Unnatural amino acid synthesis by thermostable O-phospho-L-serine sulfhydrylase from hyperthermophilic archaeon Aeropyrum pernix K1

Article abstract of DOI:10.1246/cl.170822

O-Acetyl-L-serine sulfhydrylase (OASS) from plants and bacteria synthesizes cysteine and unnatural amino acids that are important building blocks for active pharmaceuticals and agrochemicals. A thermostable O-phospho-L-serine sulfhydrylase from hyperthermophilic archaeon Aeropyrum pernix K1 (OPSSAp) exhibits a function similar to OASS. In the present study, we examined the synthesis of various unnatural amino acids using OPSSAp and demonstrated OPSSAp could efficiently synthesize various sulfur-containing amino acids. OPSSAp would be useful for industrial production of biologically important unnatural amino acids.

Full text of DOI:10.1246/cl.170822

Received: August 29, 2017 | Accepted: September 29, 2017 | Web Released: October 7, 2017
CL-170822
Unnatural Amino Acid Synthesis by Thermostable O-Phospho-L-serine Sulfhydrylase
from Hyperthermophilic Archaeon Aeropyrum pernix K1
Takashi Nakamura,*1 Kohei Kunimoto,1 Toru Yuki,1 and Kazuhiko Ishikawa2
1Laboratory of Molecular Biochemistry, Nagahama Institute of Bio-Science and Technology,
1266 Tamura-cho, Nagahama, Shiga 526-0829
2Biomedical Research Institute, National Institute of Advanced Industrial Science and Technology (AIST),
1-8-31 Midorigaoka, Ikeda, Osaka 563-8577
(E-mail: t_nakamura@nagahama-i-bio.ac.jp)
NH2
O-Acetyl-L-serine sulfhydrylase (OASS) from plants and
bacteria synthesizes cysteine and unnatural amino acids that are
important building blocks for active pharmaceuticals and agro-
chemicals. A thermostable O-phospho-L-serine sulfhydrylase
from hyperthermophilic archaeon Aeropyrum pernix K1
(OPSSAp) exhibits a function similar to OASS. In the present
study, we examined the synthesis of various unnatural amino
acids using OPSSAp and demonstrated OPSSAp could
efficiently synthesize various sulfur-containing amino acids.
OPSSAp would be useful for industrial production of bio-
logically important unnatural amino acids.
NH2
S
NH2
S
HOOC
HOOC
HO
HOOC
N
O
S-phenyl-L-cysteine
S-allyl-L-cysteine
L-mimosine
NH2
HOOC
NH2
HOOC
S
S
HOOC
Keywords: O-Phospho-L-serine sulfhydrylase
| Thermostable |
OH
Unnatural amino acid synthesis
carbocysteine
fudosteine
Unnatural amino acids are attractive compounds not only
for their role as building blocks for active pharmaceuticals
and agrochemicals containing chiral centers, but also for their
biological activity. As building blocks, unnatural amino acids
can be used to stabilize peptide bond against proteolytic attack
as well as to mimic natural structural analogues with an altered
chemical property.1 One of the unnatural amino acids, S-phenyl-
L-cysteine is a building block for a drug to an anti-acquired
immune deficiency syndrome, which is one of the leading
human immunodeficiency virus-protease inhibitors in the world
market. Furthermore, biological activities of unnatural amino
acids themselves are also remarkable. For example, β-(3-
hydroxy-4-pyridon-1-yl)-L-alanine (L-mimosine) present in
Mimosa and Leucaena spp. (Leguminosae) is a thyrotoxic
amino acid and causes loss of hair in growing animals and
shows inhibitory activity toward DNA replication in mammalian
cells.2,3 Furthermore, S-allyl-L-cysteine is one of the sulfur-
containing compounds derived from garlic and is able to
attenuate oxidative stress, to exert neuroprotective property in
various neurotoxic conditions, and to ameliorate cognitive
deficits in streptozotocin-diabetic rats.4 S-(Carboxymethyl)-L-
cysteine (carbocysteine) and S-(3-hydroxypropyl)-L-cysteine
(fudosteine) are used as an expectorant and a mucolytic agent
(Figure 1).5,6 Therefore, large-scale production of bioactive
unnatural amino acids using enzymes is very important.
Figure 1. Chemical structures of unnatural amino acids.
aminoacrylate intermediate linked to PLP. The second half-
reaction involves the addition of a secondary substrate, such as
sulfide or nucleophile, to the α-aminoacrylate intermediate to
generate an external Schiff base with cysteine or the correspond-
ing unnatural amino acid. The active-site lysine reacts with this
external Schiff base to release cysteine or the unnatural amino
acid, then regenerating the internal Schiff base with lysine.
It has been well studied that plant OASS produces novel β-
substituted L-amino acids when offered unnatural nucleophiles.8
Furthermore, Maier has reported that two isozymes of OASS
from Escherichia coli, OASS-A and OASS-B, can produce
unnatural amino acids from OAS and thiol compounds such
as thiophenol and mercaptoethanol, or N-heterocycles such as
triazole and tetrazole, and that the unnatural amino acids can be
produced efficiently by metabolic engineering of the cysteine-
biosynthetic pathway of E. coli.1 Rabeh et al. and Zhao et al.
have also reported the synthesis of unnatural amino acids using
OASS.9,10
Aeropyrum pernix is a hyperthermophilic archaeon that
grows optimally within a temperature range between 90 and
95 °C. The putative product of the gene encoding OASS
(OASSAp: APE1586) was found in the genome database of
A. pernix and its physiological role was examined. It exhibits
the activity for OASS and cystathionine β-synthase (CBS), and
it has therefore been speculated that OASSAp might be an
ancestral enzyme of OASS and CBS.11 About OASS activity
of OASSAp, OAS is labile at the optimal growth temperature
of A. pernix, 90-95 °C and this suggests that A. pernix can
circumvent the thermal instability of OAS.
O-Acetyl-L-serine sulfhydrylase (OASS), which is a pyr-
idoxal phosphate (PLP)-dependent enzyme, catalyzes the syn-
thetic reactions of cysteine and various unnatural amino acids
from O-acetyl-L-serine (OAS) and sodium sulfide or nucleo-
philes (Scheme 1).7 The active site of OASS contains PLP
linked to lysine residue as an internal Schiff base. Binding with
primary substrate OAS displaces the lysine to form an external
Schiff base, initiating the first half-reaction that yields an α-
Sulfhydrylation reactions of OASSAp using some substrate
candidates were then carried out and O-phospho-L-serine (OPS)
Chem. Lett. 2017, 46, 1789–1792 | doi:10.1246/cl.170822
© 2017 The Chemical Society of Japan | 1789

Primary Substrate
O
Lys
P
O-
COO-
-O
O
OPSS:
OPS
COO-
+
Lys
NH3
NH2
COO-
NH2
X
O
O
N+
N+
OASS:
O
P
O-
H
O-
O
O
P
O-
H
P
O
OAS
O
- O-
X=
O-
-O
O
-O
O
(OPSS)
O
N
N
Internal Schiff base
COO-
External Schiff base
O
(OASS)
Nu
NH2
O
O
or
P
-O
OH
O-
O-
Unnatural Amino Acid
(Nu = SH, cysteine)
Acetate Phosphate
COO-
Secondary Substrate
Nucleophile (Nu)
COO-
Lys
Lys
NH2
Nu
N+
N+
+
O
P
O-
H
O-
O
P
O-
H
NH3
O-
-O
O
-O
O
N
N
α-Aminoacrylate Intermediate
Scheme 1. Mechanism of synthetic reaction of cysteine or unnatural amino acid catalyzed by O-acetylserine sulfhydrylase (OASS)
and O-phosphoserine sulfhydrylase (OPSS).
was found to be the best substrates of OASSAp. This enzymatic
reaction of sulfhydrylation using OPS and sulfide was recog-
nized as a novel one and the enzyme was named O-phospho-L-
serine sulfhydrylase (OPSS; EC2.5.1.65).12
OH
HS
HS
HS
OH
OH
NH2
HS
HS
O
1
2
3
SH
4
8
SH
If OPSS from A. pernix (OPSSAp) can synthesize unnatural
amino acids like OASS (Scheme 1), it may improve the yield of
some unnatural amino acids compared to that by E. coli system,
because higher temperature is favorable to chemical and
enzymatic reactions, unless enzyme is denatured at the reaction
temperature. The activity for the sulfhydrylation reactions of
OPSSAp reached the maxima at 90 °C and the rate constant
HS
5
6
7
SH
SH
NO2
NH2
N
S
S
SH
SH
9
10
11
12
CH3
¹1
(14000 s at 85 °C)12 is significantly higher than that of E. coli
N
N
N
S
S
N
N
N
¹1
OASS-A (2030 s at 25 °C).13 Furthermore, OPSSAp retained
O
O
N
H
N
H
N
H
O
16
ca. 90% of its activity after the enzyme solution had been
incubated for 6 h at 100 °C at pH 6.1 and 6.7,11 while OASS-B
from E. coli was completely inactivated at 70 °C for 1 h.10
OPSSAp also showed higher stability toward organic solvents
such as N,N-dimethylformamide (DMF) and 1,4-dioxane than
OASS-B from E. coli as reported previously.14 These properties
should make them particularly useful as industrial biocatalysis.
In this paper, we analyzed the characterization of unnatural
amino acid synthesis using OPSSAp and compared with that of
OASS-A and OASS-B from E. coli and chemical synthesis of
S-(4-hydroxyphenyl)-L-cysteine.
O-Phospho-L-serine, OPS, and 16 nucleophiles were used as
substrates to synthesize unnatural amino acids with OPSSAp.
The nucleophiles were selected in such a way that the reaction
products are commercially available and comparable to the
previous results by OASS from E. coli (Figure 2 and Table 1).
OPSSAp was expressed and purified from recombinant E. coli
Rosetta (DE3) cells according to the methods described
previously.11 In enzymatic reaction, 2.5 mM OPS and 2.5 mM
nucleophile (1 mM in the case of 7 to 10) in 50 mM potassium
phosphate (Kpi) buffer (pH 7.5) containing 0.2 mM PLP and
1 mM EDTA were incubated with OPSS (0.05-40 ¯g) at 60 or
13
14
15
Figure 2. Chemical structures of secondary substrates (nucle-
ophiles) used in this study. The entry numbers (1 to 16)
correspond to those in Table 1.
80 °C for 1-10 min. Thiol compounds were dissolved in DMF
due to their insolubility in water and higher boiling point of
DMF, and the reaction mixture contains 5% DMF and 10%
DMF in aliphatic thiol and aromatic or heterocyclic thiols,
respectively. After the reaction was terminated by 1 M HCl, the
reaction mixture was filtered using a 0.1-μm membrane filter
(Ultrafree-MC Durapore poly(vinylidene difluoride) 0.1 ¯m,
amicon, Millipore) and was analyzed by amino acid analyzer
(L-8800A, Hitachi) or high-performance liquid chromatography
system (GILSON) that was used to analyze the production of
S-(2-aminophenyl)-L-cysteine, S-sulfo-L-cysteine, and S-4-hy-
droxyphenyl-L-cysteine when 2-aminothiophenol (10), thiosul-
fate (16) and 4-hydroxythiophenol was used as a nucleophile.
The detailed reaction conditions and analysis method are
described in Table S1. Specific activity of OPSSAp toward
1790 | Chem. Lett. 2017, 46, 1789–1792 | doi:10.1246/cl.170822
© 2017 The Chemical Society of Japan

Table 1. Nucleophiles, products, and specific activity in the synthesis of unnatural amino acids catalyzed by OPSSAp
Specific activity
Entry
Nucleophiles
Products
/¯mol min¹1 mg¹1-proteind
sodium sulfidea
sodium sulfideb
L-cysteine
L-cysteine
28 « 2.4
48 « 4.5
1
2
3
4
5
3-mercapto-1-propanolb
2-aminoethanethiola
mercaptoacetic acidb
allyl mercaptaneb
2-mercaptoethanolb
1-propanethiolb
S-(3-hydroxypropyl)-L-cysteine
S-(2-aminoethyl)-L-cysteine
S-(carboxymethyl)-L-cysteine
S-allyl-L-cysteine
S-(2-hydroxyethyl)-L-cysteine
S-propyl-L-cysteine
13 « 1.3
0.98 « 0.04
8.2 « 0.05
71 « 2.5
97 « 4.8
11 « 0.7
48 « 1.2e
6
7
thiophenolc
S-phenyl-L-cysteine
8
9
4-nitrothiophenolc
4-methylbenzenethiolc
2-aminothiophenolc
1,3-thiazole-2-thiolc
2-mercaptothiophenec
pyrazolea
S-(4-nitrophenyl)-L-cysteine
S-(4-tolyl)-L-cysteine
0.9 « 0.03
19 « 3.1
29 « 2.1
10
11
12
13
14
15
16
S-(2-aminophenyl)-L-cysteine
S-(2-thiazolyl)-L-cysteine
S-(2-thienyl)-L-cysteine
3-(1-pyrazolyl)-L-alanine
3-(1,2,4-triazol-1-yl)-L-alanine
3-(2-tetrazolyl)-L-alanine
S-sulfo-L-cysteine
0.19 « 0.005
26.5 « 1.7
0.0017 « 0.0002
0.15 « 0.05
0.20 « 0.02
2.3 « 0.09
1,2,4-triazolea
1,2,3,4-tetazolea
sodium thiosulfatea
b
aThe reaction solvent is 50 mM Kpi buffer (pH 7.5) with 0.2 mM PLP and 1 mM EDTA. The reaction solvent is 50 mM Kpi buffer
c
(pH 7.5) with 5% DMF (v/v), 0.2 mM PLP and 1 mM EDTA. The reaction solvent is 50 mM Kpi buffer (pH 7.5) with 10% DMF
(v/v), 0.2 mM PLP, and 1 mM EDTA. dValues are expressed as mean « standard deviation of three experiments. eValues are expressed
as mean « standard deviation of two experiments.
cysteine was determined by colorimetric assay with acid
ninhydrin reagent.15,16
The result showed that specific activity toward some thiol
nucleophiles such as allyl mercaptane (4) and 2-mercaptoethanol
(5) was higher than that toward sodium sulfide to produce L-
Compared to substrate (nucleophile) specificity with OASS
isozymes from E. coli, OPSSAp was more similar to OASS-B in
that OASS-B can use thiosulfate and thiophenol much better
than OASS-A, and pyrazole and triazole less than OASS-A.1,10
This result was consistent with structure-based amino acid
sequence alignment of OASS-B with OASS-A. It showed that
OASS-B contains the large residues Arg210, Arg211, and
Trp212 and a three-residue insertion at the pocket around the
active center, whereas OASS-A contains merely the small
residues Gly230, Ala231, and Gly232. In other words, OASS-B
has a three-residue insertion bulging toward the outside and thus
enlarging the active center pocket in comparison to OASS-A.17
This additional space enables OASS-B to accept thiosulfate.
OPSSAp has two arginine residues in the corresponding region
and may have similar activity to OASS-B.
Subsequently, we examined whether OPSSAp can easily
synthesize unnatural amino acid, S-(4-hydroxyphenyl)-L-cys-
teine, which is a candidate antitumor agent for melanoma.18 The
specific activity of OPSSAp toward 4-hydroxythiophenol was
17 « 1.1 ¯mol min¹1 mg¹1-protein. The value is expressed as
mean « standard deviation of three experiments. This compound
was reported to be chemically synthesized by refluxing a
mixture of phenol and L-cystine in 47% aqueous HBr for 2 h,
followed by separation from a side product, S-(2-hydroxyphen-
yl)-L-cysteine by column chromatography and recrystallization
in 37% overall yield (Scheme 2a).19
cysteine (Table 1). When 5 was used as a nucleophile, OPSS
¹1
showed the highest specific activity (97 ¯mol min¹1 mg
-
protein) to produce S-2-hydroxyethyl-L-cysteine. OPSSAp could
also efficiently produce S-allyl-L-cysteine, which is a bioactive
molecule with anti-diabetic, antioxidant, and anti-inflammatory
properties as mentioned above, from OPS and 4. Therefore,
OPSSAp is a very useful biocatalyst for the production of S-
allyl-L-cysteine. Thiophenol (7) and thiophenol with electron-
donating groups (9 and 10) showed higher activity than that with
electron-withdrawing group (8) when aromatic thiols were used
as nucleophile. Furthermore, OPSSAp could synthesize more
S-propyl-L-cysteine (from 1-propanethiol, 6) and S-phenyl-L-
cysteine (from thiophenol, 7) than OASSs from E. coli, although
it cannot be simply compared due to the different substrate
concentrations.10
On the other hand, specific activity of OPSSAp toward
N-heterocycles was very low. The specific activity increased
corresponding to the number of nitrogen atoms in 5-membered
ring, such as two nitrogen atoms in pyrazole (13, 0.0017
¯mol min¹1 mg¹1-protein) to four nitrogen atoms in 1,2,3,4-
tetrazole (15, 0.2 ¯mol min¹1 mg¹1-protein). These specific
activities ranged from 1/17000 to 1/140 as compared to that
of sodium sulfide in Kpi buffer (28.5 ¯mol min¹1 mg¹1-protein)
and much less than those of OASSs from E. coli.
In enzymatic reaction, 148 mg of OPS (80 mM) and 101 mg
of 4-hydroxythiophenol (80 mM) were incubated with 125 ¯g of
OPSSAp (0.3 ¯M) at 80 °C for 3 h in 10 mL of 0.1 M Kpi buffer
(pH 7.5) solution (Scheme 2). The pH of the reaction solution
was adjusted to 6 by 1 M HCl and crystallized by cooling the
solution on ice. The crystal was separated by centrifugation
and recrystallized by aqueous ethanol. The identification of the
For example, the specific activities of OASS-A and OASS-
B from E. coli to 15 are 16.5 and 19.7 ¯mol min¹1 mg¹1-protein,
respectively.10 Furthermore, OPSSAp could use thiosulfate (16)
to synthesize S-sulfo-L-cysteine (2.3 ¯mol min¹1 mg¹1-protein).
Chem. Lett. 2017, 46, 1789–1792 | doi:10.1246/cl.170822
© 2017 The Chemical Society of Japan | 1791

HOOC
OH
NH2
The Science Research Promotion Fund. The authors thank
emeritus professor of Kyoto University, Isao Saito for reading
the manuscript and Mr. Masahiro Hashimoto (JEOL) for
measuring the mass spectrum of S-4-hydroxyphenyl-L-cysteine.
(a)
(b)
HOOC
NH2
NH2
HOOC
S
S
47%aq HBr
reflux, 2 h
S
+
+
S
HO
COOH
H2N
2.3 %
Supporting Information is available on http://dx.doi.org/
10.1246/cl.170822.
OH
37 %
HOOC
NH2
HOOC
This manuscript is dedicated to the late Professor Yoshihiko
OH
SH
NH2
OPSSAp
Ito.
0.1 M Kpi (pH 7.5)
10% DMF (v/v)
S
+
O
P
80oC, 3 h
References
O
O
O
1
2
T. H. P. Maier, Nat. Biotechnol. 2003, 21, 422.
F. Ikegami, M. Mizuno, M. Kihara, I. Murakoshi, Phyto-
chemistry 1990, 29, 3461.
OH
80 %
3
4
P. J. Mosca, H.-B. Lin, J. L. Hamlin, Nucleic Acids Res.
1995, 23, 261.
Scheme 2. Synthesis of S-(4-hydroxyphenyl)-L-cysteine by (a)
chemical reaction19 and (b) OPSSAp-catalyzed reaction.
T. Baluchnejadmojarad, Z. Kiasalari, S. Afshin-Majd, Z.
Ghasemi, M. Roghani, Eur. J. Pharmacol. 2017, 794, 69.
D. T. Brown, Drug Intell. Clin. Pharm. 1988, 22, 603.
K. Takahashi, H. Kai, H. Mizuno, T. Koda, T. Miyata, J.
Pharm. Pharmacol. 2001, 53, 911.
C. H. Tai, S. R. Nalabolu, T. M. Jacobson, D. E. Minter, P. F.
Cook, Biochemistry 1993, 32, 6433.
F. Ikegami, I. Murakoshi, Phytochemistry 1994, 35, 1089.
W. M. Rabeh, S. S. Alguindigue, P. F. Cook, Biochemistry
2005, 44, 5541.
product, S-(4-hydroxyphenyl)-L-cysteine, was confirmed by
1H NMR (JEOL, JNM-AL400, 2.5 M DCl-D2O, 3-(trimethyl-
silyl)propanoic-2,2,3,3-d4 acid as an internal reference, ¤ 3.33
(2H, d, J = 5.1 Hz, CH2), 4.10 (1H, t, J = 5.1 Hz, CH), 6.75
(2H, d, J = 7.7 Hz, C2 and C6 H), 7.33 (2H, d, J = 7.7 Hz, C3
and C5 H), Figure S1)19 and mass spectrometry (JEOL, JMS-
S3000, calcd. for C9H12NO3S [M + H]+ 214.05324, found m/z
214.05327). The yield of S-4-hydroxyphenyl-L-cysteine was
80% and higher than the reported chemical synthesis yield
(37%), as shown in Scheme 2b.
In conclusion, we found that O-phospho-L-serine sulfhy-
drylase from A. pernix (OPSSAp), could synthesize unnatural
amino acids with a similar substrate specificity to OASS-B from
E. coli. OPSSAp is very useful to produce sulfur-containing
unnatural amino acids such as S-allyl-L-cysteine, S-2-hydroxy-
ethyl-L-cysteine, S-phenyl-L-cysteine, and S-4-hydroxyphenyl-L-
cysteine. We have already determined three-dimensional struc-
tures of OPSSAp, and analyzed detailed reaction and substrate-
recognition mechanisms.16,20,21 The improvement of the yield of
various unnatural amino acids may be achieved by genetically
modified OPSSAp enzymes using computer modeling and site-
directed mutagenesis or random mutagenesis technology. Such
protein engineering research in combination with further kinetic
study on secondary substrate (nucleophile) specificity is neces-
sary for the further development of industrial process for
producing pharmaco-therapeutically important unnatural amino
acids by using OPSSAp enzyme.
5
6
7
8
9
10 C. Zhao, Y. Kumada, H. Imanaka, K. Imamura, K.
Nakanishi, Protein Expr. Purif. 2006, 47, 607.
11 K. Mino, K. Ishikawa, J. Bacteriol. 2003, 185, 2277.
12 K. Mino, K. Ishikawa, FEBS Lett. 2003, 551, 133.
13 K. Mino, T. Yamanoue, T. Sakiyama, N. Eisaki, A.
Matsuyama, K. Nakanishi, Biosci. Biotechnol. Biochem.
2000, 64, 1628.
14 T. Nakamura, S. Asai, K. Nakata, K. Kunimoto, M. Oguri,
K. Ishikawa, Biosci. Biotechnol. Biochem. 2015, 79, 1280.
15 M. K. Gaitonde, Biochem. J. 1967, 104, 627.
16 T. Nakamura, Y. Kawai, K. Kunimoto, Y. Iwasaki, K. Nishii,
M. Kataoka, K. Ishikawa, J. Mol. Biol. 2012, 422, 33.
17 M. T. Claus, G. E. Zocher, T. H. P. Maier, G. E. Schulz,
Biochemistry 2005, 44, 8620.
18 S. Miura, T. Ueda, K. Jimbow, S. Ito, K. Fujita, Arch.
Dermatol. Res. 1987, 279, 219.
19 S. Ito, S. Inoue, Y. Yamamoto, K. Fujita, J. Med. Chem.
1981, 24, 673.
20 Y. Oda, K. Mino, K. Ishikawa, M. Ataka, J. Mol. Biol. 2005,
351, 334.
This work was financially supported by The Promotion and
Mutual Aid Corporation for Private Schools of Japan through
21 E. Takeda, K. Kunimoto, Y. Kawai, M. Kataoka, K.
Ishikawa, T. Nakamura, Extremophiles 2016, 20, 733.
1792 | Chem. Lett. 2017, 46, 1789–1792 | doi:10.1246/cl.170822
© 2017 The Chemical Society of Japan