ABSTRACT
In many organisms (e.g., gram-positive eubacteria) Gln-tRNA is not formed by direct glutaminylation of tRNA
Gln
but by a specific transamidation of Glu-tRNA
Gln
. We wondered whether a similar transamidation pathway also operates in the
formation of Asn-tRNA in these organisms. Therefore we tested in S-100 preparations of
Lactobacillus bulgaricus
, a gram-positive eubacterium, for the conversion by an amidotransferase of [
14
C]Asp-tRNA to [
14
C]Asn-tRNA. As no transamidation was observed, we searched for genes for
asparaginyl-tRNA synthetase (AsnRS). Two DNA fragments (from different locations of
the
L.bulgaricus
chromosome) were found each containing an ORF whose sequence resembled that of
the
Escherichia coli
asnS
gene. The derived amino acid sequences of the two ORFs (432 amino acids) were
the same and 41% identical with
E.coli
AsnRS. When one of the ORFs was expressed in
E.coli
, it complemented the temperature sensitivity of an
E.coli asnS
mutant. S-100 preparations of this transformant showed increased charging of
unfractionated
L.bulgaricus
tRNA with asparagine. Deletion of the 3
'
-terminal region of the
L.bulgaricus
AsnRS gene led to loss of its complementation and aminoacylation properties.
This indicates that
L.bulgaricus
contains a functional AsnRS. Thus, the transamidation pathway operates only for
Gln-tRNA
Gln
formation in this organism, and possibly in all gram-positive eubacteria.
It is generally assumed that a minimum of 20 aminoacyl-tRNA synthetases exist in the cell for the synthesis of the aminoacyl-tRNAs for 21 amino acids (including selenocysteine). This is
certainly true for many organisms as evidenced by the biochemical and protein
sequence information of all
Escherichia coli
aminoacyl-tRNA synthetases (see e.g.,
1
) and the knowledge of the genomic sequences of
Haemophilus influenzae
(
2
) and
Saccharomyces cerevisiae
(
3
). However, in a large part of the living kingdom (e.g., gram-negative eubacteria, archaebacteria, in chloroplasts and mitochondria)
only 19 aminoacyl-tRNA synthetases are found as they lack glutaminyl-tRNA synthetase (GlnRS) (
4
,
5
). In these organisms/organelles the formation of Gln-tRNA
Gln
requires the presence of two enzymes, a glutamyl-tRNA synthetase (GluRS) that misacylates tRNA
Gln
with glutamate and a Glu-tRNA
Gln
-specific amidotransferase. In the presence of an amido donor this enzyme
amidates glutamate bound to tRNA to form the correct Gln-tRNA
Gln
(
6
-
8
). There is as yet no explanation why nature developed two pathways (direct aminoacylation or transamidation) for the
formation of Gln-tRNA
Gln
. There may be metabolic reasons why certain organisms prefer one route over the
other. The pathway may also have arisen to compensate for the absence of GlnRS
if this was the last aminoacyl-tRNA synthetase formed (from the duplication of the gene encoding GluRS)
in evolution (
9
).
The question of whether a similar transamidation pathway is also active in the
formation of asparaginyl-tRNA has never been examined. Although
asnS
, the gene encoding asparaginyl-tRNA synthetase (AsnRS), has been cloned from a number of organisms [e.g.,
E.coli
(
10
,
11
)], there is no information for gram-positive bacteria whether the biosynthesis of Asn-tRNA
Asn
involves direct asparaginylation or a misacylation/transamidation pathway.
Therefore we searched for the biochemical transamidation of Asp-tRNA in the gram-positive bacterium,
Lactobacillus bulgaricus
, which utilizes the transamidation pathway for Gln-tRNA formation (
12
). Here we report that we could not detect the transamidation reaction, but
instead found two genes encoding AsnRS in this organism.
Uniformly labeled [
3
H]asparagine (1 Ci/mmol) was custom- synthesized by American Radiochemicals, St Louis, MO. [
14
C]Aspartate (53.8 mCi/mmol) was from DuPont. Unfractionated
L.bulgaricus
tRNA was prepared as described (
12
). A mixture of
L.bulgaricus
aminoacyl-tRNA synthetases was a dialyzed S-100 preparation (
12
).
Lactobacillus bulgaricus
ATCC 11842 was grown in MRS medium as reported previously (
12
).
Escherichia coli
strain CGSC6340, also known as HO202 (
13
), is temperature-sensitive because of a mutation in
asnS
, and was grown in LB medium. For the growth of transformants, 100 [mu]g/ml of ampicillin (Amp) and/or 50 [mu]g/ml of kanamycin (Km) were added to the medium. Plasmids pET15-b and pGP1-2 (
14
) were used for the over- expression of the cloned gene with a coupled T7 RNA polymerase/T7
promoter. The
E.coli-Bacillus subtilis
shuttle vector pRB395 (
15
) was used for the constitutive expression of the
L.bulgaricus
asnS
gene by the
veg
II promoter of
B.subtilis
(
16
).
Cultured
L.bulgaricus
ATCC11842 (500 ml) was harvested and resuspended in 50 ml of buffer A (30 mM
HEPES-KOH, pH 7.0, 25 mM KCl, 15 mM MgCl
2
, 4 mM DTT and 0.1 mM PMSF). The resuspended cells were incubated for 15 min at
37oC after addition of 50 mg of lysozyme and disrupted with a French Press
under a pressure of 12 000 p.s.i. at the rate of ~25 ml/min. After disruption, crude lysate was centrifuged at 27 000
g
for 30 min in an SS34 rotor. The supernatant (40 ml) was mixed with 10 ml of
glycerol and stored at -20oC. For
E.coli
transformants, the frozen cell pellet from a 50 ml culture was resuspended in 5
ml of 50 mM Tris-HCl, pH 7.6 containing 10 mM MgCl
2
. Following sonication, the crude lysates were clarified by centrifugation at 27
000
g
for 30 min in an SS34 rotor.
The search for transamidation of Asp-tRNA to Asn-tRNA was based on a two-step reaction. The first step was formation of
L.bulgaricus
[
14
C]Asp-tRNA which was performed at 37oC for 30 min in 100 [mu]l of 0.1 M HEPES-KOH pH 7.5, 30 mM KCl, 10 mM MgCl
2
, 5 mM ATP and 1 [mu]mol of aspartic acid mix (100 nmol of [
14
C]Asp and 900 nmol of cold Asp), 10
A
260
U of partially fractionated tRNAs and 10 [mu]g of
L.bulgaricus
S-100 extract protein. The reaction was stopped by addition of 10 [mu]l of 3 M sodium acetate (pH 5.0) and 100 [mu]l of phenol (pH 5.0). The charged tRNAs were precipitated with 2.5
vol of cold ethanol and washed twice with 70% ethanol.
The precipitated Asp-tRNAs were then dissolved in 40 [mu]l of HEPES-KOH pH 7.5, 2.5 mM asparagine (or glutamine or ammonium
chloride) as amino group donor, and 3 mM ATP. The putative transamidation
reaction was started by adding 10 [mu]g of undialyzed S-100 extract protein and then incubated for up to 60 min at 37oC. The reaction was stopped by phenol extraction, and the
radioactive aminoacyl-tRNAs precipitated with ethanol and washed well with ethanol. The tRNA was
deacylated dissolving the precipitate with 50 [mu]l of 0.01 N NaOH and incubation at room temperature for 10 min. The
resulting solution was dried completely, taken up in 3 [mu]l of distilled water and spotted onto a cellulose thin layer plate.
Chromatography was performed in methanol-chloroform-ammonia-water (6:6:2:1) and the
14
C-labeled amino acids on the dried TLC plate were visualized on a
phosphoimager (Fusix Bas 2000). The R
f
values were: Asn (0.51) and Asp (0.43). Control reactions were run without tRNA
or amino donors.
Asparaginyl tRNA synthetase activity was measured in a reaction mixture (0.1 ml)
consisting of 50 mM Tris-HCl, pH 7.6, 10 mM MgCl
2
, 0.2 mM spermidine and 15
A
260
U of unfractionated
L.bulgaricus
tRNA, 2 mM ATP and 60 [mu]M of asparagine (10 [mu]M of [
3
H]Asn and 50 [mu]M of cold Asn).
The reaction mixtures and sample extracts were both preincubated for 10 min at
43oC to destroy the AsnRS activity of the
E.coli
host cell. Then, the aminoacylation reaction was started by the addition of ~5 [mu]g protein (the enzyme was in a dialyzed S-100 fraction of
E.coli
strain transformed with the
L.bulgaricus
asnS
gene). Samples were removed at the times indicated and spotted on Whatman 3MM
filter discs, which were then washed and the radioactivity measured by liquid
scintillation counting (
17
).
In analogy to the reactions elaborated in the transamidation of Glu-tRNA
Gln
we charged unfractionated
L.bulgaricus
tRNA with [
14
C]aspartate using a
L.bulgaricus
aminoacyl-tRNA synthetase preparation. The isolated Asp-tRNA was then added to a crude S-100 extract in the presence of the known amido group donors
(asparagine, glutamine or ammonium chloride). As described in Materials and
Methods, after the reaction the aminoacyl-tRNA was isolated and then deacylated with KOH to release the bound amino
acid(s). Instead of using Dowex chromatography (
5
), the amino acid mixture was then separated by thin layer chromatography. The
results showed very clearly that there was no conversion of the radioactive Asp
into Asn. As a control the reaction was run with Glu-tRNA; as observed earlier, a facile transamidation was seen in the
presence of glutamine as nitrogen donor. While these results do not prove the
absence of a transamidation reaction for Asn-tRNA formation, we wanted to see if the organism could directly attach Asn
to tRNA.
In a random sequence analysis study of
L.bulgaricus
ATCC11842 DNA two clones (5.1 kb
Bam
HI-
Sma
I DNA fragment for
asnS1
and 2.5 kb
Bam
HI DNA fragment for
asnS2
) were found whose ORFs showed great sequence homology to AsnRS of
E.coli
. The corresponding DNA sequences are deposited in EMBL data library (accession
no. X89438 for
asnS1
and X89439 for
asnS2
). The nucleotide sequences of the two ORFs differ only in 2 nt and they make an
identical derived 432 amino acid sequence (mol. wt 49 971 Da). Using an
internal probe, Southern blot experiments with whole genomic DNA showed the
presence of two
asnS
genes in
L.bulgaricus
. This is in line with several other examples of two genes for the same
aminoacyl-tRNA synthetase in gram-positive organisms (
18
), in contrast with the situation in gram-negative eubacteria (
1
,
4
) where there is usually only one gene. The two
asnS
genes may be differently regulated, as one of them may be transcribed and
regulated together with asparagine synthetase by a transcription
antitermination mechanism (
19
).
The deduced amino acid sequence of
L.bulgaricus
AsnRs has high homology (41% identity) with AsnRS of
E.coli
(
10
,
11
) and a few other bacteria (Fig.
1
). It is most closely related (57% identity) to
B.subtilis
AsnRS (based on a recently deposited nucleotide sequence (
20
). The identities with the corresponding enzyme from other organism is:
H.influenzae
(39%) (
2
),
Thermus thermophilus
(56%) (
21
),
Synechocystis
(36%) (
22
),
Mycoplasma genitalium
(35%) (
3
) and
Brugia malayi
(31%) (
23
,
24
). In addition, as is well known (
1
), the AsnRS also shows great homology to aspartyl- and lysyl-tRNA synthetases (e.g.,
1
).
To examine whether the protein expressed from the cloned
L.bulgaricus
gene can complement the temperature-sensitive mutation of AsnRS in
E.coli
CGSC6340, this strain was transformed with three plasmids, pRBLN (containing
the complete
L.bulgaricus
asnS
gene), pRB395 (vector only) and pRBL[Delta]K (truncation of 20% of the
L.bulgaricus
asnS
gene resulting in loss of the 3'-terminal part). Only the transformant CGSC6340/pRBLN could grow at
42oC (data not shown). These results demonstrate that the complete
L.bulgaricus
asnS
gene provides the
E.coli
strain with thermostable AsnRS activity, while a mutant
asnS
gene or the absence of the gene will not do.
When cell extracts were prepared from the three transformed strains and AsnRS
activity was determined, the results in Figure
2
were obtained. Significant activity was only found in the extract of
CGSC6340/pRBLN whereas the other extracts had about background activity. As
mentioned in Materials and Methods, the activity of the endogenous
E.coli
CGSC6340 mutant AsnRS was destroyed by preincubation at 43oC. These results demonstrate that the cloned
L.bulgaricus
asnS
gene encodes an active enzyme.
Our inability to demonstrate conversion of Asp-tRNA to Asn-tRNA and the finding of two
asnS
genes in
L.bulgaricus
makes it clear that Asn-tRNA in this organism is formed by direct aminoacylation of tRNA. This may
also occur in other gram-positive eubacteria; this idea is supported as
B.subtilis
also possesses an
asnS
gene based on a recently published DNA sequence (
20
). While there is still the possibility that other organisms may use tRNA-dependent amino acid transformations as a way to generate Asn-tRNA, these results clearly demonstrated that transamidation is not
an
a priori
requirement for the formation of amide-aminacyl tRNAs in
Bacilli
.
We are grateful to Michael Ibba for discussions and Andrea Pfeifer for her
encouragement. This work was supported by a grant from NIH.
REFERENCES
Return