New bioactive rosigenin analogues and aromatic polyketide metabolites from the freshwater aquatic fungus Massarina tunicata

Article abstract of DOI:10.1021/np020342d

Four new rosigenin analogues (massarigenins A-D; 1-4) and two new aromatic polyketide-derived secondary metabolites (massarinins A and B; 6, 7) and have been isolated from the freshwater aquatic fungus Massarina tunicata. The structures of these compounds were determined primarily by analysis of NMR data, and that of compound 1 was verified by X-ray crystallography. The known compound 4-(2-hydroxybutynoxy)benzoic acid (11) was also obtained, and its absolute stereochemistry was assigned. Several of these metabolites showed antibiotic activity against Gram-positive bacteria.

Full text of DOI:10.1021/np020342d

J . Nat. Prod. 2003, 66, 73-79
73
New Bioa ctive Rosigen in An a logu es a n d Ar om a tic P olyk etid e Meta bolites fr om
th e F r esh w a ter Aqu a tic F u n gu s Ma ssa r in a tu n ica ta
Hyuncheol Oh,† Dale C. Swenson,† J ames B. Gloer,*,† and Carol A. Shearer‡
Department of Chemistry, University of Iowa, Iowa City, Iowa 52242, and Department of Plant Biology,
University of Illinois, Urbana, Illinois 61801
Received J uly 25, 2002
Four new rosigenin analogues (massarigenins A-D; 1-4) and two new aromatic polyketide-derived
secondary metabolites (massarinins A and B; 6, 7) and have been isolated from the freshwater aquatic
fungus Massarina tunicata. The structures of these compounds were determined primarily by analysis
of NMR data, and that of compound 1 was verified by X-ray crystallography. The known compound 4-(2-
hydroxybutynoxy)benzoic acid (11) was also obtained, and its absolute stereochemistry was assigned.
Several of these metabolites showed antibiotic activity against Gram-positive bacteria.
Freshwater aquatic fungi remain relatively unexplored
as sources of biologically active natural products. Our prior
chemical investigations of the freshwater aquatic fungus
Massarina tunicata Shearer & Fallah (A-25-1; Lophiosto-
mataceae) have resulted in the isolation of three new
sequiterpenoids1 and two polyketide-derived lactones with
novel ring systems.2 Studies of M. tunicata scale-up cul-
tures have afforded several additional new compounds, in-
cluding four new rosigenin analogues (massarigenins A-D;
1-4) and two new aromatic polyketide metabolites (mas-
sarinins A and B; 6 and 7). Details of the isolation and
characterization of these metabolites are presented here.
assigned as shown in 1 by single-crystal X-ray diffraction
analysis. The four-bond heteronuclear correlation between
H-9 and C-1 could be ascribed to a W-type conformational
relationship between these nuclei, as evident in the final
X-ray model (Figure 1). Compound 1 is closely related to
the known fungal metabolite rosigenin (8)3 and other
previously reported compounds.4-10 Although the relative
stereochemistry of 1 was provided by the X-ray structure,
NOESY data were acquired to assist in stereochemical
comparisons with compounds 2-4 (see below). As might
be expected for the 3D structure depicted in Figure 1, a
strong correlation was observed between H-6 and H-10.
Weak correlations were also observed between H3-12 and
both of the C-11 protons.
Resu lts a n d Discu ssion
The molecular formula of massarigenin A (1) was
established as C11H14O5 on the basis of HRFABMS and 13C
NMR data. The 1H, 13C, and DEPT NMR spectra for 1
(Tables 1 and 2) revealed the presence of a -CHCH3
moiety, an ester carbonyl, three hydroxy groups, three sp3
oxymethine units, a non-oxygenated quaternary carbon, a
1,2-disubstituted olefin, and an oxygenated terminal olefin
unit. The remaining two degrees of unsaturation required
by the molecular formula indicated that 1 must be bicyclic.
Analysis of the results from a series of 1H-1H homonuclear
decoupling experiments permitted the establishment of two
isolated spin-systems corresponding to the C3/C4/C11 and
C6-C10/C12 portions of structure 1.
HMBC results (Table 3) revealed connections of these
subunits to the ester carbonyl and the quaternary carbon,
and identified the location of one hydroxy group. HMBC
correlations of H-9 and H3-12 to the quaternary carbon C-5,
together with correlations of H-6 and H-10 to C-1 and C-4,
suggested the connection of C-1, C-4, C-6, and C-10 to C-5,
although the observation of a (selective INEPT) correlation
of H-9 with the carboxy carbon C-1 initially complicated
the unambiguous positioning of this group. The chemical
shift of C-4 (δ 68.5), along with HMBC correlations of the
hydroxy proton signal at δ 4.02 to C-3 and C-5, was
consistent with attachment of a hydroxy group at this
position. Although the other hydroxy proton signals were
too broad to show HMBC correlations, structure 1 appeared
most likely based on chemical shift and structural consid-
erations. Fortunately, suitable crystals were obtained, and
its complete structure and relative stereochemistry were
† University of Iowa.
‡ University of Illinois.
10.1021/np020342d CCC: $25.00
© 2003 American Chemical Society and American Society of Pharmacognosy
Published on Web 12/17/2002

74 J ournal of Natural Products, 2003, Vol. 66, No. 1
Oh et al.
Ta ble 1. 1H NMR Data [δ (mult.; J H-H in Hz)] for Massarigenins A-D (1-4)
C#
1a
2b
3a
4a
4
6
7
5.51 (dt; 3.9, 3.0)
2.58 (m)
5.67 (dd; 9.9, 1.7)
4.82 (br t; 2.1)
2.59 (m)
1.90 (dq; 4.7, 13)
2.03 (m)
5.07 (br t; 2)
3.34 (m)
6.74 (dd; 10, 2.1)
5.06 (br t; 2.4)
3.01 (m)
6.98 (dd; 10, 5.1)
8
9
5.77 (ddd; 9.9, 5.1, 3.2)
2.54 (m)
2.66 (ddd; 16, 4.5, 2.1)
6.03 (dd; 10, 3.3)
6.00 (dd; 10, 0.9)
4.29 (m)
10
11
4.11 (dd; 3.3, 3.6)
4.67 (dd; 2.5, 2.4)
4.51 (t; 2.3)
4.28 (d; 1.2)
4.47 (s)
4.70 (s)
4.86 (dd; 2.8, 2.7)
4.63 (dd; 2.8, 2.1)
0.98 (d; 6.8)
4.75 (t; 2.4)
4.55 (dd; 2.4, 1.8)
1.16 (d; 7.5)
not observed
4.62 (dd; 2.7, 2.4)
4.55 (dd; 2.4, 2.1)
1.38 (d; 7.2)
12
1.21 (d; 7.5)
4-OH
9-OH
10-OH
4.02 (d; 3.9)c
4.72 (d; 5.8)c
4.86 (br d; 3.3)c
not observed
not observed
not observed
not observed
not observed
a
Acetone-d6; 300 MHz. b CDCl3; 300 MHz. c These OH signals and their coupling effects were variable in appearance and were sometimes
not observed.
Ta ble 2. 13C NMR Data for Massarigenins A-D (1-4)
position
1a
2b
3a
4a
1
3
4
δ 176.2
159.1
68.5
δ 173.3
158.2
69.2
δ 174.4
159.8
69.0
δ 174.2
160.0
70.5
5
55.8
58.6
61.1
57.1
6
39.6
34.4
37.1
37.1
7
8
9
10
11
12
135.2
125.9
66.7
70.1
86.6
28.9
37.6
207.2
77.8
88.8
151.4
127.8
196.1
77.2
88.7
15.7
153.3
125.6
197.4
73.1
86.5
15.5
16.3
15.7
a
b
Acetone-d6. CDCl3.
Ta ble 3. HMBC and Selective INEPT Data for Massarigenins
A-C (1-3)
1a
2b
3c
1H
HMBC
selective INEPT
selective INEPT
signal correlations (C#) correlations (C#) correlations (C#)
4
6
7
8
3, 6, 10, 11
1, 4, 5, 7, 8, 12
5, 6, 8, 9, 12
6, 7, 9, 10
1,e 5, 7, 8, 10
1, 4, 5, 6
3, 5, 6, 10
d
5, 6, 8, 9, 12
6, 7, 9, 10
d
1, 4, 5, 9
3, 4
5, 6, 7
3, 5, 6, 10, 11
1, 4, 5, 7, 8, 10, 12
d
6, 10
d
1, 4, 5, 6, 9
3, 4
5, 6, 7
F igu r e 1. Final X-ray model of massarigenin A (1). The ellipsoid
probability level is 35%.
9
10
11
12
3, 4
5, 6, 7
as was the case in 1, and was consistent with observation
of a large J -value (13 Hz) between H-7ax and H-6.
However, NOESY correlations of H-4 with H3-12 and with
H-7ax in 2 revealed that these protons are all spatially close.
Given the relative configurations at C-6 and C-10, the latter
correlations would not both be expected if the stereochem-
istry at C-4 in 2 were the same as in 1. In fact, they would
only be likely for a system with the relative stereochemistry
at C-4 and C-5 shown in 2.
a
Data recorded at 600 MHz (1H dimension) in acetone-d6
solution. Data recorded at 300 MHz in CDCl3 solution. c Data
b
d
recorded at 300 MHz in acetone-d6 solution. Experiments were
not conducted for these proton signals. e This four-bond correlation
was not present in the HMBC spectrum, but was observed in a
selective INEPT experiment.
Compound 2 (massarigenin B) was determined to be an
isomer of 1 on the basis of HRFABMS and 13C NMR data.
The 1H and 13C NMR spectra of 2 (Tables 1 and 2) were
very similar to those of 1, except that the data for 2 contain
signals for two mutually coupled methylene units and a
ketone carbonyl in place of the signals for the 1,2-disub-
stituted olefin and one of the sp3-hybridized oxymethines
The molecular formula of massarigenin C (3) was found
to be C11H12O5 on the basis of HRFABMS and NMR data.
The NMR spectra for 3 (Tables 1 and 2) were again similar
to those of 1 except for the presence of a ketone carbonyl
signal and the absence of signals for one of the oxygenated
methine units (δ 4.29/66.7) observed in the spectra for 1.
The chemical shifts of the ketone and 1,2-disubstituted
olefin carbon signals in the 13C NMR spectrum of 3 (δ 196.1,
151.4, 127.8) suggested the presence of an R,â-unsaturated
ketone moiety. The locations of two hydroxy groups in 3
were verified by 1H NMR analysis of diacetate 5 formed
upon treatment of 3 with acetic anhydride, as the signals
for both H-4 and H-10 were shifted downfield by ap-
proximately 1.5 ppm in the diacetate. A series of selective
INEPT experiments (Table 3) led to confirmation of the
gross structure as shown in 3.
observed in the spectra of 1. H-1H homonuclear decou-
1
pling experiments conducted in DMSO-d6 solution con-
firmed the coupling of the two exchangeable hydroxy proton
signals with H-4 and H-10, respectively, and placed the
hydroxy groups at these positions. Analysis of selective
INEPT data (Table 3) allowed unambiguous assignment
of the gross structure as shown in 2. It was initially
anticipated that the stereochemistry would be analogous
to that of 1. Indeed, a NOESY correlation of H-6 with H-10
required a cis-1,3-diaxial-type orientation of these protons,

Bioactive Rosigenin Analogues from Massarina
J ournal of Natural Products, 2003, Vol. 66, No. 1 75
The possible spatial relationships among substituents
of the more conformationally restricted cyclohexenone ring
in 3 are significantly different from those of the corre-
sponding rings in 1 and 2. NOESY correlations of H-4 with
H3-12 and of H-6 with H-10 were again noted. Although
the latter correlation was much weaker than in 1, these
data suggested that 3 possesses the same relative stereo-
chemistry as in 2. This assignment was supported by
NOESY data for diacetate 5, which showed the same two
key correlations, including a more robust H6/H10 interac-
tion.
[4,5]decane skeleton,5,9,14 compounds 1-4 may be derived
from a pentaketide precursor, or a tetraketide precursor
and a malic acid unit.
A compound very similar to 3 has been previously
reported and assigned a diastereomeric structure differing
only in the relative configuration at C-4.10,11 The physical
and spectral properties described were similar to those of
3, though there were some limited differences in the NMR,
mp, [R]D, and UV data. NOE data reported for this
metabolite included an H-6/H-10 interaction, but, as would
be expected from the proposed structural difference, an
H-4/H3-12 correlation was not noted for the literature
compound.
HRFABMS and NMR data revealed that massarigenin
D (4) is an isomer of 3. The NMR data for 4 (Tables 1 and
2) were very similar to those of 3 except for chemical shift
differences associated mostly with the protons of the
cyclohexenone unit. These data suggested a diastereomeric
relationship between 3 and 4. In the case of 4, the H3-12-
to-H-4 NOESY correlation was still present, but the H-6/
H-10 and H-4/H-10 correlations were completely absent.
Instead, additional correlations were observed between H-4
and H-6 and between H-10 and H3-12. This result led to
the conclusion that the stereochemistry at C-10 is inverted
in 4 relative to 1-3.
The unusual 2-oxaspiro[4,5]decane skeleton found in
1-4 has been reported for fungal secondary metabolites
from Mycosphaerella rosigena,3 Arthropsis truncata,4,5 Co-
niothyrium sporulosum,6 a mutant strain of Rhodotorula
glutinis,7,8 and a Microsphaeropsis sp. known primarily for
production of the macrosphelides.10 A number of fungal
metabolites containing the similar, nitrogen-containing
2-azaspiro[4,5]decane skeleton have been isolated from
Drechslera tritici-repentis9 and from both wild-type and
mutant strains of Staphylotrichum coccosporum (spirostaph-
ylotrichins).12-15 The numbering system employed for 1-4
is based on that originally reported for 8-9 and for the
spirostaphylotrichins.
The molecular formula of massarinin A (6) was assigned
as C21H20O6 (12 unsaturations) on the basis of HREIMS
and 13C NMR data. The 1H and 13C NMR data (Table 4)
contained signals for pentasubstituted and 1,2,3,4-tetra-
substituted benzene rings, a 1,2-disubstituted olefin, an
aryl methyl group, two coincident or identical methyl
groups linked to an oxygenated quaternary carbon, a
methoxy group, a ketone carbon, an aldehyde moiety, and
two phenolic OH groups with shifts (δ 9.71 and 12.77)
suggesting differing levels of intramolecular hydrogen
bonding. Analysis of HMQC and HMBC data for 6 (Table
4) provided evidence for a 2,2-dimethylchromene unit
substituted with the hydrogen-bonded phenolic OH group
at δ 12.77. A search of the literature led to the hypothesis
that this compound is closely related to arugosin E (10), a
polyketide metabolite from Aspergillus silvaticus originally
identified by X-ray crystallography.16 Comparison of the
chemical shift data with those of 10 supported this conclu-
sion, but because other possible substitution patterns for
the pentasubstituted aromatic ring could be envisioned,
analysis of the 2D NMR data was necessary to verify the
structure and to unambiguously locate the hydroxy and
methoxy groups.
Massarigenins A-D (1-4) also bear close resemblance
to massarilactone A (9),2 a compound we previously re-
ported from M. tunicata that contains an additional ether
bridge linking O-4 to C-7. The relative stereochemistry of
9 is analogous to that of 1. The absolute stereochemistry
depicted in 9 was determined by X-ray crystallographic
analysis of its p-bromobenzoate ester. Although the abso-
lute configurations of 1-4 were not directly determined,
the stereostructures shown for 1-4 seem most likely by
analogy. However, in related compounds bearing a ketone
functionality at C-4 (presumed precursors of 1-4),4 it has
been proposed that epimerization can occur at positions 5
and/or 10 via a retroaldol process involving opening and
re-closing of the six-membered ring at the C5-C10 bond.4,9
Thus, the absolute stereochemistry of the compounds
isolated in this instance (particularly 2-4) remains uncer-
tain.
The biosynthetic origin of this class of compounds has
not been clearly demonstrated. On the basis of biogenetic
pathways proposed for rosigenin,3 and for metabolites
containing the oxaspiro[4,5]decane skeleton or the azaspiro-
The methoxy group was located on the basis of an HMBC
correlation of H3-21 with C-6. HMBC correlations of H3-20

76 J ournal of Natural Products, 2003, Vol. 66, No. 1
Oh et al.
Ta ble 4. NMR Data for Massarinins A (6) and B (7)
6a
7b
C#
δ
Η (mult., J )c
δC
HMBC (C#)d
δ
Η (mult., J )c
δC
sel. INEPT (C#)c
1
2
200.6
128.9
149.0
124.3
152.0
119.6
198.8
45.1
3 (OH)
4
9.71 (s)
7.04 (s)
2, 3
1, 2, 3, 6, 20
2.70 (dd; 17, 3.6)
2.40 (dd; 17, 10)
2.29 (m)
2, 3, 5, 6, 20
5
6
7
132.6
154.1
126.8
162.6
108.8
133.5
112.5
154.3
111.2
121.4
126.8
76.8
26.7
26.7
190.0
15.1
63.4
39.1
69.1
3, 6, 7, 20
2, 4, 5, 7, 19, 20
4.59 (br d; 7.5)
128.8
156.8
112.5
129.4
115.7
153.0
109.5
121.9
128.3
77.0
27.7
27.8
138.5
17.2
-
8 (OH)
9
12.77 (s)
6.44 (d; 8.1)
7.18 (d; 8.1)
8, 9, 13
8, 11, 13
8, 12, 14
7.96 (s)
6.62 (d; 8.1)
6.98 (d; 8.4)
8, 9, 13
10
11
12
13
14
15
16
17
18
19
20
21
6.24 (d; 9.9)
5.42 (d; 9.9)
10, 11, 12, 16
11, 16
6.26 (d; 9.6)
5.48 (d; 9.9)
0.86 (s)
0.86 (s)
10.15 (s)
2.28(s)
15, 16
15, 16
2, 7
4, 5, 6
6
1.37 (s)
1.37 (s)
7.5 (d; 1.2)
1.15 (d; 6.3)
1, 2, 7
3.80 (s)
a
b
d
DMSO-d6. CDCl3. c 300 MHz. 600 MHz (1H dimension).
with C-4, C-5, and C-6 placed the aryl methyl group ortho
to both the isolated aromatic proton and the methoxy group
(i.e., at C-5). A correlation of the aldehyde proton H-19
with C-7 appeared as a pair of satellite peaks reflecting a
1H-13C coupling constant of 23 Hz, which is consistent with
correlation with H-19, together with its chemical shift. C-7
and C-2 were also connected by virtue of selective INEPT
correlations of each with both H-5 and H-19. These con-
nections completed a second subunit consisting of a furan
ring fused to a cyclohexanone unit. On the basis of chemical
shift considerations, the remaining OH group (indicated
by the formula and DEPT data) must be located at C-6,
and C-1 and C-13 must be connected to complete the gross
structure of massarinin B (7).
The 1H NMR J -values for Hax-4 and Heq-4 with H-5 (10.2
and 3.6 Hz, respectively) suggested that Hax-4 and H-5
adopt a near trans-diaxial relationship, placing the H3-20
methyl group in an equatorial (or pseudoequatorial) posi-
tion with respect to the cyclohexanone ring. A strong
NOESY correlation between H-6 and Hax-4 indicated that
these groups are spatially close, suggesting that these
groups have a cis-1,3-diaxial relationship. On the basis of
these results, the relative stereochemistry of massarinin
B was proposed as shown in 7.
Massarinins A and B (6-7) share a skeleton that is most
likely derived from the polyketide pathway. The structure
of massarinin A (6) is analogous to that of arugosin E (10),16
and the numbering system shown for 6 and 7 was chosen
to be consistent with the numbering system for arugosins
A-C.18 It has been suggested that the arugosins are
derived from cleavage of an anthraquinone/anthrone pre-
cursor,19,20 and biosynthetic studies of similar compounds
support this hypothesis.21,22 Although several representa-
tives of this class have been reported,16,18-20 arugosin E (10)
is the only previously described member that possesses a
2,2-dimethylchromene unit as found in 6-7.16 The isoben-
zofuran skeleton in massarinin B (7) could be envisioned
as forming via adjustment of oxidation state and cyclization
reactions of the aldehyde (C-19) and ketone (C-1) groups
present in 6. Isobenzofuran (and tetrahydroisobenzofura-
none) units are precedented among natural products,23 but
are not common.
1
the expected value for the two-bond H-13C coupling in an
aldehyde group.17 This observation allowed direct connec-
tion of the aldehyde carbonyl C-19 to C-7. The remaining
phenolic proton (δ 9.71) showed correlations to C-2 and C-3,
placing this OH group at C-3, and ortho to C-2. Correlations
of H-19 and H-4 with C-2 further supported arrangement
of the substituents as shown in 6. The ketone carbon (C-1)
was connected to C-2 and to C-13 to complete the structure
assignment for 6 based on a weak four-bond HMBC
correlation between H-4 and C-1, together with the re-
quirement that OH-8 be involved in a strong intramolecu-
lar hydrogen bond.
NMR data and HREIMS analysis of massarinin B (7)
indicated that it has the molecular formula C20H20O5. The
1H and 13C NMR spectra of 7 (Table 4) again contained
signals for a hydroxy-substituted 2,2-dimethylchromene
unit. In this instance, the 8-OH proton appeared at δ 7.96,
suggesting the absence of an ortho ketone group as is
present in 6, although a ketone resonance (δ 198.8) was
observed. In addition, only four aromatic/olefinic 13C NMR
resonances were unaccounted for by the 2,2-dimethyl-
chromene unit, indicating the absence of a second benzene
1
ring in 7. Interestingly, the H NMR shifts of the methyl
groups of the 2,2-dimethylchromene unit are significantly
downfield-shifted compared to those in 6. This observation
is consistent with the absence of a benzene ring anisotropic
shielding effect in 7 that is present in 6. 1H-1H decoupling
and selective INEPT (Table 4) data revealed the presence
of a spin-system corresponding to the C4-C6/C20 portion
of 7. The ketone carbon (C-3) was linked to C-4 on the basis
of selective INEPT correlations with H-5 and H2-4. One of
the remaining double bonds (C-7/C-19) was linked to
oxymethine C-6 on the basis of correlations of H-5 with C-7
and of H-6 with C-19. Another sp2 carbon (δ 119.6; C-2)
was linked to the ketone carbon on the basis of correlations
observed with H2-4. The only remaining carbon signal
(C-1; δ 152.0) must be paired in a double bond with C-2
and was connected to C-19 via oxygen on the basis of a
The known compound 11 [4-(2-hydroxybutynoxy)benzoic
acid] was also isolated from M. tunicata in this investiga-
tion. This compound has been previously reported only as
a metabolite of an unidentified soil fungus.24 The structure
was independently confirmed by NMR, MS, and IR analy-
sis, although it afforded distinctive results in some of the

Bioactive Rosigenin Analogues from Massarina
J ournal of Natural Products, 2003, Vol. 66, No. 1 77
fractions. The material was eluted with 1 L of CH2Cl2, followed
by 1.5 L of 1% CH3OH in CH2Cl2, 500 mL each of 2%-7%
CH3OH in CH2Cl2, 300 mL each of 10% CH3OH in CH2Cl2,
20% CH3OH in CH2Cl2, and 30% CH3OH in CH2Cl2, and 500
mL of 50% CH3OH in CH2Cl2. Fractions of similar composition
were pooled on the basis of TLC analysis (9:1 CH2Cl2-
CH3OH). The fraction that eluted at 2% CH3OH (320 mg)
contained 1 and 7 and was subjected to Sephadex LH-20
column chromatography with a step-gradient elution sequence
of hexane-CH2Cl2 (1:4), CH2Cl2-acetone (4:1), and CH2Cl2-
acetone (2:3). Massarigenin A (1) eluted in the CH2Cl2-acetone
(2:3) subfraction, and white crystals of 1 were obtained upon
slow evaporation of the solvent. The fraction containing 7 was
eluted at CH2Cl2-hexane (4:1), and this fraction was purified
by semipreparative reversed-phase HPLC using a gradient
from 50 to 100% CH3CN in H2O over 30 min (Alltech 100 HS
BDS C18 column; 1.0 × 25 cm; 8 µm particle size; 2 mL/min;
UV detection at 215 nm) to yield 7 (9.6 mg).
The fraction that eluted with 1% CH3OH in CH2Cl2 from
silica gel VLC (540 mg) contained 2, 3, 4, and 6 and was
further separated by Sephadex LH-20 column chromatography
with the same solvent step-gradient elution sequence used in
isolating 1. A subfraction containing massarigenins B and C
(2 and 3; 71 mg) eluted with CH2Cl2-acetone (4:1) and was
further separated by reversed-phase HPLC as above using a
gradient from 40 to 53% CH3CN in H2O over 10 min to yield
2 (11 mg) and 3 (8.8 mg). An adjoining subfraction con-
taining 6 was also eluted at CH2Cl2-acetone (4:1), and this
fraction was subjected to reversed-phase HPLC as above using
a gradient program of 40-47% CH3CN in H2O over 10 min,
47-55% CH3CN in H2O over 5 min, and 55-62% CH3CN in
H2O over 10 min to yield 6 (3.8 mg). A fraction containing
massarigenin D (4; 34 mg) was eluted with hexane-CH2Cl2
(1:4) and was similarly subjected to reversed-phase HPLC
using a gradient from 35 to 65% CH3CN in H2O over 35 min
to yield 4 (11 mg).
NMR experiments (e.g., initially misleading DEPT data)
by virtue of the large (52 Hz) heteronuclear 2J -value
observed between H-4′ and C-3′. To determine the absolute
configuration at C-2′of 11, the R- and R/ S-phenylbutyrate
ester derivatives of its corresponding methyl ester were
prepared. Treatment of 11 with trimethylsilyldiazomethane
in CH3OH yielded methyl ester 12. Separate reactions of
this product with R- and R/ S-phenylbutyric acid in the
presence of 1,3-dicyclohexylcarbodiimide and 4-N,N-di-
methylaminopyridine produced the desired derivatives, as
1
confirmed by HRFABMS data and H NMR results, includ-
ing a downfield shift of H-2′ from δ 4.77 to 5.75. It was
possible to assign the relevant proton signals of the S-2-
1
phenylbutyrate ester by comparison of the H NMR data
for the product mixture with the data for the independently
prepared R-ester. A significant downfield shift of the H-4′
resonance and upfield shifts of the H2-1′, H-2, and H-3
signals for the S-derivative relative to those of the R-
derivative were observed. On the basis of Helmchen’s
rules,25 the S-configuration was assigned to C-9.
In standard disk assays, compounds 1, 3, 4, 6, 7, and 11
were active against Bacillus subtilis (ATCC 6051), causing
zones of inhibition of 11, 9, 14, 17, 23, and 15 mm,
respectively, at 200 µg/disk. Compound 2 was not active
in this assay. Compounds 6 and 7 also showed activity
against Staphylococcus aureus (ATCC 29213), affording
zones of inhibition of 7 and 12 mm, respectively, at the
same concentration. None of the compounds showed sig-
nificant activity against Candida albicans (ATCC 14053),
Aspergillus flavus (NRRL 6541), or Fusarium verticillioides
(NRRL 25457) at the same level.
Exp er im en ta l Section
The material that eluted from the silica gel VLC column
with 4% CH3OH in CH2Cl2 (335 mg) was subjected to Sephadex
LH-20 column chromatography as described above. Slow
evaporation of a subfraction eluted with 4:1 CH2Cl2-acetone
afforded 40 mg of 4-(2-hydroxybutynoxy)benzoic acid (11) as
a white powder.
Gen er a l Exp er im en ta l P r oced u r es. NMR spectra were
recorded using CDCl3, DMSO-d6, or acetone-d6 solutions, and
chemical shifts were referenced relative to the corresponding
residual solvents signals (δH 7.24/δC 77.0, δH 2.49/δC 39.5, and
δH 2.04/δC 29.8, respectively). Carbon multiplicities were
established by DEPT experiments. Selective INEPT NMR
experiments were carried out at 75 MHz on a Bruker AC-300
spectrometer. HMQC and HMBC experiments were optimized
Ma ssa r igen in A (1): white crystals; mp 169-171 °C; [R]D
-5.6° (c 0.26 g/dL; 24 °C; CH3OH); UV (CH3OH) 210 (ꢀ 2000);
IR 3590, 2928, 1806 (ester group), 1682 (vinyl ether group),
1605, 1275, 1174, 1103, 1027 cm-1; EIMS (70 eV) m/z 208 [(M
- H2O)+, rel int 9], 193 (36), 175 (15), 152 (25), 140 (32), 125
n
for J CH ) 135 and 8 Hz, respectively, and conducted using a
Bruker AMX-600 spectrometer. FABMS, HRFABMS, and
HREIMS data were obtained on a VG ZAB-HF mass spec-
trometer, and EIMS data were obtained on a VG TRIO 1
quadrupole instrument at 70 eV. Reagents for chemical
reactions were purchased from Aldrich Chemical Co. Other
general procedures and instrumentation employed have been
described previously.26
1
(29), 105 (29), 91 (32), 84 (100), 77 (34), 55 (66), 43 (56); H
NMR data, Table 1; 13C NMR data, Table 2; HMBC data, Table
3; NOESY data (acetone-d6, H-# T H-#) H-6 T H-10; H2-11 T
H3-12; HRFABMS (NaI/3-NBA matrix) obsd m/z 249.0735,
calcd for C11H14O5+Na, 249.0739.
Ma ssa r igen in B (2): white solid; mp 78-79 °C; [R]D -9.3°
(c 0.35 g/dL; 24 °C; CH3OH); HPLC tR 8.8 min; UV (CH3OH)
212 (ꢀ 2700); IR 3509, 2970, 1802, 1727, 1678, 1190, 1153,
1111, 1039 cm-1; EIMS (70 eV) m/z 226 (M+, rel int 69), 208
(7), 190 (5), 156 (49), 137 (48), 127 (92), 111 (61), 97 (51), 81
(75), 69 (38), 55 (99), 43 (100); 1H NMR data, Table 1; 13C NMR
data, Table 2; selective INEPT data, Table 3; NOESY data
(CDCl3, H-# T H-#) H-4 T H-7ax; H-4 T H3-12; H-6 T H-10;
H-6 T H-8ax; HRFABMS (LiI/3-NBA matrix) obsd m/z
233.0987, calcd for C11H14O5+Li, 233.1001.
Ma ssa r igen in C (3): white crystals; mp 81-83 °C; [R]D
-123° (c 0.34 g/dL; 24 °C; CH3OH); HPLC tR 8.0 min; UV (CH3-
OH) 238 (ꢀ 2900); IR 3538, 2977, 1798, 1698, 1677, 1381, 1179,
1147, 1090, 1042 cm-1; EIMS (70 eV) m/z 224 (M+, rel int 2),
141 (7), 123 (5), 107 (4), 95 (4), 82 (100), 67 (4), 54 (14), 43
(15); 1H NMR data, Table 1; 13C NMR data, Table 2; selective
INEPT data, Table 3; NOESY data (acetone-d6, H-# T H-#)
H-4 T H3-12; H-6 T H-10; HRFABMS (NaI/3-NBA matrix)
obsd m/z 247.0575, calcd for C11H12O5+Na, 247.0582.
Acetyla tion of Ma ssa r igen in C (3). A solution of 3 (1.1
mg), 4-N,N-dimethylaminopyridine (0.5 mg), and acetic an-
hydride (0.5 mL) in acetone (2 mL) was stirred for 20 h at room
temperature. The solvent was then evaporated under N2 flow.
Isola tion , Cu ltiva tion , a n d Extr a ction of M. tu n ica ta .
The strain of Massarina tunicata used in this study (culture
number A25-1 ) ATCC 201760) was isolated from a decorti-
cated submerged twig collected from the Lemonweir River,
Adams County, WI, on J uly 31, 1992, by C.A.S. This fungus
is classified in the Lophiostomataceae, Pleosporales, Loculoas-
comycetes. The cultivation procedure was analogous to that
reported earlier,1 except that it was carried out on a larger
scale. Twenty flasks, each containing 400 mL of potato
dextrose broth (Difco) which had been sterilized at 120 °C for
15 min and then cooled to room temperature, were individually
inoculated with 1 cm2 agar plugs taken from stock cultures of
M. tunicata. Flask cultures were inoculated at 25-28 °C and
aerated by agitation on an orbital shaker at 150 rpm for period
of 30 days. Extraction of the filtered broth with EtOAc (5 × 1
L) provided an organic phase which was dried with MgSO4
and then concentrated using a rotary evaporator to yield 4.0
g of crude extract.
Isola tion of Meta bolites 1-4, 6, 7, a n d 11. The EtOAc
extract (4.0 g) was subjected to silica gel VLC using a stepwise
gradient elution of CH3OH in CH2Cl2, collecting 200-400 mL

78 J ournal of Natural Products, 2003, Vol. 66, No. 1
Oh et al.
The residue was redissolved in 1.5 mL of EtOAc and extracted
with H2O (2 × 2 mL). The organic phase was dried (MgSO4)
and evaporated to yield diacetate 5 (1.3 mg, 86% yield): 1H
NMR data (CDCl3, 300 MHz) δ 1.27 (d; J H-H ) 7.2 Hz; H3-12),
1.99 (s, COCH3-4*), 2.15 (s, COCH3-10*), 3.45 (m, H-6), 4.53
(dd; 3.3, 2.0; H-11a), 4.93 (dd; 3.3, 2.4; H-11b), 5.72 (s, H-10),
6.00 (dd; 10.2, 3.2; H-8), 6.20 (t; 2.1; H-4), 6.60 (dd; 10.2, 2.1;
H-7); NOESY data (acetone-d6, H-# T H-#) H-4 T H3-12; H-6
T H-10; H-6 T H-7; H-7 T H3-12; *these assignments are
interchangeable.
Meth yl 4-(2-h yd r oxybu tyn oxy)ben zoa te (12). To a solu-
tion of 2.8 mg (14 µmol) of 11 in 1 mL of CH3OH was added a
2 M solution of trimethylsilyldiazomethane (TMSCHN2) in
hexane (140 µL) until the solution stayed yellow. After stirring
for 3 h, the solution was concentrated under N2 flow to give
methyl ester 12 (2.8 mg, 93%): 1H NMR (CDCl3, 300 MHz) δ
7.99 (distorted d, J H-H ) 8.5 Hz, H-2/6), 6.94 (distorted d, 8.5,
H-3/5), 4.77 (m, H-2′), 4.19 (dd, 9.6, 3.8, H-1′), 4.12 (dd, 9.6,
6.9, H-1′), 3.87 (s, OCH3), 2.53 (d, 2.4, H-4′); HREIMS obsd
m/z 220.0734, calcd for C12H12O4, 220.0736.
R-2-P h en ylbu tyr a te Ester of 12. To a solution of 1,3-
dicyclohexylcarbodiimide (5.4 mg, 26 µmol) in THF (3 mL) were
added R-2-phenylbutyric acid (4 µL, 26 µmol), compound 12
(1.8 mg, 8.1 µmol), and a catalytic amount of 4-N,N-dimeth-
ylaminopyridine (0.5 mg). After the mixture was stirred for
48 h, TLC analysis confirmed the disappearance of starting
material and the solvent was evaporated under N2 flow. The
residue was then redissolved in 1.5 mL of Et2O, and the
resulting solution was extracted sequentially with 2%
CH3COOH (2 mL), 3% NaHCO3 (2 mL), and H2O (2 × 2 mL).
The organic phase was dried (MgSO4) and evaporated to give
a crude reaction product. This material was subjected to
semipreparative reversed-phase HPLC (Beckman Ultrasphere
C8 column; 1.0 × 25 cm; 5 µm particle size; 2 mL/min; UV
detection at 215 nm) using a gradient elution from 80 to 100%
CH3OH in H2O over 30 min to yield 2.4 mg of the R-2-
phenylbutyrate ester of 12: HPLC tR 14.0 min (under the
Ma ssa r igen in D (4): white solid; mp 117-121 °C; [R]D -96°
(c 0.54 g/dL; 23 °C; CH3OH); HPLC tR 9.0 min (under the
conditions above); UV (CH3OH) 244 (ꢀ 3100); IR 3482, 2976,
1803, 1695, 1681, 1373, 1168, 1136, 1101, 1051 cm-1; EIMS
(70 eV) m/z 224 (M+, rel int 0.9), 141 (13), 123 (7), 107 (3), 95
1
(3), 82 (100), 67 (4), 54 (16), 43 (20); H NMR data, Table 1;
13C NMR data, Table 2; NOESY data (acetone-d6, H-# T H-#)
H-4 T H3-12; H-4 T H-6; H-10 T H3-12; HRFABMS (NaI/
3-NBA matrix) obsd m/z 247.0570, calcd for C11H12O5+Na,
247.0582.
X-r a y Cr yst a llogr a p h ic An a lysis of Ma ssa r igen in A
(1).27 A crystal of 1 (0.45 × 0.25 × 0.09 mm) was orthorhombic
(space group P212121) with cell dimensions a ) 8.251(1), b )
16.147(2), c ) 8.229(1) Å. Data were collected on an Enraf-
Nonius CAD4 diffractometer (Mo KR radiation; graphite
monochromator) at 203 K (N2 cold gas stream) using θ-2θ
scans. The structure was solved using a MULTAN direct
methods program and refined using full-matrix least-squares
with the XL computer program from the SHELXTL v5.0
package. Computer programs from the MoLEN package were
used for data reduction. The 2872 measurements yielded 1879
independent reflections (201 parameters) after equivalent data
were averaged and Lorentz and polarization corrections were
applied. The final refinement gave R1 ) 0.0318, wR2 ) 0.0680.
1
conditions above); H NMR data (CDCl3, 300 MHz) δ (mult,
J H-H in Hz, H-#) 7.97 (distorted d, 8.5, H-2/6), 7.21-7.29 (ov.
m, Ar-H5), 6.87 (distorted d, 8.5, H-3/5), 5.75 (ddd, 6.6, 5.0,
2.3, H-2′), 4.23 (ov. m, H2-1′), 3.87 (s, OCH3), 3.49 (dd, 8.3, 7.5,
H-2′′), 2.44 (d, 2.3, H-4′), 2.08 (m, Ha-3′′), 1.79 (m, Hb-3′′), 0.88
(dd, 7.5, 7.5, H3-4′′); HRFABMS (thioglycerol matrix) obsd m/z
367.1527, calcd for C22H22O5+H, 367.1545.
R/S-2-P h en ylbu tyr a te Ester s of 12. (R/ S)-2-Phenylbu-
tyric acid (2.5 mg, 15 µmol), compound 12 (1.2 mg, 5.5 µmol),
and catalytic amount of 4-N,N-dimethylaminopyridine (0.5 mg)
were added to a solution of 1,3-dicyclohexylcarbodiimide (3.1
mg, 15 µmol) in THF (3 mL). The reaction mixture was stirred
for 7 days and was then worked up as described above to yield
a mixture of R- and S-derivatives. Chemical shift assignments
for the S-phenylbutyrate ester were made by comparison of
the 1H NMR data for the mixture with those of the indepen-
dently prepared R-derivative (see above): 1H NMR data (S-
phenylbutyrate ester; CDCl3, 300 MHz) δ (mult, J H-H ) Hz,
H-#) 7.91 (distorted d, 8.5, H-2/6), 6.73 (distorted d, 8.5, H-3/
5), 5.75 (m, H-2′), 4.14 (m, H2-1′), 3.86 (s, OCH3), 2.51 (d, 2.4,
H-4′).
Ma ssa r in in A (6): pale yellow oil; HPLC tR 27.4 min (BDS
column conditions described above); UV (CH3OH) 222 (ꢀ
14000), 274 (ꢀ 12000), 355 (ꢀ 3400); IR 3567, 3063, 2978, 2931,
1685, 1652, 1616, 1474, 1244, 1206, 1119 cm-1; EIMS (70 eV)
m/z 368 (M+, rel int 25), 353 (49), 339 (43), 319 (18), 193 (15),
187 (57), 161 (100), 77 (19); 1H, 13C, and HMBC data (DMSO-
1
d6), Table 4; H NMR data (CDCl3, 300 MHz) δ 7.00 (s, H-4),
7.01 (d; J H-H ) 6.9, H-10), 6.49 (d; J H-H ) 6.9, H-9), 6.11 (d;
J H-H ) 8.3, H-14), 5.26 (d; J H-H ) 8.3, H-15), 0.93 (s, H3-17,-
18), 10.22 (s, H-19), 2.33 (s, H3-20), 3.82 (s, H3-21), 12.41 (s,
8-OH); 13C NMR data (CDCl3, 75.5 MHz) δ 200.3 (C-1), 128.5
(C-2), 147.8 (C-3), 125.2 (C-4), 134.0 (C-5, 10), 155.5 (C-6), 127.8
(C-7), 163.5 (C-8), 109.6 (C-9), 112.7 (C-11), 154.5 (C-12), 111.7
(C-13), 121.8 (C-14), 126.4 (C-15), 77.3 (C-16), 27.3 (C-17, 18),
189.3 (C-19), 15.5 (C-20), 63.6 (C-21); FABMS (3-NBA matrix)
(M + H)+ m/z 369, (M + Na)+ m/z 391; HREIMS obsd m/z
368.1256, calcd for C21H20O6, 368.1260.
Ack n ow led gm en t. We thank the National Institutes of
Health (R01 GM 60600) for financial support, and Mr. Ricardo
Reategui for technical assistance.
Ma ssa r in in B (7): yellow solid; mp 123-125 °C; [R]D -204°
(c 0.6 g/dL; 25 °C; CH3OH); HPLC tR 19.3 min (under the
conditions above); UV (CH3OH) 229 (ꢀ 8500), 291 (ꢀ 5200); IR
3601, 3575, 3948, 2978, 2930, 1656, 1605, 1537, 1480, 1125
cm-1; EIMS (70 eV) m/z 340 (M+, rel int 18), 325 (100), 307
(26), 279 (6), 251 (6), 187 (14), 161 (10), 127 (12), 115 (16), 105
(14); 1H, 13C, and selective INEPT NMR data, Table 4;
HREIMS obsd m/z 340.1314, calcd for C20H20O5, 340.1311.
Su p p or tin g In for m a tion Ava ila ble: 1H and 13C NMR spectra
for massarigenins A-D (1-4) and massarinins A and B (6, 7), key
selective INEPT data for massarinin B (7), and tables of X-ray data
for massarigenin A (1). This material is available free of charge via
the Internet at http://pubs.acs.org.
Refer en ces a n d Notes
(1) Oh, H.; Shearer, C. A.; Gloer, J . B. J . Nat. Prod. 1999, 62, 497-501.
(2) Oh, H.; Swenson, D. C.; Gloer, J . B.; Shearer, C. A. Tetrahedron Lett.
2001, 42, 975-977.
4-(2-Hyd r oxybu tyn oxy)ben zoic a cid (11): white powder;
mp 126-127 °C; [R]D +80° (c 0.4 g/dL; 24 °C; CH3OH); UV
(CH3OH) 213 (ꢀ 2900), 253 (ꢀ 6800); IR 3306, 3024, 3012, 2125,
1710, 1607, 1425, 1363, 1255, 1170 cm-1; EIMS (70 eV) m/z
206 (M+, rel int 23), 151 (38), 138 (34), 121 (100), 105 (20), 93
(17), 65 (75); 1H NMR (CDCl3, 300 MHz) δ 7.99 (distorted d, J
) 7.5 Hz, H-2/6), 7.06 (distorted d, 7.5, H-3/5), 4.74 (ddd, 6.7,
4.7, 2.2, H-2′), 4.20 (dd, 9.7, 4.7, H-1′), 4.16 (dd, 9.7, 6.7, H-1′),
2.96 (d, 2.2, H-4′); 13C NMR (CDCl3, 75 MHz) δ 167.3 (C-7),
163.5 (C-4), 132.6 (2C, C-2/6), 124.1 (C-1), 115.3 (2C, C-3/5),
83.5 (C-3′), 74.8 (C-4′), 72.8 (C-1′), 61.3 (C-2′); HMBC (600
MHz) H-2/6 f C-6/2, 4, 7; H-3/5 f C-1, 4, 5/3; H2-1′ f C-2′,
3′, 4; H-2′ f C-1′, 3′, 4′; H-4′ f C-1′, 2′, 3′; FABMS (3-NBA
matrix) (M + H)+ m/z 207; HREIMS obsd m/z 206.0583, calcd
for C11H10O4, 206.0579.
(3) Albinati, A.; Bruckner, S.; Camarda, L.; Nasini, G. Tetrahedron Lett.
1980, 36, 117-121.
(4) Ayer, W. A.; Craw, P. A.; Neary, J . Can. J . Chem. 1992, 70, 1338-
1347.
(5) Ayer, W. A.; Craw, P. A. J . Can. J . Chem. 1992, 70, 1348-1355.
(6) Guiraud, P.; Steiman, R.; Seigle-Murandi, F.; de Gusmao, N. B. J .
Nat. Prod. 1999, 62, 1222-1224.
(7) Doi, J .; Hirota, A.; Nakagawa, M.; Sakai, H.; Isogai, A. Agric. Biol.
Chem. 1985, 49, 2247-2248.
(8) Isogai, A.; Nakagawa, S.; Hirota, A.; Doi, A.; Nakanishi, O. Chem.
Abstr. 1988, 108, 30050a.
(9) Sugawara, F.; Takahashi, N.; Strobel, G. A.; Strobel, S. A.; Lu, H. S.
M.; Clardy, J . J . Am. Chem. Soc. 1988, 110, 4086-4087.
(10) Fukami, A.; Taniguchi, Y.; Nakamura, T.; Rho, M.-C.; Kawaguchi,
K.; Hayashi, M.; Komiyama, K.; Oh mura, S. J . Antibiot. 1999, 52, 501-
504.

Bioactive Rosigenin Analogues from Massarina
J ournal of Natural Products, 2003, Vol. 66, No. 1 79
(11) Roll, D. M.; Tischler, M.; Williamson, R. T.; Carter, G. T. J . Antibiot.
2002, 55, 520-523.
(22) Nishida, H.; Tomoda, H.; Okuda, S.; Oh mura, S. J . Org. Chem. 1992,
57, 1271-1274.
(12) Sandmeier, P.; Tamm, C. Helv. Chim. Acta 1989, 72, 774-783.
(13) Sandmeier, P.; Tamm, C. Helv. Chim. Acta 1989, 72, 784-792.
(14) Sandmeier, P.; Tamm, C. Helv. Chim. Acta 1989, 72, 1107-1120.
(15) Sandmeier, P.; Tamm, C. Helv. Chim. Acta 1990, 73, 975-984.
(16) Kawahara, N.; Nozawa, K.; Nakajima, S.; Kawai, K. J . Chem. Soc.,
Perkin Trans. 1 1988, 907-911.
(23) Turner, W. B.; Aldridge, D. C. Fungal Metabolites II; Academic
Press: New York, 1983; pp 55-70.
(24) Yang, J .; Qi, Xiu L.; Li, W.; Shao, G.; Yao, X. S.; Kitanaka, S. Chin.
Chem. Lett. 1998, 9, 539-540.
(25) Helmchen, G. Tetrahedron Lett. 1974, 1527-1530.
(26) Ho¨ller, U.; Gloer, J . B.; Wicklow, D. T. J . Nat. Prod. 2002, 65, 876-
882.
(27) Crystallographic data for compound 1 have been deposited with the
Cambridge Crystallographic Data Centre (deposition number CCDC
193811). Copies of the data can be obtained, free of charge, on
application to the Director, CCDC, 12 Union Road, Cambridge CB2
1EZ, UK (fax: +44-(0)1223-336033 or e-mail: deposit@ccdc.cam.ac.uk).
(17) A¨ yra¨s, P.; Laatikainen, R.; Lo¨tjo¨nen, S. Org. Magn. Reson. 1980, 13,
387-390.
(18) Ballantine, J . A.; Ferrito, V.; Hassall, C. H.; J enkins, M. L. J . Chem.
Soc., Perkin Trans. 1 1973, 1825-1830.
(19) Turner, W. B.; Aldridge, D. C. Fungal Metabolites II; Academic
Press: New York, 1983; pp 156-167.
(20) Chexal, K. K.; Holker, J . S. E.; Simpson, T. J . J . Chem. Soc., Perkin
Trans. 1 1975, 549-554.
(21) Holker, J . S. E.; Lapper, R. D.; Simpson, T. J . J . Chem. Soc., Perkin
Trans. 1 1974, 2135-2140.
NP020342D