Requirement of catalytic-triad and related amino acids for the acyltransferase activity of Tanacetum Cinerariifolium gdsl lipase/esterase tcglip for ester-bond formation in pyrethrin biosynthesis

Article abstract of DOI:10.1271/bbb.130143

We have recently discovered that a GDSL lipase/ esterase (TcGLIP) in Tanacetum cinerariifolium catalyzed acyltransferase activity to form an ester bond in the natural insecticide, pyrethrin. TcGLIP contained Ser40 in Block I, Gly64 in Block II, Asn168 in Block III and Asp318 and His321 in Block V, suggesting underlying hydrolase activity, although little is known about their role in acyltransferase activity. We expressed TcGLIP here in Esherichia coli as a fusion with maltosebinding protein (MBP), part of the fusion being cleaved with a protease to obtain MBP-free TcGLIP. A kinetic analysis revealed that the MBP moiety scarcely influenced the kinetic parameters. The effects on acyltransferase activity of mutations of Gly64, Asn168, Asp318 and His321 were investigated by using MBP-fused TcGLIP. Mutations of these amino acids markedly reduced the acyltransferase activity, suggesting their critical role in the production of pyrethrins.

Full text of DOI:10.1271/bbb.130143

Biosci. Biotechnol. Biochem., 77 (9), 1822–1825, 2013
Requirement of Catalytic-Triad and Related Amino Acids
for the Acyltransferase Activity of Tanacetum cinerariifolium
GDSL Lipase/Esterase TcGLIP for Ester-Bond Formation
in Pyrethrin Biosynthesis
1
2
2
2
Yukio KIKUTA, Gen YAMADA, Tomonori MITSUMORI, Takayuki TAKEUCHI,
1
1
3
2;y
Koji NAKAYAMA, Yoshio KATSUDA, Akikazu HATANAKA, and Kazuhiko MATSUDA
1
Dainihon Jochugiku Co., Ltd., 1-1-11 Daikoku-cho, Toyonaka, Osaka 561-0827, Japan
Department of Applied Biological Chemistry, Faculty of Agriculture, Kinki University,
2
3
327-204 Nakamachi, Nara 631-8505, Japan
Department of Biological Chemistry, Yamaguchi University, 1677-1 Yoshida, Yamaguchi 753-8515, Japan
3
Received February 22, 2013; Accepted June 17, 2013; Online Publication, September 7, 2013
[doi:10.1271/bbb.130143]
We have recently discovered that a GDSL lipase/
and the biochemistry of pyrethrins remains elusive. We
therefore examined the biosynthetic pathway to pyreth-
esterase (TcGLIP) in Tanacetum cinerariifolium cata-
lyzed acyltransferase activity to form an ester bond in
the natural insecticide, pyrethrin. TcGLIP contained
Ser40 in Block I, Gly64 in Block II, Asn168 in Block III
and Asp318 and His321 in Block V, suggesting under-
lying hydrolase activity, although little is known about
their role in acyltransferase activity. We expressed
TcGLIP here in Esherichia coli as a fusion with maltose-
binding protein (MBP), part of the fusion being cleaved
with a protease to obtain MBP-free TcGLIP. A kinetic
analysis revealed that the MBP moiety scarcely influ-
enced the kinetic parameters. The effects on acyltrans-
ferase activity of mutations of Gly64, Asn168, Asp318
and His321 were investigated by using MBP-fused
TcGLIP. Mutations of these amino acids markedly
reduced the acyltransferase activity, suggesting their
critical role in the production of pyrethrins.
1
3
rins by using [1- C] D-glucose as the precursor. We
found that the acid moiety was biosynthesized by the
non-mevalonate pathway, whereas the alcohol moiety
was biosynthesized by the oxylipin pathway, sharing
part of the pathway with the plant hormone, jasmonic
3
)
acid.
The acid and alcohol moieties in pyrethrin biosyn-
thesis are biosynthesized by different pathways which
join to yield pyrethrins in the final step. Since an
acyltransferase was postulated to catalyze this esterifi-
cation step (Fig. 1A), chrysanthemoyl CoA and (S)-
pyrethrolone were added to a crude enzyme from the
flower bud, resulting in the production of pyrethrin I.
Purifying this enzyme and cloning the gene unexpect-
edly led to the discovery that the isolated enzyme
showed similarities to GDSL lipases/esterases which
4
)
have been named after the Gly-Asp-Ser-Leu motif.
This family consists of many members with such diverse
Key words: acyltransferase activity; GDSL lipase;
pyrethrin; Tanacetum cinerariifolium
5
)
6)
functions as ceramidase, sinapine esterase, trans-
9)
7
)
8)
esterification, cutin synthesis and cutinase activities
1
0)
Pyrethrins are natural insecticides that are biosynthe-
sized by Tanacetum cinerariifolium in the Asteraceae
family.1 Pyrethrins consist of six esters: pyrethrin I,
pyrethrin II, cinerin I, cinerin II, jasmolin I and jasmolin
II, resulting from the esterification of two acid moieties
in plants. We confirmed that the recombinant enzyme
expressed in Escherichia coli as a fusion with maltose-
binding protein (MBP) showed acyltransferase activity
for pyrethrin synthesis. The enzyme displayed high
substrate specificity for the acid and alcohol moieties,
recognizing the absolute configuration of the three
asymmetric carbons in the substrates. Gene expression
was also highly correlated with the pyrethrin content in
T. cinerariifolium, suggesting that the acyltransferase,
referred to as TcGLIP, underlaid the ester-bond for-
)
(
chrysanthemic acid and pyrethric acid) with three
alcohol moieties (pyrethrolone, jasmololone and ciner-
olone). Pyrethrins act by keeping voltage-sensitive
sodium channels open, thereby inducing hyperexcita-
tion, conduction blocking, or both, in the nervous
system.2 In respect of their selective neurotoxicity,
pyrethrins are used for household and pet pest control,
and also for agricultural pest control. However, since
pyrethrins are readily degraded in the field and their
production is climate-sensitive, synthetic analogs
)
4)
mation in pyrethrin biosynthesis.
GDSL lipases/esterases generally contain serine in
Block I, glycine in Block II and asparagine in Block III,
as well as aspartate and histidine in Block V as the
1
0,11)
amino acids involved in hydrolase activity.
Histi-
(
pyrethroids) with longer persistency and higher potency
dine in Block V deprotonates serine in Block I with the
assistance of aspartate in Block V, and activated serine
attacks the carbonyl carbon of substrates in the ‘‘cata-
have been developed. Pesticide research has conse-
quently been redirected from pyrethrins to pyrethroids,
y
To whom correspondence should be addressed. Tel: +81-742-43-7153; Fax: +81-742-43-1445; E-mail: kmatsuda@nara.kindai.ac.jp
Abbreviations: TcGLIP, Tanacetum cinerariifolium GDSL lipase/esterase; MBP, maltose-binding protein; SDS–PAGE, sodium dodecyl sulfate-
polyacrylamide gel electrophoresis

Amino Acids Underlying Pyrethrin Synthesis by TcGLIP
1823
Table 1. DNA Primers for Generating Mutant cDNAs of TcGLIP
A
B
0 0
Sequence (5 -3 )
Mutation
G64A
Primer
Forward CTTTTGGCCATATGCTCAATCCTACTTCAG
Reverse CTGAAGTAGGATTGAGCATATGGCCAAAAG
Forward GATTCTCTGATGCCCGTATCATCCCTG
Reverse CAGGGATGATACGGGCATCAGAGAATC
Forward TAGCTGTGGAGGTGCCGACTACCAAAGCCC
Reverse CTTTGGTAGTCGGCACCTCCACAGCTAAAC
Forward GTATTTTTTCTTTGCCCCTTTTCATCCTAATG
Reverse GGATGAAAAGGGGCAAAGAAAAAATAC
Forward CTTTGACCCTTTTGCTCCTAATGAATTG
Reverse CCAATTCATTAGGAGCAAAAGGGTCAAAG
G78A
N168A
D318A
H321A
USA) was expressed in E. coli as previously described.4) When
preparing the cDNA vector, the enterokinase cleavage site (DDDDK)
was replaced by the PreScission Protease (GE Healthcare Life
Sciences, Uppsala, Sweden) cleavage site (LEVLFQGP) to facilitate
removal of the MBP and protease. The mutant cDNAs were prepared
from wild-type cDNA by using a QuikChange kit (Agilent Technol-
ogies, Santa Clara, CA, USA) with the primers listed in Table 1,
according to the manufacturer’s protocol. The entire DNA sequences
of the mutant cDNAs were determined by automated sequencing with a
3
100 genetic analyzer (Life Technologies, Carlsbad, CA, USA).
The wild-type and mutant cDNAs were expressed in E. coli BL21
DE3) pLysS and purified with Amylose Resin (New England Biolabs)
(
to obtain the recombinant TcGLIP as an MBP-fused protein. We
confirmed that the wild-type MBP-fused TcGLIP purified with
Amylose Resin always showed acyltransferase activity. The wild-type
and N168A mutant were further purified by using an A¨ KTAexplorer
Fig. 1. Acyltransferase Activity for Pyrethrin I Synthesis (A) and
Primary Structure of TcGLIP (B).
Ser40 in Block I, Gly64 in Block IIA, Gly68 in Block IIB,
Asn168 in Block III, and Asp318 and His321 in Block V are
enclosed by squares (amino acid numbers are counted from Met1).
The signal peptide cleavage site is indicated by an arrow.
10S with a Mono Q 5/50 GL column (GE Healthcare Life Sciences) to
determine their specific activity, whereas the other mutants were not
purified with the Mono Q column because they had no acyltransferase
activity, as described later. The purity was checked by 10% sodium
dodecyl sulfate (SDS)-polyacrylamide gel electrophoresis (PAGE),
and the protein bands were stained by the Coomassie Brilliant Blue
reagent, using a SimplyBlueSafeStain kit (Life Technologies).
Some of the Mono-Q-purified MBP-fused TcGLIP was treated at
lytic triad.’’ Backbone NH of serine in Block I and that
of glycine in Block II, as well as the amide side chain of
ꢀ
1
1)
4 C with PreScission Protease (GE Healthcare Life Sciences) in a
0 mM Tris–HCl buffer (pH 7.5) to cleave the MBP–TcGLIP linkage.
asparagine in Block III serve as oxyanion acceptors.
All these structural features are preserved as Ser40,
Gly64, Asn168, Asp318 and His321 in TcGLIP
5
The digest was purified with a mixture of Amylose Resin and
Glutathione Sepharose 4B (GE Healthcare Life Sciences) at a ratio of
(
Fig. 1B). A GDSL lipase-like acyltransferase has been
2
:1 to remove the MBP and the glutathione S-transferase-tagged
found in tomato to exhibit glucarate caffeoyltransferase
activity for phenylpropanoid synthesis. In respect of this
enzyme, mutations of the catalytic-triad serine in Block
I, as well as of the aspartate and histidine in Block V had
protease, and to obtain MBP-free TcGLIP as an unabsorbed protein
in the 50 mM Tris–HCl buffer (pH 7.5).
Enzyme assay. The enzyme assay for pyrethrin synthesis was
conducted in a 100-mL solution consisting of (1R,3R)-chrysanthemoyl
CoA, (S)-pyrethrolone and TcGLIP as indicated in the Results section
in 50 mM Tris–HCl (pH 7.5). Kinetic analyses were conducted by
fixing the concentration of one substrate, while that of the other
substrate was increased. The reaction mixture was incubated for 10 min
7
)
only a small impact on the activity. However, this may
not be the case for the pyrethrin-forming acyltransferase
activity of TcGLIP, since we have observed that the
4
)
S40A mutation eliminated the activity.
We prepared in this present study MBP-fused and
MBP-free TcGLIPs to examine the effects of MBP on
the acyltransferase activity. Having observed no signifi-
cant impact of the MBP moiety on the activity, we
investigated the effects of mutations on the acyltransfer-
ase activity of several amino acids that are presumably
involved in hydrolase activity.
ꢀ
at 25 C and stopped by adding 10 mL of acetic acid. The reaction
mixture was extracted with 100 mL of hexane, and 10 mL of the extract
was subjected to an HPLC analysis of pyrethrin I, using a Class VP
system (Shimadzu, Kyoto, Japan) at 230 nm with a Cadenza CD-C18
column (4:6 ꢁ 100 mm; Imtakt, Kyoto, Japan) and an acetonitrile–
ꢂ1
water mixture (4:1) at a flow rate of 1 mL min
.
Results
Materials and Methods
Effects of MBP on the kinetics of pyrethrin synthesis
Wild-type TcGLIP was expressed in E. coli as an
MBP fusion protein and purified in a Mono Q column to
yield a single band by 10% SDS–PAGE (Fig. 2). We
measured the specific activity of MBP-fused wild-type
TcGLIP in the presence of 1 mM (S)-pyrethrolone with
various concentrations of (1R,3R)-chrysanthemoyl CoA
to determine the kinetic parameters for the CoA
Chemicals. The substrates, (1R,3R)-chrysanthemoyl CoA, (1R,3R)-
pyrethroyl CoA and (S)-pyrethrolone, were synthesized and their
chemical structures were confirmed as described in the supplementary
information of Ref. 3.
Site-directed mutagenesis and enzyme preparation. The cDNA of
wild-type and mutant TcGLIPs (GenBank accession no. JN418994)
cloned in the pMAL-c4E vector (New England Biolabs, Ipswich, MA,

1824
Y. KIKUTA et al.
Table 2. Kinetic Parameters of the Acyltransferase Activity of TcGLIP Determined in the Presence and Absence of the MBP Moiety
ꢂ1
ꢂ1
ꢂ1 ꢂ1
kcat=Km (s M )
Substrates (TcGLIP status)
Km (mM)
Vmax [nkat (mg protein)
]
kcat (s
)
Chrysanthemoyl CoA (MBP-fused)
Pyrethrolone (MBP-fused)
751 ꢃ 101
29:5 ꢃ 2:0
832 ꢃ 105
33:7 ꢃ 3:5
1:53 ꢃ 0:08
1:17 ꢃ 0:03
3:24 ꢃ 0:17
2:63 ꢃ 0:09
0:125 ꢃ 0:007
0:095 ꢃ 0:002
0:123 ꢃ 0:007
0:100 ꢃ 0:004
167
3230
147
Chrysanthemoyl CoA (MBP-free)
Pyrethrolone (MBP-free)
2960
ꢀ
Data were determined by 250 ng of protein at 25 C and pH 7.5, and are indicated as the mean ꢃ standard error of the mean (n ¼ 3).
Fig. 2. 10% SDS–PAGE Images for Recombinant TcGLIP Before
and After Cleavage by PreScission Protease.
Each sample is indicated above the gel picture.
substrate (Fig. 3A). A non-linear regression analysis of
the isotherm resulted in respective Km and Vmax values
Fig. 3. Specific Acyltransferase Activity for Pyrethrin I Synthesis of
Recombinant MBP-Fused and MBP-Free TcGLIP Saturated with an
ꢀ
ꢂ1
Increase in Substrates at a Constant Temperature of 25 C.
(
of 751 ꢃ 101 mM and 1:53 ꢃ 0:08 nkat (mg of protein)
A) Specific activity of MBP-TcGLIP saturated with an increase
(
n ¼ 3, mean ꢃ standard error of the mean) (Table 2).
of (1R,3R)-chrysanthemoyl CoA in the presence of 1 mM (S)-
pyrethrolone. (B) Specific activity of MBP-TcGLIP saturated with
an increase of (S)-pyrethrolone in the presence of 2 mM (1R,3R)-
chrysanthemoyl CoA. (C) Specific activity of MBP-free TcGLIP
saturated with an increase of (1R,3R)-chrysanthemoyl CoA in the
presence of 1 mM (S)-pyrethrolone. (D) Specific activity of MBP-
free TcGLIP saturated with an increase of (S)-pyrethrolone in the
presence of 2 mM (1R,3R)-chrysanthemoyl CoA. Data are plotted as
the mean ꢃ standard error of the mean (n ¼ 3).
These values and the molecular weight (81,689 Da) of
MBP-fused TcGLIP were used to calculate the respec-
tive kcat and k =K values for (1R,3R)-chrysanthemoyl
cat
m
ꢂ1
ꢂ1 ꢂ1
M (Table 2).
Similarly, the K , V , k and k =K values for (S)-
CoA of 0:125 ꢃ 0:007 s and 167 s
m
max
cat
cat
m
pyrethrolone were also respectively determined in the
presence of 2 mM (1R,3R)-chrysanthemoyl CoA by
fitting the isotherm (Fig. 3B) to be 29:5 ꢃ 2:0 mM,
ꢂ1
ꢂ1
1
:17 ꢃ 0:03 nkat (mg of protein) , 0:095 ꢃ 0:002 s
it was easier to prepare the MBP-fused protein than the
MBP-free protein as already described, we employed
MBP-fused TcGLIP to examine the effects on acyl-
transferase activity of the mutations to alanine of Gly64
in Block II, Asn168 in Block III, and Asp318 and
His321 in Block V which have been proposed as
underlying the hydrolase activity of GDSL-lipases/
ꢂ1 ꢂ1
and 3230 s
We cleaved MBP-fused TcGLIP by PreScission
M
(Table 2).
Protease and purified the digest with an Amylose Resin
and Glutathione Sepharose 4B resin mixture, resulting in
a single band of MBP-free TcGLIP with a molecular
weight of about 40 kDa by SDS–PAGE (Fig. 2). We
determined the kinetic parameters for the MBP-free
enzyme by fitting the isotherm shown in Fig. 3C and D.
1
1)
esterases. When assayed with 2 mM (1R,3R)-chrysan-
themoyl CoA and 1 mM (S)-pyrethrolone as substrates,
the G64A, D318A and H321A mutants did not show any
detectable acyltransferase activity (Fig. 4), as was the
ꢂ1
The Km (832 ꢃ 105 mM), kcat (0:123 ꢃ 0:007 s ) and
ꢂ1 ꢂ1
kcat=Km values (147 s
moyl CoA resembled those obtained for the MBP
M
) for (1R,3R)-chrysanthe-
4
)
case for the S40A mutant. Of those tested, only the
N168A mutant showed acyltransferase activity, although
its level was 18.8% of that of the wild-type enzyme
(Fig. 4).
ꢂ1
fusion. The Km (33:7 ꢃ 3:5 mM), kcat (0:100 ꢃ 0:004 s
)
) values for (1S)-pyrethrolone
ꢂ1 ꢂ1
and kcat=Km (2960 s
M
were also similar to those obtained for MBP-fused
TcGLIP (Table 2).
Discussion
Effects of site-directed mutagenesis on acyltransfer-
ase activity for the production of pyrethrin I
Since the MBP moiety had no significant effect on the
kinetic parameters of the acyl and alcohol substrates and
We expressed in this study MBP-fused TcGLIP in
E. coli and digested part of the MBP-fused enzyme with
PreScission Protease to prepare a soluble form of MBP-

Amino Acids Underlying Pyrethrin Synthesis by TcGLIP
1825
loss of hydrolase activity4) appear to suggest that the
acyltransferase and hydrolase activities of TcGLIP
shared a common mechanism. Nevertheless, we cannot
rule out the possibility that some mutations may change
the conformation of the active site, thereby eliminating
the activities. Further investigations are therefore
needed to better understand the underlying mechanisms
involved.
Acknowledgments
This study was supported by the Grant-in-Aid for
Scientific Research (S; no. 19101009) from the Japan
Society for the Promotion of Science, and by the
matching fund subsidy of the strategic project to support
the formation of research bases at private universities
Fig. 4. Effects of Mutations of the Catalytic Triad-Forming Amino
Acids on the Acyltransferase Activity of Recombinant TcGLIPs
Expressed in E. coli.
Bar graphs indicate the mean ꢃ standard error of the mean
(
n ¼ 3). ND, no detectable activity.
(no. S1101035) from the Ministry of Education, Culture,
Sports, Science and Technology of Japan.
free TcGLIP; we then compared the kinetic parameters
for the pyrethrin I-generating acyltransferase activity of
the MBP-fused and MBP-free enzymes. (S)-Pyrethro-
lone had 25-fold higher affinity than (1R,3R)-chrysan-
themoyl CoA in terms of the Km value previously
reported, whereas the kcat value of the acyl group donor
was about 1.32-fold higher than that of the alcohol
References
1) Matsuda K, Top. Curr. Chem., 314, 73 (2012).
2) Narahashi T, Mini Rev. Med. Chem., 2, 419 (2002).
3
)
Matsuda K, Kikuta Y, Haba A, Nakayama K, Katsuda Y,
Hatanaka A, and Komai K, Phytochemistry, 66, 1529 (2005).
Kikuta Y, Ueda H, Takahashi M, Mitsumori T, Yamada G,
Sakamori K, Takeda K, Furutani S, Nakayama K, Katsuda Y,
Hatanaka A, and Matsuda K, Plant J., 71, 183 (2012).
4
)
(
Table 2). The catalytic efficiency (kcat=Km) for the
alcohol moiety eventually became greater than that for
the acyl moiety (Table 2), suggesting the rate-limiting
substrate to be the acyl donor in the production of
pyrethrin I. A comparison of the kinetic parameters for
TcGLIP determined in the presence and absence of MBP
5) Pata MO, Wu BX, Bielawski J, Xiong TC, Hannun YA, and Ng
CK, Plant J., 55, 1000 (2008).
6)
7)
8)
Clauß K, Baumert A, Nimtz M, Milkowski C, and Strack D,
Plant J., 53, 802 (2008).
Teutschbein J, Gross W, Nimtz M, Milkowski C, Hause B, and
Strack D, J. Biol. Chem., 285, 38374 (2010).
(Table 2) led to the finding that the MBP moiety had no
significant effect on catalytic activity.
Girard AL, Mounet F, Lemaire-Chamley M, Gaillard C,
Elmorjani K, Vivancos J, Runavot JL, Quemener B, Petit J,
Germain V, Rothan C, Marion D, and Bakan B, Plant Cell, 24,
3119 (2012).
The role of ‘‘SGNH’’ amino acids has also been
studied for tomato (Solanum lycopersicum) glucarate
caffeoyltransferase underlying phenylpropanoid biosyn-
thesis. The D328A and H331A mutations, and also even
the S27A mutation, only had a small impact on the
9
)
Takahashi K, Shimada T, Kondo M, Tamai A, Mori M,
Nishimura M, and Hara-Nishimura I, Plant Cell Physiol., 51,
1
23 (2010).
7
)
glucarate caffeoyltransferase activity. In the case of
TcGLIP, however, mutations of the corresponding
amino acids markedly reduced the pyrethrin-synthesiz-
ing acyltransferase activity. The results and our previous
observation of the S40A mutation-induced complete
1
1
0) Chepyshko H, Lai CP, Huang LM, Liu JH, and Shaw JF, BMC
Genomics, 13, 309 (2012).
1) Akoh CC, Lee GC, Liaw YC, Huang TH, and Shaw JF, Prog.
Lipid Res., 43, 534 (2004).