Stereoselective Total Synthesis of Bisorbicillinoid Natural Products by Enzymatic Oxidative Dearomatization/Dimerization

Article abstract of DOI:10.1002/anie.201705976

Natural products are a virtually inexhaustible source of small molecules with spectacular molecular architectures and biomedical potential. Their structural complexity generates formidable challenges to total synthesis but often also precludes time- and resource-efficient, stereoselective synthetic access. Biosynthetically, nature frequently uses dimerization and oligomerization reactions to produce highly challenging frameworks from simple starting materials. Impressive examples are the bisorbicillinoids, a family of fungal natural products thought to originate from the polyketide precursor sorbicillin. Utilizing the recombinant oxidoreductase SorbC from the sorbicillin biosynthetic gene cluster, a robust, fully stereoselective synthesis of bisorbicillinoid natural products and unnatural side-chain analogues was developed.

Full text of DOI:10.1002/anie.201705976

A Journal of the Gesellschaft Deutscher Chemiker
Angewandte
Chemie
International Edition
www.angewandte.org
Accepted Article
Title: Stereoselective Total Synthesis of Bisorbicillinoid Natural
Products by Enzymatic Oxidative Dearomatization / Dimerization
Authors: Anna Sib and Tobias Alexander Marius Gulder
This manuscript has been accepted after peer review and appears as an
Accepted Article online prior to editing, proofing, and formal publication
of the final Version of Record (VoR). This work is currently citable by
using the Digital Object Identifier (DOI) given below. The VoR will be
published online in Early View as soon as possible and may be different
to this Accepted Article as a result of editing. Readers should obtain
the VoR from the journal website shown below when it is published
to ensure accuracy of information. The authors are responsible for the
content of this Accepted Article.
To be cited as: Angew. Chem. Int. Ed. 10.1002/anie.201705976
Angew. Chem. 10.1002/ange.201705976
Link to VoR: http://dx.doi.org/10.1002/anie.201705976
http://dx.doi.org/10.1002/ange.201705976

10.1002/anie.201705976
Angewandte Chemie International Edition
COMMUNICATION
Stereoselective Total Synthesis of Bisorbicillinoid Natural
Products by Enzymatic Oxidative Dearomatization / Dimerization
Anna Sib and Tobias A. M. Gulder*
Abstract: Natural products are a virtually inexhaustible source of
small molecules with spectacular molecular architectures and
biomedical potential. Their structural complexity generates formidable
challenges to total synthesis but often also precludes a time- and
resource-efficient, stereoselective synthetic access. Biosynthetically,
Nature frequently uses dimerization and oligomerization reactions to
produce highly challenging frameworks from simple starting materials.
An impressive example are the bisorbicillinoids, a family of fungal
natural products thought to originate from the polyketide precursor
sorbicillin. Utilizing the recombinant oxidoreductase SorbC from the
sorbicillin biosynthetic gene cluster, a robust, fully stereoselective
synthesis of bisorbicillinoid natural products and unnatural side-chain
analogs was developed.
by subsequent dimerization events via Michael addition,
ketalization, and cycloaddition chemistry (Figure 1).
Ever since the discovery of penicillin, specialized metabolites
from fungi continue to fascinate natural product chemists,
medicinal chemists and biochemists alike. A particularly intriguing
class of fungal compounds are the sorbicillinoids.[1] This family
comprises a large number of dimeric natural products with
beautiful,
stereochemically
complex
three-dimensional
architectures and promising biological properties. Examples
include the efficient radical scavenger bisorbicillinol (1),[2] the
cage-like trichodimerol (2)[3] that inhibits production of TNF-α by
targeting prostaglandin H synthase-2,[4] and the β-1,6-glucan
biosynthesis inhibitor[5] bisvertinolone (3)[6] (Figure 1). Trimeric
compounds such as trisorbicillinone A (4)[7] as well as divergently
functionalized sorbicillinol derivatives such as the antileukemic
sorbicillactone A (5)[8] further extend the structural variability of
this intriguing natural product class.
Figure 1. Structures of selected sorbicillinoid natural products 1-5 putatively
formed after oxidation of 6 to the reactive sorbicillinol (7) with subsequent
structural diversification by Diels-Alder cycloaddition (red bonds), Michael-
Addition (green bonds) and ketalization (blue bonds) reactions.
The bisorbicillinoids are all thought to biosynthetically
originate from dimerization of just one building block, sorbicillin (6).
It has long been under debate whether 6 rather constitutes a
stable biosynthetic intermediate on the way to bisorbicillinoids, or
a metabolic end point derived from the reactive intermediate
sorbicillinol (7).[8a,9] Recent work by the Cox group firmly
established the order of biosynthetic events.[10] They conclusively
showed that sorbicillin (6) is the direct product of a polyketide
synthase (PKS). Compound 6 is the substrate of the enzyme
SorbC that catalyzes oxidative dearomatization to 7 as the final
enzyme-derived product of this pathway. Sorbicillinol (7) might
then serve as a common reactive precursor to all bisorbicillinoids
Owing to their structural complexity and biomedical potential,
bisorbicillinoids have attracted broad interest from the synthetic
community, particularly bisorbicillinol (1) and trichodimerol (2)
(Figure 2). The first independent total syntheses of 1 by Corey[11]
and Nicolaou[12] et al. both relied on a biomimetic strategy with
oxidative dearomatization of 6 using Pb(OAc)4 as the key step to
give acetate rac-8. Resolution of this material by HPLC on a chiral
phase and subsequent O-deacetylation under suitable conditions
yielded 1 or 2. This resulted in impressively short routes to these
highly complex compounds. However, the lack of good regio- and
any stereocontrol in the oxidative dearomatization reaction led to
low overall yields. The first progress towards a stereoselective
synthesis of 1 was reported by Pettus et al.[13] Using a chiral tether
in 9 during oxidative dearomatization triggered by hypervalent
iodine chemistry, sorbicillinol synthon 10 was obtained as a
mixture of diastereomers that was without separation further
converted into 1 with 51 % ee. The only truly enantioselective
synthesis of 1, as published by Deng et al.,[14] introduces the
stereochemical information into ketone 11 in the first step of the
synthesis by enantioselective cyanosilylation to quantitatively
deliver 12 in good 92% ee. However, this strategy required a
subsequent eight-step sequence to deliver the PMB-protected
sorbicillinol 13, which, in analogy to above syntheses, was
dimerized to 1 upon PMB removal. In summary, the impressively
[a]
A. Sib, Prof. Dr. T.A.M. Gulder
Biosystems Chemistry, Department of Chemistry and Center for
Integrated Protein Science Munich (CIPSM)
Technical University of Munich
Lichtenbergstraße 4, 85748 Garching (Germany)
E-mail: tobias.gulder@ch.tum.de
Supporting information for this article is given via a link at the end of
the document.
This article is protected by copyright. All rights reserved.

10.1002/anie.201705976
Angewandte Chemie International Edition
COMMUNICATION
transformation. Racemic
8
was prepared by oxidative
dearomatization of sorbicillin (6) utilizing hypervalent iodine
chemistry[15] (see Supporting Information). The material was used
to establish an analytical separation method by HPLC on
CHIRALCEL OD-RH material, leading to the expected 1:1 ratio of
the two chromatographically resolved enantiomers (Figure 2B).
Scheme 1. Total synthetic routes to 1 and 2 by A. Corey (a1: Pb(OAc)4,
AcOH/CH2Cl2 (5:1), 23°C, 30 min., 38 % rac-8  max. 19 % (S)-8 isolated by
HPLC on a chiral phase; a2: NaOMe, MeOH, then NaH2PO4 x H2O, HCl/MeOH,
10%) and Nicolaou (b1: Pb(OAc)4, AcOH/CH2Cl2 (1:1), 0°C, 30 min., 40 % rac-
8  max. 20 % (S)-8 isolated by HPLC on a chiral phase, b2: CsOH x H2O,
MeOH, then NaH2PO4 x H2O, 16%; b3: KOH, THF/H2O (9:1), then 1N HCl, 40%),
B. Pettus and C. Deng et al.
Analysis of
8
derived from the enzymatic oxidative
dearomatization revealed virtually perfect stereocontrol by the
enzyme with an ee > 99.5 %.
We next turned our attention to the synthesis of the
bisorbicillinoid target structures 1 and 2. During the initial test
assays with SorbC compound 7 seemed to be rather stable in the
aqueous buffer system. This is in sharp contrast to the observed
high reactivity of 7 under the conditions reported in the total
syntheses of 1 and 2 published so far (e.g., its dimerization to 1
in THF/H2O = 9:1, cf. Scheme 1). It was thus expected that the
reactivity of 7 towards dimerization will dramatically be increased
by changing solvent polarity. Indeed, when the enzymatic
reactions were quenched with a large excess of dichloromethane
(50:1), the selective formation of a single new product peak at the
cost of 7 was observable by HPLC analysis. Sorbicillinol (7) was
fully consumed within 40 min (Figure 2A). Isolation of the product
by HPLC gave bisorbicillinol (1) in 27% isolated yield (Scheme 2)
based on recovered starting material.[16] Dimerization by Diels-
Alder cycloaddition between two units of 7 thus proved to be very
fast and did not permit formation of any other dimeric analogs. We
thus suspected that by choice of a suitable polar organic co-
solvent during the enzymatic oxidation, the slower formation of
alternative dimerization products might become possible. In fact,
using DMF as organic co-solvent in the biocatalytic process with
subsequent extraction with dichloromethane indeed allowed the
production of trichodimerol (2) in 27% isolated yield as the main
product, with additional 20% of 1. Intriguingly, the product
spectrum was further extendable by extraction of reaction set-ups
containing DMF after 4 hours using dichloromethane, the fast
evaporation of the latter and the addition of pyridine to the
remaining DMF followed by heating. This not only resulted in the
formation of 1 (16%) and 2 (13%) but also gave sorbiquinol (14)[17]
in 7% yield, arising from Diels-Alder reaction of 7 with the side-
chain of sorbicillin (6). To the best of our knowledge, this
constitutes the first total synthesis of 14. By simple variation of the
short biomimetic approaches by Corey, Nicolaou and Pettus
suffer from their lack in stereocontrol, while the enantioselective
Deng strategy results in a comparably long synthetic route.
To combine high stereoselectivity with a most streamlined access
we set out to develop a biocatalytic approach using the cognate
enzyme SorbC in the synthetic key step. The gene sorbC, as
identified in the sorbicillactone A producing strain Penicillium
chryosgenum E01-10/3 by Cox et al.,[10] was used in a codon-
optimized version for expression in E. coli. The gene was cloned
into pET-28a(+) to facilitate heterologous production of SorbC
equipped with an N-terminal hexa-histidine tag, enabling its batch
purification using Ni-NTA affinity chromatography (see Supporting
Information). An initial screening of suitable conditions for the
biocatalytic transformation revealed that 50 mM phosphate buffer
at pH = 8.0 allowed for high enzymatic activity of SorbC. Substrate
6 for the envisioned enzymatic oxidation was available in a three-
step synthetic sequence[12] from 2-methyl resorcinol (see
Supporting Information). Due to the low solubility of 6 in the
aqueous buffer system, acetone (up to 20% of the final volume)
was used to dissolve the substrate prior to its addition to the
reaction mixture. In addition, NADPH was shown to be
replaceable by the cheaper NADH in vitro. To our delight, SorbC
efficiently and selectively formed a single product with a
significantly altered UV spectrum when compared to that of 6, as
expected for an oxidative dearomatization (Figure 2A). LC-MS
analysis revealed a molecular mass consistent with 7, in line with
the functional data reported by Cox et al.[9] To firmly establish
product identity, this reactive intermediate was also intercepted
by O-acylation with acetyl chloride to give the corresponding
stable acetate 8. Having enzymatically produced 8 in hands
facilitated to probe the enantioselectivity of the enzymatic
This article is protected by copyright. All rights reserved.

10.1002/anie.201705976
Angewandte Chemie International Edition
COMMUNICATION
Figure 2. HPLC-UV traces of the Sorb C-catalyzed synthesis of 1 by oxidative
dearomatization of sorbicillin (6, black) to sorbicillinol (7, red) with subsequent
quenching of the reaction using CH2Cl2 to give 1 (green and blue, 20 and 40
min after quenching, respectively). B. Stereoselectivity of the oxidative
dearomatization reaction monitored by HPLC-UV on a chiral phase (OD-RH):
separation of synthetic rac-8 and analysis of enzymatically derived (S)-8.
Scheme 2. Developed methodology for the stereoselective enzymatic total
synthesis of the complex bisorbicillinoids bisorbicillinol (1), trichodimerol (2) and
sorbiquinol (14), also establishing formal routes to bisorbibutenolide (15) and
bisorbicillinolide (16).
product formed rather corresponded to a novel structural
framework containing the bisvertinol core (cf. Figure 1) formed by
an initial Michael-Addition / ketalization sequence with an
additional Michael-Addition of an oxygen function onto the α,β-
unsaturated side-chain to deliver the cyclic ether portion in 21 as
a single diastereomer in 6 % yield. While not yet described in the
literature, similar cyclic ethers might also be formed in the natural
producers. In conclusion, we developed a stereoselective, one-
pot biocatalytic total synthesis of complex bisorbicillinoids using
an enzymatic oxidative dearomatization / dimerization cascade.
This facilitates the direct preparation of sorbiquinol (14),
trichodimerol (2), and bisorbicillinol (1), with additional formal total
syntheses of bisorbibutenolide (15) and bisorbicillinolide (16) by
single rearrangement reactions from 1. The approach combines
the strengths of the land-mark biomimetic syntheses of Corey[11]
and Nicolaou[12] with the high stereoselectivity provided by the
biocatalyst SorbC. It is therefore the most efficient synthetic
access towards bisorbicillinoid natural products yet developed.
The biocatalytic transformations were generally conducted at a
20-80 mg substrate scale, delivering up to 10 mg of the target
molecules in high purity. The methodology is thus suitable for the
fast preparation of mg quantities of the target compounds and
unnatural analogs thereof sufficient to permit future in-depth
structure-activity relationship studies. The uncatalyzed,
spontaneous formation of bisorbicillinoids in vitro strongly
suggests that these compounds are also assembled by non-
work-up of the enzymatic oxidation of 6, three structurally diverse
bisorbicillinoids formed by Michael-Addition / ketalization (2) as
well as core-to-side-chain (14) or core-to-core (1) Diels-Alder
cycloaddition are thus readily accessible in one-pot reactions from
6 (Scheme 2). Compound 1 can furthermore be rearranged to
bisorbibutenolide (15) upon treatment with KHMDS followed by
HCl,[12] or to bisorbicillinolide (16) by exposure to MeOH,[14] thus
additionally concluding formal enantioselective total syntheses of
these natural products.
To test the versatility of this biocatalytic approach, the
preparation of unnatural analogs of 1 and 2 was attempted.
Following the identical synthetic procedure as for the synthesis of
6 but using hexanoic instead of sorbic acid in the Friedel-Crafts
acylation reaction (Supporting Information), the fully saturated
sorbicillin side-chain analog 17 was prepared. To our delight,
SorbC accepted 17 as a substrate, permitting the synthesis of the
respective side-chain saturated bisorbicillinol derivative 18 in 20%
isolated yield when using acetone as the organic co-solvent
(Scheme 3). In analogy to the conversion with the natural
substrate 6, the respective trichodimerol derivative 19 was
accessible in 12% yield when utilizing DMF as a co-solvent during
oxidative dearomatization. Interestingly, when applying the (E)-2-
hexenoic acid-derived sorbicillin derivative 20 as a substrate, the
formation of the expected side-chain analogs of 1 and 2 was
never observed, although substrate turnover to the oxidatively
dearomatized reactive intermediate was clearly visible. The
This article is protected by copyright. All rights reserved.

10.1002/anie.201705976
Angewandte Chemie International Edition
COMMUNICATION
enzymatic reactions in the natural producers. As the product
spectrum of bisorbicillinoids reported from different fungi varies,
this raises questions on how Nature potentially directs the
selective formation of specific dimers. Investigations into the full
synthetic potential and substrate promiscuity of SorbC and into
the elucidation of factors that may govern product selectivity in
Nature are currently in progress in our laboratory.
Keywords: biocatalysis • enzyme catalysis • sorbicillinoids •
natural products • total synthesis
[1]
[2]
A. M. Harned, K. A. Volp, Nat. Prod. Rep. 2011, 28, 1790-1810.
N. Abe, T. Murata, A. Hirota, Biosci. Biotechnol. Biochem. 1998, 62, 661-
666.
[3]
a) R. Andrade, W. A. Ayer, P. P. Mebe, Can. J. Chem., 1992, 70, 2526-
2535. b) Q. Gao, J. E. Leet, S. T. Thomas, J. A. Matson, D. P. Bancroft,
J. Nat. Prod. 1995, 58, 1817-1821.
[4]
[5]
C. E. Mazzucco, G. Warr, J. Leukoc. Biol. 1996, 60, 271-277.
M. Kontani, Y. Sakagami, S. Marumo, Tetrahedron Lett. 1994, 35, 2577-
2580.
[6]
[7]
[8]
L. S. Trifonov, H. Hilpert, P. Floersheim, A. S. Dreiding, D. M. Rast, R.
Skrivanova, L. Hoesch, Tetrahedron 1986, 42, 3157-3179.
D. Li, F. Wang, X. Xiao, Y. Fang, T. Zhu, Q. Gu, W. Zhu, Tetrahedron
Lett. 2007, 48, 5235-5238.
a) G. Bringmann, G. Lang, T. A. M. Gulder, H. Tsuruta, J. Mühlbacher,
K. Maksimenka, S. Steffens, K. Schaumann, R. Stöhr, J. Wiese, J. F.
Imhoff, S. Perovic-Ottstadt, O. Boreiko, W. E. G. Müller, Tetrahedron
2005, 61, 7252-7265. b) G. Bringmann, T. A. M. Gulder, G. Lang, S.
Schmitt, R. Stöhr, J. Wiese, K. Nagel, J. F. Imhoff, Mar. Drugs 2007, 5,
23-30.
[9]
N. Abe, T. Arakawa, K. Yamamoto, A. Hirota, Biosci. Biotechnol.
Biochem. 2002, 66, 2090-2099.A. Al-Fahad, A. Abood, K. M. Fisch, A.
Osipow, J. Davison, M. Avramovic, C. P. Butts, J. Piel, T. J. Simpson, R.
J. Cox, Chem. Sci. 2014, 5, 523-527.
[10] A. Al-Fahad, A. Abood, K. M. Fisch, A. Osipow, J. Davison, M.
Avramovic, C. P. Butts, J. Piel, T. J. Simpson, R. J. Cox, Chem. Sci.
2014, 5, 523-527.
Scheme 3. Chemo-enzymatic synthesis of 18, 19 and 21, unnatural derivatives
of 1 and 2 accessible by using sorbicillin side-chain analogs 17 and 20 as
enzymatic substrates.
[11] D. Barnes-Seeman, E. J. Corey, Org. Lett. 1999, 1, 1503-1504.
[12] a) K. C. Nicolaou, R. Jautelat, G. Vassilikogiannakis, P. S. Baran, K. B.
Simonsen, Chem. Eur. J. 1999, 5, 3651-3665. b) K. C. Nicolaou, G.
Vassilikogiannakis, K. B. Simonsen, P. S. Baran, Y.-L. Zhong, V. P.
Vidali, E. N. Pitsinos, E. A. Couladouros, J. Am. Chem. Soc. 2000, 122,
3071-3079.
Acknowledgements
We thank Elke Duell for initial cloning and functional expression
of sorbC, Stephanie Kohlhepp (AG Tanja Gulder, TUM) for help
with the HPLC analyses on a chiral phase, and Christine Schwarz
for measuring NMR spectra. We are very grateful to the DFG for
generous financial support of this work (GU 1233/1-1 and the
Center for Integrated Protein Science Munich CIPSM).
[13] L. H. Pettus, R. W. Van De Water, T. R. R. Pettus, Org. Lett. 2001, 3,
905-908.
[14] R. Hong, Y. Chen, L. Deng, Angew. Chem. Int. Ed. 2005, 44, 3478-3481.
[15] C. Qi, T. Qin, D. Suzuki, J. A. Porco, J. Am. Chem. Soc. 2014, 136, 3374-
3377.
[16] As the biocatalytic transformations were conducted with an excess of
substrate, all yields are given based on recovered starting material.
Product isolation was readilly achieved by semipreperative HPLC.
[17] R. Andrade, W. A. Ayer, L. S. Trifonov, Can. J. Chem. 1996, 74, 371-
379.
Conflict of interest
The authors declare no conflict of interest.
This article is protected by copyright. All rights reserved.

10.1002/anie.201705976
Angewandte Chemie International Edition
COMMUNICATION
Entry for the Table of Contents (Please choose one layout)
Layout 2:
COMMUNICATION
Anna Sib and Tobias A. M. Gulder*
Page No. – Page No.
Stereoselective Total Synthesis of
Bisorbicillinoid Natural Products by
Enzymatic Oxidative Dearomatization
/ Dimerization
An enzyme for diversity. A first chemo-enzymatic route to structurally challenging
bisorbicillinoid natural products based on the highly stereoselective oxidative
dearomatization of sorbicillin with SorbC is presented. The streamlined methodology
opens efficient access to a broad range of architecturally diverse compounds of this
biomedically interesting natural product family.
This article is protected by copyright. All rights reserved.