Revelation of the Balanol Biosynthetic Pathway in Tolypocladium ophioglossoides

Article abstract of DOI:10.1021/acs.orglett.8b01543

A cryptic gene cluster, bln, was activated by genome mining in Tolypocladium ophioglossoides. This activation led to the production of balanol and eight other metabolites. Gene disruption and metabolite profile analysis showed that the biosynthesis of balanol involved the convergence of independent PKS and NRPS pathways, and a biosynthetic pathway for balanol was proposed.

Full text of DOI:10.1021/acs.orglett.8b01543

Letter
Cite This: Org. Lett. XXXX, XXX, XXX−XXX
pubs.acs.org/OrgLett
Revelation of the Balanol Biosynthetic Pathway in Tolypocladium
ophioglossoides
Xian He, Min Zhang, Yuan-Yang Guo, Xu-Ming Mao, Xin-Ai Chen,* and Yong-Quan Li*
Institute of Pharmaceutical Biotechnology, Zhejiang University, Hangzhou 310058, China
*
S Supporting Information
ABSTRACT: A cryptic gene cluster, bln, was activated by
genome mining in Tolypocladium ophioglossoides. This
activation led to the production of balanol and eight other
metabolites. Gene disruption and metabolite profile analysis
showed that the biosynthesis of balanol involved the
convergence of independent PKS and NRPS pathways, and
a biosynthetic pathway for balanol was proposed.
alanol (4) was first isolated from Tolypocladium ophioglos-
1
2−4
B
soides in 1977 as ophiocordin.
This compound is
known to be a potent natural protein kinase C (PKC) inhibitor
that can bind to the ATP-binding pockets of PKC and protein
kinase A (PKA) with an affinity 3000 times greater than that of
ATP due to its three-dimensional structural resemblance to
5
ATP. As a hybrid polyketide-nonribosomal peptide (PK-NRP)
compound, balanol consists of a benzophenone, an azepine, and
a p-hydroxybenzamide unit, and thus, balanol is unlikely to be
synthesized by commonly integrated PKS-NRPS enzymes. Since
the yield of balanol is very low when extracted from the fungus
Figure 1. Comparison of the balanol (bln) and monodictyphenone
mdp) gene clusters from T. ophioglossoides.
(
3
under laboratory conditions, many studies have focused on the
6
−13
chemical synthesis of balanol and its analogues.
The
16
ing, and heterologous expression. BlnR is a pathway-specific
Zn Cys transcriptional regulator in the bln gene cluster and is
chemical synthesis and structural modification of balanol are
complex and challenging, and the biosynthetic pathway of
balanol has not yet been reported.
2
6
homologous to mdpE from the mdp cluster. Since replacing the
promoter of mdpE with a regulatable promoter has allowed the
Recently, we sequenced the genome of the previously isolated
15,17
14
induction of expression of the mdp cluster in A. nidulans,
we
T. ophioglossoides strain. Bioinformatic analysis indicates that
the genome has the potential to generate 13 polyketides, 8
nonribosomal peptides, 6 hybrid PK-NRP metabolites and 4
terpenes (Tables S2−S5). Among these biosynthetic gene
clusters, we found an intriguing cluster, namely, bln, which
contains a PKS and two NRPS genes. Some genes, such as the
PKS gene, in the bln cluster exhibited relatively high homology
with genes from the monodictyphenone biosynthetic cluster in
constructed a vector for the overexpression of bln R, which was
under the control of the strong promoter ApTEF. The obtained
transformants (blnROE) were verified by RT-PCR (Figure S2).
The culture broth of the blnROE strain was a darker shade of
yellow than that of the wt strain (Figure S1), and HPLC analysis
of the blnROE culture showed a distinctly different metabolite
profile from that of the wt strain (Figure 2). To characterize the
new compound produced by the blnROE strain, large-scale
fermentation was carried out. Ninemetabolites were isolatedand
identified (Figure 3), including two nonribosomal peptides (1,
1
5
Aspergillus nidulans (Figure 1). As balanol contains moieties
derived from amino acids and a benzophenone moiety that is
structurally similar to monodictyphenone (Figure 1), we
speculated that this gene cluster could be associated with
balanol biosynthesis. Transcriptional analysis showed that the
expression of genes within the bln cluster could not be detected
2
), five polyketide compounds (3, 5, 7, 8, 9), and two
compounds containing both polyketide and amino acid derived
moieties (4, 6). Among these compounds, compound 7 has not
been reported to date, and 1, 2, 3, and 6 were isolated from
nature for the first time.
(
data not shown), which suggests that bln was a cryptic gene
cluster under the laboratory conditions used.
Techniques have been developed to activate silent gene
clusters, including variation in growth conditions, manipulation
of regulatory genes, epigenetic perturbation, promoter engineer-
Received: May 17, 2018
©
XXXX American Chemical Society
A
DOI: 10.1021/acs.orglett.8b01543
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
was further elucidated by extensive analysis of 1D and 2D NMR
spectroscopic data (Table S13 and Figures S23−S26).
Compound 6 was identified to be a known analog of balanol,
which has been chemically synthesized previously for examina-
1
9
tion of its inhibitory activity toward protein kinases.
Compound 4 can be synthesized by esterification of 1 and 3.
To elucidate the biosynthetic pathway of balanol, we further
determined the borders of the bln gene cluster via analysis of the
expression of putative genes by qRT-PCR based on
20
antiSMASH prediction. Compared with the expression in
the wt strain, the expression of 18 genes in the blnROE strain,
from blnA to blnQ, was significantly upregulated, while there was
no significant change in the expression of other genes (Figure 4).
Figure 2. HPLC profiles (λ = 254 nm) of the culture broths of the wt
and blnROE strains of T. ophioglossoides.
Figure 4. Determination of the borders of the balanol cluster by qRT-
PCR.
This result demonstrates that the bln biosynthetic cluster spans
the region from blnA to blnQ, comprising 18 genes
Figure 3. Structures of compounds 1−9 from blnROE.
(
approximately 42 kilobases in length), including one PKS
The molecular formula of compound 1 was established as
gene and two NRPS genes; the detailed annotations are shown in
the Supporting Information (Table S6).
In addition to several genes of the bln cluster exhibiting
similarity with genes from the mdp cluster of A. nidulans, five
monodictyphenone analogs (3, 5, 7, 8, and 9) have been isolated
from the blnROE strain. Since the mdp cluster had been
C H N O by high-resolution mass spectrometry (HRMS)
1
3
18
2
3
(
Table S7). The structure of 1 was elucidated by extensive
analysis of one- and two-dimensional (2D) NMR spectroscopic
data (Table S8 and Figures S3−S6). Compound 1 was
18
previously synthesized to confirm the structure of balanol.
The molecular formula of 2 was established as C H N O by
15
1
3
18
2
2
elucidated, we focused on the genes in the bln cluster that were
−
HRMS (m/z 233.1299 [M − H] ). Compared to the known
compound 1, compound 2 lacked a hydroxyl group. Therefore,
the planar structure of 2 was further determined via combination
of the H, C, and 2D NMR spectroscopic data (Table S9 and
Figures S7−S10).
not homologous to the mdp cluster. To further investigate the
functions of these genes, several genes in the bln cluster were
disrupted by homologous recombination in the blnROE
background. The genotypes of the deletion strains were
confirmed by PCR (Figure S34−S40). The effect of gene
deletion on the production of compounds 1−9 in fermented
broth was analyzed by HPLC (Figure S41) and LC−MS (Table
S17 and Figures S42−S51), as summarized in Table 1.
1
13
The structures of 3, 5, 8, and 9 were elucidated by
spectroscopic analysis (Tables S7, S10, S12, S15, and S16 and
Figures S11−S14, S19−S22, and S31−S33). Compound 3 was
formerly synthesized together with 1 to confirm the structure of
1
8
balanol. The molecular formula of 7 was established as
Table 1. Effects of the Deletion of bln Genes on the
Production of Compounds 1−9
−
C H O by HRMS (m/z 299.0189 [M - H] ), and the UV
a
1
5
8
7
absorption data suggested that compounds 5, 7, and 9 had the
same xanthone skeleton (Table S7). Compound 7 seems to have
a carboxyl group instead of a hydroxymethyl group at the C-6
position, in contrast to 5. In addition, the structure of 7 was
effect on the production of compounds 1−9
strain
bln ROE
1
2
3
4
5
6
7
8
9
+
+
+
+
+
+
+
−
+
+
+
+
+
−
+
+
+
+
−
+
+
−
+
+
+
+
+
+
+
+
+
+
−
+
+
+
+
+
1
13
identified by H, C, and 2D NMR spectroscopic data (Table
S14 and Figures S27−S30) as a previously unreported structure.
The structure of 4 (balanol) was identified by HRMS and 1D
and 2D NMR spectroscopy (Table S11 and Figures S15−S18).
Compound 4 has been previously purified from T. ophioglos-
bln ROEΔbln E
bln ROEΔbln F
bln ROEΔbln J
bln ROEΔbln L
bln ROEΔbln N
bln ROEΔbln O
+
+
+
+
+
−
−
+
−
−
−
−
+
−
−
−
R
R
+
+
+
−
+
−
+
2
soides, but weobtainedthis compound onlyafter overexpression
R
−
+
of blnR, as shown in the HPLC analysis (Figure 2). The
a
molecular formula of 6 was deduced by HRMS (m/z 531.1409
+ indicates production. − indicates absence of production. R
indicates reduced production level.
−
[
M − H] ) as C H N O , and the structure of this compound
2
8
24
2
9
B
DOI: 10.1021/acs.orglett.8b01543
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
Figure 5. Putative biosynthetic pathway of balanol in T. ophioglossoides.
Bln E is a hypothetical protein-encoding gene. Deletion of bln
E led to the absence of xanthones 5, 7, and 9, suggesting that bln
E is involved in the cyclodehydration of upstream benzophe-
nones to produce xanthones. The fact that 6 could still be
produced in the absence of 7 suggests that 6 is more likely
produced by cyclization of 4 than by an esterification reaction
between 7 and 1. When the PKS gene was disrupted, production
of the nonribosomal peptides 1 and 2 remains unaffected,
indicating the existence of a relatively independent NRPS
pathway during balanol biosynthesis.
Bln L is a cytochrome-P450-encoding gene, and deletion of bln
L abolished the production of 3, 4, 5, 6, and 7. Since 3 is an
essential precursor of 4, the absence of 3 seemed to result in
elimination of the hybrid PK-NRP compounds 4 and 6.
However, the xanthone compound 9 was still produced,
implying that Bln L is involved in catalyzing the generation of
major role in the synthesis of 1 and 2. There are six NRPS
domains in Bln O, namely, A -T -C -A -T -TE, as revealed by
1
1
2
2
2
analysis ofthe amino acidsequence ofthis protein. TheA andA
1
2
domains are predicted to activate Phe/Tyr/Arg/Leu and Arg/
Glu/Val/Leu/Trp/Lys/Gln, respectively (Tables S18 and S19).
Based on the structures of 1 and 2, it can be deduced that 2 likely
originates from intramolecular cyclization of L-lysine linked via
an amide bond with p-hydroxybenzoic acid, which is followed by
carbonyl reduction. This speculation was supported by feeding
1
3
experiment of labeled C -L-lysine in culture broth of bln ROE
6
strain, which led to the formation of labeled 4 with a mass shift of
6 Da to higher masses (Figure S55). The additional hydroxyl
group in 1 compared with 2 suggests that 1 is produced by
oxidation of 2. While the production of the nonribosomal
peptides 1 and 2 was eliminated by bln O deletion, the
biosynthesis of polyketides was not affected, suggesting that the
biosyntheses of polyketides and nonribosomal peptides from the
bln cluster are processed in parallel.
3
from the upstream benzophenone of 9 via oxidation of the
methyl group to a carboxylic acid group.
There are two NRPS genes in the bln cluster, namely, blnN and
blnO. BlnN encodes three NRPS domains, namely, A -T -C , in
Bln F is a predicted dioxygenase. In the blnROEΔblnF strain,
while the biosynthesis of 2 was not affected, the hydroxylated
derivative 1 was absent, which eliminated the production of 4
and 6 due to loss of ester bond formation. This result suggests
that Bln F could catalyze the hydroxylation of 2 to produce 1. To
further verify our speculation, blnF was cloned and expressed in
E. coli. Formation of 1 is observed when Bln F was incubated with
2, α-KG, Fe(NH ) (SO ) and L-ascorbate (Figure S53).
3
3
3
which A is predicted to activate Hyv/Orn/Leu/Gly/Glu (Table
3
S20). Upon deletion of blnN, the hybrid PK-NRP compounds 4
and 6 could not be detected, and the biosynthesis of other
compounds was not affected. Both 4 and 6 have an
intermolecular ester bond that connects the PK and NRP
moieties. It has been reported that NRPS can catalyze ester bond
4
2
4 2
21−24
formation by utilizing thioesterase (TE) or C domains.
Bln J is predicted to be a phenylalanine/tyrosine ammonia-
lyase. In the bln ROEΔbln J strain, the production of compounds
1 and 2 was significantly reduced, which suggests that Bln J might
be related to the biosynthesis of p-hydroxybenzoic acid as the
substrate of Bln O. We expressed bln J in E. coli and performed in
vitro biochemical reaction, the result verified that Bln J could
catalyze the deamination of tyrosine to produce p-coumaric acid
(Figure S54), and p-coumaric acid could be further catabolized
Taken together, our data show that NRPS Bln N is responsible
for catalyzing ester bond formation between the polyketide 3 and
the nonribosomal peptide 1 to produce 4.
Upon deletion of bln O, the production of compounds 1, 2, 4,
and 6 was not observed. The absence of 4 and 6 could be a result
of 1 deficiency. This result, together with the fact that the
biosynthesis of the nonribosomal peptides 1 and 2 are not
affected in the bln ROEΔbln N strain, suggests that Bln O plays a
2
5
to generate p-hydroxybenzoic acid. The production of
C
DOI: 10.1021/acs.orglett.8b01543
Org. Lett. XXXX, XXX, XXX−XXX

Organic Letters
Letter
compounds 1 and 2 was only reduced but not eliminated since p-
REFERENCES
1) Quandt, C. A.; Kepler, R. M.; Gams, W.; Arauj
Evans, H. C.; Hughes, D.; Humber, R.; Hywel-Jones, N.; Li, Z. IMA
■
hydroxybenzoic acid could also be originated from the shikimic
(
́
o, J. P.; Ban, S.;
2
6
acid pathway.
These results show that the biosynthesis of balanol involves
two convergent biosynthetic pathways, in which PKS is
responsible for biosynthesis of benzophenone and other
polyketides and the NRPS Bln O produces the dipeptides 1
and 2. Coupling of the parallel products is achieved by
esterification of 3 and 1 to generate balanol (4), which is
catalyzed by the NRPS Bln N. This result is consistent with the
fungus 2014, 5, 121−134.
(2) Kneifel, H.; Konig, W. A.; Loeffler, W.; Muller, R. Arch. Microbiol.
̈
1
(
977, 113, 121−130.
3) Kulanthaivel, P.; Hallock, Y. F.; Boros, C.; Hamilton, S. M.; Janzen,
W. P.; Ballas, L. M.; Loomis, C. R.; Jiang, J. B.; Katz, B. J. Am. Chem. Soc.
993, 115, 6452−6453.
4) Boros, C.; Hamilton, S. M.; Katz, B.; Kulanthaivel, P. J. Antibiot.
994, 47, 1010−1016.
5) Pande, V.; Ramos, M. J.; Gago, F. Anti-Cancer Agents Med. Chem.
1
(
1
(
27
hypothesis introduced by Molnar. Based on a combination of
́
15
the studies conducted on the homologous mdp cluster and the
findings of our study, a biosynthetic pathway for balanol was
proposed (Figure 5). The mechanism for the removal of the
ketone of the amide bond during balanol biosynthesis will
require further study.
2008, 8, 638−645.
(6) Lampe, J. W.; Hughes, P. F.; Biggers, C. K.; Smith, S. H.; Hu, H. J.
Org. Chem. 1994, 59, 5147−5148.
(7) Nicolaou, K.; Bunnage, M. E.; Koide, K. J. Am. Chem. Soc. 1994,
1
(
16, 8402−8403.
8) Adams, C.; Fairway, S.; Hardy, C.; Hibbs, D.; Hursthouse, M.;
Morley, A.; Sharp, B.; Vicker, N.; Warner, I. J. Chem. Soc., Perkin Trans. 1
995, 2355−2362.
9) Lai, Y.-S.; Mendoza, J. S.; Jagdmann, G. E.; Menaldino, D. S.;
In summary, we identified the balanol biosynthetic gene
cluster bln for the first time by genome mining. By over-
expression of the pathway-specific regulatory gene blnR from the
bln cluster, the biosynthesis of balanol and other intermediate
compounds was activated. By gene deletion and analysis of the
metabolite profile, the biosynthetic pathway of balanol,
containing independent PKS and NRPS pathways, was
proposed. Convergence of these pathways is achieved by
intramolecular ester bond formation between parallel products,
which is catalyzed by the NRPS Bln N. The discovery of the
balanol biosynthetic pathway will provide improved access to
balanol and its analogues and could also improve our
understanding of the mechanisms of fungal hybrid PK-NRP
metabolite biosynthesis.
1
(
Biggers, C. K.; Heerding, J. M.; Wilson, J. W.; Hall, S. E.; Jiang, J. B.;
Janzen, W. P. J. Med. Chem. 1997, 40, 226−235.
(10) Miyabe, H.; Torieda, M.; Kiguchi, T.; Naito, T. Synlett 1997,
1997, 580−582.
(
11) Tanner, D.; Tedenborg, L.; Almario, A.; Pettersson, I.; Cso
I.; Kelly, N. M.; Andersson, P. G.; Hogberg, T. Tetrahedron 1997, 53,
857−4868.
12) Miyabe, H.; Torieda, M.; Inoue, K.; Tajiri, K.; Kiguchi, T.; Naito,
T. J. Org. Chem. 1998, 63, 4397−4407.
13) Denieul, M.-P.; Laursen, B.; Hazell, R.; Skrydstrup, T. J. Org.
̈
regh,
̈
4
(
(
Chem. 2000, 65, 6052−6060.
(14) Xu, Q.; Lu, L.; Chen, S.; Zheng, J.; Zheng, G.; Li, Y. Chin. J. Chem.
Eng. 2009, 17, 278−285.
(15) Chiang, Y.-M.; Szewczyk, E.; Davidson, A. D.; Entwistle, R.;
ASSOCIATED CONTENT
■
Keller, N. P.; Wang, C. C.; Oakley, B. R. Appl. Environ. Microb. 2010, 76,
2067−2074.
*
S
Supporting Information
(
16) Rutledge, P. J.; Challis, G. L. Nat. Rev. Microbiol. 2015, 13, 509−
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.orglett.8b01543.
523.
(17) Bok, J. W.; Chiang, Y.-M.; Szewczyk, E.; Reyes-Dominguez, Y.;
Davidson, A. D.; Sanchez, J. F.; Lo, H.-C.; Watanabe, K.; Strauss, J.;
Oakley, B. R.; Wang, C. C. C.; Keller, N. P. Nat. Chem. Biol. 2009, 5,
Supplementary methods, figures, tables, compound
characterization, and NMR spectra of compounds 1-9
4
(
4
(
62−464.
18) Nicolaou, K.; Koide, K.; Bunnage, M. E. Chem. - Eur. J. 1995, 1,
54−466.
19) Lampe, J. W.; Biggers, C. K.; Defauw, J. M.; Foglesong, R. J.; Hall,
(PDF)
AUTHOR INFORMATION
S. E.; Heerding, J. M.; Hollinshead, S. P.; Hu, H.; Hughes, P. F.;
■
Jagdmann, G. E. J. Med. Chem. 2002, 45, 2624−2643.
Corresponding Authors
(20) Blin, K.; Medema, M. H.; Kazempour, D.; Fischbach, M. A.;
*
E-mail: lyq@zju.edu.cn.
Breitling, R.; Takano, E.; Weber, T. Nucleic Acids Res. 2013, 41, W204−
W212.
*E-mail: biolab@zju.edu.cn.
(
21) Kohli, R. M.; Walsh, C. T. Chem. Commun. 2003, 297−307.
ORCID
(
22) Zaleta-Rivera, K.; Xu, C.; Yu, F.; Butchko, R. A.; Proctor, R. H.;
Hidalgo-Lara, M. E.; Raza, A.; Dussault, P. H.; Du, L. Biochemistry 2006,
5, 2561−2569.
23) Lin, S.; Van Lanen, S. G.; Shen, B. Proc. Natl. Acad. Sci. U. S. A.
009, 106, 4183−4188.
24) Li, W.; Fan, A.; Wang, L.; Zhang, P.; Liu, Z.; An, Z.; Yin, W. B.
Chem. Sci. 2018, 9, 2589−2594.
25) Jung, D. H.; Kim, E. J.; Jung, E.; Kazlauskas, R. J.; Choi, K. Y.; Kim,
B. G. Biotechnol. Bioeng. 2016, 113, 1493−1503.
26) Heleno, S. A.; Martins, A.; Queiroz, M. J.; Ferreira, I. C. Food
Yong-Quan Li: 0000-0001-6013-4068
Notes
4
(
2
(
The authors declare no competing financial interest.
ACKNOWLEDGMENTS
(
■
We thank Jian-Yang Pan (Zhejiang University, China) for NMR
data collection and Dr. Sheng-Hua Ying (Zhejiang University,
China) for providing the SUR gene. We thank the editors at ACS
ChemWorx Authoring Services for their English language
editing of the manuscript. This work was supported by NSFC
projects (31520103901, 31370064) and by the Science and
Technology Planning Project of Zhejiang Province, China
(
Chem. 2015, 173, 501−513.
(
27) Molnar
́
, I.; Gibson, D. M.; Krasnoff, S. B. Nat. Prod. Rep. 2010, 27,
1241−1275.
(
2017C33137).
D
DOI: 10.1021/acs.orglett.8b01543
Org. Lett. XXXX, XXX, XXX−XXX