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
Trang 1revealed 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: igloi@biologie.uni-freiburg.de
(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-tRNAArgwhen using RNA transcribed from the jack bean tRNAACGgene The corresponding kcat⁄ KM 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 tRNAArg 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.
Trang 2resembles 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 eukprok-aryotes (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 tRNAArg
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 tRNAACG (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 tRNAArgACG 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 KMvalue that is equivalent to that of
HN
O
L-Arginine
NH2
NH2
NH2
NH2
O
OH O
L-Canavanine
Fig 1 Structures of L -arginine and its guanidinooxy analogue,
L -canavanine.
Trang 3tRNA as measured by aminoacylation (data not shown)
Using either [14C]-canavanine in the conventional aminoacylation assay, or unlabelled canavanine together with [32P]-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 [32P]-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 10 12 14 16
–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 l M ( ) or 30 l M (r) transcript tRNA
or 12 l M (d) periodate-oxidized jack bean transcript tRNA using jack
bean arginyl-tRNA synthetase PPi, tetrasodium pyrophosphate.
Trang 4l-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 (KM for soybean
56 lm; KMfor 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, (kcat⁄
KM)Arg⁄ (kcat⁄ KM)Cav, of 44 (Table 1) A similar factor
was obtained when assayed with non-radioactive
cana-vanine using the [32P]-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 [32P]-labelled tRNA assay and higher concen-trations of canavanine, a KM for this substrate of 1.3 mm was determined, and the relative magnitude of the kcat⁄ KM 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-tRNAArg by the
L
A
-
g r
i n
i n
e
L
H
-
o
m
o
g r
i n
e
L
h
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
i t r u
l l i n e
L
h
i o
c
i t r u
l l i n e
L
L -
s
i n
e
L
- A
l b
z
i i n e
L
C-an a a in e
Jack bean enzyme Soybean enzyme
0
20
40
60
80
100
120
NH
NH
N
2
NH 2
OH
O
NH
O
NH
N
2
NH 2
OH
NH
O
S
N
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
2
O
NH2
NH2
NH2
OH O OH
NH O
NH 2
N
N
2
NH 2
O
OH
O
tRNA
Origin Aminoacyl-A76
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 m M 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.
Trang 5jack bean enzyme Cav-tRNAArg 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-tRNAArg even in the absence of either enzyme
In contrast, the half life of Arg-tRNAArg 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 tRNAArgACG (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 [14C]-labelled amino acids or [32P]-labelled tRNA.
Source of
enzyme
Assay method
Aminoacylation of transcript tRNA with
[14C]-labelled amino acid
Aminoacylation of [ 32 P]-labelled transcript tRNA
Discrimination factor(kcat⁄ K M )Arg⁄ (k cat ⁄ K M )Cav
KM(l M )
kcat⁄ K M ( M )1Æmin)1) K
kcat⁄ K M ( M )1Æmin)1) K
kcat⁄ K M ( M )1Æmin)1) K
kcat⁄ K M ( M )1Æmin)1)
58 ([ 32 P]-labelled tRNA)
a ND, not determined because of the impracticality of using large amounts of [ 14 C]-Cav.
Time (min)
0
0 5 10 15 20 25 30 35
20
40
60
80
100
120
Fig 5 Stability of canavanyl-tRNA Jack bean transcript tRNAArg
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
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.
Trang 6E colitRNA 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 tRNAArg E coli transcript Jack bean transcript Soybean transcript
Aminoacylation
(pmol ArgÆpmol)1
Aminoacylation (pmol ArgÆpmol)1
Aminoacylation (pmol ArgÆpmol)1
Aminoacylation (pmol ArgÆpmol)1
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-Arg
E.coli
transcript
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.
Trang 7bean 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 pKaof the guanidine group, lowering the
value from 12.5 by more than 5 pKaunits [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
tRNAArg
ACGgenes 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 KMfor 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 kcat⁄ KMvalues result
in discrimination factors of approximately 40 and 485
for the respective enzymes
However, in a heterologous system using either
native E coli tRNAArgICG 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 tRNAArg 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-tRNAArgby the jack bean enzyme Cav-tRNAArg 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 tRNAArgin vitroand incorporation of this allelochemi-cal into proteins in vivo [30,43] However, the several hundred-fold discrimination measured for the jack
Trang 8bean enzyme is considerably lower than the factor of
104 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 kcat⁄ KM
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 vivoprocesses 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
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
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
Karlsruhe, Germany) Sequences were identified by blast comparison (http://www.ncbi.nlm.nih.gov/blast/Blast.cgi)
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
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)
Trang 9Gene for tRNAArgACGfrom 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
(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 lL of
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
5 min, followed by RNase H treatment (GE Healthcare)
amplified using the universal M13 primer, which binds to
the 3¢ tail of the RNA, and an 18-mer based on the 5¢
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
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
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
HCl, 4% polyethylene glycol 8000, 0.002% Triton X-100),
5 mm 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
filtration (GE Healthcare), phenol extraction and ethanol precipitation Its homogeneity, as determined by denaturing polyacrylamide gel electrophoresis, was greater than 80%
fol-lowed by slow cooling in the presence of 25 mm Tris-HCl,
Colorimetric detection of canavanine Canavanine detection and quantification were achieved by following its colour reaction with pentacyanoamidoferroate
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 ofL-canaline Synthesis of radioactive canavanine from l-canaline was
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
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
Synthesis of [14C]-L-canavanine
described previously [60] from 46 lmol canaline free base
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
Trang 10silica (EtOH : AcOH : H2O, 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
Pyrophosphate exchange
together with amino acid, tRNA and enzyme in 50 lL
reac-tions Radioactivity incorporated into ATP was quantified
by spotting 10 lL aliquots of the reaction onto 25 mm
diameter charcoal-impregnated filters (Type 69K)
(Munk-tell, Ba¨renstein, Germany) [62] Filters were washed for
followed by rinsing with water, before being dried under
infrared lamps Scintillation counting was performed using
Rotiszint (Roth, Karlsruhe, Germany)
Aminoacylation
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
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
(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
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
polyethyleneimine cellulose TLC plates (Macherey & Nagel,
Du¨ren, Germany) that were developed in AcOH : 1 m
aminoacyl-tRNA link under the acidic conditions of nuclease treat-ment was confirmed by separate experitreat-ments Radioactivity
Plus; Bio-Rad, Munich, Germany), and quantified using
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, Cheˆnevert 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