Tài liệu Báo cáo khoa học: Amino acid discrimination by arginyl-tRNA synthetases as revealed by an examination of natural specificity variants doc

Amino acid discrimination by arginyl-tRNA synthetases as
revealed by an examination of natural specificity variants
Gabor L. Igloi and Elfriede Schiefermayr
Institute of Biology, University of Freiburg, Germany
The accuracy of protein biosynthesis is critically
dependent on the fidelity with which aminoacyl-tRNA
synthetases (EC 6.1.1.x) recognize their cognate amino
acid and tRNA substrates [1]. The mechanism(s) by
which the family of aminoacyl-tRNA synthetases
maintains the accuracy of protein biosynthesis has
been the subject of intensive research for some years
[2]. To discriminate between structurally similar amino
acids, whose binding energy difference is insufficient to
guarantee the required distinction [3], some aminoacyl-
tRNA synthetases possess an additional proofreading
or editing activity [4–8] that actively hydrolyses mis-
acylated products. For others that are specific for
structurally idiosyncratic amino acids, no active editing
may be required. In the case of glutamyl- and glutami-
nyl-tRNA synthetases, which together with arginyl-
tRNA synthetase form a subgroup of enzymes that
require tRNA for amino acid activation, the potential
for misrecognition of related amino acids has been
investigated [9–13] and modulated by amino acid
replacements and active site redesign [14]. A mecha-
nism that does not rely on hydrolytic editing but
Keywords
arginyl-tRNA synthetase;
L-canavanine;
discrimination; jack bean; soybean
Correspondence

G. L. Igloi, Institute of Biology, University of
Freiburg, Scha
¨
nzlestr. 1, D-79104 Freiburg,
Germany
Fax: +49 761 203 2745
Tel: +49 761 203 2722
E-mail:
(Received 22 September 2008, revised 17
December 2008, accepted 19 December
2008)
doi:10.1111/j.1742-4658.2009.06866.x
l-Canavanine occurs as a toxic non-protein amino acid in more than 1500
leguminous plants. One mechanism of its toxicity is its incorporation into
proteins, replacing l-arginine and giving rise to functionally aberrant poly-
peptides. A comparison between the recombinant arginyl-tRNA synthetases
from a canavanine producer (jack bean) and from a related non-producer
(soybean) provided an opportunity to study the mechanism that has evolved
to discriminate successfully between the proteinogenic amino acid and its
non-protein analogue. In contrast to the enzyme from jack bean, the
soybean enzyme effectively produced canavanyl-tRNA
Arg
when using RNA
transcribed from the jack bean tRNA
ACG
gene. The corresponding k
cat
⁄ K
M
values gave a discrimination factor of 485 for the jack bean enzyme. The

arginyl-tRNA synthetase does not possess hydrolytic post-transfer editing
activity. In a heterologous system containing either native Escherichia coli
tRNA
Arg
or the modification-lacking E. coli transcript RNA, efficient dis-
crimination between l-arginine and l-canavanine by both plant enzymes
(but not by the E. coli arginyl-tRNA synthetase) occurred. Thus, interaction
of structural features of the tRNA with the enzyme plays a significant role
in determining the accuracy of tRNA arginylation. Of the potential amino
acid substrates tested, apart from l-canavanine, only l-thioarginine was
active in aminoacylation. As it is an equally good substrate for the
arginyl-tRNA synthetase from both plants, it is concluded that the higher
discriminatory power of the jack bean enzyme towards l-canavanine does
not necessarily provide increased protection against analogues in general,
but appears to have evolved specifically to avoid auto-toxicity.
Abbreviations
L-Cav, L-canavanine; PCAF, pentacyanoamidoferroate.
FEBS Journal 276 (2009) 1307–1318 ª 2009 The Authors Journal compilation ª 2009 FEBS 1307
resembles an induced fit type of substrate selection,
including the participation of tRNA structural
features, has been proposed [14]. The specificity of
arginyl-tRNA synthetase (EC 6.1.1.19) towards amino
acids for which a similar discriminatory mechanism
may be required has not been studied systematically.
Research regarding the accuracy of protein biosyn-
thesis has, in the past, been largely devoted to prok-
aryotes and lower eukaryotes (yeast). With isolated
exceptions in the early literature, aminoacyl-tRNA
synthetases from plants, which must not only discrimi-
nate between the 20 common amino acids but must

also contend with related potentially toxic natural ana-
logues [15,16], have been ignored. This challenge faced
by plants offers a natural alternative to targeted muta-
genesis or rational redesign of the active site of the
enzymes to elucidate the mechanism by which fidelity
of amino acid selection is maintained. We have focused
our attention on a pair of species-specific enzyme vari-
ants, one of which is said to be evolutionarily adapted
to reject a naturally occurring toxic arginine analogue
[17], while the other lacks this ability. l-Canavanine
[18,19] [l-2-amino-4-(guanidinooxy)butyric acid], the
guanidino-oxy structural analogue of arginine (Fig. 1)
occurs as a toxic non-protein amino acid in more than
1500 leguminous plants. One mechanism of its toxicity
is its incorporation into proteins, replacing l-arginine
and giving rise to functionally aberrant polypeptides
[20–22]. A comparison between the recombinant argi-
nyl-tRNA synthetases from a canavanine producer
(jack bean, Canavalia ensiformis) and from a related
non-producer (soybean, Glycine max) provides an
opportunity to gain insight into the mechanism of
amino acid recognition in the arginine system.
Results
On the basis of the annotated Arabidopsis genome, we
established the cDNA sequence of the argS gene of
jack bean (accession number AM950325) [23] and of
soybean (accession number FM209045). The derived
proteins comprise 597 (soybean) and 595 (jack bean)
amino acids, with molecular masses of 68.2 and
67.4 kDa, respectively. The genes for arginyl-tRNA

synthetase from jack bean and soybean were cloned
into the bacterial expression vector pET32a and trans-
formed into Escherichia coli BL21 cells. Despite their
sequence similarity (Fig. 2), the enzyme from soybean
proved much more resistant to soluble expression than
the one from jack bean [23]. The yield from jack bean
(10 mgÆL
)1
cell culture) compares with 1.2 mgÆL
)1
cul-
ture for soybean. Removal of the His-tag ⁄ thioredoxin
fusion by cleavage at the enterokinase site provided by
the vector was unsuccessful. However, the thrombin
site, located 30 amino acids upstream of the native
synthetase sequence, was accessible to proteolysis. A
predicted internal thrombin site (position 130 of the
native protein) in the soybean arginyl-tRNA synthe-
tase was not targeted by this protease. The position of
the cleavage was confirmed by N-terminal protein
sequencing. The results reported here were obtained
using thrombin-treated preparations of arginyl-tRNA
synthetases that retained a 3.2 kDa N-terminal exten-
sion compared to the native enzyme.
Sequence analysis of the tRNA
Arg
ACG
gene from
Canavalia ensiformis established its identity to the Ara-
bidopsis sequence (accession number NR_023294). The

subsequent appearance in the NCBI trace archives of a
sequence corresponding to the gene of soybean
tRNA
ACG
(accession number gnl|ti|1583039205) con-
firmed its similarity to the jack bean sequence with a
single base difference from A (jackbean) to G (soy-
bean) at position 37. The chemically synthesized gene
for tRNA
Arg
ACG
from jack bean was cloned, and the
full-length tRNA was generated by in vitro transcrip-
tion. The transcript could be aminoacylated with argi-
nine to a level of approximately 0.05 pmol amino
acid ⁄ pmol tRNA. The corresponding soybean tran-
script had an arginine acceptance level of approxi-
mately 0.1 pmol amino acid ⁄ pmol tRNA.
As is the case for arginyl-tRNA synthetases from
other sources [24–26], the pyrophosphate exchange
reaction is absolutely dependent on the presence of
aminoacylatable tRNA. Periodate-oxidized tRNA,
which has been shown to be inactive in aminoacyla-
tion, did not stimulate pyrophosphate exchange
(Fig. 3). The tRNA concentration dependence of this
reaction gives a K
M
value that is equivalent to that of
HN
O

NH OH
L-Arginine
H
2
NN
NH
2
NH
2
NH
2
NH
2
O
OH
O
L-Canavanine
Fig. 1. Structures of L-arginine and its guanidinooxy analogue,
L-canavanine.
Arginyl-tRNA synthetase amino acid discrimination G. L. Igloi and E. Schiefermayr
1308 FEBS Journal 276 (2009) 1307–1318 ª 2009 The Authors Journal compilation ª 2009 FEBS
tRNA as measured by aminoacylation (data not
shown).
Using either [
14
C]-canavanine in the conventional
aminoacylation assay, or unlabelled canavanine
together with [
32
P]-labelled jack bean transcript tRNA,

it was observed that the soybean enzyme effectively
transferred this amino acid to the transcript tRNA,
but it was a much poorer substrate for the jack bean
enzyme (Fig. 4, inset). To examine whether the argi-
nyl-tRNA synthetases from the two plants show differ-
ent specificities towards other arginine analogues, the
[
32
P]-labelled tRNA assay was used to screen a selec-
tion of amino acids, including ones that have previ-
ously been shown not to be substrates for the enzyme
from other sources. l-thiocitrulline and the naturally
occurring l-homoarginine, l-citrulline, l-homocitrul-
line and l-albizziine (l-2-amino-3-ureidopropanoic
acid) were, at 1 mm concentration, if at all, extremely
poor substrates for both plant enzymes (Fig. 4), and
Fig. 2. Alignment of derived arginyl-tRNA synthetase primary structures from jack bean (Ce, Canavalia ensiformis), soybean (Gm,
Glycine max) and yeast (Sc, Saccharomyces cerevisiae). Shading in black indicates identity in all three sequences; shading in grey indicates
identity in two sequences.
Time (min)
0 2 4 6 8 10121416
PPi exchange in to ATP (Pmol)
–20
0
20
40
60
80
100
120

Fig. 3. Dependence of the pyrophosphate exchange reaction on
tRNA. The pyrophosphate exchange reaction was carried out in the
absence (
) or the presence of 3 lM ( )or30lM (r) transcript tRNA
or 12 l
M (d) periodate-oxidized jack bean transcript tRNA using jack
bean arginyl-tRNA synthetase. PPi, tetrasodium pyrophosphate.
G. L. Igloi and E. Schiefermayr Arginyl-tRNA synthetase amino acid discrimination
FEBS Journal 276 (2009) 1307–1318 ª 2009 The Authors Journal compilation ª 2009 FEBS 1309
l-lysine charging was barely detectable. The synthetic
arginine analogue, l-thioarginine, recently introduced
as a substrate for arginase [27], was extensively trans-
ferred to tRNA by both enzymes (K
M
for soybean
56 lm; K
M
for jack bean 81 lm).
In order to quantify the discrimination exhibited by
the plant enzymes with respect to canavanine, kinetic
parameters for aminoacylation were determined using
the tRNA transcript derived from the jack bean gene.
Radioactive canavanine was efficiently transferred to
the plant tRNA transcript by the arginyl-tRNA
synthetase from soybean. In this case, the kinetic para-
meters correspond to a discrimination factor, (k
cat
⁄
K
M

)
Arg
⁄ (k
cat
⁄ K
M
)
Cav
, of 44 (Table 1). A similar factor
was obtained when assayed with non-radioactive cana-
vanine using the [
32
P]-labelled tRNA assay [28]. For the
jack bean enzyme, a distinct discrimination between
arginine and canavanine for aminoacylation of the plant
tRNA transcript was observed when using [
14
C]-canava-
nine. At 0.4 mm canavanine, less than 10% of the tRNA
was aminoacylated compared to arginine transfer. This
low but significant level of mischarging is the result of a
relatively modest degree of discrimination. Using the
sensitive [
32
P]-labelled tRNA assay and higher concen-
trations of canavanine, a K
M
for this substrate of
1.3 mm was determined, and the relative magnitude of
the k

cat
⁄ K
M
parameters for arginine and canavanine
charging revealed a discrimination factor of 485; a fac-
tor of 10 greater than for the soybean enzyme (Table 1).
The discrimination based on catalytic efficiency may
in itself be insufficient to guarantee survival of the
canavanine-producing plant. An additional classic
post-transfer proofreading mechanism [7,29] would
require the rapid deacylation of Cav-tRNA
Arg
by the
L
A -
g r
i
n
i
n
e
L
H
-
o
m
o
a
g r
i

n
i
n
e
L
h T
-
i
o
a
r
g i
n
i
n
e
L
C -
u
t
u
r
l
l
i
n
e
L
H -
o

m
o c
i
t
r
u
l
l
i
n e
L
h T -
i
o
c
i
t
r
u
l
l
i
n
e
L
L -
y
s
i
n

e
L
-
A
l
b i
z
z
i
i
n
e
L
C
-
an
a
v
a
n
i
n
e
Jack bean enzyme
Soybean enzyme
0
20
40
60
80

100
120
NH
NH
N H
2
NH
2
OH
O
NH
O
NH
N H
2
NH
2
OH
NH
O
S
N H
2
NH
2
OH
NH
2
O
NH

O
NH
2
OH
NH
2
NH
O
NH
2
OH
O
NH
2
O
NH
S
NH
2
OH
N H
2
O
NH
2
NH
2
NH
2
OH

O
OH
NH
O
NH
2
N
N H
2
NH
2
O
OH
O
c a o n i m A y n o i t a l
(
r a % g n i y n o i t a l
)

tRNA
Origin
Aminoacyl-A
76
n a e b k c a J
n a e b y o S
g r A
v a
C
g
r

A
v a C
Fig. 4. Quantitative comparison of amino acid utilization by the plant arginyl-tRNA synthetases. The aminoacylation level attained in the pres-
ence of
L-arginine was compared to that in the presence of 1 mM of the analogue indicated, using [
32
P]-labelled jack bean transcript tRNA.
Inset: Activity of arginyl-tRNA synthetase from jack bean and soybean with
L-canavanine, under the above conditions. Aminoacylation is char-
acterized by the liberation of labelled aminoacyl-A76 after nuclease P1 treatment.
Arginyl-tRNA synthetase amino acid discrimination G. L. Igloi and E. Schiefermayr
1310 FEBS Journal 276 (2009) 1307–1318 ª 2009 The Authors Journal compilation ª 2009 FEBS
jack bean enzyme. Cav-tRNA
Arg
was prepared by
canavanylation of the jack bean tRNA transcript using
arginyl-tRNA synthetase from soybean. The stability
of the isolated charged tRNA was compared in the
presence of arginyl-tRNA synthetase from soybean or
jack bean (Fig. 5). The first-order decay curves corre-
spond to a half life of only approximately 5 min for
Cav-tRNA
Arg
even in the absence of either enzyme.
In contrast, the half life of Arg-tRNA
Arg
is 46 min.
Addition of arginyl-tRNA synthetase from jack bean
does not further decrease the stability of the canavany-
lated species.

The role of tRNA as a cofactor for aminoacylation
in those aminoacyl-tRNA synthetases that require
tRNA for amino acid activation is well documented
[9], and the determinants within the tRNA that are
required for arginine activation by a mammalian
enzyme have been established using various constructs,
including tRNA chimeras comprising domains from
yeast [26]. If or how these structural elements are
involved in amino acid discrimination was not speci-
fied. Using the pair of plant arginyl-tRNA synthetases
characterized here, it is possible to investigate how
alterations in the tRNA structure manifest themselves
in terms of misaminoacylation. As a first approach, we
screened a number of heterologous tRNA ⁄ enzyme
pairs for aminoacylation. tRNAs from a number of
sources, when compared to the activity with E. coli
arginyl-tRNA synthetase, proved to be arginylated by
the plant enzymes (Fig. 6). In absolute terms, tran-
scripts of tRNA genes were poorly arginylated by their
respective enzymes (Table 2). Remarkably, the soybean
enzyme was no longer able to attach canavanine to
E. coli tRNA
Arg
ACG
(Fig. 7) despite the fact that
Table 1. Quantification of discrimination between L-arginine and L-canavanine using jack bean transcript tRNA. Assays were based on the
aminoacylation reaction using either [
14
C]-labelled amino acids or [
32

P]-labelled tRNA.
Source of
enzyme
Assay method
Aminoacylation of transcript tRNA with
[
14
C]-labelled amino acid
Aminoacylation of [
32
P]-labelled transcript
tRNA
Discrimination
factor(k
cat
⁄ K
M
)
Arg
⁄ (k
cat
⁄ K
M
)
Cav
Arg Cav Arg Cav
K
M
(lM)
k

cat
⁄ K
M
(M
)1
Æmin
)1
) K
M
(lM)
k
cat
⁄ K
M
(M
)1
Æmin
)1
) K
M
(lM)
k
cat
⁄ K
M
(M
)1
Æmin
)1
) K

M
(lM)
k
cat
⁄ K
M
(M
)1
Æmin
)1
)
Jack bean 19.6 1.0 ND
a
ND 7.8 0.82 1320 0.0017 482
Soybean 2.2 9.2 45.3 0.27 3.0 3.2 45.4 0.055 34 ([
14
C]-labelled amino acid);
58 ([
32
P]-labelled tRNA)
a
ND, not determined because of the impracticality of using large amounts of [
14
C]-Cav.
Time (min)
) % ( g n i n i a m e r A N R t - l y c a o n i m A
0
0 5 10 15 20 25 30 35
20
40

60
80
100
120
Fig. 5. Stability of canavanyl-tRNA. Jack bean transcript tRNA
Arg
that had been aminoacylated with [
14
C]-L-canavanine was incubated
in the absence of enzyme (
), or in the presence of jack bean
(d) or soybean (,) arginyl-tRNA synthetase, and the amount of
aminoacyl-tRNA remaining after a given time was quantified.
Alternatively, [
14
C]-L-arginyl-tRNA was incubated in the absence of
enzyme ()).
0.00
20.00
40.00
60.00
80.00
100.00
120.00
140.00
Jack bean
transcript
Soybean
transcript
Wheat

germ total
tRNA
Bovine
liver total
tRNA
E.coli
native
tRNA-Arg
E.coli
transcript
Source of tRNA
level noitalynigrA
htiw deniatta %( iloc.E )emyzne
Fig. 6. Interspecies arginylation. tRNA from the sources indicated
were arginylated in the presence of arginyl-tRNA synthetase from
jack bean (diagonal shading) or soybean (vertical shading), and the
level of charging was compared with that in the presence of the
E. coli enzyme.
G. L. Igloi and E. Schiefermayr Arginyl-tRNA synthetase amino acid discrimination
FEBS Journal 276 (2009) 1307–1318 ª 2009 The Authors Journal compilation ª 2009 FEBS 1311
E. coli tRNA is a good substrate for arginylation. The
presence of E. coli tRNA, irrespective of whether
native or the modification-lacking transcript, caused
‘evolution’ of a discriminatory soybean enzyme that
could, in contrast to the E. coli enzyme, reject canav-
anylation as efficiently as the jack bean enzyme. The
jack bean enzyme did not charge either its cognate
tRNA or the transcript corresponding to the soybean
sequence with canavanine.
Discussion

The evidence that the arginyl-tRNA synthetase of a
canavanine producer, e.g. jack bean (Canavalia ensifor-
mis), can discriminate between l-arginine and its ana-
logue is indirect. It relies on the observation that jack
bean plants injected with radioactive l-canavanine do
not incorporate the label into their proteins, compared
to soybean plants, which do [30]. In a previous study,
‘somewhat indefinite’ conclusions regarding activation
of canavanine by the arginyl-tRNA synthetase from
Canavalia ensiformis [17] were reported. However, the
pyrophosphate exchange assay, in the absence of the
absolutely required tRNA [24], was used to study sub-
strate specificity. The apparent arginine activation
described may be due to a co-purified lysyl-tRNA syn-
thetase (as characterized in the same publication), that
does not require tRNA for pyrophosphate exchange
and can accept arginine [31,32]. While the subsequent
discovery of a corrective proofreading activity of
several aminoacyl-tRNA synthetases [6–8] provides a
reasonable basis for assuming an evolution of a
discriminating function by the jack bean enzyme, we
considered that investigation of a natural, discriminat-
ing ⁄ non-discriminating pair of enzymes would provide
further insight into this process.
The translated gene sequences proved to be 85% iden-
tical to each other but had only 25% identity to the
yeast enzyme, the only eukaryotic arginyl-tRNA synthe-
tase whose 3D structure has been elucidated to date [33].
Despite this limited similarity and the fact that arginyl-
tRNA synthetases from fungi are considered to belong

to a distinct class [34], certain features that have been
identified in yeast as being involved in substrate binding
[35] are conserved in the plant enzymes.
In the case of tRNA recognition, G(483:Y), which is
part of the so-called X loop and is said to form a
molecular switch [33], is conserved (Fig. 2). [We refer
here to comprises the one-letter amino acid followed by
its position in the sequence of the organisms whose
name is abbreviated after the colon, i.e. Y, yeast; C,
Canavalia ensiformis (jack bean); G, Glycine max (soy-
bean)]. Other residues participating in hydrophobic
interactions, such as F(109:Y) and L(70:Y), are also
conserved, and may align with F(100:C), F(102:G) and
L(59:C), L(61:G), respectively. On the other hand,
R(66:Y), R(75:Y) and K(102:Y) do not align with any
charged residues in the jack bean or soybean, leaving
one to speculate on the source of the interaction with
the sugar–phosphate backbone. Correct positioning of
the essential Ade76 of the tRNA has been ascribed to
residues E(294:Y), Y(347:Y) and N(153:Y) [35], all of
which are conserved at corresponding positions in jack
Table 2. Arginine acceptance by homologous and heterologous tRNAs. Arginine acceptance by native E. coli tRNA
Arg
was compared with
that of modification-lacking tRNA transcripts using E. coli or plant arginyl-tRNA synthetases. ND, not determined.
Source of
enzyme
E. coli native tRNA
Arg
E. coli transcript Jack bean transcript Soybean transcript

Aminoacylation
(pmol ArgÆpmol
)1
tRNA) K
M
(lM)
Aminoacylation
(pmol ArgÆpmol
)1
tRNA) K
M
(lM)
Aminoacylation
(pmol ArgÆpmol
)1
tRNA) K
M
(lM)
Aminoacylation
(pmol ArgÆpmol
)1
tRNA) K
M
(lM)
E. coli 0.64 ND 0.07 ND 0.1 ND 0.11 ND
Jack bean 0.61 0.86 0.05 1.2 0.06 ND 0.09 ND
Soybean 0.46 1.5 0.06 1.1 0.05 ND 0.11 ND
0
10
20

30
40
50
60
70
80
90
100

Jack bean
transcript
Soybean
transcript
Wheat germ
total tRNA
Bovine liver
total tRNA
E.coli native
tRNA-Ar
g

E.coli
transcript
Canavanylation (% Arg incorportation)
Fig. 7. Comparison of canavanine incorporation. The amount of
L-canavanine transferred by arginyl-tRNA synthetase from E. coli
(waved shading), jack bean (vertical dashes) and soybean (diagonal
shading) to the tRNA species indicated was quantified using
0.4 m
M [

14
C]-L-canavanine relative to the corresponding arginine
incorporation.
Arginyl-tRNA synthetase amino acid discrimination G. L. Igloi and E. Schiefermayr
1312 FEBS Journal 276 (2009) 1307–1318 ª 2009 The Authors Journal compilation ª 2009 FEBS
bean and soybean. When binding of arginine in the
presence of tRNA was investigated, some changes in
the binding architecture were observed [35], in that
N(153:Y), in addition to interacting with the a-carbox-
ylate, also associates with the 2¢O of Ade76. Similarly,
Y(347:Y) recognizes the guanidinium g-N but also
comes into contact with the adenosine ring of Ade76.
There is a general consensus that tRNA binding is not
required for arginine binding [33], although arginine
binding is a prerequisite for correct positioning of the
CCA end, mediated through movement of a conserved
tyrosine [Y(347:Y)] to a different conformation [26],
allowing ATP to bind productively. Although arginine
and canavanine are stereochemically similar, the pres-
ence of the oxygen atom in canavanine dramatically
influences the pK
a
of the guanidine group, lowering the
value from 12.5 by more than 5 pK
a
units [36,37], lock-
ing the molecule in an imino-oxy tautomer (Fig. 1) and
resulting in a largely uncharged side chain at physiolog-
ical pH.
Transcripts derived from the sequences of the

tRNA
Arg
ACG
genes from jack bean and soybean were
arginylated to only 6–10% of the theoretical acceptance
by the arginyl-tRNA synthetases from both jack bean
and soybean, although the K
M
for the jack bean tRNA
resembles that of native tRNA (Table 2). In general, the
efficiency of transcript aminoacylation may be close to
100% [38,39] but can be substantially less [40–42]. It has
been proposed that the presence of base modifications
leads to reduced flexibility of the tRNA molecule [38],
whereas G:U base pairs are responsible for the tRNA
flexibility required for arginylation in a mammalian sys-
tem [26]. Despite the low level of arginine acceptance by
the transcripts, there was a clear distinction between the
two enzymes when it came to canavanine incorporation.
The enzyme from jack bean produces only low levels of
canavanyl-tRNA with both its cognate and the soybean
tRNA. In contrast, the soybean enzyme effectively
linked the analogue to both plant tRNAs. Examination
of the kinetics of the reaction revealed a significantly
higher affinity of the soybean synthetase for canavanine
(69 lm) compared with that of the jack bean enzyme
(1.3 mm), and the corresponding k
cat
⁄ K
M

values result
in discrimination factors of approximately 40 and 485
for the respective enzymes.
However, in a heterologous system using either
native E. coli tRNA
Arg
ICG
or a transcript of the corre-
sponding gene, we observed how the structure of the
tRNA itself can modulate the efficiency of discrimina-
tion. Whereas these tRNAs are arginylated efficiently
by the synthetases from E. coli, jack bean and soy-
bean, and although canavanylation to a high level is
achieved by the E. coli enzyme, the soybean enzyme
reveals a discriminatory ability that has characteristics
approaching those of the jack bean enzyme.
In view of the distinct role of conformational changes
that accompany the catalytic cycle of the mammalian
enzyme [26], one should consider the possibility that the
amino acid-dependent positioning of the tRNA (or the
CCA end) in a functional configuration, mediated by
global conformational changes in the protein, could be a
further factor in preventing the formation of misacyl-
ated tRNA. For arginyl-tRNA synthetase, rearrange-
ment of the enzyme active site appears to rely on
additional discriminatory elements within the tRNA
structure to ensure accurate formation of aminoacyl-
tRNA. This is reminiscent of the glutamyl- and gluta-
minyl-tRNA synthetases of E. coli. For glutamyl-tRNA
synthetase, the presence of tRNA eliminates non-speci-

fic binding of d-glutamic acid and l-aspartic acid to the
enzyme [9,10]. Detailed analysis of glutaminyl-tRNA
synthetase has led to the proposal of an induced-fit type
of active site rearrangement that plays a role in enzyme
specificity [11–13], and the concept of discriminatory ele-
ments in tRNA that participate in amino acid selection
has been proposed [14]. It would then be consistent with
our observations for jack bean tRNA
Arg
to trigger an
active site rearrangement in the jack bean enzyme that
provides the means to enhance amino acid discrimina-
tion. The fact that the association of the same tRNA
with the soybean enzyme promotes both arginylation
and canavanylation, while in the heterologous system
the soybean enzyme is unable to canavanylate the E. coli
tRNA, is an indication of the subtlety of this structural
interplay, that requires further investigation.
An additional classic post-transfer proofreading
mechanism [6,29], that is not observed in the glutamine
or glutamic acid systems [9,12] but that would enhance
the overall accuracy, would require rapid, specific
deacylation of Cav-tRNA
Arg
by the jack bean enzyme.
Cav-tRNA
Arg
prepared by canavanylation of the jack
bean tRNA transcript using arginyl-tRNA synthetase
from soybean is highly unstable, being rapidly hydroly-

sed at neutral pH even in the absence of added
enzyme. This instability (half life of approximately
5 min) compared to arginyl-tRNA (half life of 46 min)
may be attributed to the electronic charge distribution
of the canavanyl ester that promotes rapid degra-
dation. However, as no additional enzyme-specific
destabilization was observed, post-transfer hydrolytic
proofreading may be ruled out.
The low discrimination factor achieved by the
soybean enzyme leads to efficient canavanylation of
tRNA
Arg
in vitro and incorporation of this allelochemi-
cal into proteins in vivo [30,43]. However, the several
hundred-fold discrimination measured for the jack
G. L. Igloi and E. Schiefermayr Arginyl-tRNA synthetase amino acid discrimination
FEBS Journal 276 (2009) 1307–1318 ª 2009 The Authors Journal compilation ª 2009 FEBS 1313
bean enzyme is considerably lower than the factor of
10
4
normally expected from systems that rely on an
active proofreading process to correct misrecognized
substrates [8]. Nevertheless, physiological evidence
indicates that canavanine producers do not incorporate
this toxic analogue into their proteins. A discrimina-
tion factor between leucine and isoleucine of similarly
modest magnitude (approximately 600) has been
described for leucyl-tRNA synthetase from E. coli [44].
In that case, it was suggested that an evolutionary
balance between catalytic efficiency and specificity can

lead to sacrifices in both these parameters. This may
be reflected in the 5–10-fold reduced relative k
cat
⁄ K
M
for the jack bean enzyme compared to the soybean
synthetase. Additionally, to what extent low levels
of mischarged tRNA can be tolerated [45] or other
in vivo processes such as discrimination at the stage of
elongation factor ⁄ aminoacyl-tRNA complex formation
[2,46,47], competition between various cellular levels of
the amino acids, or metabolic processes competing for
canavanine utilization [48] contribute to the overall
avoidance of auto-toxicity remains to be seen.
The ability of the jack bean enzyme to distinguish
between the secondary metabolite canavanine and its
intended substrate arginine appears to have evolved
specifically. Other arginine analogues such as l-orni-
thine, l-a-amino-c-guanidinobutyric acid, l-citrulline,
l-homocitrulline or l-homoarginine have been assessed
as substrates for arginyl-tRNA synthetases from various
non-plant sources [49–51], and have at best been weak
inhibitors but are generally not incorporated into pro-
teins [20,52]. Of the potential substrates that we have
tested, apart from l-canavanine, only l-thioarginine [27]
was activated significantly. In contrast to l-canavanine,
it is the bridging N of the guanidine group that is
replaced by the heteroatom in l-thioarginine, locking
the guanidino nitrogens into the arginine-like tauto-
meric form. As we have shown that l-thioarginine is an

effective and equally good substrate for the arginyl-
tRNA synthetases from both plants, we conclude that
the higher discriminatory power of the jack bean
enzyme towards canavanine is a specific evolutionary
property that may not necessarily provide increased
protection against analogues in general.
Experimental procedures
Primers were designed using oligo 5.0 (MedProbe, Oslo,
Norway) or gap4 of the Staden Package [53], synthesized
using an ABI3948 nucleic acid synthesis and purification
system (Applied Biosystems, Foster City, CA, USA) by the
Freiburg Institute of Biology core facility. DNA sequence
analysis was performed using BigDye version 1.1 chemicals
(Applied Biosystems) in combination with an ABI Prism 310
genetic analyser. Contigs were assembled using the Staden
Package [53]. Native nucleotidyl transferase from yeast
originated from the stocks of H. Sternbach (formerly Max-
Planck-Institute, Go
¨
ttingen), while that from E. coli in
recombinant form was provided by A. Weiner (University of
Washington School of Medicine, Seattle, WA, USA). [
14
C]-
l-arginine (12.8 GBqÆmmol
)1
) was purchased from Perkin-
Elmer (Waltham, MA, USA). l-homoarginine, l-citrulline
and l-thiocitrulline were obtained from Acros Organics
(Geel, Belgium). The source of other chemicals was as

follows: l-homocitrulline (Advanced Asymmetrics, Millstadt,
IL, USA), l-albizziine (2-amino-3-ureidopropanoic acid)
(Bachem, Bubendorf, Switzerland), l-canavanine (Sigma,
Munich, Germany) and l-thioarginine (l-2-amino-5-isothio-
ureidovaleric acid) (Cayman Chemical, Tallinn, Estonia). An
extract from E. coli, active for aminoacylation, was obtained
by depleting an S30 bacterial supernatant of endogenous
nucleic acids by fractionation on a DEAE-cellulose column.
Bulk tRNAs from wheat germ and from calf liver were pur-
chased from Sigma. E. coli tRNA enriched in tRNA
Arg
ACG
to an arginine acceptance of 760 pmol ⁄ A
260
was obtained
from an expression construct provided by G. Eriani (Institut
de Biologie Mole
´
culaire et Cellulaire, Strasbourg, France)
and E D. Wang (Shanghai Institutes for Biological Sciences,
China) [54].
DNA and RNA isolation
Total RNA was isolated from 100 to 200 mg leaf tissue
from 3 to 4-week-old soybean (Soybean UK, Southampton,
UK) or jack bean (Sigma) plants using RNeasy plant mini
kits (Qiagen, Hilden, Germany). cDNA was prepared using
aT
17
-mer and Superscript reverse transcriptase (Invitrogen,
Karlsruhe, Germany). Sequences were identified by blast

comparison ( />Gene for arginyl-tRNA synthetase
The gene for the enzyme from jack bean has been charac-
terized recently [23] (accession number AM950325). For
the soybean sequence (accession number FM209045), the
translated cDNA sequence of Arabidopsis arginyl-tRNA
synthetase (accession numbers NM_118763 and NM_
105324) was aligned with the corresponding sequences in
other eukaryotes. Soybean EST fragments mined from the
databases were compiled to identify conserved regions,
reverse-translated and used to design primers for cDNA
amplification. The longest PCR fragment obtained by
combining the gene-specific probes with a T
17
primer and
whose sequence could be identified as being that of arginyl-
tRNA synthetase was used to generate primers for stepwise
5¢ RACE elongation of the sequence [55]. PCR products
were purified using Montage cartridges (Millipore, Esch-
born, Germany).
Arginyl-tRNA synthetase amino acid discrimination G. L. Igloi and E. Schiefermayr
1314 FEBS Journal 276 (2009) 1307–1318 ª 2009 The Authors Journal compilation ª 2009 FEBS
Gene for tRNA
Arg
ACG
from jack bean
Total tRNA from jack bean was obtained from cellular
RNA by extraction with 1 m NaCl, and purified by DEAE-
Sephadex chromatography as described previously [56].
tRNA (1 lg) was ligated to 20 pmol of a 5¢-phosphory-
lated, 3¢-periodate-oxidized hybrid RNA ⁄ DNA oligonucleo-

tide [5¢p-rCrCd(CCTCCTTTTATTcactggccgtcgttttacTC)r
A
ox
] synthesized on an ABI 394 DNA ⁄ RNA synthesizer
(Applied Biosystems). The oligonucleotide was designed to
permit efficient ligation through its 5¢-ribonucleotides,
enable the use of the universal M13 primer for reverse tran-
scription (binding region in lower case), and prevent self-
ligation after periodate oxidation of the 3¢-terminal ribose.
Ligation was performed in HCC buffer [57] using 50 units
of T4 RNA ligase (GE Healthcare, Munich, Germany) in a
total volume of 50 lL. For reverse transcription, 1 lLof
the ligation product was annealed to 1 pmol of universal
M13 primer, and the reaction was performed under stan-
dard conditions using 15 units of Thermoscript reverse
transcriptase (Invitrogen). After incubation for 1 h at
56 °C, the reaction was terminated by heating to 85 °C for
5 min, followed by RNase H treatment (GE Healthcare)
for 20 min at 37 °C. The gene specific for tRNA
Arg
was
amplified using the universal M13 primer, which binds to
the 3¢ tail of the RNA, and an 18-mer based on the 5¢
terminus of tRNA
Arg
ACG
from Arabidopsis (accession
number AT1G13010). The amplicon was sequenced using
the M13 primer to give the Canavalia ensiformis 3¢-terminal
55-base sequence. The remaining 5¢ region was assembled

taking into account conserved D-loop bases and the
base-pairing requirement of the D-loop and acceptor
stems, while bearing in mind that none of the 14 plant
tRNA
Arg
ACG
sequences available in the databases possess a
G:U base pair in the acceptor stem (data not shown).
Protein expression
Cloning and bacterial expression of the His-tagged soybean
enzyme was performed as described for jack bean [23].
Thrombin treatment to remove the His tag was performed
as described previously [23]. In the case of the soybean
enzyme, an additional cleaning step comprised adsorption
on Source15Q (GE Healthcare) followed by an 80 mm
NaCl wash and elution at 0.3 m NaCl. The homogeneity of
the preparation was monitored by SDS–PAGE, and the
identity of the protein was confirmed by N-terminal
sequencing.
In vitro transcription
The genes for jack bean and soybean tRNA
Arg
ACG
were
synthesized as a single-stranded oligonucleotide and then
amplified by PCR using appropriate primers bearing a T7
promoter extension. Transcription at a 0.5 mL scale was
performed in T7 RNA polymerase buffer (40 mm Tris ⁄ HCl
pH 8, 12 mm MgCl
2

,5mm dithiothreitol, 1 mm spermidine
HCl, 4% polyethylene glycol 8000, 0.002% Triton X-100),
5mm NTP, 20 mm GMP, 0.1 units of inorganic pyrophos-
phatase (Sigma), 0.7 nmol template DNA, and 52 nm T7
RNA polymerase prepared from the recombinant pAR1219
expression plasmid [58]. Incubation was performed for 4 h
at 37 °C, and was followed by purification by NAP-5 gel
filtration (GE Healthcare), phenol extraction and ethanol
precipitation. Its homogeneity, as determined by denaturing
polyacrylamide gel electrophoresis, was greater than 80%.
The tRNA was refolded by heating to 70 °C in water, fol-
lowed by slow cooling in the presence of 25 mm Tris-HCl,
pH 7.5, 250 mm NaCl, 5 mm MgCl
2
.
Colorimetric detection of canavanine
Canavanine detection and quantification were achieved by
following its colour reaction with pentacyanoamidoferroate
(PCAF) (ICN Biomedicals, Aurora, OH, USA) [59]
using an ND-1000 photometer (NanoDrop Technologies,
Wilmington, DE, USA). To the canavanine-containing
sample in 10 lL was added 10 lL of 200 mm potassium
phosphate pH 7.5, 2 lL 1% potassium persulphate and
5 lL 1% PCAF in water. The colour was allowed to
develop for 40 min at room temperature and the absorbance
at 530 nm was measured.
Preparation of L-canaline
Synthesis of radioactive canavanine from l-canaline was
based on a previously described procedure [60] using [
14

C]-
cyanamide as a guanylating reagent. As l-canaline is no
longer commercially available, l-canavanine sulphate was
converted to l-canaline by arginase treatment, essentially as
described previously [61]. The arginase required for this
was obtained as a crude extract from the leaves of Canava-
lia brasiliensis. The extract enriched in arginase was used
immediately for preparative-scale conversion of canavanine
to canaline. Canaline was recovered from the reaction mix-
ture as its picrate salt, and converted to the free base as
described previously [61]. Elemental analysis indicated C
35.81% (calculated 35.82%), H 7.66% (calculated 7.51%),
N 19.43% (calculated 20.88%). Canaline was stored desic-
cated at )20 ° C.
Synthesis of [
14
C]-L-canavanine
[Guanidino-
14
C]-l-canavanine was synthesized essentially as
described previously [60] from 46 lmol canaline free base
and 2 mCi barium [
14
C]-cyanamide (57.5 mCiÆmmol
)1
,
34.8 mmol; Moravek, Brea, CA, USA). The required
pH adjustments were made using a micro pH electrode
(Metrohm, Filderstadt, Germany). Analysis by TLC on
G. L. Igloi and E. Schiefermayr Arginyl-tRNA synthetase amino acid discrimination

FEBS Journal 276 (2009) 1307–1318 ª 2009 The Authors Journal compilation ª 2009 FEBS 1315
silica (EtOH : AcOH : H
2
O, 65 : 1 : 34) gave a single
PCAF-reactive spot with 95% isotopic homogeneity, and
the canavanine-specific PCAF reaction showed the presence
of canavanine at 20 mm concentration containing a total of
1.4 mCi radioactivity (50 mCiÆ mmol
)1
). The stock solution
was stored at )70 °C in the presence of 2% ethanol.
Pyrophosphate exchange
Pyrophosphate exchange was monitored at 30 °C in the
presence of 50 mm Hepes ⁄ KOH pH 7.5, 10 mm MgCl
2
,
1.5 mm ATP, 80 lm [
32
P]-tetrasodium pyrophosphate
(Perkin-Elmer; specific activity in assay 5–10 cpmÆpmol
)1
),
together with amino acid, tRNA and enzyme in 50 l L reac-
tions. Radioactivity incorporated into ATP was quantified
by spotting 10 l L aliquots of the reaction onto 25 mm
diameter charcoal-impregnated filters (Type 69K) (Munk-
tell, Ba
¨
renstein, Germany) [62]. Filters were washed for
10 min in 1.5% perchloric acid ⁄ 40 mm pyrophosphate,

followed by rinsing with water, before being dried under
infrared lamps. Scintillation counting was performed using
Rotiszint (Roth, Karlsruhe, Germany).
Aminoacylation
Aminoacylation was performed at 30 °C in a volume of
50 lL containing 50 mm Hepes ⁄ KOH pH 7.5, 10 mm
MgCl
2
,4mm ATP and the appropriate amount of [
14
C]-
amino acid, tRNA and arginyl-tRNA synthetase. Amino
acid incorporation was followed using 3 MM filter discs
(Whatman, Dassel, Germany) that had been pretreated with
50 lL 5% trichloroacetic acid (to reduce non-specific back-
ground, particularly when using [
14
C]-canavanine) and dried.
Aliquots were spotted onto the discs which were then washed
with two changes of 5% trichloroacetic acid and once with
ethanol (10 min each), before being dried and quantified by
scintillation counting. Preparative aminoacylation reactions,
scaled to 100 lL, were allowed to reach a plateau, rapidly
extracted with phenol, and the aminoacylated tRNA was
collected by ethanol precipitation at pH 4.8.
Alternatively, the procedure described by Wolfson and
Uhlenbeck [28] to detect the incorporation of unlabelled
amino acids into [
32
P]-labelled tRNA was used. The tRNA

transcript was labelled with [a-
32
P]-ATP (111 TBqÆmmol
)1
)
(Perkin-Elmer) in the presence of yeast or E. coli tRNA
nucleotidyl transferase. Approximately 0.1 lCi tRNA and
0.35 nmol unlabelled tRNA was aminoacylated in a 10 lL
total volume containing 50 mm Hepes pH 7.5, 10 mm
MgCl
2
and 2.5 mm ATP together with aminoacyl-tRNA
synthetase and amino acids as indicated in the text. Aliqu-
ots (1 lL) were transferred to 4 lL 200 mm NaOAc pH 5
containing 0.4 units of nuclease P1 (Roche, Mannheim,
Germany). Digestion proceeded at room temperature for
15 min, after which 1 lL was spotted onto H
2
O-prewashed
polyethyleneimine cellulose TLC plates (Macherey & Nagel,
Du
¨
ren, Germany) that were developed in AcOH : 1 m
NH
4
Cl : H
2
O, 5 : 10 : 85. The stability of the aminoacyl-
tRNA link under the acidic conditions of nuclease treat-
ment was confirmed by separate experiments. Radioactivity

was detected by phosphorimager analysis (PharosFX
Plus; Bio-Rad, Munich, Germany), and quantified using
quantityone software (Bio-Rad). Kinetic constants were
calculated using sigmaplot (Systat, San Jose
´
, CA, USA).
Acknowledgements
This work was supported in part by the Deutsche
Forschungsgemeinschaft (Ig9 ⁄ 4). We thank Dr Gerald
Rosenthal for advice on the synthesis of [
14
C]-l-cana-
vanine.
References
1 Ibba M & So
¨
ll D (2000) Aminoacyl-tRNA synthesis.
Annu Rev Biochem 69, 617–650.
2 Geslain R & Ribas de Pouplana L (2004) Regulation of
RNA function by aminoacylation and editing? Trends
Genet 12, 604–610.
3 Pauling L (1958) The probability of errors in the pro-
cess of synthesis of protein molecules. In Festschrift
Arthur Stoll, pp. 597–602. Birkha
¨
user, Basel.
4 Baldwin AN & Berg P (1966) Transfer ribonucleic acid-
induced hydrolysis of valyladenylate bound to isoleucyl
ribonucleic acid synthetase. J Biol Chem 241, 839–845.
5 Eldred EW & Schimmel PR (1972) Rapid deacylation

by isoleucyl transfer ribonucleic acid synthetase of iso-
leucine-specific transfer ribonucleic acid aminoacylated
with valine. J Biol Chem 247, 2961–2964.
6 Fersht AR & Kaethner MM (1976) Enzyme hyperspeci-
ficity. Rejection of threonine by the valyl-tRNA synthe-
tase by misacylation and hydrolytic editing.
Biochemistry 15, 3342–3346.
7 von der Haar F & Cramer F (1976) Hydrolytic action
of aminoacyl-tRNA synthetases from baker’s yeast:
‘chemical proofreading’ preventing acylation of
tRNA(Ile) with misactivated valine. Biochemistry 15,
4131–4138.
8 Ataide SF & Ibba M (2006) Small molecules: big play-
ers in the evolution of protein synthesis. ACS Chem
Biol 1, 285–297.
9 Sekine S, Shichiri M, Bernier S, Cheˆ nevert R, Lapointe
J & Yokoyama S (2006) Structural bases of transfer
RNA-dependent amino acid recognition and activation
by glutamyl-tRNA synthetase. Structure 14, 1791–1799.
10 Hara-Yokoyama M, Yokoyama S & Miyazawa T
(1986) Conformation change of tRNAGlu in the com-
plex with glutamyl-tRNA synthetase is required for the
specific binding of l-glutamate. Biochemistry 25,
7031–7036.
Arginyl-tRNA synthetase amino acid discrimination G. L. Igloi and E. Schiefermayr
1316 FEBS Journal 276 (2009) 1307–1318 ª 2009 The Authors Journal compilation ª 2009 FEBS
11 Sherlin LD & Perona JJ (2003) tRNA-dependent active
site assembly in a class I aminoacyl-tRNA synthetase.
Structure 11, 591–603.
12 Uter NT, Gruic-Sovulj I & Perona JJ (2005) Amino

acid-dependent transfer RNA affinity in a class I ami-
noacyl-tRNA synthetase. J Biol Chem 280 , 23966–
23977.
13 Uter NT & Perona JJ (2006) Active-site assembly in
glutaminyl-tRNA synthetase by tRNA-mediated
induced fit. Biochemistry 45, 6858–6865.
14 Bullock TL, Rodrı
´
guez-Herna
´
ndez A, Corigliano EM &
Perona JJ (2008) A rationally engineered misacylating
aminoacyl-tRNA synthetase. Proc Natl Acad Sci USA
105, 7428–7433.
15 Fowden L, Lewis D & Tristram H (1967) Toxic amino
acids: their action as antimetabolites. Adv Enzymol
Relat Areas Mol Biol 29, 89–163.
16 Wink M (2006) Importance of plant secondary metabo-
lites for protection against insects and microbial infec-
tions. Adv Phytomed 3, 251–267.
17 Fowden L & Frankton JB (1967) The specificity of ami-
noacyl-sRNA synthetases with special reference to argi-
nine activation. Phytochemistry 7, 1077–1086.
18 Kitagawa M & Tomiyama T (1929) A new amino-
compound in the jack bean and a corresponding new
ferment. Biochem J 75, 618–620.
19 Rosenthal GA (1990) Metabolism of l-canavanine and
l-canaline in leguminous plants. Plant Physiol 94, 1–3.
20 Rosenthal GA (1991) The biochemical basis for the
deleterious effects of l-canavanine. Phytochemistry 30,

1055–1058.
21 Rosenthal GA, Lambert J & Hoffmann D (1989)
Canavanine incorporation into the antibacterial pro-
teins of the fly, Phormia terranovae (Diptera), and its
effect on biological activity. J Biol Chem 264, 9768–
9771.
22 Rosenthal GA & Dahlman DL (1991) Studies of
l-canavanine incorporation into insectan lysozyme.
J Biol Chem 266, 15684–15687.
23 Hogg J, Schiefermayr E, Schiltz E & Igloi GL (2008)
Expression and properties of arginyl-tRNA synthetase
from jack bean (Canavalia ensiformis). Protein Expr
Purif 61, 163–167.
24 Mitra K & Mehler AH (1966) The role of transfer ribo-
nucleic acid in the pyrophsphate exchange reaction of
arginine-transfer ribonucleic acid synthetase. J Biol
Chem 241, 5161–5162.
25 Lazard M, Agou F, Kerjan P & Mirande M (2000) The
tRNA-dependent activation of arginine by arginyl-
tRNA synthetase requires inter-domain communication.
J Mol Biol 302, 991–1004.
26 Guigou L & Mirande M (2005) Determinants in tRNA
for activation of arginyl-tRNA synthetase: evidence that
tRNA flexibility is required for the induced-fit mecha-
nism. Biochemistry 44, 16540–16548.
27 Han S, Moore RA & Viola RE (2002) Synthesis and
evaluation of alternative substrates for arginase. Bioorg
Chem 30, 81–94.
28 Wolfson AD & Uhlenbeck OC (2002) Modulation of
tRNAAla identity by inorganic pyrophosphatase. Proc

Natl Acad Sci USA 99, 5965–5970.
29 Igloi GL, von der Haar F & Cramer F (1978) Amino-
acyl-tRNA synthetases from yeast. The generality of
chemical proofreading in the prevention of misaminoa-
cylation of tRNA. Biochemistry 17, 3459–3468.
30 Rosenthal GA, Berge MA, Bleiler JA & Rudd TP
(1987) Aberrant, canavanyl protein formation and the
ability to tolerate or utilize l-canavanine. Experientia
43, 558–561.
31 Jakubowski H (1999) Misacylation of tRNALys with
noncognate amino acids by lysyl-tRNA synthetase.
Biochemistry 38, 8088–8093.
32 Levengood J, Ataide SF, Roy H & Ibba M (2004)
Divergence in noncognate amino acid recognition
between class I and class II lysyl-tRNA synthetases.
J Biol Chem 279, 17707–17714.
33 Cavarelli J, Delagoutte B, Eriani G, Gangloff J &
Moras D (1998) l-Arginine recognition by yeast argi-
nyl-tRNA synthetase. EMBO J 17, 5438–5448.
34 Geslain R, Bey G, Cavarelli J & Eriani G (2003)
Limited set of amino acid residues in a class Ia amino-
acyl-tRNA synthetase is crucial for tRNA binding.
Biochemistry 42, 15092–15101.
35 Delagoutte B, Moras D & Cavarelli J (2000) tRNA am-
inoacylation by arginyl-tRNA synthetase: induced con-
formations during substrates binding. EMBO J 19,
5599–5610.
36 Boyar A & Marsh RE (1982) l-Canavanine, a paradigm
for the structures of substituted guanidines. J Am Chem
Soc 104, 1995–1998.

37 Andriole EJ, Colyer KE, Cornell E & Poutsma JC
(2006) Proton affinity of canavanine and canaline,
oxyanalogues of arginine and ornithine, from the
extended kinetic method. J Phys Chem A 110, 11501–
11508.
38 Sissler M, Giege
´
R & Florentz C (1996) Arginine
aminoacylation identity is context-dependent and
ensured by alternate recognition sets in the anticodon
loop of accepting tRNA transcripts. EMBO J 15, 5069–
5076.
39 Farrow MA, Nordin BE & Schimmel P (1999) Nucleo-
tide determinants for tRNA-dependent amino acid dis-
crimination by a class I tRNA synthetase. Biochemistry
38, 16898–16903.
40 Ling C, Yao Y, Zheng YG, Wei H, Wang L, Wu XF &
Wang E-D (2005) The C-terminal appended domain of
human cytosolic leucyl-tRNA synthetase is indispens-
able in its interaction with arginyl-tRNA synthetase in
the multi-tRNA synthetase complex. J Biol Chem 280,
34755–34763.
G. L. Igloi and E. Schiefermayr Arginyl-tRNA synthetase amino acid discrimination
FEBS Journal 276 (2009) 1307–1318 ª 2009 The Authors Journal compilation ª 2009 FEBS 1317
41 Sylvers LA, Rogers KC, Shimizu M, Ohtsuka E & So
¨
ll
D (1993) A 2-thiouridine derivative in tRNAGlu is a
positive determinant for aminoacylation by Escherichia
coli glutamyl-tRNA synthetase. Biochemistry 32, 3836–

3841.
42 Wu XR, Kenzior AZ, Topin A, He SH, Acevedo A &
Folk WR (2007) Altered expression of plant lysyl tRNA
synthetase promotes tRNA misacylation and transla-
tional recoding of lysine. Plant J 50, 627–636.
43 Rosenthal GA (1977) The biological effects and mode
of action of l-canavanine, a structural analogue of
l-arginine. Q Rev Biol 52, 155–178.
44 Lue SW & Kelley SO (2007) A single residue in
leucyl-tRNA synthetase affecting amino acid
specificity and tRNA aminoacylation. Biochemistry
46, 4466–4472.
45 Ruan B, Palioura S, Sabina J, Marvin-Guy L, Kochhar
S, Larossa RA & So
¨
ll D (2008) Quality control despite
mistranslation caused by an ambiguous genetic code.
Proc Natl Acad Sci USA 105, 16502–16507.
46 Dale T & Uhlenbeck OC (2005) Amino acid specificity
in translation. Trends Biochem Sci 30, 659–665.
47 Cathopoulis TJ, Chuawong P & Hendrickson TL
(2008) Conserved discrimination against misacylated
tRNAs by two mesophilic elongation factor Tu ortho-
logs. Biochemistry 47, 7610–7616.
48 Rosenthal GA, Hughes CG & Janzen DH (1982) l-Ca-
navanine, a dietary nitrogen source for the seed predator
Caryedes brasiliensis (Bruchidae). Science 217, 353–355.
49 Allende CC & Allende JE (1964) Purification and sub-
strate specificity of arginyl-ribonucleic acid synthetase
from rat liver. J Biol Chem 239, 1102–1106.

50 Mitra SK & Mehler AH (1967) The arginyl transfer
ribonucleic acid synthetase of Escherichia coli. J Biol
Chem 242, 5490–5494.
51 Nazario M & Evans JA (1974) Physical and kinetic
studies of arginyl transfer ribonucleic acid ligase of
Neurospora. A sequential ordered mechanism. J Biol
Chem 249, 4934–4936.
52 Rosenthal GA & Harper L (1996) l-Homoarginine
studies provide insight into the antimetabolic properties
of l-canavanine. Insect Biochem Mol Biol 26, 389–394.
53 Bonfield JK, Smith KF & Staden R (1995) A new
DNA sequence assembly program. Nucleic Acids Res
23, 4992–4999.
54 Wu JF, Wang ED, Wang YL, Eriani G & Gangloff J
(1999) Gene cloning, overproduction and purification of
Escherichia coli tRNA(Arg)(2). Acta Biochim Biophys
Sin (Shanghai) 31, 226–232.
55 Sørensen AB, Duch M, Jørgensen P & Pedersen FS
(1993) Amplification and sequence analysis of DNA
flanking integrated proviruses by a simple two-step poly-
merase chain reaction method. J Virol 67, 7118–7124.
56 Steinmetz A & Weil J-H (1986) Isolation and character-
ization of chloroplast and cytoplasmic tRNAs. Methods
Enzymol 118, 212–231.
57 Tessier DC, Brousseau R & Vernet T (1986) Ligation
of single-stranded oligodeoxyribonucleotides by T4
RNA ligase. Anal Biochem 158, 171–178.
58 Davanloo P, Rosenberg AH, Dunn JJ & Studier FW
(1984) Cloning and expression of the gene for bacterio-
phage T7 RNA polymerase. Proc Natl Acad Sci USA

81, 2035–2039.
59 Rosenthal GA (1977) Preparation and colorimetric
analysis of l-canavanine. Anal Biochem 77, 147–151.
60 Ozinskas AJ & Rosenthal GA (1986) l-Canavanine syn-
thesis by zinc-mediated guanidination of l-canaline with
cyanamide. Bioorg Chem 14, 157–162.
61 Bass M, Harper L, Rosenthal GA, Phuket SN &
Crooks PA (1995) Large-scale production and chemical
characterization of the higher plant allelochemicals:
l-canavanine and l-canaline. Biochem Syst Ecol 23,
717–721.
62 Simlot MM & Pfaender P (1973) Amino acid dependent
ATP-
32
PPi exchange measurement. A filter paper disk
method. FEBS Lett 35, 201–203.
Arginyl-tRNA synthetase amino acid discrimination G. L. Igloi and E. Schiefermayr
1318 FEBS Journal 276 (2009) 1307–1318 ª 2009 The Authors Journal compilation ª 2009 FEBS