Unexpected Metabolic Versatility in a Combined Fungal Fomannoxin/Vibralactone Biosynthesis

Article abstract of DOI:10.1021/acs.jnatprod.6b00147

The secondary metabolome of an undescribed stereaceous basidiomycete (BY1) was investigated for bioactive compounds. Along with a known fomannoxin derivative and two known vibralactones, we here describe three new compounds of these natural product famili

Full text of DOI:10.1021/acs.jnatprod.6b00147

Article
pubs.acs.org/jnp
Unexpected Metabolic Versatility in a Combined Fungal
Fomannoxin/Vibralactone Biosynthesis
Daniel Schwenk,† Philip Brandt,† Robert A. Blanchette,‡ Markus Nett,§,⊥ and Dirk Hoffmeister*,†
†Department of Pharmaceutical Microbiology at the Hans-Knoll-Institute, Friedrich-Schiller-Universitat, Beutenbergstrasse 11a,
̈
̈
07745 Jena, Germany
‡Plant Pathology, University of Minnesota, 1991 Upper Buford Circle, Saint Paul, Minnesota 55108, United States
§Leibniz Institute for Natural Product Research and Infection Biology, Hans-Knoll-Institute, Beutenbergstrasse 11a, 07745 Jena,
̈
Germany
⊥Department of Biochemical and Chemical Engineering, Technical Biology, Technical University Dortmund, Emil-Figge-Strasse 66,
44227 Dortmund, Germany
S
* Supporting Information
ABSTRACT: The secondary metabolome of an undescribed stereaceous basidiomycete (BY1)
was investigated for bioactive compounds. Along with a known fomannoxin derivative and two
known vibralactones, we here describe three new compounds of these natural product families,
whose structures were elucidated using 1D and 2D NMR spectroscopy and high-resolution mass
spectrometry. The new compound vibralactone S (4) shows a 3,6-substituted oxepin-2(7H)-one
ring system, which is unprecedented for the vibralactone/fomannoxin class of compounds.
Stable isotope labeling established a biosynthetic route that is dissimilar to the two published
cascades of oxepinone formation. Another new compound, the antifungal methyl seco-
fomannoxinate (6), features a 2-methylprop-1-enyl ether moiety, which is only rarely observed
with natural products. The structure of 6 was confirmed by total synthesis. 13C-labeling
experiments revealed that the unusual 2-methylprop-1-enyl ether residue derives from an
isoprene unit. The diversity of BY1’s combined fomannoxin/vibralactone metabolism is
remarkable in that these compound families, although biosynthetically related, usually occur in
different organisms.
he order Russulales represents an evolutionary lineage of
the basidiomycetes that includes saprotrophic, symbiotic,
rearranged Δ6-protoilludene-derived products.9 Besides sesqui-
terpenes, Heterobasidion produces fomannoxins, such as 5
(Figure 1). These are phytotoxic benzofurans and their
derivatives10−13 that support the infection process, as they are
present in infected conifer heartwood.14 Fomannoxin produc-
tion was also elicited through coculturing of H. annosum with
spruce cells in vitro.11 A hybrid biosynthetic route is taken to
elaborate the fomannoxins, as the shikimic acid pathway
supplies the aromatic portion to which is then added a prenyl
moiety.15 We recently investigated the metabolism of
basidiomycete BY1 that was isolated from dead aspen wood
showing signs of initial white rot.16 Although the taxonomy of
this fungus has not fully been clarified, DNA sequencing of the
internal transcribed spacer (ITS) region of its genome
suggested an evolutionary relationship to the Russulales and,
more specifically, to the family Stereaceae. Initially, the
rationale behind our research into the BY1 secondary
metabolome was more profound insight in its antifungal and
antilarval effects. The latter could be assigned to methyl-
branched polyene polyketides. However, neither does this class
T
and parasitic species. As white-rotting lignocellulose degraders
or heterotrophic partners in mycorrhizal tree-fungus symbioses,
they play a critical role in global carbon cycling and the healthy
functioning of forest ecosystems. The Russulales also comprise
the Heterobasidion annosum species complex, i.e., detrimental
forest pathogens that negatively impact conifer forests in the
northern hemisphere, as they cause a root and butt rot.1,2 Given
the ecological relevance of Heterobasidion and the Russulales in
general, the secondary metabolism of these fungi has
extensively been investigated and revealed intriguing com-
pounds, such as the vibralactones 1 and 2 from Boreostereum
vibrans (Figure 1).3,4 Very recently, vibralactone oximes have
been described from the same fungus.5 For molecules of this
class of compounds, various bioactivities have been recognized,
among them inhibition of pancreatic lipase3 as well as modest
inhibition of mammalian 11β-hydroxysteroid dehydrogenases.4
Also, 1 covalently binds and, thus, inhibits both caseinolytic
proteases ClpP1 and ClpP2 of the human pathogenic
bacterium Listeria monocytogenes.6
As a severe phytopathogen, H. annosum (sensu lato) received
particular interest regarding identification of its small-molecule
toxins, among them sesquiterpenes, such as the fomannosins7
and the fomajorins8 along with other highly functionalized and
Received: February 19, 2016
© XXXX American Chemical Society and
American Society of Pharmacognosy
A
DOI: 10.1021/acs.jnatprod.6b00147
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
vibralactones and fomannoxins by the same species in
conjunction with the above newly described derivatives
underscores an unusually diversified metabolism.
RESULTS AND DISCUSSION
■
Isolation and Structure Elucidation of Natural
Products. Ethyl acetate extracts of three-week-old BY1
cultures displayed antifungal activity in the agar diffusion
assay. Bioactivity-guided chromatographic fractionation led to
the isolation of various compounds, among them vibralactone 1
and vibralactone acetate 2, which were identified by comparison
of their spectroscopic data with literature values.3,17 The
capacity of BY1 to produce vibralactones that were previously
isolated from Boreostereum and Stereum species underscores its
phylogenetic relationship within the Stereaceae family of
basidiomycetes.3,18 This taxonomic assignment was already
suggested from an analysis of the ITS region of BY1.16 Two
further metabolites with antifungal properties, now referred to
as vibralactones R (3) and S (4) (Figure 1), were identified.
HRESIMS of 3 indicated a molecular formula of C11H14O2,
which corresponds to five degrees of unsaturation. A strong
absorption band at 1730 cm−1 in the IR spectrum suggested the
presence of an ester or aldehyde moiety. The latter possibility
was ruled out after inspecting the 13C NMR spectrum, as the
only carbonyl resonance was observed at 171.4 ppm (Table 1,
Figures S2 and S4). Six additional carbon atoms (C-3, C-6, C-7,
C-8, C-9, C-10) are sp2-hybridized according to their chemical
shifts. Thus, compound 3 must feature three carbon−carbon
double bonds and one ring structure to comply with the
Figure 1. Chemical structures of small molecules isolated from the
basidiomycete BY1: (A) vibralactone (1), vibralactone acetate (2),
vibralactone R (3), and vibralactone S (4); (B) 5-formyl-2-(isopropyl-
1′-ol)-2,3-dihydrobenzofuran (5) and methyl seco-fomannoxinate (6).
of compounds allow chemotaxonomic inferences, nor did the
polyenes exert antifungal effects.
We here describe reinvestigation of this fungus’s secondary
metabolome (i) to support the taxonomic placement on the
grounds of natural product biosyntheses and (ii) to identify the
antifungal principle of BY1’s mycelial extracts. Along with the
three known compounds vibralactone (1), vibralactone acetate
(2), and 5-formyl-2-(isopropyl-1′-ol)-2,3-dihydrobenzofuran
(5), three new fomannoxin- and vibralactone-type natural
products are described (Figure 1), whose structures were
elucidated by interpretation of 1D and 2D NMR data and high-
resolution mass spectrometry. The newly described molecules
include the γ-lactone vibralactone R (3), the 2(7H)-oxepinone
vibralactone S (4), which features an unprecedented sub-
stitution pattern in this family of compounds, and the
antifungal methyl seco-fomannoxinate (6). The production of
1
required degrees of unsaturation. The signals in the H NMR
spectrum (Figure S1) were attributed to their directly attached
carbon atoms by heteronuclear single-quantum coherence
(HSQC, Figure S6). Proton−proton correlation spectroscopy
(COSY, Figure S5) revealed two discrete spin systems. The first
spin system included four olefinic protons (H-6, H-7, H-8, H-
9), while the second spin system was composed of two
methylene groups, i.e., CH2-4 and CH2-5. The two spin systems
a
Table 1. NMR Spectroscopic Data for Vibralactone R (3), Vibralactone S (4), and Methyl seco-Fomannoxinate (6)
3
4
6
no.
δC, mult
δH, M (J in Hz)
no.
δC, mult
δH, M (J in Hz)
no.
δC, mult
δH, M (J in Hz)
1
2
3
4
5
6
7
1
2
3
4
5
6
7
1
2
3
4
5
6
1′
122.6, C
128.3, CH
130.9, C
157.4, C
112.4, CH
129.5, CH
166.0, C
171.4, C
168.9, C
8.06, d (2.0)
123.6, C
132.9, C
b
25.2, CH2
65.3, CH2
135.2, CH
125.7, CH
2.96, dt (7.5, 2.1)
4.34, t (7.5)
135.5, CH
126.9, CH
6.66, d (6.0)
6.22, dd (6.0, 1.5)
7.07, d (8.6)
b
7.07, dt (11.7, 2.1)
6.27, dd (14.5, 11.7)
143.2, C
7.84, dd (8.6, 2.0)
66.3, CH2
35.1, CH2
(a) 4.44, d (12.7)
(b) 4.37, d (12.7)
(a) 2.86, m
b
8
138.2, CH
7.02, dd (14.5, 11.7)
6.05, d (11.7)
8
3′
51.9, CH3
134.1, CH
3.83, s
6.54, s
(b) 2.46, m
9
125.4, CH
9
38.3, CH
178.3, C
2.84, m
2″
3″
4″
5″
b
b
10
11
12
142.1, C
10
11
12
118.6, C
18.6, CH3
26.1, CH3
1.84, s
1.84, s
19.0, CH3
15.1, CH3
1.69, s
1.66, s
66.3, CH2
28.0, CH2
21.9, CH3
(a) 4.30, dt (8.8, 3.0)
(b) 4.15, dt (8.8, 6.7)
(a) 2.31, dddd (12.3, 8.6, 6.7, 3.0)
(b) 1.90, m
b
13
14
1‴
57.3, CH2
4.58, d (5.7)
5.29, t (5.7)
2.02, s
1‴-OH
a
DMSO-d6, compound 3: 500 MHz for 1H NMR, 125 MHz for 13C NMR; compounds 4 and 6: 600 MHz for 1H NMR, 150 MHz for 13C NMR.
b
Signal increase upon [1-13C]acetate addition to the culture.
B
DOI: 10.1021/acs.jnatprod.6b00147
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
Scheme 1. Synthetic Route toward Methyl seco-Fomannoxinate (6) and Respective Intermediates/Analogues 7−9
could be connected via C-3 through heteronuclear multiple-
Compound 5 of the basidiomycete BY1 was identified as (S)-
5-formyl-2-(isopropyl-1′-ol)-2,3-dihydrobenzofuran,19 based on
1
bond correlation (HMBC, Figure S7) spectroscopy. H,13C
1
polarimetric analyses, HRESIMS, and H and 13C NMR data.
long-range interactions from H-4, H-5, and H-6 to C-2
established the γ-lactone moiety of 3. Furthermore, HMBC
data allowed a linkage of the two remaining methyl groups
CH3-11 and CH3-12 via C-10 with the rest of the molecule.
The positions of CH3-11 and CH3-12 were assigned according
to literature data.3 The E configuration of the Δ3,6 double bond
was deduced from a strong NOE between H2-4 and H-7
(Figure S8). A large coupling constant of 14.5 Hz between H-7
and H-8 indicated an E-configured double bond Δ7,8.
The spectroscopic data of another compound (6), however, did
not match any previously isolated fomannoxin or vibralactone.
The empirical formula of 6 was assigned to be C13H16O4 by
HRESIMS, which corresponds to six degrees of unsaturation.
Strong stretching vibrations at 3410 and 1713 cm−1 indicated
the presence of hydroxy and carbonyl moieties, respectively.
Inspection of the 1H NMR and COSY spectra (Figures S16 and
S20) revealed two discrete spin systems. Three resonances in
the aromatic region occur as an isolated AMX spectrum (δA
7.07, δM 7.84, δX 8.06, JAM = 8.6 Hz, JMX = 2.0 Hz),
characteristic of a 1,2,4-trisubstituted benzene derivative. This
assumption is in agreement with 13C, DEPT 135, HSQC, and
HMBC data (Figures S17, S19, S21, and S22, respectively),
which identified six 13C resonances as constituents of the
postulated benzene moiety (Table 1). The second spin system
corresponded to a hydroxymethylene residue, which could be
connected to the aromatic core of 6 due to long-range
interactions from H2-1‴ to C-2, C-3, and C-4 and from H-2 to
C-1‴. HMBC data also allowed for the identification of the two
remaining substituents, that is, a methyl ester group at C-1 and
a (2-methylprop-1-en-1-yl)oxy group at C-4. Hereafter, this
compound is referred to as methyl seco-fomannoxinate. A (2-
methylprop-1-en-1-yl)oxy group is rarely observed in natural
products but represented by 4-(1-hydroxyethyl)-1-O-(2-meth-
ylprop-1-en-1-yl)salicylaldehyde, which had been reported from
Osteospermum muricatum (Compositae, sunflower family).20
Synthesis of Methyl seco-Fomannoxinate. To confirm
the structural proposal and to also secure a sufficient supply for
biological testing, compound 6, now referred to as methyl seco-
fomannoxinate, was synthesized (Scheme 1). 4-Hydroxyben-
zoic acid methyl ester was regioselectively formylated at C-3,
following a published procedure.21 In our hands, the strategy by
Ma et al.,22 to directly install an isobut-1-enyl group at the
phenolic OH via a nucleophilic substitution with 1-bromo-2-
methylprop-1-ene, repeatedly failed to yield the desired
product. Attempts to introduce the isobut-1-enyl group as a
triflate, as proposed,23 did not prove successful either, perhaps
due to the phenolic group being deactivated by the ortho-
positioned aldehyde functionality. A successful alternative
included a two-step approach: O-alkylation of the phenol
with 3-bromo-2-methylprop-1-ene followed by a subsequent
isomerization of the double bond resulted in the desired vinyl
HRESIMS data of 4 (m/z 223.0967 [M + H]+) suggested a
sum formula of C12H14O4 and six double-bond equivalents. The
optical rotation value of 4 corroborated a biosynthetic origin.
The IR spectrum of 4 displayed a strong band at 1733 cm−1,
corresponding to a carbonyl functionality. Since no absorption
bands associated with O−H bond stretching were observed, 4
does not contain carboxylic acid moieties. In the 13C NMR
spectrum (Figure S10), only 11 discrete resonances were
resolved. Two ester functions (δC‑2 = 168.9 and δC‑10 = 178.3
ppm) as well as two carbon−carbon double bonds (δC‑3
=
=
132.9 ppm, δC‑4 = 135.5 ppm, δC‑5 = 126.9 ppm, and δC‑6
1
143.2 ppm) were readily identified. Subsequently, H NMR,
DEPT135, and HSQC experiments (spectra in Figures S9, S12,
and S14) allowed the assignment of two methylenoxy groups
(CH2-7 and CH2-12) that exhibited overlapping chemical shifts
(δC = 66.3 ppm), thus providing an explanation for the
apparently missing resonance. The remaining signals in the 13
C
NMR spectrum were attributed to a methine (CH-9), a methyl
(CH3-14), and two methylene groups (CH2-8 and CH2-13).
The chemical shifts of CH3-14 were suggestive of a methyl
group connected to an sp2-hybridized carbon. This assumption
was confirmed in the HMBC (Figure S15) spectrum, which
indicated C-6 as the directly bonded carbon atom. On the basis
of further 1H,13C long-range interactions as well as spin
coupling between H-4 and H-5, a 3-substituted 6-methylox-
epin-2-(7H)-one partial structure was elucidated. COSY data
(Figure S13) combined with an interpretation of the HMBC
spectrum then led to the identification of a (2-oxotetrahy-
drofuran-3-yl)methylene moiety, which was finally linked to the
oxepin-2-(7H)-one ring due to long-range interactions from H-
8 to C-2, C-3, and C-4. Compound 4 was hence identified as 6-
methyl-3-((2-oxotetrahydrofuran-3-yl)methyl)oxepin-2(7H)-
one, for which we propose the name vibralactone S.
C
DOI: 10.1021/acs.jnatprod.6b00147
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
a
Table 2. NMR Spectroscopic Data for Synthetic Products 7−9
7
8
9
no.
δC, mult
δH, M (J in Hz)
δC, mult
δH, M (J in Hz)
δC, mult
δH, M (J in Hz)
1
123.0, C
122.6, C
123.6, C
2
130.5, CH
124.6, C
8.48, d (1.8)
130.0, CH
129.3, C
7.99, d (2.2)
131.6, CH
115.1, CH
161.4, C
7.97, d (8.8)
6.96, d (8.8)
3
4
163.9, C
160.0, C
5
112.7, CH2
137.0, CH
166.0, C
7.00, d (8.7)
110.8, CH
131.0, CH
166.8, C
6.86, d (8.6)
115.1, CH
131.6, CH
166.8, q
6.96, d (8.8)
7.97, d (8.8)
6
8.19, dd (8.7, 1.8)
7.94, dd (8.6, 2.2)
1′
3′
2″
3″
4″
52.1, CH3
72.4, CH2
139.3, C
3.85, s
4.60, s
51.9, CH3
71.8, CH2
139.9, C
3.85, s
4.50, s
51.9, CH3
134.1, CH
119.7, C
3.88, s
6.23, m
113.8, CH2
(a) 5.11, s
(b) 5.05, s
1.85, s
113.2, CH2
(a) 5.07, s
(b) 5.00, s
1.80, s
15.3, CH3
1.70, s
1.70, s
5″
1‴
19.3, CH3
188.7, CH
19.4, CH3
61.5, CH2
19.5, CH3
10.50, s
4.70, s
a
Chloroform-d1, 500 MHz for H NMR, 125 MHz for 13C NMR.
1
Figure 2. Proposed and simplified routes of the combined vibralactone and fomannoxin biosynthesis. Vibralactone, vibralactone D, and fomannoxin
formation according to published data.15,28 Asterisks indicate 13C-enriched carbons after adding [1-13C]acetate to the medium. The 2-methylprop-1-
enyl ether moiety of methyl seco-fomannoxinate (6) is highlighted in blue,; oxepanone/oxepinone rings are highlighted in red.
ether. O-Alkylation proved efficient. Since the leaving group of
the alkylating agent was attached to an sp3-configured carbon
atom, the first step (nucleophilic substitution) could be
performed smoothly according to established procedures,24
with virtually quantitative turnover of educts, leading to
intermediate 7 (1H and 13C NMR data in Table 2, NMR
spectra in Figures S23−S28). Reduction of the aldehyde with
methanolic NaBH4 yielded compound 8 in near-quantitative
D
DOI: 10.1021/acs.jnatprod.6b00147
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
yield (1H and 13C NMR data in Table 2, NMR spectra in
Figures S29−S34). Its conversion into the corresponding vinyl
ether followed the procedure of Okamoto et al.,24 who describe
a rhodium-catalyzed route for dihydrobenzofuran derivatives
through an olefin isomerization and enantioselective intra-
molecular Alder-ene reaction cascade. The combination of
Rh(ndb)2BF4 and (R,R)QuinoxP in anhydrous (CH2Cl)2 as
catalyst and ligand performed only modestly with regard to the
intramolecular Alder-ene reaction, but supported reasonable
turnover in the olefin isomerization. Consequently, this
procedure was modified, and CH2Cl2 was used as solvent,
which led to near-quantitative conversion of 8 into 6.
Reduction of the aldehyde function prior to isomerization
proved crucial, as 7 did not isomerize into the respective vinyl
propose that 1,5-seco-vibralactone as the above-mentioned
central intermediate undergoes hydrolysis of its oxepin-
2(7H)-one ring as well as hydroxylation at the distal position
of the isoprene unit, followed by dual intermolecular
esterification (via C-7 and C-2 and via C-12 and C-10) as
well as double-bond migration and reduction of the γ-lactone
(Figure 2).
Compound 6 appeared structurally somewhat related to the
fomannoxins despite the absence of a benzofurane core and an
overall ambiguous biosynthetic origin. Fomannoxin-type
isobut-1-enyl ethers have never been described. Analogously
to the assumed route for the above-mentioned Osteospermum
salicylaldehyde derivative,20 we hypothesized that the biosyn-
thesis of 6 may proceed from fomannoxic acid with the pathway
being extended to the (as yet undescribed) 3-hydroxyfoman-
noxic acid methyl ester. Subsequently, it undergoes cleavage of
the dihydrofuran ring, followed by the reduction of the
aldehyde (Figure 2). As described above, fungi produce
DMAPP exclusively via the mevalonate pathway. The benzylic
carbon atom found in ortho-position to the phenolic group (C-
1‴) as well as the sp2-hybridized carbon atom of the isobut-1-
enyl moiety (C-3″) of 6 would therefore derive from C-1 and
C-3 of DMAPP, respectively. Hence, feeding of [1-13C]acetate
should lead to enriched 13C isotopic abundance at these
positions. To test the hypothesis, 6 was purified from BY1
cultures grown in the presence of 24.4 mM sodium
[1-13C]acetate. The 13C spectrum of 6 showed selective
enrichment of the signals at δC = 57.3 ppm and δC = 118.6
ppm (Figure S18), which correspond to positions C1‴ and
C3″, respectively (Table 1). These findings verify our
hypothesis that both carbon atoms originate from an isoprene
unit whose transfer ortho to the alcohol functionality of 4-
hydroxybenzaldehyde or 4-hydroxybenzoic acid initiates
fomannoxin biosynthesis.15,28 Our results also verify the
fragmentation step as the key biosynthetic event toward 6
and make the postulated route for the Osteospermum
salicylaldehyde isobut-1-enyl ether plausible.20 In return, this
raises the question if this compound is truly a plant metabolite
or perhaps the product of an endophytic fungus.
1
ether. H and 13C NMR spectra of the final synthetic product
were identical to those of 6, isolated from the fungal source.
Analogously, 9 was synthesized from 4-hydroxybenzoic acid
1
methyl ester without an initial formylation step (Scheme 1, H
and 13C NMR data in Table 2, NMR spectra in Figures S35−
S40).
Biosynthetic Studies. The new γ-lactone 3 was also
isolated from cultures of BY1 grown on [1-13C]sodium acetate-
enriched medium. 13C NMR data revealed a selective increase
of the signals at δC = 138.2 ppm and δC = 142.1 ppm (Figure
S3). These signals correspond to C-8 and C-10, respectively
(Figure 1 and Table 1). Fungi synthesize γ,γ-dimethylallyldi-
phosphate (DMAPP) exclusively from acetyl-CoA via the
mevalonate pathway.25 Consequently, upon routing [1-13C]-
acetate into this metabolism, C-1 and C-3 of DMAPP will carry
the stable isotope label, which correspond to carbon atoms C-8
and C-10 in 3. As a common structural feature, 3 shares the γ-
lactone moiety with vibralactones G−J, L, and M. These
compounds presumably derive from 1,5-seco-vibralactone by
ester hydrolysis, decarboxylation in the case of vibralactones G
and H, and re-esterification as principal steps.26,27 Our results
are perfectly compatible with this hypothesis, confirm this
biosynthetic route and also emphasize the central role of 1,5-
seco-vibralactone as an intermediate toward bicyclic β- and γ-
lactones.
Compound 4 represents a new 2(7H)-oxepinone that results
from BY1’s highly diverse fomannoxin/vibralactone metabo-
lism. Precedence for seven-membered, oxygen-containing ring
systems in the vibralactone family stems, for example, from 1,5-
seco-vibralactone (Figure 2), which is the 3,5-branched
immediate precursor to vibralactone.28 Vibralactone D (Figure
2) features a 2-oxepane moiety and is considered a follow-up
product of vibralactone that underwent β-lactone cleavage,
reduction of the cyclopentene, and re-esterification.4 Collec-
tively, these data support two routes of heterocycle formation,
which both maintain the prenyl group as an exocyclic
substituent. Intriguingly, 4 has a γ-lactone substituent and
possesses a 3,6-substituted oxepin-2(7H)-one core, which is
incompatible with the above routes that result in different
substitution patterns. This finding implies a third, independent
biogenesis for seven-membered vibralactone-type heterocycles
in the basidiomycete BY1, which was investigated by [1-13C]-
acetate stable isotope labeling. The 13C NMR spectra of 4
showed two selectively enriched signals (δC = 135.5 ppm and
δC = 143.2 ppm, respectively, corresponding to C-4 and C-6
(Figure S11)). This result demonstrates that the exocyclic
prenyl group, likely of the intermediate 1,5-seco-vibralactone,
has been integrated into a seven-membered heterocycle to yield
the branched vibralactone ε-lactone ring system of 4. We
Antifungal Bioactivity. The pure compounds 1−9 were
tested for antifungal activity against the human pathogenic
fungi Candida albicans, Aspergillus fumigatus, and Arthroderma
benhamiae, as well as against Penicillium notatum. C. albicans was
not inhibited by any compound. Likewise, 1−4 did not inhibit
A. fumigatus. However, inhibition was detected for A.
benhamiae (MIC of 12.5 μg/mL for 3, 50 μg/mL for 1 and
2, Table 3) and P. notatum (MIC: 25 μg/mL). 6 was
particularly active against A. benhamiae (MIC: 6.25 μg/mL).
The synthetic analogues 7−9 were included in the test as well
(MIC: 6.25 μg/mL for 7 and 25 μg/mL for 8, respectively, but
100 μg/mL for 9). As A. fumigatus was less inhibited, or not at
all, divergent structural features may be required for activity
against this species, or differences exist with regard to transport
into or out of the cell.
None of the vibralactones 1−4 were active against C. albicans
and A. fumigatus at concentrations of ≤100 μg/mL (Table 3).
However, 1−3 inhibited P. notatum at 25 μg/mL. A more
heterogeneous situation was observed with A. benhamiae, as this
pathogen was not inhibited by the oxepinone 4, but clearly by 3
(MIC 12.5 μg/mL), whereas vibralactones 1 and 2 showed
MIC values of 50 μg/mL. The fomannoxins are typical
metabolites for Heterobasidion (Bondarzewiaceae), while the
vibralactones are found in Stereumx and Boreostereum species
E
DOI: 10.1021/acs.jnatprod.6b00147
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
below. HRESIMS data were recorded on an Exactive Orbitrap
instrument (Thermo Scientific) using the direct injection port.
Analytical and semipreparative HPLC was performed on an Agilent
1200 instrument fitted with Zorbax Eclipse XDB C18 columns (5 μm
particle size, 150 × 4.6 mm or 250 × 9.6 mm, respectively).
Chromatograms were recorded at λ = 254 and 280 nm; the respective
diode array detectors covered the wavelength range of λ = 190−600
nm. Chemicals, solvents, and media components were purchased from
Deutero, Roth, Sigma-Aldrich, and VWR, and sodium [1-13C]acetate
was from Cortecnet.
Table 3. Antifungal Activity of Described and New Natural
Products 1−6 and Synthetic Compounds 7−9 (MIC in μg/
mL)
Aspergillus
fumigatus
Penicillium
notatum
Arthroderma
benhamiae
compound
Vibralactones
1
2
3
4
>100
>100
>100
>100
25
25
25
25
50
50
12.5
>100
Microbiological Methods. The fungus BY116 was routinely
maintained on solid HA medium (4 g/L D-glucose, 4 g/L yeast extract,
10 g/L malt extract, 18 g/L agar, pH = 6.5). For metabolite
production, modified liquid HA medium (1 L per 4 L penicillin flask)
was used (D-glucose and yeast extract increased to 10 g/L each) and
cultivated as stationary culture for 21 days at room temperature in the
dark or until HPLC analysis showed secondary metabolite production.
Stable isotope labeling was carried out by adding filter-sterilized
sodium [1-13C]sodium acetate (2 g/L, 24.4 mM final) upon
inoculation.
Compound Isolation and Chromatography. The culture broth
was separated by filtration from the mycelia and subsequently
extracted three times with equal volumes of ethyl acetate. The organic
phase was filtered, and the solvent was removed under reduced
pressure using a rotary evaporator. The resulting crude extract was
dissolved in 25% MeOH in water (v/v) and prefractionated on Waters
Sep-Pak C18 cartridges (5 g), conditioned equally prior to sample
loading. A step gradient of 25, 50, 75, and 100% MeOH was used for
elution. The respective fractions were subjected to preparative
chromatography running a linear water/acetonitrile (ACN) gradient
from 15% to 100% ACN within 18 min at a flow rate of 20 mL/min.
Final purification was accomplished through semipreparative chroma-
tography applying an isocratic flow of 4 mL/min. The water/ACN
ratio was chosen based on the polarity of the compounds to be
separated and ranged between 25% and 50% ACN. Analytical HPLC
of the crude extract, fractions, and pure substances was run using the
following water/ACN gradient (1 mL/min): 10% ACN, held constant
for 10 min, increase to 90% within 22 min, increase to 100% ACN
within 1 min.
Synthesis of 6. The synthesis of 3-formyl-4-hydroxybenzoic acid
methyl ester from commercially available 4-hydroxybenzoic acid
methyl ester followed a published protocol with 81% yield after
workup.21 The O-alkylation of phenolic groups to yield 7 was carried
out as described.24 A 0.5 mmol amount of phenol was mixed with 0.6
mmol of 3-bromo-2-methylprop-1-ene and 1 mmol of potassium
carbonate in 3 mL of ACN. The reaction was stirred for 15 h at room
temperature and then stopped by adding an equal volume of water.
Products were extracted with ethyl acetate, the organic phases were
combined, and volatiles were removed under reduced pressure in a
rotary evaporator. Subsequent purification accomplished by silica gel
column chromatography was performed as described above, but using
a step gradient of ethyl acetate/cyclohexane (9:1, 8:2, 7:3, v/v) as
mobile phase. A near-quantitative turnover (99% yield) of the phenol
educt into 2-methylprop-1-en-3-yl ether 7 was observed. The aldehyde
of 7 was then reduced with sodium borohydride in MeOH at room
temperature. After 2 h, the reaction had proceeded quantitively to
yield 8. Isomerization of the allylic ether into the respective vinylic
ether of 6 was completed following a published procedure24 in
modified form: 8 was mixed in a glass vial with 0.1 molar equiv of
Rh(ndb)2BF4 and (R,R)QuinoxP in anhydrous CH2Cl2. The reaction
was sealed and heated to 80 °C for 16 h with constant stirring. The
reaction was cooled to room temperature, quenched by addition of
water, and subsequently extracted with an equal volume of ethyl
acetate. Volatile compounds of the combined ethyl acetate phases were
removed, and traces of remaining catalysts were removed during
HPLC as described above to yield product 6 (86% yield after
chromatography). An analogous synthesis route (except initial
formylation) was taken to synthesize 9 (Scheme 1).
Methyl seco-Fomannoxinate and Its Synthetic Analogues
6
7
8
9
50
50
25
25
25
25
6.25
6.25
25
>100
>100
100
(Stereaceae). Both fungal families belong to the Russulales.29
Given that members of these two natural product classes,
known from phylogenetically distinct families, occur side by
side in the BY1 fungus, its metabolism is intriguing and strongly
underscores its relationship to the Russulales from the
chemotaxonomic perspective. This relationship was previously
based solely on genetic evidence.16 However, the ecological
relevance of BY’s simultaneous and greatly diversified
fomannoxin/vibralactone metabolism remains shrouded. Fo-
mannoxin is known to exert phytotoxic effects, with the 2-
isopropenyl side chain being relevant for phytotoxicity.30
Activity against E. coli and, to a lesser degree, against the
basidiomycete Radulomyces confluens was found as well.10,11
The antifungal activity of vibralactone was established during
this study. BY1 is a fast-growing pioneer colonist that causes a
white rot of woody biomass. Given their antifungal activity,
these metabolites may confer a selective advantage for the
fungus that helps it thwart off and compete with other fast-
growing, ubiquitous soil and plant surface inhabitants, such as
Penicillium species, although phytotoxic fomannoxins may not
be primarily required to secure growth in its habitat. BY1
combines (i) the ability to colonize wood in an early stage of
decay and (ii) the capacity to break down lignin with (iii) a
metabolism that produces a diversity of bioactive small
molecules. Therefore, this fungus may represent a future
model organism for technologies such as biopulping and
preprocessing of lignocellulosic material for biobased ethanol
production.
EXPERIMENTAL SECTION
■
General Experimental Procedures. Optical rotation was
measured on a Jasco P-1020 polarimeter, with methanol (MeOH) as
solvent, in a 1 mL cell (path length d = 5.0 cm, λ = 589 nm, T = 20
°C). UV−vis spectra were recorded on a ScanDrop instrument
(Analytik Jena) using Winaspect software. Samples were dissolved in
MeOH. IR spectra were recorded on a Jasco FT/IR 4100 instrument.
1D and 2D NMR spectra of compounds 1−8 were recorded in
DMSO-d6 and chloroform-d1, respectively, at 300 K. Chemical shifts
were referenced relative to internal residual nondeuterated solvent
traces (DMSO-d5: δH 2.50 ppm; δC 39.5 ppm, chloroform δH 7.24
ppm; δC 77.2 ppm). NMR spectra were recorded on Bruker Avance III
500 and 600 MHz spectrometers. Preparative HPLC was accom-
plished on an Agilent 1260 instrument equipped with a Phenomenex
Luna C18 column (250 × 21.2 mm, 10 μm particle size). Routine mass
spectrometry was performed on an Agilent 1260 chromatograph with a
Zorbax Eclipse XDB C18 column (150 × 4.6 mm, 3.5 μm particle size),
coupled to an Agilent 6130 mass detector using electrospray ionization
in positive and negative mode and applying the gradient described
Bioactivity Tests. The pure compounds (Table 3) were dissolved
in MeOH (1 mg/mL) for antimicrobial tests. Minimal inhibitory
F
DOI: 10.1021/acs.jnatprod.6b00147
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
concentration (MIC) values against Candida albicans, Aspergillus
nidulans, Aspergillus flavus, and Penicillium notatum were determined
with the broth dilution assay according to a published procedure.31
Vibralactone R (3): colorless oil; UV (MeOH) λmax 325 (log ε =
3.98) nm; IR(neat) 1730 (s), 1648 (w), 1190 (s), 1083 (s) cm−1; 1H and
13C NMR data (DMSO-d6, 500 and 125 MHz, respectively), see Table
1; HRESIMS m/z 179.1068 [M + H]+ calcd for C11H15O2 179.1067.
Hogberg, N.; James, T. Y.; Karlsson, M.; Kohler, A.; Kues, U.; Lee,
̈
̈
Y. H.; Lin, Y. C.; Lind, M.; Lindquist, E.; Lombard, V.; Lucas, S.;
Lunden, K.; Morin, E.; Murat, C.; Park, J.; Raffaello, T.; Rouze, P.;
Salamov, A.; Schmutz, J.; Solheim, H.; Stahlberg, J.; Velez, H.; de
́
́
̊
́
̈
Vries, R. P.; Wiebenga, A.; Woodward, S.; Yakovlev, I.; Garbelotto, M.;
Martin, F.; Grigoriev, I. V.; Stenlid, J. New Phytol. 2012, 194, 1001−
1013.
Vibralactone S (4): colorless oil; [α]20 +8 (c 0.1, MeOH); UV
(2) Asiegbu, F. O.; Adomas, A.; Stenlid, J. Mol. Plant Pathol. 2005, 6,
395−409.
D
(MeOH) λmax 275 (log ε = 3.38) nm; IR(neat) 1733 (s), 1652 (w), 1194
1
(s) cm−1; H and 13C NMR data (DMSO-d6, 600 and 150 MHz,
(3) Liu, D. Z.; Wang, F.; Liao, T. G.; Tang, J. G.; Steglich, W.; Zhu,
H. J.; Liu, J. K. Org. Lett. 2006, 8, 5749−5752.
(4) Jiang, M. Y.; Zhang, L.; Dong, Z. J.; Yang, Z. L.; Leng, Y.; Liu, J.
K. Chem. Pharm. Bull. 2010, 58, 113−116.
respectively), see Table 1; HRESIMS m/z 223.0967 [M + H]+, calcd
for C12H15O4 223.0965.
Methyl seco-fomannoxinate (6): colorless oil; UV (MeOH) λmax
263 (log ε = 4.62) nm; IR(neat) 3410 (b), 1713 (s), 1606 (s), 1388 (s),
(5) Chen, H. P.; Zhao, Z. Z.; Li, Z. H.; Dong, Z. J.; Wei, K.; Bai, X.;
Zhang, L.; Wen, C. N.; Feng, T.; Liu, J. K. ChemistryOpen 2016, DOI:
10.1002/open.201500198.
1
1248 (s), 1132 (s), 767 (s) cm−1; H and 13C NMR data (DMSO-d6,
600 and 150 MHz, respectively), see Table 1; HRESIMS m/z
237.1124 [M + H]+, calcd for C13H17O4 237.1121.
(6) Zeiler, E.; Braun, N.; Bottcher, T.; Kastenmuller, A.; Weinkauf,
̈
̈
Methyl 4-(2-methylallyloxy)-3-formylbenzoate (7): colorless oil;
UV (MeOH) λmax 233 (log ε = 4.59) and 256 (log ε = 4.52) nm;
S.; Sieber, S. A. Angew. Chem., Int. Ed. 2011, 50, 11001−11004.
(7) Kepler, J. A.; Wall, M. E.; Mason, J. E.; Basset, C.; McPhail, A. T.;
Sim, G. A. J. Am. Chem. Soc. 1967, 89, 1260−1261.
(8) Donnelly, D. M. X.; O’ Reilly, J.; Polonsky, J.; van Eijk, G. W.
Tetrahedron Lett. 1982, 23, 5451−5452.
1
IR(neat) 1715 (s), 1686 (s), 1607 (s), 1124 (s), 766 (s) cm−1; H and
13C NMR data (CDCl3, 500 and 125 MHz, respectively), see Table 2;
HRESIMS m/z 235.0962 [M + H]+, calcd for C13H15O4 235.0965.
Methyl 4-(2-methylallyloxy)-3-(hydroxymethyl)benzoate (8): col-
orless oil; UV (MeOH) λmax 258 (log ε = 4.45) nm; IR(neat) 3429 (b),
(9) Hansson, D.; Menkis, A.; Himmelstrand, K.; Thelander, M.;
Olson, K.; Stenlid, J.; Karlsson, M.; Broberg, A. Phytochemistry 2012,
82, 158−165.
1
1712 (s), 1606 (s), 1249 (s), 1127 (s), 1132 (s) cm−1; H and 13C
NMR data (CDCl3, 500 and 125 MHz, respectively), see Table 2;
HRESIMS m/z 235.0977 [M − H]−, calcd for C13H15O4 235.0976.
Methyl 4-(2-methylprop-1-enyloxy)benzoate (9): colorless oil; UV
(MeOH) λmax 255 (log ε = 3.16) nm; IR(neat) 1716 (s), 1603 (s), 1244
(10) Hirotani, M.; O’Reilly, J.; Donnelly, D. M. Tetrahedron Lett.
1977, 7, 651−652.
(11) Sonnenbichler, J.; Bliestle, I. M.; Peipp, H.; Holdenrieder, O.
Biol. Chem. Hoppe-Seyler 1989, 370, 1295−1303.
1
(s), 1109 (s), 769 (s) cm−1; H and 13C NMR data (CDCl3, 500 and
(12) Sonnenbichler, J.; Peipp, H.; Dietrich, J. Biol. Chem. Hoppe-
Seyler 1993, 374, 467−473.
125 MHz, respectively), see Table 2; HRESIMS m/z 207.1019 [M +
H]+, calcd for C12H15O3 207.1016.
(13) Hansson, D.; Wubshet, S.; Olson, A.; Karlsson, M.; Staerk, D.;
Broberg, A. Phytochemistry 2014, 108, 243−251.
ASSOCIATED CONTENT
* Supporting Information
(14) Heslin, M. C.; O’Reilly, J.; Donnelly, D. M. X. Eur. J. Forest
Pathol. 1983, 13, 11−23.
■
S
(15) Hansson, D.; Menkis, A.; Olson, K.; Stenlid, J.; Broberg, A.;
Karlsson, M. Phytochemistry 2012, 84, 31−39.
The Supporting Information is available free of charge on the
ACS Publications website at DOI: 10.1021/acs.jnat-
prod.6b00147.
(16) Schwenk, D.; Nett, M.; Dahse, H. M.; Horn, U.; Blanchette, R.
A.; Hoffmeister, D. J. Nat. Prod. 2014, 77, 2658−2663.
(17) Jiang, M. Y.; Wang, F.; Yang, X. L.; Fang, L. Z.; Dong, Z. J.; Zhu,
H. J.; Liu, J. K. Chem. Pharm. Bull. 2008, 56, 1286−1288.
1D and 2D NMR spectra of new natural products and
synthetic compounds (PDF)
(18) Aqueveque, P.; Ces
O. Z. Naturforsch., C: J. Biosci. 2015, 70, 97−102.
́ ́
pedes, C. L.; Becerra, J.; Davila, M.; Sterner,
AUTHOR INFORMATION
Corresponding Author
*Phone: +49 3641 949850. Fax: +49 3641 949852. E-mail: dirk.
hoffmeister@hki-jena.de.
■
(19) Donnelly, D. M. X.; Fukuda, N.; Koung, I.; Martin, M.; O’
Reilly, J. Phytochemistry 1988, 27, 2709−2713.
(20) Bohlmann, F.; Wallmeyer, M.; Jakupovic, J.; Ziesche, J.
Phytochemistry 1983, 22, 1645−1651.
(21) Phan, D. H.; Kim, B.; Dong, V. M. J. Am. Chem. Soc. 2009, 131,
15608−15609.
Notes
The authors declare no competing financial interest.
(22) Ma, D.; Cai, Q.; Xie, X. Synlett 2005, 11, 1767−1770.
(23) Willis, M. C.; Taylor, D.; Gillmore, A. T. Chem. Commun. 2003,
2222−2223.
ACKNOWLEDGMENTS
■
D.S. and P.B. acknowledge doctoral fellowships from the
International Leibniz Research School (ILRS Mibintact) and
the Jena School for Microbial Communication (JSMC),
respectively. We thank C. Weigel, A. Perner, and H. Heinecke
(24) Okamoto, R.; Okazaki, E.; Noguchi, K.; Tanaka, K. Org. Lett.
2011, 13, 4894−4897.
(25) Eisenreich, W.; Rohdich, F.; Bacher, A. Trends Plant Sci. 2001, 6,
78−84.
(26) Wang, G. Q.; Wie, K.; Feng, T.; Li, Z. H.; Zhang, L.; Wang, Q.
A.; Liu, J. K. J. Asian Nat. Prod. Res. 2012, 14, 115−120.
(27) Wang, G. Q.; Wie, K.; Zhang, L.; Li, Z. H.; Wang, Q. A.; Liu, J.
K. J. Asian Nat. Prod. Res. 2014, 16, 447−452.
(28) Zhao, P. J.; Yang, Y. L.; Du, L.; Liu, J. K.; Zeng, Y. Angew. Chem.,
Int. Ed. 2013, 52, 2298−2302.
(Hans-Knoll-Institute Jena) for bioactivity tests and for
recording mass and NMR spectra, respectively. Professors F.
̈
Oberwinkler and W. Steglich (Eberhard-Karls-Univeristat
̈
Tubingen and Ludwig-Maximilians-Universitat, Munchen,
̈
̈
̈
respectively) are gratefully acknowledged for insightful advice.
(29) Miller, S. L.; Larsson, E.; Larsson, K. H.; Verbeken, A.;
Nuytinck, J. Mycologia 2006, 98, 960−970.
REFERENCES
■
(30) Horlacher, N.; Nachtigall, J.; Schulz, D.; Sussmuth, R. D.;
(1) Olson, A.; Aerts, A.; Asiegbu, F.; Belbahri, L.; Bouzid, O.;
̈
Hampp, R.; Fiedler, H. P.; Schrey, S. D. J. Chem. Ecol. 2013, 39, 931−
Broberg, A.; Canback, B.; Coutinho, P. M.; Cullen, D.; Dalman, K.;
̈
941.
Deflorio, G.; van Diepen, L. T.; Dunand, C.; Duplessis, S.; Durling,
M.; Gonthier, P.; Grimwood, J.; Fossdal, C. G.; Hansson, D.;
Henrissat, B.; Hietala, A.; Himmelstrand, K.; Hoffmeister, D.;
(31) National Committee for Clinical Laboratory Standards. Methods
for dilution antimicrobial susceptibility tests for bacteria that grow
G
DOI: 10.1021/acs.jnatprod.6b00147
J. Nat. Prod. XXXX, XXX, XXX−XXX

Journal of Natural Products
Article
aerobically, Approved standard, 4th ed.; Document M7−A4; Villanova,
PA, USA, 1997.
H
DOI: 10.1021/acs.jnatprod.6b00147
J. Nat. Prod. XXXX, XXX, XXX−XXX