore, we present here the apo structure of
S. aureus EF-G.
Overall structure and comparison with previous
EF-G structures
All five domains of S. aureus EF-G are ordered in our
structure. The overall conformation of the two EF-G
molecules in the asymmetric unit is very similar (rmsd
of 0.58 A
˚
for 660 C

a
atoms), with only a slight differ-
ence in the orientation of domains III and IV. Thus,
when the two molecules are superimposed on the basis
of domains I and II, the maximum difference at the
edge of domain III is 2.3 A
˚
.
Comparison of S. aureus EF-G with the previously
solved T. thermophilus EF-G structures shows that the
individual domains are highly similar. However,
domains III, IV and V are in a different orienta-
tion relative to domains I and II in comparison to
previous EF-G structures (Fig. 1C). Between all the
A
B
C
Fig. 1. EF-G structure. (A) Overall structure
and structural domains of S. aureus EF-G
(PDB 2xex). The switch regions are shown
in black, with switch II facing domains II
and III, and switch I behind the G-domain.
(B). Crystal structure of EF-G bound to the
T. thermophilus 70S ribosome with GDP
and FA (PDB 2wri [22]). FA (left) and GDP
(right) are shown in black. Numbers indicate
ribosomal contact areas: 1, decoding centre;
2, 23S RNA 2475 loop; 3, 23S RNA
1067 ⁄ 1095 loops; 4, ribosomal protein L6;
5, C-terminal domain of ribosomal protein

L12; 6, 23S RNA 2660 loop (from the back);
7, ribosomal protein S12. Thickness of lines
indicates closeness to the viewer.
(C) Comparison of apo-EF-G from S. aureus
(PDB 2xex, magenta) and T. thermophilus
(PDB 1elo [13], grey). Superposition is
based on domains I and II.
Y. Chen et al. Crystal structure of Staphylococcus aureus EF-G
FEBS Journal 277 (2010) 3789–3803 ª 2010 The Authors Journal compilation ª 2010 FEBS 3791
T. thermophilus EF-G structures, wild type and various
mutants, domains III, IV and V display a movement
relative to domains I and II, resulting in a shift of the
tip of domain IV of up to 8 A
˚
[14,16]. The ribosome-
bound structure of EF-G as visualized by crystallogra-
phy [22] shows a larger movement in the same direc-
tion, measuring about 27 A
˚
. The equivalent
comparison with S. aureus EF-G shows a shift of
domains III, IV and V by approximately 25 A
˚
in the
perpendicular direction (Fig. 2). The hinge region for
this conformational change consists of residues 400–
405, as previously observed in T. thermophilus EF-G
[16]. Conformational changes of EF-G in more than
one direction were suggested in early cryo-EM studies
of ribosome–EF-G complexes [20], but because of the

low resolution (17–20 A
˚
), the structural interpretation
is not very reliable.
The P-loop (residues 12–27) has the same conforma-
tion as in the apo structure of T. thermophilus EF-G
[13], and upon crystal soaking with GDP, partial occu-
pancy of the peptide-flipped structure is observed, in
agreement with structures of T. thermophilus EF-G
[30] (data not shown).
The switch I region consists of residues 39–63. The
ordered part is a helix from residues 46 to 56 that
packs against helix A
G
so that Trp52 makes hydropho-
bic interactions with Leu31, Tyr32 and Ile37, and
Met53 interacts with Glu28 (Fig. 3A). With the
exception of Ile37, all of these residues are conserved in
EF-G from different species. In contrast to the
situation in EF-G, the switch I region is fully ordered
in structures of EF-Tu with GDP [31] and a GTP
analogue [32], as well as in the structure of the EF-G
homologue EF-G-2 with GTP [18]. In EF-Tu, the
switch I region forms a short helix followed by a b-hair-
pin reaching away from the nucleotide-binding site in
the GDP state, whereas in the GTP state, it forms two
short helices just before the conserved Thr that coordi-
nates a magnesium ion and the c phosphate. The helix
that we observe is longer than in any of these structures,
and has a different orientation (Fig. 3B). It is too far

away from the nucleotide-binding site to allow the inter-
action of the conserved Thr62 with magnesium and c
phosphate that should occur in the GTP state. Superpo-
sition of the current structure with the ribosome-bound
EF-G [22] shows that the observed switch I conforma-
tion would be compatible with ribosome binding and
located in the intersubunit space at a distance of 10 A
˚
from residue 2655 of 23S RNA, 15 A
˚
from ribosomal
protein L14, and 10 A
˚
from residue 342 of 16S RNA.
The switch II region of S. aureus EF-G has electron
density for all residues, including side chains, except
for Gly84 (Fig. 3C). It has a different conformation
compared to the T. thermophilus EF-G structures
(Fig. 3D). The hydrogen bond Asp87–Arg659 stabi-
lizes the switch II region and the current domain
38 Å
ABC
Fig. 2. Conformational space of EF-G on and off the ribosome. (A) Superposition of S. aureus EF-G with the ribosome-bound T. thermophilus
EF-G (PDB 2wri [22], dark blue) and the T. thermophilus apo-EF-G structure (PDB 1elo [13], yellow), based on domains I and II. The arrow
indicates the direction of projection to the circle in (C). (B) As (A), view 90° away. (C) Comparison of the positions of the tip of domain IV in
all available EF-G structures in the PDB. The structures were superimposed on the basis of domains I and II. Looking from the direction of
the arrow in (A) and (B), the coordinates of His572 at the tip of domain IV are roughly in one plane, and were manually covered with a col-
oured dot. 1, E. coli EF-G + GMPPNP + 70S (PDB 2om7 [18], cryo-EM); 2, T. thermophilus EF-G + 70S + FA + GDP (PDB 2wri [22]); 3,
E. coli EF-G + GDP + 70S + FA (PDB 1jqm [21], cryo-EM); 4, T. thermophilus EF-G T84A + GMPPNP (PDB 2bv3 [15]); 5, T. thermophilus
EF-G + GDP (dimer, PDB 1ktv); 6, T. thermophilus EF-G apo (PDb 1elo [13]); 7, T. thermophilus EF-G T84A + GDP, FA-resistant (PDB 2bm0

[14]); 8, T. thermophilus EF-G G16V + GDP, FA-hypersensitive (PDB 2bm1 [14]); 9, T. thermophilus EF-G H573A + GDP (PDB 1fnm [16]); 10,
S. aureus EF-G (PDB 2xex).
Crystal structure of Staphylococcus aureus EF-G Y. Chen et al.
3792 FEBS Journal 277 (2010) 3789–3803 ª 2010 The Authors Journal compilation ª 2010 FEBS
A
C
B
D
E
Fig. 3. The switch regions of EF-G. (A) Switch I region of S. aureus EF-G. The ordered switch I helix packs against helix A
G
. The 2F
o
)F
c
map is contoured at 1r. (B) Comparison of switch I in EF-G and EF-Tu. The structures were superimposed on the basis of the equivalent
parts of domains G and II. The switch I region of S. aureus apo-EF-G (PDB 2xex, magenta), EF-Tu in complex with GDP (PDB 1tui, yellow
[31]) and EF-Tu in complex with GDPNP (PDB 1eft, green, Mg
2+
, green sphere and GDPNP, in cpk [32]) are shown together with the EF-G
structure (grey). The first, shorter helix shown in green is identical in the GDP and GTP forms of EF-Tu. (C) F
o
)F
c
omit map of the switch II
region of S. aureus EF-G contoured at 3r. Omitted residues are shown in yellow stick representation. (D) Comparison of the switch II region
from different EF-G crystal structures. Superposition based on helix B
G
(Val90-Asp100) of S. aureus EF-G (PDB 2xex, magenta); T. thermo-
philus EF-G wild type (PDB 1elo, yellow [13]); T. thermophilus EF-G H573A (PDB 1fnm, orange [16]); T. thermophilus EF-G T84A (PDB

2bm0, light blue [14]); T. thermophilus EF-G G16V (PDB 2bm1, red [14]); T. thermophilus EF-G T84A with GDPNP (PDB 2bv3, green [15]);
T. thermophilus EF-G–GDP–FA complex with the ribosome (PDB 2wri, dark blue [22]). Residues 20–200 are shown, but only the switch II
region and the side chain of Phe88 are coloured. (E) Domain III and the FA-binding site. Switch II regions of S. aureus EF-G (magenta) and
FA-bound EF-G at the ribosome (PDB 2wri [22], switch II in dark blue, FA in yellow ⁄ red) superposed on the basis of domain III (grey,
residues 407–474), showing how the FA-binding site in the S. aureus EF-G structure is blocked by the switch II region.
Y. Chen et al. Crystal structure of Staphylococcus aureus EF-G
FEBS Journal 277 (2010) 3789–3803 ª 2010 The Authors Journal compilation ª 2010 FEBS 3793
arrangement, with a larger contact area between
domains II and V than that in any other structure of
EF-G. The switch II region and the conserved Phe88
(Phe90 in T. thermophilus) displays many different ori-
entations in the available crystal structures, and none
of the isolated EF-G structures displays a switch II
conformation identical to that of EF-G in complex
with the 70S ribosome and FA [22] (Fig. 3D). Whereas
Phe88 in the ribosome-bound structure is exposed at
the surface of EF-G and forms part of the FA-binding
site, it points to the opposite direction in our structure,
interacting with Glu93 in helix B
G
and Tyr126 in helix
C
G
in the core of the domain, and blocking the FA-
binding site (Fig. 3E). Despite its many different con-
formations, the switch II region also blocks the FA-
binding site in all structures of free T. thermophilus
EF-G (not shown). However, we do not believe that
the alternative switch II conformations observed in
structures of wild-type and mutant T. thermophilus

EF-G are responsible for FA sensitivity and resistance,
respectively [14]. Rather, the switch II region only
adopts its FA-stabilized conformation when bound to
the ribosome in the presence of the drug, and several
FA resistance mutations in the switch II region influ-
ence direct contacts with FA (discussed further below).
Conformational space of EF-G
The present structure of S. aureus EF-G shows that
EF-G, when not bound to the ribosome, can acquire
AB
CD
Fig. 4. FA resistance mutations. (A) All known FA resistance mutation sites (Table 2) mapped on the S. aureus EF-G structure. The mutation
sites are displayed as side chains and located in domain III, domain V and the interface of domains G, III and V. (B) Mutation sites in domain
III that may affect the FA-binding pocket. Mutation sites are shown with yellow carbons, and influence the packing between helix A
III
and
helix B
III
. To the left is the interface with domains I, II and V; to the right is the connection to domain IV. (C) Mutation sites in helices A
V
and
B
V
at the surface of domain V. Gly621 and Gly617 are in the area of contact with the 1095 and 2473 regions of 23S RNA. The two helices
are facing the ribosome, and the four-stranded b-sheet is facing the other domains of EF-G. (D) Linker region between domains I–II and
domains III–V. The four sites of FA resistance mutations in this region are shown with side chains.
Crystal structure of Staphylococcus aureus EF-G Y. Chen et al.
3794 FEBS Journal 277 (2010) 3789–3803 ª 2010 The Authors Journal compilation ª 2010 FEBS
a conformation that is distinctly different from what
has previously been observed for T. thermophilus

EF-G. The fact that the two molecules in the asym-
metric unit show identical conformations, despite
being involved in different crystal contacts, suggests
that this is not a crystallographic artefact.
In the many different crystal structures of T. thermo-
philus EF-G, only smaller conformational differences
have been observed [12–16]. An explanation for this
could be that the interdomain movements of T. ther-
mophilus EF-G are limited by crystal contacts between
domains G¢ and IV, which form an extended b-sheet
between the two domains [15,33]. By crystallizing
EF-G from another species, S. aureus, we have
obtained a new crystal packing arrangement that
avoids this problem.
Furthermore, the only existing data on the confor-
mation of EF-G in solution come from small-angle
scattering measurements on T. thermophilus EF-G
[23]. These measurements resulted in radii of gyra-
tion in the range 30.2–32.9 A
˚
for EF-G bound to
different nucleotides, and the difference between the
different states was judged to be nonsignificant as
compared with the experimental errors. The calcu-
lated radius of gyration from the S. aureus EF-G
crystal structure is 30.6 A
˚
, whereas the corresponding
value for T. thermophilus EF-G [Protein Data Bank
(PDB) 1fnm] [16] is 30.5 A

˚
, agreeing equally well
with those measurements. In conclusion, EF-G may
display larger interdomain flexibility in solution than
previously thought. Our new conformation is signifi-
cant, as it demonstrates the size of the conforma-
tional space of EF-G when not bound to the
ribosome.
The active conformation of EF-G is the one that
occurs on the ribosome. So far, only post-transloca-
tional states, where domain IV of EF-G has entered
the A-site, have been visualized on the ribosome
[18–22]. The ribosome-bound EF-G conformations in
the presence of GMPPNP or GDP and FA differ by
approximately 6 A
˚
in position of the tip of domain
IV when the G-domains are superimposed (Fig. 2C,
points 1 and 2). However, there is, at present, no
structural information regarding the initial binding of
EF-G to a presumably ratcheted ribosome where
the 30S A-site is still occupied by the peptidyl
tRNA. Most likely, ribosome binding induces a
somewhat stable but transient conformation of EF-G
that is compatible with a tRNA in the 30S A-site,
and we can only speculate that this conformation of
EF-G may be more similar to either of the confor-
mations observed in the crystal structures of free
EF-G.
FA resistance mutations

FA binds to EF-G on the ribosome and prevents its
dissociation after GTP hydrolysis and translocation. In
the recent crystal structure of a 70S–EF-G complex
[22] (Fig. 1B), it is shown that FA allows the switch I
region to change from a GTP to a GDP conformation,
whereas the switch II region is prevented from adopt-
ing its GDP conformation. This, in turn, stops the glo-
bal conformational change of EF-G to the GDP state
that would leave the ribosome. In other words, FA
locks EF-G in a conformation between its GTP and
GDP forms that cannot dissociate from the ribosome.
FA resistance mutations belong to three classes:
fusA mutants, with mutations in the EF-G gene; fusB,
fusC and fusD mutants, which express a resistance pro-
tein that somehow protects the cell from FA inhibi-
tion; and fusE mutants, with mutations in ribosomal
protein L6 [27].
There are, in total, 42 positions in EF-G where
point mutations of the fusA class have been reported
to cause FA resistance [27,34,35] (Fig. 4A). The previ-
ous analysis of these [16] was performed without accu-
rate knowledge of the FA-binding site [22], and, in
addition, new mutations have recently been identified
[27]. On the basis of analysis of the S. aureus EF-G
structure together with the recent T. thermophilus EF-
G–70S complex structure with FA [22], we can now
classify the mutations into four groups, A–D (Table 2).
These perturb four critical parameters for locking EF-
G to the ribosome: drug binding (A), ribosome–EF-G
interactions (B), EF-G conformation (C), and EF-G

stability (D). Several mutations seem to affect more
than one of these parameters; for example, EF-G con-
formation and stability are intimately linked to FA
binding as well as ribosome binding.
Group A mutations involve residues in direct con-
tact with FA as well as residues that shape the drug-
binding pocket. These resistance mutations will directly
alter drug–EF-G interactions, probably lowering the
affinity of FA for the ribosome-bound EF-G. The
switch II loop directly contributes to the FA-binding
site, where both Thr82 and Phe88 are in direct contact
with FA in the ribosome complex structure [22]. Muta-
tion of the corresponding residues in T. thermophilus
(Thr84 and Phe90) also leads to resistance [36].
One edge of the FA-binding pocket is formed by
domain III [22]. A cluster of mutation sites is located
in this area, where the C-terminal end of helix A
III
packs against the central part of helix B
III
(Fig. 4B).
Asp434 and Thr436 in helix A
III
both form hydrogen
bonds with His457 in helix B
III
. Thus, the mutations
P435Q, T436I, H457Y and P435Q will change this
Y. Chen et al. Crystal structure of Staphylococcus aureus EF-G
FEBS Journal 277 (2010) 3789–3803 ª 2010 The Authors Journal compilation ª 2010 FEBS 3795

Table 2. Structural interpretations of all known FA resistance mutations. Residues are numbered according to the S. aureus EF-G sequence. Helices are alphabetically ordered within each
domain, and labeled with the domain subscript. b-Strands are numbered within each domain, and labeled with the domain subscript. A, mutations influence the FA-binding site; B, muta-
tions influence EF-G–ribosome interactions; C, mutations influence interdomain interactions in EF-G; D, mutations influence the stability of EF-G.
Residue Species Mutation Group Location Interpretation of mutation
Ala66 E. coli [34,49] Val C Domain G, at the C-terminal end of the switch I
region, at the interface to domain II
Mutation can push apart domains I and II or possibly
affect the switch I conformation
Thr82 Salmonella typhimurium [34]
T. thermophilus [36]
Ala A Domain G, in the switch II region. In the ribosome
complex structure, at the interface of domain G and
domain II, and contacts FA [22]
Mutation affects the FA-binding site [22]
Phe88 S. aureus [35]
T. thermophilus [36]
Leu A Domain I, in the switch II region. In the ribosome
complex structure, in contact with FA [22]
Mutation affects the FA-binding site [22]
Val90
a
S. aureus [35] Ile C Domain G, in the switch II region. The equivalent
position in T. thermophilus, S. typhimurium and
E. coli is Ile. In T. thermophilus complex [22],
packing against domain III
Mutation may affect domain I–III interactions
Ala102 S. typhimurium [34] Glu D Domain G, in b-strand 5
G
. Packing against Met16 Mutation will induce steric hindrance, and may
disturb the structure of domain G

Leu106
a
S. typhimurium [34] Ser D Domain G, in b-strand 5
G
. Involved in hydrophobic
interactions with Val112, Val132 and Leu151. Phe
in T. thermophilus, Tyr in
Mutation can disturb the hydrophobic core of
domain G
Pro114 S. aureus [27] His B, C Domain G, in the turn before helix C
G
, at the
interface with domain V. Packing against Gly664.
In the ribosome complex [22], close to 2660 of
23S RNA
Mutation would change conformational properties,
and may influence the domain G–V interaction as
well as ribosome interactions
Gln115 S. aureus [27,35]
E. coli [49]
Leu B Domain G, in helix C
G
, hydrogen bonding to P-loop
His18. In the ribosome complex [22], hydrogen
bonds to His85 and Thr118, close to 2660 of 23S
RNA
Mutation will probably affect the ribosome-binding
surface
Thr118 S. typhimurium [34] Ile B, C Domain G, in helix C
G

, at the interface with domain
V. In the ribosome complex [22], hydrogen bonds
to His85 in switch II, close to 2660 of 23S RNA
Mutation may influence the domain G–V interaction
as well as ribosome interactions
Val119 S. typhimurium [34] Leu D Domain G, in helix C
G
, packing against Met16 in the
P-loop. In the ribosome complex [22], contacting
Val104
Mutation will make a slight change to the core of
domain G
Gln122 S. typhimurium [34] His C Domain G, hydrogen bonding to Thr667 in domain V;
close to Phe88. This interaction is not present in
the ribosome complex [22]
Mutation may affect interactions with domain V
Val132
a
S. typhimurium [34] Thr D In the hydrophobic core of domain G, close to
Leu106, Leu151 and Val256. Ala in T. thermophilus
and S. typhimurium
Mutation may disturb the hydrophobic core of
domain G
Leu155 S. typhimurium [34] Pro C, D At the C-terminal end of helix D
G
, hydrophobic
interaction swith Trp120 in helix C
G
[16]. Close to
the interface with domain V. This interaction is not

present in the ribosome complex [22]
Mutation will break the helix and may affect
interactions with domain V
Crystal structure of Staphylococcus aureus EF-G Y. Chen et al.
3796 FEBS Journal 277 (2010) 3789–3803 ª 2010 The Authors Journal compilation ª 2010 FEBS
Table 2. (Continued).
Residue Species Mutation Group Location Interpretation of mutation
Thr385 S. aureus [27] Asn C Domain II, packing against the C-terminal end of
helix A
III
in domain III. This interaction is not
present in the ribosome complex [22]
Mutation could influence the domain II–III interaction
Pro404 S. aureus [27,35] Leu
Arg
Gln
C In the linker region between domains II and III, the
main hinge region for conformational change of EF-G
Mutation will influence the linker conformation,
which could affect the relative orientation between
domains I, II and III and thereby the FA-binding site
Pro406 S. aureus [27,35]
S. typhimurium [34]
Leu C In the linker region between domains II and III, the
main hinge region for conformational change of EF-G
Mutation will influence the linker conformation,
which could affect the relative orientation between
domains I, II and III and thereby the FA-binding site
Val407 S. aureus [35] Phe C Domain III, at the interface with domain V. Packing
against the linker to domain IV. In the ribosome-

bound state [22], this linker has flipped away to
create the 40 A
˚
shift of domain IV relative to
domain III
Mutation to a larger side chain may change the
relative positions of domain III and V [16] as well as
the position of domain IV in the free state
Ala426 S. typhimurium [34] Asp B, D Domain III, in the middle of helix A
III
in the
hydrophobic core. In the ribosome complex [22],
this helix binds to 16S RNA and S12
Mutation to a large, nonhydrophobic residue leads to
a steric clash, lowers the stability of domain III, and
may affect nearby ribosome interactions
Leu430 S. typhimurium [34] Gln B, D Domain III, in the middle of helix A
III
in the
hydrophobic core. In the ribosome complex [22],
this helix binds to 16S RNA and S12
Mutation to a nonhydrophobic residue lowers the
stability of domain III and may affect nearby
ribosome interactions
Asp434 S. aureus [27,35] Asn A Domain III, at the C-terminal end of helix A
III
.
Hydrogen bonding to His457 in helix B
III
. In the

ribosome complex, in contact with FA [22]
Mutation would affect the surface in the FA-binding
pocket
Pro435 S. typhimurium [34] Gln A Domain III, in a turn after helix A
III
. In the ribosome
complex [22], the previous residue is in contact
with FA
Mutation will change the turn conformation, affecting
the interactions of Asp434 and Thr436 and the FA-
binding site
Thr436 S. aureus [27,35] Ile A Domain III, in the turn after helix A
III
. Hydrogen
bonding to His457 in helix B
III
. In the ribosome
complex, lining the FA-binding site [22]
Mutation would affect the surface in the FA-binding
pocket
His438
a
S. aureus [27] Asn C Domain III, in the turn after helix A
III
. Packing against
Pro406 in the linker between domains II and III. Arg
in T. thermophilus
Mutation may influence the linker conformation and
relative orientation between domains I, II and III
Gln447 S. typhimurium [34] His C, D Domain III, at the C-terminus of the loop in the b-

sheet; hydrogen bonding with Ser411 in next
strand. In the ribosome complex structure [22], in
the area of contact with the linker to domain IV
Mutation may affect interdomain interactions
Gly451
a
S. aureus [35] Val C, D Domain III, packing against Pro406 in the linker to
domain IV. Ser in T. thermophilus
Side chain may influence domain III stability and ⁄ or
interdomain interactions
Gly452 S. aureus [27,35] Ser
Cys
Val
A, C, D Domain III, close to helix B
III
and to the linker to
domain IV
Side chain would clash with His457 in helix B
III
, and
may affect the FA-binding pocket. May push
His457 to the interface of domain G and III [16]
Y. Chen et al. Crystal structure of Staphylococcus aureus EF-G
FEBS Journal 277 (2010) 3789–3803 ª 2010 The Authors Journal compilation ª 2010 FEBS 3797
Table 2. (Continued).
Residue Species Mutation Group Location Interpretation of mutation
Met453 S. typhimurium [34] Ile A Domain III, in the turn before helix B
III
at the
interface with domains G and V. In the ribosome

complex structure, in hydrophobic interactions with
Phe88 [22]
Mutation probably affects the FA-binding pocket
Leu456 S. aureus [27,35] Phe A, B Domain III, at the interface with domains G and V. In
the ribosome complex [22], in contact with A2662
of 23S RNA and lining the FA-binding pocket
Mutation may affect FA binding and ⁄ or ribosome
interactions
His457 S. aureus [27,35] Tyr A Domain III, in helix B
III
; forming hydrogen bonds to
Thr436 and Asp434; stabilizing domain III. In the
ribosome complex [22], lining the FA-binding site
Mutation would affect the surface in the FA-binding
pocket
Leu461
a
S. aureus [35] Ser A, D Domain III, in helix B
III
; involved in hydrophobic
interactions with Leu430, Phe437 and Ile450. Ile in
T. thermophilus. In the ribosome complex [22], the
following residue packs against FA
Mutation may change the position of helix B
III
and
affect FA binding
Arg464 S. aureus [27,35]
S. typhimurium [34]
Ser

His
Leu
Cys
A Domain III, in helix B
III
; making hydrogen bonds with
Glu433 in helix A
III
. In the ribosome complex [22],
contributing to the FA-binding pocket
Mutation probably affects the FA-binding pocket
Gly617 S. aureus [27] Asp B Domain V, helix A
V
, contacts second molecule. In
the ribosome complex [22], packing against A1095
of 23S RNA. The next residue interacts with L6
Mutation may disturb ribosome interactions
Gly621 S. typhimurium [34] Cys B Domain V, helix A
V
, contact with second molecule.
In the ribosome complex [22], packing against
U2473 of 23S RNA
Mutation may disturb ribosome interactions
Gly628 S. aureus [27] Val B, D Domain V, at the N-terminus of strand 2
V
, pointing
towards the neighbouring b-strand 3
V
. In the
ribosome complex, close to interaction with U2473

of 23S RNA [22]
Mutation to introduce a side chain would induce a
steric clash and conformational change, and affect
the nearby ribosome contact
Gly632 S. typhimurium [34] Asp B Domain V, in the b-strand that packs with a second
molecule. In the ribosome-bound structure, the
previous residue interacts with 1067 of 23S RNA
[22]
Mutation will change the domain V surface, probably
affecting ribosome interactions
Pro647 S. typhimurium [34] Gln C Domain V, at the interface with domain IV. The
neighbouring residue 648 packs against domain III.
This interaction between domains III, IV and V is
conserved in the ribosome-bound structure [22]
Mutation will change the backbone conformation and
create a steric clash, affecting the interdomain
interactions
Ala655 S. aureus [27] Glu C Domain V, in helix B
V
at the interface with domain G.
In the ribosome complex, packing against domain
III [22]
Mutation to a larger side chain would disrupt the
interaction and may influence the domain
arrangement or shift the position of helix B
V
,
thereby influencing ribosome interactions
Crystal structure of Staphylococcus aureus EF-G Y. Chen et al.
3798 FEBS Journal 277 (2010) 3789–3803 ª 2010 The Authors Journal compilation ª 2010 FEBS

interaction and affect the FA-binding site, whereas the
D434N mutation probably only exposes a new amino
group to the drug. Similarly, Glu433 in helix A
III
hydrogen bonds with Arg464 in helix B
III
, and the
mutations R464C, R464H, R464L and R464S will
cause changes to the packing between the two helices,
affecting the FA-binding pocket. The interactions
between helix A
III
and helix B
III
are conserved between
the S. aureus EF-G structure and the ribosome-bound
EF-G structure [22]. The same interactions of His457
are found in the T. thermophilus structure, but not
those of Arg464 [16].
Group B mutations alter the EF-G–ribosome inter-
actions, most likely affecting how efficiently EF-G can
be locked to the ribosome by FA. These mutations
occur either on the EF-G surface, at direct interaction
points in the EF-G–70S complex structure [22]
(Gly617, Gly621 and Arg659 in domain V), or at sites
in domains I and V that are structurally important for
the presentation of a normal contact surface of EF-G
to the ribosome (Pro114, Gln115, Thr118, Gly628,
Gly632, Ala655, Ser660 and Gly664).
The mutations of this class in domain G are all situ-

ated at the N-terminal end of helix C
G
. The mutations
P114H, Q115L and T118I will modify the surface of
EF-G exposed for interaction with the 2660 region of
23S RNA.
Several group B mutation sites occur in helices A
V
and B
V
at the surface of domain V (Fig. 4C). Helix A
V
is packed against the second molecule in our structure,
whereas in the structure of EF-G in complex with the
ribosome [22], this helix interacts with the 1095 and
2473 regions of 23S RNA. Helix B
V
packs against
domains G and III, and in complex with the ribosome
the C-terminal turn of the helix (659–663) interacts
with the 2660 region of 23S RNA as well as with
the C-terminus of ribosomal protein L6 [22]. Resis-
tance mutations in this region presumably either
disturb helix B
V
(S660P) or the following turn
(G664A ⁄ G664S), or make changes to the outer surface
of the helix (A655E, R659C ⁄ R659H ⁄ R659S), affecting
the surface in contact with the ribosome. The contacts
of domain V with domains G and III undergo large

changes between the free and ribosome-bound states
of EF-G, and thus the same mutations can also affect
the conformational dynamics of EF-G.
Gly617 and Gly621, situated in helix A
V
, are in the
contact area between the two molecules in the present
structure. On the ribosome (PDB 2wri [22]), Gly617 is
close to A1095 of 23S RNA, whereas Gly621 is close
to U2473. Mutation of any of these to introduce
large side chains would change the interaction with the
ribosome.
Table 2. (Continued).
Residue Species Mutation Group Location Interpretation of mutation
Arg659 S. aureus [27] His
Cys
Ser
B, C Domain V; making a hydrogen bond with Asp87 in
the switch II region. In the ribosome complex, the
backbone carbonyl interacts with A2660 of 23S
RNA [22]
Mutation may affect the domain G–V interaction as
well as ribosome interaction
Ser660 S. aureus [27] Pro B Domain V, in helix B
V
at the surface close to the
interaction with 2660 of 23S RNA and L6 [22]
Mutation to Pro will disturb the helix and may affect
nearby ribosome interactions
Gly664 S. aureus [27] Ser

Ala
B, C Domain V, at the interface with domain G, in contact
with Pro114. In the ribosome complex [22], this
interface changes; now close to the interaction
with 2660 of 23S RNA. Backbone conformation not
accessible for other residues
Mutation will change the backbone conformation,
may influence the domain G–V interaction as well
as ribosome interactions
Gly666 S. aureus [27] Val C Domain V, in strand 4
V
, at the interface with
domain G
Mutation may influence the domain G–V interaction
Met670 S. typhimurium [34] Ile C Domain V, in strand 4
V
, at the interface with
domain III
Mutation may influence the domain III–V interaction
a
Residue is not conserved between S. aureus and T. thermophilus.
Y. Chen et al. Crystal structure of Staphylococcus aureus EF-G
FEBS Journal 277 (2010) 3789–3803 ª 2010 The Authors Journal compilation ª 2010 FEBS 3799
The only ribosomal mutations conferring resistance
against FA occur as truncations or frameshift muta-
tions in ribosomal protein L6 [27]. Probably, they have
an effect similar to the group B mutations of EF-G,
because, in the FA complex of EF-G with the ribo-
some, the C-terminus of ribosomal protein L6 is in
direct contact with domain V of EF-G [22].

Group C mutations are related to the interdomain
orientations in EF-G, thereby probably affecting the
FA-binding pocket as well as the conformational
dynamics and FA locking of EF-G. These mutations
occur in domains I, III and V. Our prediction is that
these mutants will have a lower propensity to adopt
the FA-locked domain orientations, and may display
lower drug affinity as well as faster dissociation from
the ribosome.
Several of these mutation sites occur in the linker
between domains II and III, the only physical connec-
tion between domains G and II and domains III, IV
and V. In the S. aureus EF-G structure, this loop has
a bent conformation, and packs against domain III
and interacts with the linker between domains III and
IV (Fig. 4D). Pro404 and Pro406 are critical for this
conformation. His438, within hydrogen-bonding dis-
tance of Glu405 and the carbonyl group of Pro404,
also contributes to this particular linker conformation.
In the FA-locked conformation (PDB 2wri [22]), the
loop stretches out to reposition domains III–V. The
mutations T385N, P404L, P404Q, P404R, P406L,
V407F and H438N can affect the conformational
dynamics of EF-G through their effect on the structure
and dynamics of the linker, but can also affect the
FA-binding pocket at the domain interface.
Residues 666 and 670, within one b-strand of
domain V, are also important for interdomain interac-
tions. The N-terminal part of the strand interacts with
the N-terminal part of helix C

G
in domain G, and the
C-terminal part interacts with the turn before helix B
III
in domain III. The G664A, G666V and M670I muta-
tions could possibly affect these interactions. In the
S. aureus structure, the T118I mutation in domain G
would affect the same interaction, and in domain III,
the 438 region, containing several mutation sites, is
close in space. Both interactions of this b-strand
change upon ribosome binding (compare PDB 2wri
[22]).
The switch II mutation V90I is puzzling, as the addi-
tion of a single methyl group in S. aureus provides
32-fold FA resistance [27], whereas several other
species, such as T. thermophilus, have Ile in the
corresponding position without being resistant. In the
FA-locked conformation of T. thermophilus EF-G [22],
Ile92 is packed against domain III. As the directly
interacting residues in domain III are conserved in
S. aureus, the resistance is probably caused by the resi-
due at position 92 in combination with the sequence
variation at the surface of domain III that interacts
with domain I, motivating the classification of this
mutation as group C.
Finally, group D mutations seem to decrease the
structural stability of domains I and III of EF-G. This
is probably another way of preventing the locking of
EF-G, and these mutants with lower stability may dis-
sociate faster than the wild type from the ribosome,

and perhaps display lower drug affinity. In the hydro-
phobic core of domain G, the mutations A102E,
L106S and V132T would disturb the hydrophobic
core. This may, in turn, affect the ribosome interac-
tions of domain G. In domain III, A426D and L430Q
may have similar effects.
In conclusion, the new crystal structure of S. aureus
EF-G demonstrates the flexibility of EF-G in the free
state, and provides new details that may be useful in
the design of new and better antibiotics targeting EF-
G function in translation.
Experimental procedures
Cloning, overexpression and purification
A pET30 construct of N-terminally His
6
-tagged S. aureus
EF-G [37] was transformed into Escherichia coli BL21(DE3),
and protein expression was induced by 1 mm isopropyl-thio-
b-d-galactoside at a D
595 nm
of 0.5–0.6. The soluble fraction
of S. aureus His
6
–EF-G was purified by affinity chromatog-
raphy on a His-trap FF column (GE Healthcare, Uppsala,
Sweden) equilibrated with buffer A (50 mm Tris ⁄ HCl,
pH 7.5, 200 mm NaCl). The column was washed with 5 mm
imidazole in buffer A, and eluted with a 10–200 mm
imidazole gradient in buffer A. The purified His
6

–EF-G was
immediately dialysed against buffer A. Further purification
was performed on a Hiload 16 ⁄ 60 Superdex-75 (GE Health-
care) gel filtration column equilibrated with buffer A. The
purified protein was concentrated to 4 mgÆmL
)1
with an Am-
icon Ultra 10 kDa cut-off centrifugal filter unit (Millipore,
Billerica, MA, USA) and stored at )80 °C after shock freez-
ing in liquid nitrogen. Before crystallization, the buffer was
exchanged for 20 mm Tris ⁄ HCl (pH 7.5) and 200 mm NaCl
in a 10 kDa cut-off Vivaspin concentrator.
Crystallization
Initial microcrystals of S. aureus EF-G grew in the Index
screen (Hampton Research, Aliso Viejo, CA, USA) with a
reservoir solution containing 100 mm Tris ⁄ HCl (pH 8.5),
200 mm NaCl and 25% (w ⁄ v) poly(ethylene glycol) 3350.
In attempts to reproduce these crystals, they appeared only
Crystal structure of Staphylococcus aureus EF-G Y. Chen et al.
3800 FEBS Journal 277 (2010) 3789–3803 ª 2010 The Authors Journal compilation ª 2010 FEBS
after 1 month and in few of the drops. Therefore, S. aureus
EF-G crystals were grown in streak-seeded sitting drop
vapour diffusion experiments. Two microlitres of a
4mgÆmL
)1
protein solution was mixed with 2 lL of reser-
voir solution [20–25% (w ⁄ v) poly(ethylene glycol) 3350,
200 mm Tris ⁄ HCl, pH 8.7, 200 mm NaCl] and equilibrated
against 400 lL of reservoir solution at 295 K. Crystals
appeared after 1–2 weeks, and grew to 100–150 l m in size

in 4 weeks. The crystals were gradually transferred to a
cryoprotecting solution (reservoir solution supplemented
with 20% 2-methyl-2,4-pentanediol) and vitrified by plung-
ing into liquid nitrogen.
Crystallography
A first 2.1 A
˚
resolution apo dataset was collected at the
PXII beamline, Swiss Light source (Villigen, Switzerland),
with a Pilatus detector. Later, a 1.9 A
˚
resolution dataset
was collected at beamline ID14-1, ESRF (Grenoble,
France), with an ADSC Q210 CCD detector. Data were
collected at 100 K and 0.9334 A
˚
. Data were processed and
scaled with the xds package [38] (Table 1). The crystals
belong to space group P2
1
, with the following cell dimensions:
a = 47.16 A
˚
, b = 137.34 A
˚
, c = 125.36 A
˚
, a = c =90°,
and b = 94.93°.
The structure was solved by molecular replacement with

phaser [39], using domains I and II from T. thermophilus
EF-G as a search model (PDB 1fnm [16]). Domains V, III
and IV were sequentially placed in the difference maps with
coot [40]. The structure was improved by the use of auto-
matic model building in arp ⁄ warp [41]. Further structure
rebuilding was performed in coot [40], and the structure
was refined, with cns [42,43] and refmac [44,45], to R
work
and R
free
of 18.7% and 22.4%, respectively (Table 1). No
noncrystallographic symmetry constraints were used, but
a TLS model was applied in the last refinement runs in
refmac, with one TLS group for each domain of EF-G.
The structure contains two molecules in the asymmetric
unit, resulting in 51.1% solvent content.
Structure analysis
Calculations of radii of gyration were performed with
moleman2 [46] from the corresponding coordinates depos-
ited in the PDB. Structure superposition was performed
with o [47].
Acknowledgements
We thank the SLS and ESRF for beam time and sup-
port during data collection, and K. Ba
¨
ckbro and C. S.
Koh for comments on the manuscript. This work was
supported by individual grants from the Swedish
Research Council, the Wenner Gren foundation and
Carl Trygger’s Foundation to M. Selmer and S. San-

yal, the Go
¨
ran Gustafsson Foundation to S. Sanyal,
and Magnus Bergvall’s foundation and the Swedish
Foundation for Strategic Research to M. Selmer.
References
1 Schmeing TM & Ramakrishnan V (2009) What recent
ribosome structures have revealed about the mechanism
of translation. Nature 461 , 1234–1242.
2 Steitz TA (2008) A structural understanding of the
dynamic ribosome machine. Nat Rev Mol Cell Biol 9,
242–253.
3 Shoji S, Walker SE & Fredrick K (2009) Ribosomal
translocation: one step closer to the molecular mecha-
nism. ACS Chem Biol 4, 93–107.
4 Wintermeyer W, Savelsbergh A, Semenkov YP,
Katunin VI & Rodnina MV (2001) Mechanism of
elongation factor G function in tRNA translocation
on the ribosome. Cold Spring Harb Symp Quant Biol
66, 449–458.
5 Hirashima A & Kaji A (1973) Role of elongation
factor G and a protein factor on the release of
ribosomes from messenger ribonucleic acid. J Biol Chem
248, 7580–7587.
6 Peske F, Rodnina MV & Wintermeyer W (2005)
Sequence of steps in ribosome recycling as defined by
kinetic analysis. Mol Cell 18, 403–412.
7 Zavialov AV, Hauryliuk VV & Ehrenberg M (2005)
Splitting of the posttermination ribosome into subunits
by the concerted action of RRF and EF-G. Mol Cell

18, 675–686.
8 Mohr D, Wintermeyer W & Rodnina MV (2002)
GTPase activation of elongation factors Tu and G on
the ribosome. Biochemistry 41, 12520–12528.
9 Helgstrand M, Mandava CS, Mulder FA, Liljas A,
Sanyal S & Akke M (2007) The ribosomal stalk binds
to translation factors IF2, EF-Tu, EF-G and RF3 via a
conserved region of the L12 C-terminal domain. J Mol
Biol 365, 468–479.
10 Rodnina MV, Savelsbergh A, Katunin VI & Winter-
meyer W (1997) Hydrolysis of GTP by elongation fac-
tor G drives tRNA movement on the ribosome. Nature
385, 37–41.
11 Savelsbergh A, Rodnina MV & Wintermeyer W (2009)
Distinct functions of elongation factor G in ribosome
recycling and translocation. RNA 15, 772–780.
12 Czworkowski J, Wang J, Steitz TA & Moore PB (1994)
The crystal structure of elongation factor G complexed
with GDP, at 2.7 angstrom resolution. EMBO J 13,
3661–3668.
13 Aevarsson A, Brazhnikov E, Garber M, Zheltonosova
J, Chirgadze Y, Al-Karadaghi S, Svensson LA & Liljas
A (1994) Three-dimensional structure of the ribosomal
Y. Chen et al. Crystal structure of Staphylococcus aureus EF-G
FEBS Journal 277 (2010) 3789–3803 ª 2010 The Authors Journal compilation ª 2010 FEBS 3801
translocase: elongation factor G from Thermus thermo-
philus. EMBO J 13, 3669–3677.
14 Hansson S, Singh R, Gudkov AT, Liljas A & Logan
DT (2005) Structural insights into fusidic acid resistance
and sensitivity in EF-G. J Mol Biol 348, 939–949.

15 Hansson S, Singh R, Gudkov AT, Liljas A & Logan
DT (2005) Crystal structure of a mutant elongation fac-
tor G trapped with a GTP analogue. FEBS Lett 579,
4492–4497.
16 Laurberg M, Kristensen O, Martemyanov K, Gudkov
AT, Nagaev I, Hughes D & Liljas A (2000) Structure of
a mutant EF-G reveals domain III and possibly the
fusidic acid binding site. J Mol Biol 303, 593–603.
17 Vetter IR & Wittinghofer A (2001) The guanine nucleo-
tide-binding switch in three dimensions. Science 294,
1299–1304.
18 Connell SR, Takemoto C, Wilson DN, Wang H,
Murayama K, Terada T, Shirouzu M, Rost M, Schuler
M, Giesebrecht J et al. (2007) Structural basis for
interaction of the ribosome with the switch regions of
GTP-bound elongation factors. Mol Cell 25, 751–764.
19 Agrawal RK, Penczek P, Grassucci RA & Frank J
(1998) Visualization of elongation factor G on the
Escherichia coli 70S ribosome: the mechanism of
translocation. Proc Natl Acad Sci USA 95, 6134–6138.
20 Stark H, Rodnina MV, Wieden HJ, van Heel M &
Wintermeyer W (2000) Large-scale movement of elon-
gation factor G and extensive conformational change of
the ribosome during translocation. Cell 100, 301–309.
21 Agrawal RK, Linde J, Sengupta J, Nierhaus KH &
Frank J (2001) Localization of L11 protein on the
ribosome and elucidation of its involvement in
EF-G-dependent translocation. J Mol Biol 311, 777–787.
22 Gao YG, Selmer M, Dunham CM, Weixlbaumer A,
Kelley AC & Ramakrishnan V (2009) The structure of

the ribosome with elongation factor G trapped in the
posttranslocational state. Science 326, 694–699.
23 Czworkowski J & Moore PB (1997) The conformational
properties of elongation factor G and the mechanism of
translocation. Biochemistry 36, 10327–10334.
24 Hauryliuk V, Mitkevich VA, Eliseeva NA, Petrushanko
IY, Ehrenberg M & Makarov AA (2008) The pretrans-
location ribosome is targeted by GTP-bound EF-G in
partially activated form. Proc Natl Acad Sci USA 105,
15678–15683.
25 Bodley JW, Zieve FJ, Lin L & Zieve ST (1969) Forma-
tion of the ribosome–G factor–GDP complex in the
presence of fusidic acid. Biochem Biophys Res Commun
37, 437–443.
26 Lannergard J, Norstrom T & Hughes D (2009) Genetic
determinants of resistance to fusidic acid among clinical
bacteremia isolates of Staphylococcus aureus. Antimicrob
Agents Chemother 53, 2059–2065.
27 Norstrom T, Lannergard J & Hughes D (2007) Genetic
and phenotypic identification of fusidic acid-resistant
mutants with the small-colony-variant phenotype in
Staphylococcus aureus. Antimicrob Agents Chemother
51, 4438–4446.
28 Ticu C, Nechifor R, Nguyen B, Desrosiers M & Wilson
KS (2009) Conformational changes in switch I of EF-G
drive its directional cycling on and off the ribosome.
EMBO J 28, 2053–2065.
29 Jorgensen R, Ortiz PA, Carr-Schmid A, Nissen P,
Kinzy TG & Andersen GR (2003) Two crystal struc-
tures demonstrate large conformational changes in the

eukaryotic ribosomal translocase. Nat Struct Biol
10,
379–385.
30 Al Karadaghi S, Aevarsson A, Garber M, Zheltonos-
ova J & Liljas A (1996) The structure of elongation
factor G in complex with GDP: conformational
flexibility and nucleotide exchange. Structure 4,
555–565.
31 Polekhina G, Thirup S, Kjeldgaard M, Nissen P, Lipp-
mann C & Nyborg J (1996) Helix unwinding in the
effector region of elongation factor EF-Tu-GDP. Struc-
ture 4, 1141–1151.
32 Kjeldgaard M, Nissen P, Thirup S & Nyborg J (1993)
The crystal structure of elongation factor EF-Tu from
Thermus aquaticus in the GTP conformation. Structure
1, 35–50.
33 Liljas A, Kristensen O, Laurberg M, Al-Karadaghi S,
Gudkov A, Martemyanov K, Hughes D & Nagaev I
(2000) The states, conformational dynamics, and fusidic
acid-resistant mutants of elongation factor G. In The
Ribosome: Structure, Function, Antibiotics and Cellular
Interactions (Garrett RA, Douthwaite SR, Liljas A,
Matheson AT, Moore PB & Noller HF eds), pp. 359–
365. ASM Press, Washington, DC.
34 Johanson U & Hughes D (1994) Fusidic acid-resistant
mutants define three regions in elongation factor G of
Salmonella typhimurium. Gene 143, 55–59.
35 Nagaev I, Bjorkman J, Andersson DI & Hughes D
(2001) Biological cost and compensatory evolution in
fusidic acid-resistant Staphylococcus aureus. Mol

Microbiol 40, 433–439.
36 Martemyanov KA, Liljas A, Yarunin AS & Gudkov
AT (2001) Mutations in the G-domain of elongation
factor G from Thermus thermophilus affect both its
interaction with GTP and fusidic acid. J Biol Chem 276,
28774–28778.
37 Li JJ, Venkataramana M, Wang AQ, Sanyal S, Janson
JC & Su ZG (2005) A mild hydrophobic interaction
chromatography involving polyethylene glycol immobi-
lized to agarose media refolding recombinant Staphylo-
coccus aureus elongation factor G. Protein Expr Purif
40, 327–335.
38 Kabsch W (1993) Automatic processing of rotation
diffraction data from crystals of initially unknown
symmetry and cell constants. J Appl Crystallogr 26 ,
795–800.
Crystal structure of Staphylococcus aureus EF-G Y. Chen et al.
3802 FEBS Journal 277 (2010) 3789–3803 ª 2010 The Authors Journal compilation ª 2010 FEBS
39 McCoy AJ, Grosse-Kunstleve RW, Adams PD, Winn
MD, Storoni LC & Read RJ (2007) Phaser crystallo-
graphic software. J Appl Crystallogr 40, 658–674.
40 Emsley P & Cowtan K (2004) Coot: model-building
tools for molecular graphics. Acta Crystallogr D Biol
Crystallogr 60, 2126–2132.
41 Perrakis A, Morris R & Lamzin VS (1999) Automated
protein model building combined with iterative struc-
ture refinement. Nat Struct Biol 6, 458–463.
42 Brunger AT (2007) Version 1.2 of the Crystallography
and NMR system. Nat Protoc 2, 2728–2733.
43 Brunger AT, Adams PD, Clore GM, DeLano WL,

Gros P, Grosse-Kunstleve RW, Jiang JS, Kuszewski J,
Nilges M, Pannu NS et al. (1998) Crystallography &
NMR system: a new software suite for macromolecular
structure determination. Acta Crystallogr D Biol
Crystallogr 54, 905–921.
44 Murshudov GN, Vagin AA & Dodson EJ (1997)
Refinement of macromolecular structures by the
maximum-likelihood method. Acta Crystallogr D Biol
Crystallogr 53, 240–255.
45 CCP4 (1994) The CCP4 suite: programs for protein
crystallography. Acta Crystallogr D Biol Crystallogr 50,
760–763.
46 Kleywegt GJ (1999) Experimental assessment of
differences between related protein crystal
structures. Acta Crystallogr D Biol Crystallogr 55,
1878–1884.
47 Jones TA, Zou J-Y, Cowan SW & Kjeldgaard M (1991)
Improved methods for building protein models in
electron density maps and the location of errors in
these models. Acta Crystallogr A 47 , 110–119.
48 Diederichs K & Karplus PA (1997) Improved R-factors
for diffraction data analysis in macromolecular
crystallography. Nat Struct Biol 4, 269–275.
49 Richter Dahlfors AA & Kurland CG (1990) Novel
mutants of elongation factor G. J Mol Biol 215,
549–557.
Y. Chen et al. Crystal structure of Staphylococcus aureus EF-G
FEBS Journal 277 (2010) 3789–3803 ª 2010 The Authors Journal compilation ª 2010 FEBS 3803