Characterization of CYP76AH4 clarifies phenolic diterpenoid biosynthesis in the Lamiaceae

Article abstract of DOI:10.1039/c3ob41885e

Miltiradiene (1) is the precursor of phenolic diterpenoids such as ferruginol (2), requiring aromatization and hydroxylation. While this has been attributed to a single cytochrome P450 (CYP76AH1), characterization of the rosemary ortholog CYP76AH4 led to the discovery that these CYPs simply hydroxylate the facilely oxidized aromatic intermediate abietatriene (3). The Royal Society of Chemistry 2013.

Full text of DOI:10.1039/c3ob41885e

Organic &
Biomolecular Chemistry
View Article Online
View Journal | View Issue
COMMUNICATION
Characterization of CYP76AH4 clarifies phenolic
diterpenoid biosynthesis in the Lamiaceae†
Cite this: Org. Biomol. Chem., 2013, 11,
650
7
Jiachen Zi and Reuben J. Peters*
Received 16th September 2013,
Accepted 4th October 2013
DOI: 10.1039/c3ob41885e
www.rsc.org/obc
Miltiradiene (1) is the precursor of phenolic diterpenoids such as an olefin, such as 1, which requires further elaboration to
ferruginol (2), requiring aromatization and hydroxylation. While produce bioactive natural products, particularly the incorpor-
this has been attributed to a single cytochrome P450 (CYP76AH1), ation of oxygen catalysed by cytochromes P450 (CYPs). While
characterization of the rosemary ortholog CYP76AH4 led to the these heme thiolate mono-oxygenases typically catalyse
6
discovery that these CYPs simply hydroxylate the facilely oxidized hydroxylation reactions, members of this enzymatic super-
aromatic intermediate abietatriene (3).
family are known to catalyse more complex reactions, includ-
ing aromatization and multiple reaction cycles with certain
substrates.
Rosemary (Rosmarinus officinalis) has long been used as a fla-
voring agent, and more recently as a nutritional supplement,
as well as in cosmetics, due to its antioxidant properties. Car-
nosic acid (4) and carnosol (5) are the major phenolic diterpe-
noid constituents of rosemary, are responsible for much of its
antioxidant activity, and exhibit a variety of other effects,
7,8
Previous work has identified two diterpene synthases from
S. miltiorrhiza that together cyclize the general diterpenoid pre-
cursor (E,E,E)-geranylgeranyl diphosphate into the tricyclic
olefin 1, with initial bicyclization catalysed by DsCPS, and sub-
1
9
sequent cyclization and rearrangement catalysed by DsKSL.
2,3
including acting as anti-HIV and anticancer agents. More-
over, 4 and 5 are likely intermediates in the biosynthesis of
more elaborated phenolic diterpenoids, such as the tans-
hinones (e.g., 6 and 7; Fig. 1) that make up the bioactive lipo-
philic constituents of the widely used Chinese medicinal herb
This abietane contains a planar cyclohexa-1,4-diene ring that
is poised for aromatization, and recently was shown to be the
precursor of the phenolic diterpenoids 2 and 6 via labelling
10
studies. In addition, via an RNA-Seq approach, the S. miltior-
rhiza CYP76AH1 was identified and reported to catalyse both
4
Danshen (Salvia miltiorrhiza). Accordingly, there is significant
10
aromatization and hydroxylation of 1 to form 2. However, it
interest in elucidation of phenolic diterpenoid biosynthesis.
Phenolic diterpenoids such as 2–7 fall within the labdane-
related superfamily of natural products, whose biosynthesis is
initiated by a sequential pair of cyclization and/or rearrange-
was unclear in what order aromatization and hydroxylation
occur (Scheme 1).
Although phenolic diterpenoid biosynthesis is widely dis-
tributed in the Lamiaceae plant family (e.g., the production of
5
ment reactions. This most often results in the production of
4
and 5 in rosemary), it was unclear how applicable the S. mil-
tiorrhiza results were to other species from this family. Fortui-
tously, RNA-Seq data for a number of medicinal plants,
including rosemary, has recently become available (http://
medicinalplantgenomics.msu.edu). Thus, it is possible to simply
carry out BLAST searches of the rosemary transcriptome,
which is of quite good quality. Of particular interest here,
using CYP76AH1 as the query sequence, four full-length
Fig. 1 Representative phenolic diterpenoids.
Department of Biochemistry, Biophysics & Molecular Biology, Iowa State University,
Ames, IA 50011, USA. E-mail: rjpeters@iastate.edu
†
Electronic supplementary information (ESI) available: Description of materials
and methods. ESI figures and table. Sequences of synthetic genes for CYP76AH1
1
0
&
4–7. See DOI: 10.1039/c3ob41885e
Scheme 1 Reactions putatively catalysed by CYP76AH1.
7
650 | Org. Biomol. Chem., 2013, 11, 7650–7652
This journal is © The Royal Society of Chemistry 2013

View Article Online
Organic & Biomolecular Chemistry
Communication
homologs, CYP76AH4–7, were found. These share 60–85%
amino acid (aa) sequence identity with CYP76AH1 and
5
9–92% aa sequence identity with each other (Fig. S1†), with
CYP76AH4 exhibiting the highest identity with CYP76AH1.
The role of CYP76AH4, as well as the other CYP76AH sub-
family members from rosemary (i.e., CYP76AH5–7), in pro-
duction of 2 was investigated via a synthetic biology approach.
In particular, we have previously demonstrated that whole
gene codon optimization, as well as replacement of the
N-terminal membrane anchor region by a 10 amino acid long
lysine- and serine-rich leader peptide, can be required for func-
1
1–14
tional heterologous expression of plant P450s in E. coli.
Accordingly, this approach was applied to all four candidate Fig. 3 GC-MS chromatograms of in vitro assays of CYP76AH4 with either (A)
purified 1 (selected ions m/z = 272 + 286) or (B) 3 (selected ions m/z = 255 +
CYP76AH sub-family members from rosemary via gene syn-
286).
thesis. These were then co-expressed in E. coli with the requi-
site CYP reductase, along with overexpression of key genes
from the endogenous upstream isoprenoid precursor biosyn-
series of studies were carried out, including use of an N-term-
inally modified and codon optimized gene construct, yielding
the same results – i.e., CYP76AH1 similarly catalyses only
hydroxylation of 3, and does not appear to catalyse aromatiza-
tion of 1 (Fig. S9†).
1
5
thetic pathway, a GGPP synthase, and DsCPS and DsKSL for
production of the putative substrate 1, using a previously
1
6
described modular metabolic engineering system.
As
expected, is easily detectable in cultures expressing
2
CYP76AH4, indicating that this is the rosemary ortholog to the
S. miltiorrhiza CYP76AH1 (Fig. 2).
To determine if it would be possible to separate the facile
spontaneous oxidation and enzymatic hydroxylation, we investi-
gated the mechanism by which aromatization occurs. In par-
ticular, the role of molecular oxygen (O2), which was
investigated by incubation of 1 in phosphate buffer in the pres-
ence or absence of O (removed via bubbling N through the
Having demonstrated that the CYP76AH sub-family plays a
broader role in phenolic diterpenoid biosynthesis in the
Lamiaceae, we began investigating the order of the presumed
dual aromatization and hydroxylation reactions. In particular,
it was possible to easily convert 1 (3 mg obtained from meta-
bolically engineered E. coli) to 3, as verified by NMR analysis
2
2
solution and capping in a gas-tight vial). Significant conver-
sion to 3 was only observed in the presence of O (Fig. S10†).
2
(Fig. S2–7 and Table S1†), simply via exposure to UV-
Given the requirement for O2 in CYP catalysed reactions as
well, it is then not possible to completely separate these
transformations.
irradiation. In vitro assays indicated that CYP76AH4 readily
converted 3 to 2 (Fig. 3A), with KM = 25 ± 6 µM. Surprisingly, a
significant amount of 3 was found in the preparations of 1,
due the presence of 3 in the metabolic engineering system
To further investigate the potential role of 3 in plant pheno-
lic diterpenoid biosynthesis, we carried out phytochemical
analysis of both rosemary and S. miltiorrhiza to determine if 3
could be found, along with 1 (which also has not yet been
reported from rosemary) and 2. All three compounds were
found in hexane extracts of both aerial tissues of rosemary and
hairy root cultures of S. miltiorrhiza (Fig. 4 and S11†). This is
consistent with the hypothesis that 3 is the relevant intermedi-
ate for conversion from 1 to 2. In planta, 3 is present at higher
concentrations than 1, which is the inverse of the ratio
observed in our bacterial metabolic engineering system (cf.,
Fig. 2 and 4), which suggests that aromatization of 1 to 3 may
be enzymatically catalysed in planta.‡
(
Fig. 2), even in the absence of any CYP (Fig. S8†). Moreover,
even following careful purification of 1, increasing amounts of
was always observed upon storage, indicating that 3 can arise
3
from facile spontaneous oxidation. In addition, with fresh
preparations of pure 1, CYP76AH4 was unable to produce sig-
nificant amounts of 2, with the trace amounts that were found
likely attributable to the remaining 3 and/or spontaneous oxi-
dation of 1 to 3 occurring during the course of the enzymatic
reaction (Fig. 3B). Thus, it appears that CYP76AH4 catalyses
only hydroxylation of 3. To determine if this also was true for
1
0
the originally reported activity of CYP76AH1, an analogous
Fig. 2 GC-MS chromatogram of extract from E. coli co-expressing CYP76AH4
and enzymes for the production of 1.
Fig. 4 GC-MS chromatogram of rosemary extract.
Org. Biomol. Chem., 2013, 11, 7650–7652 | 7651
This journal is © The Royal Society of Chemistry 2013

View Article Online
Communication
Organic & Biomolecular Chemistry
another means by which these plants protect themselves from the UV-
irradiation.
1
2
W. Zheng and S. Y. Wang, J. Agric. Food Chem., 2001, 49,
165–5170.
A. Paris, B. Strukelj, M. Renko, V. Turk, M. Pukl, A. Umek
and B. D. Korant, J. Nat. Prod., 1993, 56, 1426–1430.
J. J. Johnson, Cancer Lett., 2011, 305, 1–7.
X. Wang, S. L. Morris-Natschke and K.-H. Lee, Med. Res.
Rev., 2007, 27, 133–148.
R. J. Peters, Nat. Prod. Rep., 2010, 27, 1521–1530.
P. R. Ortiz de Montellano, Chem. Rev., 2010, 110, 932–
5
Scheme 2 Actual role of CYP76AH sub-family members in plant phenolic
diterpenoid biosynthesis.
3
4
Conclusions
5
6
In summary, our results clarify the biosynthesis of phenolic
diterpenoids, confining the role of the characterized CYP76AH
sub-family members to C12-hydroxylation of the aromatic
intermediate 3 (Scheme 2). The presence of 3, along with 1
and 2, in both rosemary and S. miltiorrhiza, is consistent with
such a role for 3 in biosynthesis of phenolic diterpenoids.
Thus, the initially formed olefin intermediate 1 first undergoes
aromatization to 3 prior to formation of 2. While the conver-
sion of 1 to 3 does occurs spontaneously, it seems likely that
this aromatization reaction is enzymatically catalysed in
planta, although the relevant enzyme is yet to be determined,
providing a target for future investigation.
948.
7
8
F. P. Guengerich, Chem. Res. Toxicol., 2001, 14, 611–650.
M. Mizutani and F. Sato, Arch. Biochem. Biophys., 2011, 507,
194–203.
9
W. Gao, M. L. Hillwig, L. Huang, G. Cui, X. Wang, J. Kong,
B. Yang and R. J. Peters, Org. Lett., 2009, 11, 5170–5173.
1
0 J. Guo, Y. J. Zhou, M. L. Hillwig, Y. Shen, L. Yang, Y. Wang,
X. Zhang, W. Liu, R. J. Peters, X. Chen, Z. K. Zhao and
L. Huang, Proc. Natl. Acad. Sci. U. S. A., 2013, 110, 12108–
12113.
11 Q. Wang, M. L. Hillwig, K. Okada, K. Yamazaki, Y. Wu,
S. Swaminathan, H. Yamane and R. J. Peters, J. Biol. Chem.,
2
012, 287, 6159–6168.
2 Q. Wang, M. L. Hillwig and R. J. Peters, Plant J., 2011, 65,
7–95.
Acknowledgements
1
1
1
1
8
This work was supported by a grant from the NIH (GM076324)
and funds from Iowa State University to R.J.P., who also
thanks the Alexander von Humboldt Foundation for sabbatical
fellowship support during the preparation of this manuscript.
3 Q. Wang, M. L. Hillwig, Y. Wu and R. J. Peters, Plant
Physiol., 2012, 158, 1418–1425.
4 Y. Wu, M. L. Hillwig, Q. Wang and R. J. Peters, FEBS Lett.,
2
011, 585, 3446–3451.
5 D. Morrone, L. Lowry, M. K. Determan, D. M. Hershey,
M. Xu and R. J. Peters, Appl. Microbiol. Biotechnol., 2010,
Notes and references
85, 1893–1906.
‡
Intriguingly, if non-enzymatic oxidation plays a role in the conversion of 1 to 3
1
6 A. Cyr, P. R. Wilderman, M. Determan and R. J. Peters,
J. Am. Chem. Soc., 2007, 129, 6684–6685.
17 P. J. Hidalgo, J. L. Ubera, M. T. Tena and M. Valcarcel,
in planta, our findings may provide a rationale for the observed seasonal
variation in content of 5 in rosemary, which seems to heavily depend on
1
7
photoperiod. We speculate that sunlight promotes the conversion of 1 to 3,
increasing the rate at which the natural antioxidant 5 is formed, providing
J. Agric. Food Chem., 1998, 46, 2624–2627.
7
652 | Org. Biomol. Chem., 2013, 11, 7650–7652
This journal is © The Royal Society of Chemistry 2013