A novel L-amino acid ligase from bacillus subtilis NBRC3134 catalyzed oligopeptide synthesis

Article abstract of DOI:10.1271/bbb.90649

L-Amino acid ligase catalyzes dipeptide synthesis from unprotected L-amino acids in an ATP-dependent manner. We have purified a new L-amino acid ligase, RizA, which synthesizes dipeptides whose N-terminus is Arg, from Bacillus subtilis NBRC3134, a microorganism that produces a rhizocticin peptide antibiotic. It was suggested that RizA is probably involved in rhizocticin biosynthesis. In this study, we performed sequence analysis of unknown regions around rizA, and newly identified a gene that encodes a protein that possesses an ATP-grasp motif upstream of rizA. This gene was designated rizB, and its recombinant protein was prepared. Recombinant RizB synthesized homo-oligo-mers of branched-chain L-amino acids and L-methionine consisting of two to five amino acids in an ATP-dependent manner. RizB also synthesized various heteropeptides. Further examination showed that RizB might elongate a peptide chain at the N-terminus. This is the first report on an L-amino acid ligase catalyzing oligopeptide synthesis.

Full text of DOI:10.1271/bbb.90649

Biosci. Biotechnol. Biochem., 74 (1), 129–134, 2010
A Novel L-Amino Acid Ligase from Bacillus subtilis
NBRC3134 Catalyzed Oligopeptide Synthesis
Kuniki KINO,y Toshinobu ARAI, and Daisuke TATEIWA
Department of Applied Chemistry, Faculty of Science and Engineering, Waseda University,
3-4-1 Ohkubo, Shinjuku-ku, Tokyo 169-8555, Japan
Received September 3, 2009; Accepted October 3, 2009; Online Publication, January 7, 2010
[doi:10.1271/bbb.90649]
L-Amino acid ligase catalyzes dipeptide synthesis
from unprotected L-amino acids in an ATP-dependent
manner. We have purified a new L-amino acid ligase,
RizA, which synthesizes dipeptides whose N-terminus is
Arg, from Bacillus subtilis NBRC3134, a microorganism
that produces a rhizocticin peptide antibiotic. It was
suggested that RizA is probably involved in rhizocticin
biosynthesis. In this study, we performed sequence
analysis of unknown regions around rizA, and newly
identified a gene that encodes a protein that possesses an
ATP-grasp motif upstream of rizA. This gene was
designated rizB, and its recombinant protein was
prepared. Recombinant RizB synthesized homo-oligo-
mers of branched-chain L-amino acids and L-methionine
consisting of two to five amino acids in an ATP-
dependent manner. RizB also synthesized various
heteropeptides. Further examination showed that RizB
might elongate a peptide chain at the N-terminus. This
is the first report on an L-amino acid ligase catalyzing
oligopeptide synthesis.
ligase has been proposed: (i) an amino acid serves as an
N-terminal substrate and undergoes initial phosphoryla-
tion at its carboxyl group, and an aminoacyl phosphate is
synthesized as the reaction intermediate; (ii) the amino-
acyl phosphate is then nucleophilically attacked by the
other amino acid; and (iii) phosphate is released,
resulting in the dipeptide.1) The synthesis of aminoacyl
phosphate as a reaction intermediate is the common step
among the enzymes belonging to the ATP-dependent
carboxylate-amine/thiol ligase superfamily.10,11)
Recently, we reported a new L-amino acid ligase,
RizA, from Bacillus subtilis NBRC3134 (ATCC6633), a
microorganism that produces the rhizocticin peptide
antibiotic.12–14) Depending on chemical structure, four
types of rhizocticins have been reported: rhizocticin A:
L-arginyl-L-2-amino-5-phosphono-3-cis-pentenoic acid
(Arg-APPA); rhizocticin B: L-valyl-L-arginyl-L-2-ami-
no-5-phosphono-3-cis-pentenoic acid (Val-Arg-APPA);
and rhizocticin C and D, which are the same as
rhizocticin B but contain L-isoleucine (Ile) and L-leucine
(Leu) respectively in place of Val.15) We purified RizA
from B. subtilis NBRC3134 by detecting L-arginine
hydroxamate synthesis activity. It was found that RizA
catalyzed Arg-Xaa dipeptide synthesis (Xaa, arbitrary
amino acids) but not the synthesis of tripeptides.13)
Hence RizA might be involved in the biosynthesis of
rhizocticin A, and other enzymes might be required for
the synthesis of tripeptides such as rhizocticin B, C, and
D. In addition, we analyzed the DNA sequences of
unknown regions around rizA, because we considered
that the genes involved in rhizocticin biosynthesis might
form a cluster, as observed in the case with other
antibiotics.4,16,17) We have already identified three genes
encoding proteins that might catalyze the synthesis of
C-P compounds such as APPA.13)
Key words: L-amino acid ligase; oligopeptide synthesis;
rhizocticin
L-Amino acid ligase (EC 6.3.2.28) synthesizes dipep-
tides from unprotected L-amino acids in an ATP-
dependent manner.1) The YwfE protein from Bacillus
subtilis 168 was the first reported L-amino acid ligase,2–4)
and new members of it have been reported, including
the RSp1486a protein from Ralstonia solanacearum
JCM10489,5) the BL00235 protein from Bacillus lichen-
iformis NBRC12200,6) and the PSPPH 4299 protein
from Pseudomonas syringae pv. phaseolicola 1448A.7,8)
These proteins were obtained by in silico analysis using
the ATP-grasp motif, a signature of the ATP-dependent
carboxylate-amine/thiol ligase superfamily,9) as one
of the indexes; their identities were determined by
amino acid sequence analysis of 18–26% of the protein.
L-Amino acid ligase is very efficient in the production of
peptides,3) but currently known L-amino acid ligases
synthesize only dipeptides. Hence a novel L-amino acid
ligase capable of catalyzing oligopeptide synthesis is
required to increase the variety of peptides currently
being synthesized using L-amino acid ligases.
In the present study, sequence analysis of unknown
regions around rizA was performed, and we found a
novel L-amino acid ligase, RizB, capable of catalyzing
oligopeptide synthesis. Furthermore, we examined
several characteristics of RizB.
Materials and Methods
Materials. Bacillus subtilis NBRC3134 (ATCC6633) was from the
NITE Biological Resource Center (Chiba, Japan). Escherichia coli
BL21(DE3) and pET-21a(+) vector were from Merck (Darmstadt,
Germany). DNA cassettes for sequence analysis were from Takara Bio
(Shiga, Japan). All the other chemicals used are commercially
available and were of chemically pure grade.
L-Amino acid ligase belongs to the ATP-dependent
carboxylate-amine/thiol ligase superfamily and gener-
ally requires ATP and Mg2þ for peptide synthesis. A
mechanism of dipeptide synthesis by L-amino acid
y
To whom correspondence should be addressed. Fax: +81-3-3232-3889; E-mail: kkino@waseda.jp

130
K. KINO et al.
Sequence analysis and gene annotation. DNA manipulations were
(Methyl-Leu), L-arginine-hydroxamate (Arg-NHOH), and L-phenyl-
alanine methyl ester (Phe-OMe) were used as N- and C-protected
amino acid substrates. To test reactivity with dipeptides, a reaction was
carried out with L-valyl-L-valine (Val-Val) as the substrate. In addition,
a reaction with a combination of Val and L-arginyl-L-serine (Arg-Ser)
was conducted.
performed by the methods of Sambrook et al.,18) with minor
modifications. To analyze the DNA sequences of unknown regions
around rizA, a cassette polymerase chain reaction (PCR) method was
used.13,19) Sequence analysis was performed by the primer-walking
method with primers designed on the basis of sequencing results. To
predict the function of the protein encoded by each gene, a Pfam
search20) and a BLAST search21) were performed. To predict the
transcriptional start site, reverse transcription-PCR was performed
using the ReverTra-Plus kit (Toyobo, Osaka, Japan). The following
primers were used: R1, (50-CTCTTTTATTTGAGAGAACACGTTG-
30); F1, (50-TAATGCTGACTTCAACTGAACG-30); F2, (50-GCGAT-
TAGCATTTTAATACTGAACA-30); and F3, (50-GGACAGGAAA-
TGTCTTCTGATG-30). Total RNA was extracted from B. subtilis
NBRC3134 cultivated in the production medium for 2 d.13,22)
To examine the time course of peptide synthesis, a reaction mixture
containing 0.1 mg/ml of RizB, 20 mM Val, 50 mM ATP, and 25 mM
.
MgSO4 7H2O in 100 mM Tris–HCl buffer (pH 9.0) was prepared, and
this was incubated at 37 ꢀC. The reaction mixture was analyzed by
HPLC.
To characterize RizB, ATP was replaced with GTP, CTP, or TTP,
.
.
.
and MgSO4 7H2O was replaced with MnSO4 5H2O, CoSO4 7H2O,
.
.
.
ZnSO4 7H2O, FeSO4 7H2O, or CaSO4 2H2O. Val was used as the
substrate in these reactions, and the reaction mixtures were analyzed by
LC-ESI-MS. To determine optimum pH, 100 mM Tris–HCl buffer
(pH 7.0–10.0) and 100 mM borate buffer (pH 9.0–11.0) were used, and
the reaction with Val as a substrate was performed at 30 ꢀC for 1 h. To
determine the optimum temperature, the reaction temperature was set
between 20 ꢀC and 60 ꢀC, and the reaction with Val as substrate was
performed in 100 mM Tris–HCl buffer (pH 8.0) for 1 h. In each case,
optimal conditions were determined by measuring the amounts of
L-valyl-L-valyl-L-valine (Val-Val-Val) synthesized in each reaction
mixture.
Overexpression of rizB and preparation of RizB. The gene encoding
a putative L-amino acid ligase was designated rizB (DDBJ accession
no. AB467270). rizB-P1 and rizB-P2 were amplified from genomic
DNA of B. subtilis NBRC3134 by PCR, and the following primers
were used (restriction sites are underlined): P1 sense, (NdeI, 50-
GGAATTCCATATGCTGACTTCAACTGAACG-30); P1 anti, (Bam-
HI, 50-CGCGGATCCGATTTGAGAGAACACGTTGA-30); P2 sense,
(NdeI, 50-GGAATTCCATATGAGCATTTTAATACTGAACAAAA-
CCT-30); and P2 anti, (BamHI, 50-CGCGGATCCGATTTGAGAGAA-
CACGTTGA-30). The PCR fragments were digested with appropriate
restriction enzymes and then ligated into the pET-21a(+) expression
vector. The resulting plasmid was designed to express the gene with a
C-terminal His-tag sequence under the control of the T7 promoter. The
plasmids were introduced into E. coli BL21(DE3).
Results
Search for a gene encoding L-amino acid ligase
We formulated the hypothesis that the genes involved
in rhizocticin biosynthesis form a cluster, and that a gene
in this cluster encodes a novel L-amino acid ligase
that catalyzes tripeptide synthesis. Hence the DNA
sequences of unknown regions around rizA were
analyzed. We found a novel gene, located about
9,000 bp upstream of rizA, that encodes a protein
possessing an ATP-grasp motif. The gene starts from
an ATG codon and codes for 380 amino acid residues.
This open reading frame (ORF) was subsequently
designated rizB-P1 (Fig. 1, position 1). A BLAST
search additionally showed that the amino acid sequence
of RizB-P1 had 65% identity to that of a hypothetical
protein, BL02410 (DDBJ accession no. CP000002) in
B. licheniformis ATCC14580. The BL02410 protein has
405 amino acid residues and also possesses an ATP-
grasp motif. Alignment of the two amino acid sequences
showed the high degree of sequence conservation in the
upstream region of RizB-P1 and the N-terminal region
of the BL02410 protein (Fig. 1). Thus the ORF starting
from the ATT codon, which corresponds to the initial
Met residue of the BL02410 protein, was designated
rizB-P2 (Fig. 1, position 2). ATT codons generally code
for Ile, but some strains of bacteria and phages use ATT
as a start codon.25–29)
Furthermore, transcriptional analysis was performed
to predict the transcriptional start site of this gene.
Primers were designed as follows: R1 was designed on
the basis of the common C-terminal region of rizB-P1
and rizB-P2; F1 was designed on the basis of the region
containing the ATG codon of position 1; F2 was
designed on the basis of the region containing the
ATT codon of position 2; and F3 was designed on the
basis of an approximately 1,500-bp region upstream of
rizB-P2. RT-PCR analysis, however, showed that proper
DNA fragments were amplified in all cases (data not
shown), which suggests that this gene was transcribed
with a forward gene and might be transcribed as an
operon. In addition, sequence analysis was performed to
Recombinant E. coli cells were cultured in 3 ml of Luria-Bertani
medium (1% Bacto-tryptone, 0.5% yeast extract, 1% NaCl) contain-
ing 100 mg/ml of ampicillin (final concentration) at 37 ꢀC for 5 h with
shaking at 160 rpm. Cultivated cells were transferred to 100 ml of
fresh Luria-Bertani medium that contained 100 mg/ml of ampicillin
(final concentration) and were incubated at 37 ꢀC for 1 h with shaking
at 120 rpm. Isopropyl-ꢀ-D-thiogalactopyranoside (final concentration,
0.1 mM) was then added, and cultivation was continued at 25 ꢀC
for 19 h with shaking at 120 rpm. The cells were harvested by
centrifugation (4;160 ꢁ g, 10 min, 4 ꢀC), resuspended in 100 mM Tris–
HCl buffer (pH 8.0), and then disrupted by sonication at 4 ꢀC.
Cellular debris were removed by centrifugation (20;000 ꢁ g, 30 min,
4 ꢀC), and the supernatant was collected and purified with a HisTrap
HP Ni-affinity column (GE Healthcare, Buckinghamshire, UK). The
active fraction was then desalted with
a PD-10 column (GE
Healthcare) equilibrated with 100 mM Tris–HCl buffer (pH 8.0). To
confirm the presence of protein in the solution, sodium dodecyl
sulfate-polyacrylamide gel electrophoresis (SDS–PAGE) was per-
formed by the method of Laemmli.23) The protein concentration was
determined by the Bradford method, with bovine serum albumin as
the standard.
Characterization of RizB. The L-amino acid ligase activity of RizB
was assayed as follows, unless otherwise specified: The standard
reaction mixture (total volume, 0.3 ml) contained 0.1 mg/ml of RizB,
.
12.5 mM ATP, 12.5 mM MgSO4 7H2O, and 12.5 mM amino acid
substrates in 100 mM Tris–HCl buffer (pH 8.0). The reaction was
performed at 30 ꢀC for 20 h. To detect peptide synthesis activity, the
amount of phosphate released during the reaction was determined with
the Determiner L IP kit following the manufacturer’s protocol (Kyowa
Medex, Tokyo). To confirm peptide synthesis, reaction mixtures were
analyzed by liquid chromatography-electrospray ionization-mass spec-
trometry (LC-ESI-MS) (HPLC; Agilent 1100 series, Agilent Tech-
nologies, Santa Clara, CA; ESI-MS; LCQ Deca, Thermo Scientific,
Waltham, MA) and HPLC (L-2000 series; Hitachi High Technologies,
Tokyo). The details of analytical procedures are described in our
previous publications.13,24)
To determine substrate specificity, every combination of one or two
amino acids selected from 20 proteogenic amino acids was examined.
During analysis for homopeptide synthesis, the substrate concentration
used was 25 mM. To test reactivity with D-amino acids, reactions were
carried out with D-form amino acids. To test the reactivity with amino
acid derivatives, N-formyl-L-valine (Formyl-Val), N-methyl-L-leucine

L-Amino Acid Ligase Catalyzing Oligopeptide Synthesis
131
Position 2 (ATT codon)
Position 1 (ATG codon)
Fig. 1. Alignment of Amino Acid Sequences of RizB and the BL02410 Protein.
The ATG codon at position 1 codes for the first Met residue, but a comparison of amino acid sequences showed that the ATT codon at
position 2 might be the actual start codon, since the upstream region of position 1 is highly identical with the N-terminal region of the BL02410
protein. ꢄCorresponding to stop codon.
detect a ribosome binding site (RBS), but purine-rich
regions, which are putative RBSs,30,31) were observed
6–10 bp upstream of each start codon. Therefore, we
were not able to decide whether the actual start codon
was ATG at position 1 or ATT at position 2.
tides were not detected because their concentrations
were lower than detectable (Fig. 3). Thus oligopeptide
synthesis by RizB was confirmed by LC-ESI-MS and
HPLC.
Characterization of RizB
Confirmation of L-amino acid ligase activity
To determine substrate specificity, reactions involving
a combination of Val and other proteogenic amino acids
were conducted. LC-ESI-MS analysis showed that
various heteropeptides were synthesized, although no
heteropeptides containing Tyr, Pro, Cys, Asp, or Glu
were detected (Table 1). When Leu, Ile, or Met was
used as substrate with Val, heteropeptides containing
two residues of Leu, Ile, and Met were detected. These
amino acids were also usable as substrates for homo-
peptide synthesis. In contrast, when other amino acids
that were not usable for homopeptide synthesis were
used as substrates, heteropeptides containing only
one residue of these amino acids were detected.
These results suggest that substrate recognition at
the N-terminus influenced the resulting heteropeptide
sequences. Hence, when Leu, Ile, and Met were used as
substrates, heteropeptides with various combinations of
two amino acids should be synthesized. In contrast,
when other amino acids are used, they will probably be
incorporated at only the C-termini of the heteropeptides.
Furthermore, when D-form amino acids were used as
substrates, very little phosphate was detected and no
peptides were detected by LC-ESI-MS.
Reactions using various amino acid derivatives as
substrates were then conducted. LC-ESI-MS analysis
showed that heteropeptides containing C-protected
amino acids were synthesized, but heteropeptides
containing N-protected amino acids were not. When
Arg-NHOH was used as the substrate with Val, Val-
Arg-NHOH and Val-Val-Arg-NHOH were synthesized.
In addition, when Phe-OMe was used instead of Arg-
NHOH, Val-Phe-OMe and Val-Val-Phe-OMe were
synthesized (Supplemental Table 1; see Biosci. Bio-
technol. Biochem. Web site). Furthermore, reactions
using Val-Val instead of Val yielded no heteropeptides
(no Val-Val-Arg-NHOH and Val-Val-Phe-OMe). Hence
Val-Val could not be used as the N-terminal substrate.
These results suggest that Val-Val-Arg-NHOH was
synthesized from Val and Arg-NHOH as follows: Val
was ligated to Arg-NHOH, resulting in the formation
of Val-Arg-NHOH; then other Val residues were addi-
tionally ligated to Val-Arg-NHOH, producing Val-Val-
Arg-NHOH. Val-Val-Phe-OMe appears to be synthe-
sized through the same scheme. Therefore, peptides
might be elongated at the N-terminus in a stepwise
manner.
We cloned both rizB-P1 and rizB-P2 because the
actual start codon was not determined. At the time of
cloning, ATT at position 2 was replaced with ATG.
Recombinant RizB-P1 and RizB-P2 were produced as
C-terminal His-tagged proteins using the E. coli gene
expression system. SDS–PAGE analyses showed that
RizB-P2 was obtained in soluble form, while RizB-P1
formed an inclusion body. To obtain RizB-P1 in the
soluble form, we further experimented with gene
expression conditions, other host strains, and other tag
sequences that promote protein solubility (data not
shown), but we were unable to obtain RizB-P1 in soluble
form. Hence only RizB-P2 was used to assay L-amino
acid ligase activity. RizB-P2 was renamed RizB.
To investigate the L-amino acid ligase activity of
RizB, homopeptide synthesis was examined using 20
proteogenic amino acids. First, since ATP is hydrolyzed
to ADP and phosphate during peptide bond formation,
the amounts of phosphate produced in the reaction
mixture were measured. The measurement was per-
formed twice, and the values were averaged. When Val
or Leu was used as the substrate, approximately 8.8 mM
phosphate was detected, and this was the highest value
among 20 reactions. The formation of ADP in the
reaction mixtures was additionally confirmed by HPLC
(data not shown). Next, reaction mixtures were analyzed
by LC-ESI-MS, and the results indicated that tripeptides,
tetrapeptides, and pentapeptides were synthesized in
addition to the dipeptides in the reaction mixtures
containing Val, Leu, or Met as substrate (Fig. 2). When
Ile was used, the longest peptides detected by LC-ESI-
MS were tetrapeptides. When Trp was used as the
substrate, only dipeptides were detected. In addition, no
peptides were detected when the other 15 amino acids
were used as substrates. Thus RizB synthesized tripep-
tides as expected and, surprisingly, synthesized longer
peptides. The amounts of phosphate released in the
reactions were as follows: substrate (amount of phos-
phate); Val (8.8 mM) ꢂ Leu (8.7 mM) > Met (7.0
mM) ꢃ Ile (3.3 mM) > Trp (1.9 mM). Greater amounts
of phosphate released correlated with longer peptides
being synthesized.
The time course of peptide synthesis was determined
using Val as the substrate. The amount of each peptide
synthesized increased with reaction time, but pentapep-

132
K. KINO et al.
100
90
80
70
60
50
40
30
20
10
0
[Val-Val + H]+
m/z 217
[Val-Val-Val + H]+
[Val-Val-Val-Val + H]+
A
m/z 316
m/z 415
200 220 240 260 280 300 320 340 360 380 400 420 440 460 480 500 520 540 560 580 600
m/z
100
[Val-Val-Val-Val-Val + H]+
90
80
70
60
50
40
30
20
10
0
B
m/z 514
500
505
510
515
520
525
m/z
530
535
540
545
550
Fig. 2. LC-ESI-MS Analysis of the Reaction Catalyzed by RizB.
Val was used as the substrate. Peaks assigned are as follows: m=z 217, [Val-Val þ H]þ; m=z 316, [Val-Val-Val þ H]þ; m=z 415, [Val-Val-
Val-Val þ H]þ; m=z 514, [Val-Val-Val-Val-Val þ H]þ. The region of m=z 500 to 550 is enlarged, as shown in Fig. B.
Table 1. LC-ESI-MS Analysis of Heteropeptide Synthesis Catalyzed
by RizB
20
15
10
5
Product
Heteropeptide
Substrate 1 Substrate 2
Homopeptide
.
.
V2, V3, V4, V5 V R, V2 R, V3 R
.
Arg
Lys
His
Gln
Asn
Ala
Ser
.
V2, V3, V4, V5 V2
V2, V3, V4, V5 V2 H, V3
K
.
.
.
.
H
Q
N
.
V2, V3, V4, V5 V2 Q, V3
.
V2, V3, V4, V5 V2 N, V3
.
V2, V3, V4, V5 V2
V2, V3, V4, V5 V2
V2, V3, V4, V5 V2
V2, V3, V4, V5 V2
A
S
.
.
.
Thr
Gly
T
G
. .
Val
Leu
Ile
V2, V3, L2, L3
V2, V3, I2, I3
V2, V3, M2
V2 L, V L2
. .
V2 I, V I2
.
0
.
Met
Phe
Trp
Tyr
Glu
Asp
Pro
Cys
V2 M, V M2
F
0
2
4
6
8
10
.
.
V2, V3, V4, V5 V2 F, V3
Time (h)
.
V2, V3, V4, V5 V2
V2, V3, V4, V5 n.d.
V2, V3, V4, V5 n.d.
V2, V3, V4, V5 n.d.
V2, V3, V4, V5 n.d.
V2, V3, V4, V5 n.d.
W
Fig. 3. Time Course of Peptide Synthesis Using RizB.
The reaction was performed as described in ‘‘Materials and
Methods.’’ Closed circles indicate Val; open triangles indicate Val-
Val; open diamonds indicate Val-Val-Val; and open squares indicate
Val-Val-Val-Val. No pentamer peptide was detected by HPLC.
n.d., not detected. Homopeptides are shown as Xn (X, amino acid; n, peptide
.
length). Heteropeptides are shown as Xn Zn (X and Z, amino acids; n,
peptide length [n > 1]), since their sequences were not determined. For
.
example, X2 Z means tripeptides of X-X-Z, X-Z-X, or Z-X-X.
To test reactivity with peptides, reactions using Val-
Val as sole substrate were conducted. Little phosphate
was detected, and no products were detected by LC-ESI-
MS. Thus RizB did not participate in peptide-to-peptide
ligation. In addition, a reaction involving a combination
of Val and Arg-Ser was performed. LC-ESI-MS showed
that Val-Arg-Ser and Val-Val-Arg-Ser were synthe-
sized, but no homopeptides of Val-Val and Val-Val-Val
were detected. This suggests that RizB preferentially
uses dipeptides as the C-terminal substrate.
length of peptide synthesized by RizB became shorter.
In the reaction in which Val was used as substrate,
tetramers were the longest peptides synthesized when
Mn2þ was used, and trimers were the longest peptides
synthesized when Co2þ was used.
The effects of pH and temperature on L-amino acid
ligase activity were also examined (Fig. 4). The opti-
mum pH was within a range of 9.0 to 9.5, and activity
decreased sharply above pH 10.0. The optimum temper-
ature was found to be 40 ꢀC, and a significant decrease in
activity was observed above 45 ꢀC.
RizB required ATP and Mg2þ for L-amino acid ligase
activity. ATP could not be replaced with GTP, CTP, or
TTP, but Mg2þ could be replaced with Mn2þ and Co2þ
.
However, the use of Mn2þ or Co2þ appeared to decrease
L-amino acid ligase activity, because the maximum

L-Amino Acid Ligase Catalyzing Oligopeptide Synthesis
133
A
B
1
0.8
0.6
0.4
0.2
0
1
0.8
0.6
0.4
0.2
0
20
30
40
50
60
7
8
9
10
11
Temperature (°C)
pH
Fig. 4. Effects of pH (A) and Temperature (B) on the L-Amino Acid Ligase Activity of RizB.
Activity was assayed by measuring the amount of Val-Val-Val synthesized in the reaction mixture. In Fig. A, open circles indicate 100 mM
Tris–HCl buffer (pH 7.0 to 10.0) and closed circles indicate 100 mM borate buffer (pH 9.0 to 11.0).
Discussion
phosphate of amino acid, forming a tripeptide. Further,
the resulting tripeptide nucleophilically attaches to an
aminoacyl phosphate of another amino acid, forming
a tetrapeptide. As a result, the peptide is elongated at
its N-terminus. Therefore, the prominent difference
between RizB and known L-amino acid ligases that
catalyze only dipeptide synthesis is that the former uses
dipeptides, tripeptides, and pentapeptides as C-terminal
substrates and the latter use C-terminal amino acid
substrates only.
The optimum pH for RizB activity was 9, and the
optimum temperature was 40 ꢀC, both are similar to that
required by L-amino acid ligases that catalyze only
dipeptide synthesis, e.g., RizA, YwfE, and RSp1486a.
The amino acid sequence of RizB had 19–24% identity
to that of the L-amino acid ligases that catalyze only
dipeptide synthesis, and this similarity is very low. In
addition, the sequence identities among the L-amino acid
ligases that catalyze only dipeptide synthesis were also
low, around 18–26%. Thus it is difficult to explain why
RizB synthesizes oligopeptides solely on the basis of
amino acid sequence alignment.
We have reported that RizA catalyzes the synthesis of
an Arg-Xaa dipeptide (Xaa, arbitrary amino acid) and
might be involved in rhizocticin biosynthesis.13) The
rizB gene was found to be located approximately
9,000 bp upstream of rizA; RizB catalyzes tripeptide
synthesis and has substrate specificity for N-terminal
branched-chain amino acids. No genes encoding pro-
teins homologous to RizA and RizB were identified in
the genome of B. subtilis 168 (DDBJ accession no.
AL009126), a strain that does not produce rhizocticin.
We can not confirm the role of RizA and RizB in
rhizocticin biosynthesis due to the unavailability of
APPA, but, on the basis of their substrate specificities,
we consider that RizA and RizB might both be involved
in ligation reactions leading to rhizocticin biosynthesis.
In addition, we found several genes whose products
might be involved in APPA biosynthesis in the region
between rizB and rizA.13) Sequence analysis is still in
progress, and we expect to elucidate the mechanism of
rhizocticin biosynthesis completely.
A novel L-amino acid ligase, RizB, from B. subtilis
NBRC3134 was found to catalyze oligopeptide syn-
thesis from unprotected amino acid substrates in an
ATP-dependent manner. Although only L-amino acid
ligases that catalyze dipeptide synthesis have been
reported, this is the first report on an L-amino acid
ligase that synthesizes oligopeptides. RizB exhibited the
highest activity when Val, Leu, Ile, and Met were used
as substrates, and it synthesized various oligopeptides
consisting of two to five amino acids. RizB showed
specificity toward N-terminal substrates, but the speci-
ficity for C-terminal substrates was relaxed. Amino acid
derivatives could also be used as C-terminal substrates.
Although ATG at position 1 is generally selected as
the start codon, we selected ATT as a start codon of
RizB at position 2 (Fig. 1). Alignment of the amino acid
sequences of RizB and the BL02410 protein showed a
high degree of sequence conservation, and ATT has
been reported to function as the start codon.25,27,32)
Furthermore, we confirmed that the BL02410 protein
synthesized oligopeptides when branched-chain amino
acids were used as substrates, and an N-terminally
truncated BL02410 protein, which corresponds to RizB-
P1, formed an inclusion body, similarly to the case of
RizB-P1 (data not shown). Hence the N-terminal region
is important for the BL02410 protein to be produced in
soluble form. Position 2 might be the actual translation
start site of RizB, but further examination is necessary to
clarify this issue. In addition, it has also been reported
that the ATT codon plays a role in the translational
repression of E. coli K-12.33) The ATT codon at
position 2 may therefore play the same role in B. subtilis
NBRC3134.
RizB required ATP and Mg2þ, and ATP was hydro-
lyzed to ADP and phosphate. This reaction mechanism
is fundamentally the same as that of the L-amino acid
ligases that catalyze only dipeptide synthesis. The
reaction with C-protected amino acids showed that
RizB ligated a single amino acid to a dipeptide at its
N-terminus, resulting in a tripeptide. First, a dipeptide is
synthesized from amino acids, as occurs in reactions
catalyzed by known L-amino acid ligases; the resulting
dipeptide then nucleophilically attaches to an aminoacyl
In conclusion, we have described a novel L-amino
acid ligase that catalyzes oligopeptide synthesis, and our
results indicate that peptides of varying lengths were

134
K. KINO et al.
14) Stein T, Mol. Microbiol., 56, 845–857 (2005).
15) Rapp C, Jung G, Kugler M, and Loeffler W, Liebigs Ann.
Chem., 1988, 655–661 (1988).
synthesized. We expect that unknown L-amino acid
ligases capable of catalyzing oligopeptide synthesis are
present in other microorganisms, and further studies
should focus on identifying these L-amino acid ligases
so that they can be used to synthesize a wider array of
peptides.
16) Blodgett JAV, Zhang JK, and Metcalf WW, Antimicrob. Agents
Chemother., 49, 230–240 (2005).
17) Lauer B, Russwurm R, and Bormann C, Eur. J. Biochem., 267,
1698–1706 (2000).
18) Sambrook J and Russell WD, ‘‘Molecular Cloning, a Laboratory
Manual’’ 3rd ed., Cold Spring Harbor Laboratory Press, Cold
Spring Harbor (2001).
Acknowledgments
19) Isegawa Y, Sheng J, Sokawa Y, Yamanishi K, Nakagomi O, and
Ueda S, Mol. Cell. Probes, 6, 467–475 (1992).
20) Finn RD, Tate J, Mistry J, Coggill PC, Sammut SJ, Hotz HR,
Ceric G, Forslund K, Eddy SR, Sonnhammer EL, and Bateman
A, Nucleic Acids Res., 36, D281–D288 (2008).
21) Johnson M, Zaretskaya I, Raytselis Y, Merezhuk Y,
McGinnis S, and Madden TL, Nucleic Acids Res., 36, W5–
W9 (2008).
We thank Dr. Kazuhiko Tabata of Kyowa Hakko Bio
Co., Ltd., for helpful discussion. This work was
supported in part by funds from the Global COE
program of the MEXT Center for Practical Chemical
Wisdom and in part by a Waseda University Grant for
Special Research, Project 2009B-116.
22) Kugler M, Loffler W, Rapp C, Kern A, and Jung G, Arch.
Microbiol., 153, 276–281 (1990).
References
1) Yagasaki M and Hashimoto S, Appl. Microbiol. Biotechnol., 81,
13–22 (2008).
23) Laemmli UK, Nature, 227, 680–685 (1970).
24) Arai T and Kino K, Appl. Microbiol. Biotechnol., 81, 69–78
(2008).
2) Inaoka T, Takahashi K, Ohnishi-Kameyama M, Yoshida M, and
Ochi K, J. Biol. Chem., 278, 2169–2176 (2003).
3) Tabata K and Hashimoto S, Appl. Environ. Microbiol., 73,
6378–6385 (2007).
25) Lazarevic V, Dusterhoft A, Soldo B, Hilbert H, Mauel C, and
¨
¨
Karamata D, Microbiology, 145, 1055–1067 (1999).
¨
26) Sacerdot C, Fayat D, Dessen P, Springer M, Plumbridge JA,
Grunberg-Manago M, and Blanquet S, EMBO J., 1, 311–315
(1982).
4) Tabata K, Ikeda H, and Hashimoto S, J. Bacteriol., 187, 5195–
5202 (2005).
5) Kino K, Nakazawa Y, and Yagasaki M, Biochem. Biophys. Res.
Commun., 371, 536–540 (2008).
27) Sazuka T and Ohara O, DNA Res., 3, 225–232 (1996).
28) Shultzaberger RK, Bucheimer RE, Rudd KF, and Schneider TD,
J. Mol. Biol., 313, 215–228 (2001).
6) Kino K, Noguchi A, and Yagasaki M, J. Biosci. Bioeng., 106,
313–315 (2008).
˘
29) Vaıshkunaıte RI and Nivinskas RG, Mol. Biol. (Mosk), 27, 382–
˘
7) Arai T and Kino K, Biosci. Biotechnol. Biochem., 72, 3048–
3050 (2008).
391 (1993).
30) Boni IV, Isaeva DM, Musychenko ML, and Tzareva NV,
Nucleic Acids Res., 19, 155–162 (1991).
´
8) Bender CL, Alarcon-Chaidez F, and Gross DC, Microbiol. Mol.
Biol. Rev., 63, 266–292 (1999).
31) Laursen BS, Sørensen HP, Mortensen KK, and
Sperling-Petersen HU, Microbiol. Mol. Biol. Rev., 69, 101–
123 (2005).
9) Murzin AG, Curr. Opin. Struct. Biol., 6, 386–394 (1996).
10) Fan C, Moews PC, Shi Y, Walsh CT, and Knox JR, Proc. Natl.
Acad. Sci. USA, 92, 1172–1176 (1995).
32) Osawa S, Muto A, Jukes TH, and Ohama T, Proc. Biol. Sci.,
241, 19–28 (1990).
11) Galperin MY and Koonin EV, Protein Sci., 6, 2639–2643 (1997).
12) Fields SC, Tetrahedron, 55, 12237–12273 (1999).
13) Kino K, Kotanaka Y, Arai T, and Yagasaki M, Biosci.
Biotechnol. Biochem., 73, 901–907 (2009).
33) Butler JS, Springer M, and Grunberg-Manago M, Proc. Natl.
Acad. Sci. USA, 84, 4022–4025 (1987).