Elucidating hydroxylation and methylation steps tailoring piericidin a1 biosynthesis

Article abstract of DOI:10.1021/ol4034176

The piericidin A1 (1) gene cluster was identified from the deep-sea derived Streptomyces sp. SCSIO 03032. Our in vivo and in vitro experiments verified PieE as a 4′-hydroxylase and PieB2 as a 4′-O-methyltransferase, allowing the elucidation of the post-PKS modification steps involved in 1 biosynthesis. In addition, the shunt metabolite piericidin E1 (7) was identified as a novel analogue featuring a C-2/C-3 epoxy ring.

Full text of DOI:10.1021/ol4034176

Letter
pubs.acs.org/OrgLett
Elucidating Hydroxylation and Methylation Steps Tailoring Piericidin
A1 Biosynthesis
Yaolong Chen,†,‡,§ Wenjun Zhang,†,§ Yiguang Zhu,*,† Qingbo Zhang,† Xinpeng Tian,† Si Zhang,†
and Changsheng Zhang*,†
†Key Laboratory of Tropical Marine Bio-resources and Ecology, RNAM Center for Marine Microbiology, Guangdong Key Laboratory
of Marine Materia Medica, South China Sea Institute of Oceanology, Chinese Academy of Sciences, 164 West Xingang Road,
Guangzhou 510301, China
‡University of the Chinese Academy of Sciences, Beijing 100049, China
S
* Supporting Information
ABSTRACT: The piericidin A1 (1) gene cluster was identified from the deep-sea derived Streptomyces sp. SCSIO 03032. Our in
vivo and in vitro experiments verified PieE as a 4′-hydroxylase and PieB2 as a 4′-O-methyltransferase, allowing the elucidation of
the post-PKS modification steps involved in 1 biosynthesis. In addition, the shunt metabolite piericidin E1 (7) was identified as a
novel analogue featuring a C-2/C-3 epoxy ring.
PKS module was demonstrated to hydrolyze a β,δ-diketo
iericidins are a family of α-pyridone antibiotics that are
Pisolated mainly from various Streptomyces species of thioester polyketide product.6 However, the enzymatic
terrestrial, marine, and symbiotic origins.1 Structurally,
consequences of post-PKS modifications, involving a hydrox-
ylation and two methylations to furnish the biosynthesis of 1,
have not been elucidated. In this study, we report the
characterization of PieE as a 4′-hydroxylase and PieB2 as a
4′-O-methyltransferase, allowing the dissection of the tailoring
steps for the biosynthesis of 1 in the deep-sea derived
Streptomyces sp. SCSIO 03032.
piericidins feature a pyridone core attached with variable
polyene side chains. Due to the high structure similarity to
ubiquinone, piericidin A1 (1, Figure 1) is known as a potent
inhibitor of both mitochondrial and bacterial NADH−
ubiquinone oxidoreductase (complex I).2 In addition to broad
antimicrobial and insecticidal activities, 1 is shown as a GRP78
down-regulator and displays cytotoxic activity for etoposide
resistant cancer cells during glucose deprivation, thus being
identified as an anticancer agent.3
The deep-sea derived Streptomyces sp. SCSIO 03032 was
previously shown to produce spiroindimicins, an unusual group
of spiro-containing bisindole alkaloids.8 In addition, we found
that this strain could also produce piericidin A1 (1), the mass
and NMR spectroscopic data of which are consistent with those
reported for 1 (Figure S1 and Tables S1 and S2, Supporting
Information). From the draft genome sequence of Streptomyces
sp. SCSIO 03032 (data not shown), we identified the
biosynthetic gene cluster of 1 (GenBank accession no.
KF874660), which comprises 12 biosynthetic genes that
encode putative proteins involved in the biosynthesis of 1
(Figure 1), including a transcription regulator PieR, six typical
modular PKSs PieA1-A6, a putative cyclase PieC, an
amidotransferase PieD, two methyltransferases PieB1 and
PieB2, and a monooxygenase PieE (Table S3, Supporting
Information). The biosynthetic gene cluster of 1 in Streptomyces
The total synthesis of 1 has been achieved by several groups.4
Previous feeding studies demonstrated that the carbon
backbone of 1 is derived from a linear polyketide that is
formed from four acetates and five propionates and revealed the
low incorporation of 15N-labled aspartate and serine.5 Recently,
the biosynthetic locus of 1 has been successfully identified from
Streptomyces piomogeues var. Hangzhouwanensis, revealing that
1 originates from a modular polyketide synthase (PKS)
pathway.6 In contrast to the use of a hybrid polyketide
synthase−nonribosomal peptide synthetase (PKS−NRPS)
pathway to incorporate amino acids as the nitrogen source to
construct the α-pyridone ring in the biosynthesis of other well-
studied α-pyridone natural products kirromycin and tenellin,7
an ATP-dependent amidotransferase PieD plays a key role in
introducing the nitrogen into the α-pyridone ring in piericidin.6
In addition, the terminal thioesterase (TE) of the piericidin A1
Received: November 25, 2013
Published: January 10, 2014
© 2014 American Chemical Society
736
dx.doi.org/10.1021/ol4034176 | Org. Lett. 2014, 16, 736−739

Organic Letters
Letter
Figure 1. Genetic organization of the pie gene cluster in Streptomyces sp. SCSIO 03032 and the proposed 1 biosynthetic pathway, focusing on the
post-PKS modification steps. Bold arrows indicate biosynthetic steps that were confirmed in this study.
sp. SCSIO 03032 displays the same genetic organization as that
in S. piomogeues,6 despite their distinct terrestrial and marine
origins. Given the high amino acid sequence similarities and
identities for their corresponding biosynthetic enzymes
encoded in the two biosynthetic loci of 1 (Table S4, Supporting
Information), we propose that 1 was biosynthesized in a similar
manner in both strains (Figure S2, Supporting Information).
Briefly, the six modular PKSs PieA1−A6 function to
incorporate four molecules of malonyl-CoA and five molecules
of methylmalonyl-CoA under the colinearity rule to construct a
nascent polyketide chain, which is released by the terminal TE
domain to form a C18 linear carboxylic acid 5 (Figure 1).
Subsequently, the ATP-dependent amidotransferase PieD,
belonging to the class II glutamine amidotransferases,9 transfers
a nitrogen group from amine donors (ammonia or glutamine)
to the carboxylic acid group in 5, yielding 6 (Figure 1).
Although annotated as a hypothetical protein with distant
similarity to the well-characterized polyketide cyclase TcmF2,10
PieC is proposed to catalyze an intramolecular condensation
between the amine and the δ-keto group in 6, leading to the
Figure 2. HPLC traces for metabolite profiling of different strains and
the structure of piericidin E1 (7): (i) Strptomyces sp. SCSIO 03032,
(ii) ΔpieE mutant, (iii) ΔpieB1 mutant, (iv) ΔpieB2 mutant, UV
detection at 238 nm.
cyclized α-pyridone product 2 (Figure 1).6 Finally, a
hydroxylation by PieE and two methylations by PieB1 and
PieB2 complete the biosynthesis of 1.
Functions of PieC and PieD have been investigated by in
vivo genetic experiments to support the proposed pathway for
1.6 Disruption of pieD abolished the production of 1 and the
inactivation of pieC led to the decrease of the yield of 1.6
However, the modification steps after the α-pyridone ring
formation remain elusive. To this end, we are interested in
probing the functions of PieE, PieB1, and PieB2. A PCR
approach was then utilized to screen a SuperCos 1-based
genomic library of Streptomyces sp. SCSIO 03032 to yield four
overlapping cosmids that cover the entire biosynthetic gene
cluster of 1 (Figure S3, Supporting Information). Next, the
PCR-targeted gene inactivation of pieE was carried out
according to the reported REDIRECT strategy (Figure S4,
Supporting Information).11 HPLC analysis of the ΔpieE mutant
revealed the production of two compounds that were distinct
from 1 (Figure 2). One was characterized to be identical to
previously reported Mer-A2026B (3, Figure 1),12 by HR-ESI-
1
Supporting Information). The comparison of the H and 13C
NMR spectroscopic data of 7 (Tables S1 and S2, Figure S6,
Supporting Information) and 3 revealed that two olefinic
carbon signals (δC 123.5 C-2, 134.4 C-3) in 3 were replaced by
an oxymethine (δH 4.26 H-2, δC 73.4 C-2) and an oxygenated
quaternary carbon (δC 75.3 C-3) in 7, indicating a C-2/C-3
epoxy ring structure, which was supported by HMBC
correlations from H3-14 to C-2/C-3/C-4, and from H-1 to
C-2/C-3 (Figure S6, Supporting Information). In addition, a
carbonyl group (δC 181.2 C-3′) signal was observed in 7, which
was located at C-3′ by key HMBC correlations from both H-4′
and H3-6′ to C-3′. Thus the planar structure of 7 was
established (Figure 2). The relative configuration of 7 was
assigned by analyzing the 1H−1H coupling constants (Table S1,
Supporting Information) and NOESY correlations (Figure S6,
Supporting Information). The absolute configuration at C-10 of
7 was determined as 10R by the modified Mosher’s method
(Table S5 and Figure S7, Supporting Information) and thus
also established the 9R configuration (Figure S7, Supporting
Information). However, the absolute configuration at C-2/C-3
1
MS and H and 13C NMR spectroscopic data (Tables S1 and
S2, Figure S5, Supporting Information). The other compound
7 presented a molecular formula of C24H35NO4 (m/z [M + H]+
402.2637, calcd 402.2639) deduced by HR-ESI-MS (Figure S6,
737
dx.doi.org/10.1021/ol4034176 | Org. Lett. 2014, 16, 736−739

Organic Letters
Letter
in 7 remained unassigned. Different from the presence of the
C-11/C-12 epoxy ring in piericidin C and D series,13 7 is novel
to feature a C-2/C-3 epoxy ring, and was designated as
piericidin E1. Given that 7 was isolated from the ΔpieE mutant
and PieE was the only oxygenase found in the gene cluster of 1,
we hypothesized that 7 was a shunt metabolite from the
pathway of 1 and the formation of the C-2/C-3 epoxy ring in 7
was probably catalyzed by an unknown enzyme encoded in the
genome of Streptomyces sp. SCSIO 03032.
Bioinformatic analysis shows that PieE is a putative FAD-
dependent monooxygenase. PieE shares the highest identity
(74%) to EME97622, a protein encoded by a putative locus of
1 in the genome of S. mobaraensis (Table S4, Supporting
Information), the first described 1 producer.1a In addition, PieE
exhibits similarity to 2,4-dichlorophenol 6-monooxygenase
TfdB and methylhydroquinone oxygenase MhqA encoded in
diverse bacteria genomes.14 The bioinformatic analysis of PieE,
together with the accumulation of 3 in the ΔpieE mutant,
indicated PieE as the 4′-hydroxylase modifying the α-pyridone
ring in 3. To validate this hypothesis, soluble N-His6-fused PieE
protein was produced in E. coli BL21(DE3) harboring the
pET28a-based expression plasmid pCSG601 (Table S6,
Supporting Information), and was purified to near homogeneity
via Ni-NTA chromatography (Figure S8, Supporting Informa-
tion). Consistent with the yellow color of the purified PieE,
noncovalently bound FAD was released from heat-denatured
PieE (Figure S9, Supporting Information). The subsequent
incubation of 100 μM 3 and 5 μM PieE in the presence of 2
mM NADH (or NADPH) at 28 °C led to the conversion of 3
to a new product 4 (Figure 3A, traces i and ii), the yield of
Supporting Information), which was 16 mass units higher than
that of 3, suggesting that 4 was a hydroxylated product of 3.
The 1H NMR spectrum of 4 (Tables S1 and S2 and Figure S11,
Supporting Information) was almost identical to that of 3,
except that an aromatic proton signal (δH 5.95 H-4′) in 3 was
absent in 4, indicating the location of the hydroxyl group at C-
4′ in 4 (Figure 1). The structure assignment of 4 was further
supported by HMBC correlations from H-1 to C-1′/C-2′, from
H-6′ to C-1′/C-2′/C-3′, and from H-7′ to C-5′ (Table S7 and
Figure S12, Supporting Information). Thus, the product 4 was
elucidated as 4′-O-demethyl-piericidin A1 (Figure 1), confirm-
ing PieE as the 4′-hydroxylase in biosynthesis of 1. However,
no conversion of 7 by PieE was detected (Figure 3A, trace v).
There are two methyltransferase-encoding genes pieB1 and
pieB2 in the 1 gene clusters from three different strains (Table
S4, Supporting Information). PieB1 from Streptomyces sp.
SCSIO 03032 shows high homology to S-adenosylmethionine
(SAM)-dependent methyltransferases NorI (65% identity) and
AurI (60% identity) involved in neoaureothin and aureothin
pathways,15 while PieB2 exhibits the highest identity (49%) to
3-demethylubiquinone-9 3-methyltransferase in Rhodococcus
jostii RHA1.16 To probe the exact role of PieB1 and PieB2,
the genes pieB1 and pieB2 were inactivated by PCR-targeting
strategy (Figure S4, Supporting Information). No 1-related
compounds were detected in the ΔpieB1 mutant (Figure 2,
trace iii), while the ΔpieB2 mutant produced a compound that
was identified as 4 on the basis of the same retention times and
molecular masses as the standard 4 (Figure 2, trace iv; Figure
S13, Supporting Information). To further probe functions of
PieB1 and PieB2 in vitro, both pieB1 and pieB2 were cloned
into pET28a and pGEX-6p-1 (Table S6, Supporting
Information). Soluble PieB1 and PieB2 proteins were only
achieved by the fusion with GST-tags and were purified using
GST·Bind Purification Kits (Figure S8, Supporting Informa-
tion). Subsequently, 4 was found to be converted by PieB2 to a
product coeluted with 1 (Figure 3B, traces vi and x), the yield
of which increased with longer incubation time in a time course
assay (Figure S14, Supporting Information). No conversions of
4 to 1 were observed in control assays lacking either PieB2 or
SAM (Figure 3B, traces vii, viii). No change of 4 was detected
by using PieB1 instead of PieB2 (Figure 3B, traces ix). These
data confirm that PieB2 is a specific 4′-O-methyltransferase that
tailors 1 biosynthesis. However, the lack of the putative
substrate 2 (Figure 1) prevents the in vitro functional
characterization of PieB1. Nevertheless, the abolishment of 1
production in the ΔpieB1 mutant suggested that PieB1 was
indeed involved in the biosynthesis of 1, probably responsible
for forming the 5′-O-methyl group in piericidins 3, 7, and 1.
In summary, we have identified the biosynthetic gene cluster
of 1 from the deep-sea derived Streptomyces sp. SCSIO 03032.
PieE was characterized both in vivo and in vitro as the 4′-
hydroxylase to provide the 4′-hydroxyl group in the α-pyridone
moiety of 1. PieB2 was verified as the 4′-O-methyltransferase by
in vivo genetic disruptions and in vitro biochemical assays, and
PieB1 was indicated as the requisite 5′-O-methyltransferase, in
contrast to the methylation of two adjacent hydroxyl groups on
ubiquinone by a single O-methyltransferase Coq3.17
Figure 3. (A) HPLC traces for analyzing in vitro assays of PieE: (i) a
PieE assay containing 100 μM 3, 5 μM PieE, and 2 mM NADH; (ii)
PieE assay in which NADH was replaced by NADPH; (iii) assay in (i)
minus PieE; (iv) assay in (i) minus NAD(P)H; (v) PieE assay
containing 100 μM 7, 5 μM PieE, and 2 mM NADH. A standard PieE
assay was performed in Tris−HCl buffer (50 mM, pH 8.0) at 28 °C
for 2 h. (B) HPLC traces for in vitro assays of PieB2 and PieB1: (vi)
PieB2 assay containing 100 μM 4, 10 μM PieB2, and 2 mM SAM; (vii)
assay in (vi) minus PieB2; (viii) assay in (vi) minus SAM; (ix) PieB1
assay containing 100 μM 4, 10 μM PieB1, and 2 mM SAM; (x)
standard 1. The PieB2 (or PieB1) assays were conducted in Tris−HCl
buffer (50 mM, pH 8.0) at 28 °C for 2 h.
which increased with a longer incubation time (Figure S10,
Supporting Information). Compound 3 remained unchanged in
control assays lacking either PieE or NAD(P)H (Figure 3A,
traces iii, iv). To determine its chemical structure, 10 mg of 4
was isolated and purified from a 15 mL reaction mixture
containing 15 mg of 3, 5 μM PieE, and 2 mM NADH. The
chemical formula of 4 was established as C24H35NO4 by HR-
ESI-MS (m/z [M + H]+ 402.2647, calcd 402.2639, Figure S11,
738
dx.doi.org/10.1021/ol4034176 | Org. Lett. 2014, 16, 736−739

Organic Letters
Letter
(12) (a) Kominato, K.; Watanabe, Y.; Hirano, S.; Kioka, T.;
Terasawa, T.; Yoshioka, T.; Okamura, K.; Tone, H. J. Antibiot. 1995,
48, 103−105. (b) Hoecker, J.; Gademann, K. Org. Lett. 2013, 15, 670−
673.
(13) (a) Yoshida, S.; Yoneyama, K.; Shiraishi, S.; Watanabe, A.;
Takahashi, N. Agric. Biol. Chem. 1977, 41, 855. (b) Kubota, N. K.;
Ohta, E.; Ohta, S.; Koizumi, F.; Suzuki, M.; Ichimura, M.; Ikegami, S.
Bioorg. Med. Chem. 2003, 11, 4569.
(14) (a) Tago, K.; Sato, J.; Takesa, H.; Kawagishi, H.; Hayatsu, M. J.
Biosci. Bioeng. 2005, 100, 517−523. (b) Lu, Y.; Yu, Y.; Zhou, R.; Sun,
W.; Dai, C.; Wan, P.; Zhang, L.; Hao, D.; Ren, H. Biotechnol. Lett.
2011, 33, 1159−1167.
(15) (a) Traitcheva, N.; Jenke-Kodama, H.; He, J.; Dittmann, E.;
Hertweck, C. ChemBioChem 2007, 8, 1841−1849. (b) He, J.;
Hertweck, C. Chem. Biol. 2003, 10, 1225−1232.
ASSOCIATED CONTENT
* Supporting Information
Experimental procedures and materials and characterization
data of compounds. This material is available free of charge via
the Internet at http://pubs.acs.org.
■
S
AUTHOR INFORMATION
Corresponding Authors
■
*E-mail: zhuyiguang2003@163.com.
*E-mail: czhang2006@gmail.com.
Author Contributions
§These authors contribute equally.
(16) McLeod, M. P.; Warren, R. L.; Hsiao, W. W.; Araki, N.; Myhre,
M.; Fernandes, C.; Miyazawa, D.; Wong, W.; Lillquist, A. L.; Wang, D.;
Dosanjh, M.; Hara, H.; Petrescu, A.; Morin, R. D.; Yang, G.; Stott, J.
M.; Schein, J. E.; Shin, H.; Smailus, D.; Siddiqui, A. S.; Marra, M. A.;
Jones, S. J.; Holt, R.; Brinkman, F. S.; Miyauchi, K.; Fukuda, M.;
Davies, J. E.; Mohn, W. W.; Eltis, L. D. Proc. Natl. Acad. Sci. U.S.A.
2006, 103, 15582−15587.
Notes
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
■
This work is supported in part by MOST (2012AA092104 and
2010CB833805), CAS (XDA11030403, KSCX2-EW-G-12),
NSFC (31125001, 31290233), and the Administration of
Ocean and Fisheries of Guangdong Province (GD2012-D01-
001, GD2012-D01-002). We are grateful to analytical facility at
SCSIO for recording spectroscopic data.
(17) James, A. M.; Cocheme, H. M.; Murai, M.; Miyoshi, H.;
Murphy, M. P. FEBS J. 2010, 277, 2067−2082.
REFERENCES
■
(1) (a) Takahashi, N.; Suzuki, A.; Tamura, S. J. Am. Chem. Soc. 1965,
87, 2066−2068. (b) Takahashi, N.; Suzuki, A.; Kimura, Y.; Miyamoto,
S. Tetrahedron Lett. 1967, 21, 1961−1964. (c) Matsumoto, M.; Mogi,
K.; Nagaoka, K.; Ishizeki, S.; Kawahara, R.; Nakashima, T. J. Antibiot.
1987, 40, 149−156. (d) Kroiss, J.; Kaltenpoth, M.; Schneider, B.;
Schwinger, M. G.; Hertweck, C.; Maddula, R. K.; Strohm, E.; Svatos,
A. Nat. Chem. Biol. 2010, 6, 261−263. (e) Shaaban, K. A.; Helmke, E.;
Kelter, G.; Fiebig, H. H.; Laatsch, H. J. Antibiot. 2011, 64, 205−209.
(2) (a) Hall, C.; Wu, M.; Crane, F. L.; Takahashi, H.; Tamura, S.;
Folkers, K. Biochem. Biophys. Res. Commun. 1966, 25, 373−377.
(b) Darrouzet, E.; Issartel, J. P.; Lunardi, J.; Dupuis, A. FEBS Lett.
1998, 431, 34−38.
(3) Hwang, J. H.; Kim, J. Y.; Cha, M. R.; Ryoo, I. J.; Choo, S. J.; Cho,
S. M.; Tsukumo, Y.; Tomida, A.; Shin-Ya, K.; Hwang, Y. I.; Yoo, I. D.;
Park, H. R. J. Cell Physiol. 2008, 215, 243−250.
(4) (a) Schnermann, M. J.; Boger, D. L. J. Am. Chem. Soc. 2005, 127,
15704−15705. (b) Schnermann, M. J.; Romero, F. A.; Hwang, I.;
Nakamaru-Ogiso, E.; Yagi, T.; Boger, D. L. J. Am. Chem. Soc. 2006,
128, 11799−11807. (c) Lipshutz, B. H.; Amorelli, B. J. Am. Chem. Soc.
2009, 131, 1396−1397.
(5) (a) Takahashi, N.; Kimura, Y.; Tamura, S. Tetrahedron Lett. 1968,
1968, 9. (b) Kimura, Y.; Takahashi, N.; Tamura, S. Agr. Biol. Chem.
1969, 33, 1507−1516. (c) Tanabe, M.; Seto, H. J. Org. Chem. 1970,
35, 2087−2088.
(6) Liu, Q.; Yao, F.; Chooi, Y. H.; Kang, Q.; Xu, W.; Li, Y.; Shao, Y.;
Shi, Y.; Deng, Z.; Tang, Y.; You, D. Chem. Biol. 2012, 19, 243−253.
(7) (a) Weber, T.; Laiple, K. J.; Pross, E. K.; Textor, A.; Grond, S.;
Welzel, K.; Pelzer, S.; Vente, A.; Wohlleben, W. Chem. Biol. 2008, 15,
175−188. (b) Halo, L. M.; Heneghan, M. N.; Yakasai, A. A.; Song, Z.;
Williams, K.; Bailey, A. M.; Cox, R. J.; Lazarus, C. M.; Simpson, T. J. J.
Am. Chem. Soc. 2008, 130, 17988−17996.
(8) Zhang, W.; Liu, Z.; Li, S.; Yang, T.; Zhang, Q.; Ma, L.; Tian, X.;
Zhang, H.; Huang, C.; Zhang, S.; Ju, J.; Shen, Y.; Zhang, C. Org. Lett.
2012, 14, 3364−3367.
(9) Massiere, F.; Badet-Denisot, M. A. Cell. Mol. Life Sci. 1998, 54,
205−222.
(10) Shen, B.; Hutchinson, C. R. Biochemistry 1993, 32, 11149−
11154.
(11) Gust, B.; Challis, G. L.; Fowler, K.; Kieser, T.; Chater, K. F. Proc.
Natl. Acad. Sci. U.S.A. 2003, 100, 1541−1546.
739
dx.doi.org/10.1021/ol4034176 | Org. Lett. 2014, 16, 736−739