Frenolicins C-G, Pyranonaphthoquinones from streptomyces sp. RM-4-15

Article abstract of DOI:10.1021/np400231r

Appalachian active coal fire sites were selected for the isolation of bacterial strains belonging to the class actinobacteria. A comparison of high-resolution electrospray ionization mass spectrometry (HRESIMS) and ultraviolet (UV) absorption profiles from isolate extracts to natural product databases suggested Streptomyces sp. RM-4-15 to produce unique metabolites. Four new pyranonaphthoquinones, frenolicins C-F (1-4), along with three known analogues, frenolicin (6), frenolicin B (7), and UCF76-A (8), were isolated from the fermentation of this strain. An additional new analogue, frenolicin G (5), along with two known compounds, deoxyfrenolicin (9) and UCF 13 (10), were isolated from the fermentation supplied with 18 mg/L of scandium chloride, the first example, to the best of our knowledge, wherein scandium chloride supplementation led to the confirmed production of new bacterial secondary metabolites. Structures 1-5 were elucidated on the basis of spectral analysis and chemical modification. While frenolicins are best known for their anticoccidial activity, the current study revealed compounds 6-9 to exhibit moderate cytotoxicity against the human lung carcinoma cell line (A549) and thereby extends the anticancer SAR for this privileged scaffold.

Full text of DOI:10.1021/np400231r

Article
pubs.acs.org/jnp
Frenolicins C−G, Pyranonaphthoquinones from Streptomyces sp. RM-
4-15
Xiachang Wang,† Khaled A. Shaaban,† Sherif I. Elshahawi,† Larissa V. Ponomareva,† Manjula Sunkara,‡
Yinan Zhang,† Gregory C. Copley,§ James C. Hower,§ Andrew J. Morris,‡ Madan K. Kharel,*,†
and Jon S. Thorson*,†
†Center for Pharmaceutical Research and Innovation, College of Pharmacy, University of Kentucky, 789 South Limestone Street,
Lexington, Kentucky 40536-0596, United States
‡Division of Cardiovascular Medicine, University of Kentucky, Lexington, Kentucky 40536, United States
§Center for Applied Energy Research, University of Kentucky, Lexington, Kentucky 40511, United States
S
* Supporting Information
ABSTRACT: Appalachian active coal fire sites were selected
for the isolation of bacterial strains belonging to the class
actinobacteria. A comparison of high-resolution electrospray
ionization mass spectrometry (HRESIMS) and ultraviolet
(UV) absorption profiles from isolate extracts to natural
product databases suggested Streptomyces sp. RM-4-15 to
produce unique metabolites. Four new pyranonaphthoqui-
nones, frenolicins C−F (1−4), along with three known
analogues, frenolicin (6), frenolicin B (7), and UCF76-A (8),
were isolated from the fermentation of this strain. An
additional new analogue, frenolicin G (5), along with two
known compounds, deoxyfrenolicin (9) and UCF 13 (10),
were isolated from the fermentation supplied with 18 mg/L of scandium chloride, the first example, to the best of our knowledge,
wherein scandium chloride supplementation led to the confirmed production of new bacterial secondary metabolites. Structures
1−5 were elucidated on the basis of spectral analysis and chemical modification. While frenolicins are best known for their
anticoccidial activity, the current study revealed compounds 6−9 to exhibit moderate cytotoxicity against the human lung
carcinoma cell line (A549) and thereby extends the anticancer SAR for this privileged scaffold.
HRMS) profiling of the corresponding culture extracts and
metabolite database comparison revealed the potential presence
of unique metabolites in the extract produced by Streptomyces
sp. RM-4-15. In this report, we describe the comparative
fermentation of Streptomyces sp. RM-4-15 in the absence and
presence of scandium chloride and the corresponding isolation,
structure elucidation, and biological activity of five new
frenolicin analogues (1−5) and five previously reported
analoguesfrenolicin (6), frenolicin B (7), UCF76-A (8),
deoxyfrenolicin (9), and UCF 13 (10)produced by
Streptomyces sp. RM-4-15. Consistent with previous studies
that revealed notable enhancement of known metabolite
production in the presence of scandium chloride,11 the
presence of scandium chloride in the current study led to
marked improvements in production of frenolicin (6) and
frenolicin B (7) by Streptomyces sp. RM-4-15. Interestingly,
unlike prior studies, the production of frenolicin G (5),
deoxyfrenolicin (9), and UCF 13 (10) by Streptomyces sp. RM-
4-15 was notably only observed in the presence of the heavy
The Appalachian Mountains of eastern Kentucky are known for
both their rich natural biodiversity and the production of
coal.1−5 Fires within abandoned underground mines through-
out the region are a common occurrence and contribute to the
emission of greenhouse gases and the dispersion of volatile
organic and inorganic species through surface vents.6−9 The
Ruth Mullins underground coal mine fire, which unofficially
dates back to 1960, is located in the high volatile A bituminous
Pennsylvanian-age Hazard No.7 coal bed, Perry County,
Kentucky. Recent studies of the Ruth Mullins site highlight a
notable alteration of the soil composition in and around the
fire’s thermal vents with adjacent surface temperature variation
ranging from 26.0 to 64.8 °C.10 In the context of microbial
diversity, the convergence of natural biodiversity and man-made
environmental alteration upon the Ruth Mullins coal fire site
offers an unprecedented environment for exploration. As part
of a broad natural product discovery initiative, we examined soil
samples collected from this site with a focus upon actino-
mycetes capable of producing novel secondary metabolites.
More than two dozen different actinomycetes were isolated
from a single soil sample collected from the Ruth Mullins coal
fire site. HPLC−high-resolution mass spectrometry (HPLC-
Received: March 19, 2013
Published: August 14, 2013
© 2013 American Chemical Society and
American Society of Pharmacognosy
1441
dx.doi.org/10.1021/np400231r | J. Nat. Prod. 2013, 76, 1441−1447

Journal of Natural Products
Article
Figure 1. Structures of compounds 1−10.
Table 1. NMR Data for Frenolicins C (1) and G (5)
frenolicin C (1)
frenolicin G (5)
a
b
c
d
position
δC, type
δH (J in Hz)
HMBC
δC, type
δH (J in Hz)
HMBC
1
77.0, CH
62.8, CH
29.4, CH2
3.94, d (11.1)
4.45, m
3, 4a, 10a
80.1, CH
63.1, CH
29.1, CH2
3.81, dd (9.8, 1.4)
3.89, m
3, 4a, 10a, 11, 12
14, 15
3
4α
4β
4a
2.13, t (13.7, 11.9)
1.89, d (14.2)
4a, 5, 10a
3, 4a, 5
2.11, d (13.7)
2.25, dd (15.0, 11.8)
3, 5, 10a
3, 4a, 5, 14
58.3, C
189.5, C
132.8, C
118.5, CH
137.1, CH
123.8, CH
160.5, C
114.4, C
198.6, C
75.8, C
64.6, C
191.5, C
135.0, C
121.8, CH
138.7, CH
124.9, CH
163.1, C
115.8, C
199.4, C
77.4, C
5
5a
6
7.42, d (7.4)
5, 8, 9a
5a, 9
7.77, d (6.3)
5, 8, 9a
5a, 9
7
7.73, t (7.5, 8.3)
7.28, d (8.4)
7.73, t (7.4, 8.2)
7.26, d (6.9)
8
6, 9a
6, 9, 9a
9
9a
10
10a
11
12
13
14
15
9-OH
10a-OH
1′
29.2, CH2
20.0, CH2
13.8, CH3
40.1, CH2
172.3, C
1.69, 2.32, m
1.31, 1.59, m
0.90, t (7.3)
12
31.1, CH2
21.8, CH2
14.3, CH3
41.4, CH2
1.40, 1.51, m
1.06, 1.31, m
0.63, t (7.3)
2.40, 2.52, m
12, 13
11, 13
11, 12
3, 4, 15
11, 13
11, 12
3, 15
e
2.38, 2.58, m
174.6, C
b
11.21, s
7.15, s
11.1,
7.32
s
b
1, 4a, 10
4a, 2′, 3′
1′, 3′, 4′
c
31.5, CH2
51.2, CH
171.4, C
2.44, 2.58, m
2′
3′
4.12, m
4′
169.4, C
5′
22.2, CH3
1.74, s
4′
N−H
8.16, d (6.7)
b
2′, 4′
a
c
d
Measured in DMSO-d6, 100 MHz. Measured in DMSO-d6, 500 MHz. Measured in CD3OD, 100 MHz. Measured in CD3OD, 500 MHz.
Overlap.
e
metal additive. While frenolicins are best known for their
anticoccidial activity,12 the corresponding bioactivity assess-
ment highlighted within also extends the anticancer structure−
activity relationship (SAR) of this structural family.
From a separate 10 L culture, which was supplemented with 18
mg/L of scandium chloride, compounds 6 and 7 were isolated
with higher yield. An additional new analogue, frenolicin G (5,
yield: 1.2 mg/L), along with two known compounds,
deoxyfrenolicin (9, yield: 1.5 mg/L) and UCF 13 (10, yield:
0.8 mg/L), were also isolated and characterized.
RESULTS AND DISCUSSION
■
Compound 1 was isolated as a yellow, amorphous powder.
The HRESIMS of 1 displayed masses of m/z = 510.1380 [M +
H]+ and m/z = 508.1340 [M − H]− in positive and negative
modes, respectively, suggesting a molecular formula of
C23H27O10NS with 11 degrees of unsaturation. The analysis
A four-day grown seed culture (300 mL) of Streptomyces sp.
RM-4-15 was used to inoculate 4 L of production medium.
Fermentation of the strain for six days followed by extraction,
fractionation, and resolution of components within the crude
extract mixture led to the isolation of four new compounds,
frenolicins C (1, yield: 4.25 mg/L), D (2, yield: 2.5 mg/L), E
(3, yield: 3.0 mg/L), and F (4, yield: 1.25 mg/L), in addition to
the known analogues frenolicin (6, yield: 2.25 mg/L), frenolicin
B (7, yield: 2.75 mg/L), and UCF76-A (8, yield: 0.5 mg/L).
1
of the H/13C and gHSQC NMR data suggested the presence
of two methyl, five methylene, and six methine groups and 10
quaternary carbons (Table 1). Comprehensive analysis of the
COSY spectrum of 1 established four structural fragments as
1442
dx.doi.org/10.1021/np400231r | J. Nat. Prod. 2013, 76, 1441−1447

Journal of Natural Products
Article
illustrated in Figure 2. The presence of three aromatic proton
signals at δH 7.73 (1H, t, J = 7.5, 8.3 Hz, H-7), 7.42 (1H, d, J =
correlation but displayed a COSY methine (δH 4.12, H-2′)
correlation and HMBC correlations with C-2′ (δC 51.2) and
the N-acetyl carbonyl C-4′ (δC 169.4). The observation of the
key HMBC cross-peaks [H2-1′ (δH 2.44, 2.58) and C-2′ (δC
51.2), C-3′ (δC 171.4); H-2′ (δH 4.12) and C-1′ (δC 31.5), C-3′
(δC 171.4), C-4′ (δC 169.4); and the acetate methyl H3-5′ (s,
δH 1.74) and C-4′ (δC 169.4)] as well as the presence of a sulfur
atom in the molecular formula of 1 suggested the presence of
an N-acetyl-cysteine moiety connected at the C-4a position.
This was further confirmed by the presence of a strong cross-
peak of H2-1′ and C-4a (δC 58.3) in the HMBC spectrum.
The relative configuration of 1 was established by ROESY
experiments (Figure 2). Two cross-peaks of H-1/OH-10a and
OH-10a/H-4α observed in the ROESY indicated that they
were α-oriented. Similarly, the ROESY correlations of H-3/H-
4β and H-4β/H2-1′ suggested their β-orientations. To
determine the absolute configuration of the amino acid moiety,
compound 1 was subjected to acid hydrolysis. The hydrolyti-
cally released amino acid was derivatized following Marfey’s
method, and the product was subsequently compared to amino
acid standards (derivatized in an identical manner) via HPLC.13
This analysis revealed the identity of the hydrolytic product as
L-cysteine. On the basis of this cumulative analysis, the
structure of 1 was established as depicted in Figure 1 and, as
a new member of the frenolicin family, given the name
frenolicin C.
Figure 2. COSY, select HMBC, and ROESY correlations of frenolicin
C (1).
7.4 Hz, H-6), and 7.28 (1H, d, J = 8.4 Hz, H-8) indicated the
presence of a trisubstituted benzene ring, which was further
supported by the observation of COSY cross-peaks for H-6/H-
7 and H-7/H-8 and HMBC correlations for H-6/C-9a, H-7/C-
9, and H-8/C-9a (Figure 2). The designation of the n-propyl
group at C-1 was supported by the key HMBC correlations H-
1/C-11 and H-1/C-12. The singlet proton signal at δH 11.21 (s,
OH-9) revealed HMBC correlations to C-8, C-9, and C-9a in
acetone-d6 (see Supporting Information S5), consistent with a
hydroxyl group at C-9 position chelated with the C-10 carbonyl
(δC 198.6) through an intramolecular hydrogen bond.
Comparison of the NMR data of 1 with the known analogue
frenolicin (6) indicated structural conservation with divergent
C-4a and C-10a substitution. The hydroxyl proton at δH 7.15
(s) demonstrated HMBC correlations with C-1 (δC 77.0), C-4a
(δC 58.3), and C-10 (δC 198.6), consistent with C-10a
regiospecificity. Consistent with the presence of an amide
bond, the NH proton at δH 8.16 (d, J = 6.7 Hz) lacked HSQC
HRESIMS of compounds 2 and 3 suggested both to be
1
isomers of the molecular formula C18H20O8. H and 13C NMR
of both compounds were similar to those of 1 (Table 2); both
notably lacked the N-acetyl-L-cysteine spectroscopic signatures
but displayed a new hydroxy proton signal at δH 6.36 and 6.53
in 2 and 3, respectively. Both hydroxyl protons displayed
HMBC correlations with the corresponding C-4, C-4a, and C-
Table 2. NMR Data for Frenolicins D−F (2−4)
frenolicin D (2)
frenolicin E (3)
frenolicin F (4)
a
b
a
b
c
b
position
δC, type
δH (J in Hz)
δC, type
δH (J in Hz)
δC, type
δH (J in Hz)
1
79.4, CH
65.4, CH
34.0, CH2
3.80, dd (2.7, 12.0)
4.00, m
78.8, CH
62.6, CH
32.8, CH2
3.99, d (9.8)
4.32, m
68.2, CH
61.4, CH
27.7, CH2
4.26, dd (6.3, 8.3)
3.95, m
3
4α
1.73, t (11.5)
2.00, m
1.76, dd (11.2, 14.3)
2.60, dd (4.7, 14.5)
4β
2.27, dd (3.1, 12.5)
1.76, t (12.9)
4a
75.7, C
191.8, C
134.1, C
119.8, CH
138.0, CH
124.4, CH
161.6, C
116.1, C
202.4, C
79.7, C
75.0, C
194.7, C
132.7, C
119.6, CH
137.1, CH
124.7, CH
162.5, C
115.8, C
199.1, C
78.8, C
63.7, C
67.1, CH
141.8, C
121.4, CH
137.2, CH
117.3, CH
160.7, C
112.8, C
198.0, C
59.3, C
5
4.78,d (8.6)
5a
6
7.59, d (7.5)
7.45, d (7.5)
6.95, d (7.7)
7
7.82, t (7.6, 8.2)
7.38, d (8.4)
7.75, t (7.6, 8.3)
7.31, d (8.4)
7.58, t (7.6, 8.2)
6.93, d (8.7)
8
9
9a
10
10a
11
30.9, CH2
18.8, CH2
13.1, CH3
39.6, CH2
175.7, C
0.69, 1.22, m
1.05, 1.28, m
0.64, t (7.4)
2.42, 2.58, m
28.1, CH2
20.1, CH2
13.8, CH3
40.3, CH2
175.7, C
1.66, 2.35, m
1.26, 1.58, m
0.90, t (7.4)
2.42, 2.58, m
28.9, CH2
18.1, CH2
13.8, CH3
40.9, CH2
172.0, C
1.85, m
12
1.38, 1.57, m
0.92, t (7.4)
2.29, 2.46, m
13
14
15
4a-OH
5-OH
9-OH
10a-OH
6.36, s
6.53, s
6.09, d (8.6)
11.28, s
11.96, s
6.09, s
11.58, s
6.76, s
a
b
c
Measured in CDCl3, 100 MHz. Measured in DMSO-d6, 500 MHz. Measured in DMSO-d6, 100 MHz.
1443
dx.doi.org/10.1021/np400231r | J. Nat. Prod. 2013, 76, 1441−1447

Journal of Natural Products
Article
Figure 3. Key ROESY correlations of frenolicins D (2) and E (3).
5, with a C-4a hydroxy. ROESY experiments to establish C-4a
relative stereochemistry revealed the presence of cross-peaks for
H-1/H-4α, H-1/H2-14, H-1/OH-10a, H-4α/OH-4a, and OH-
4a/OH-10a indicating all as cofacial and the C-4a and C-10a
hydroxys of compound 2 to adopt an α-orientation. In a similar
manner, ROESY cross-peaks for H-1/OH-10a, H-4α/H-10a,
and H-4α/H2-14 in compound 3 indicated a similar cofacial α-
orientation, while the ROESY correlations of H-3/H-4β, H-3/
OH-4a, H-4β/H-4a, and OH-4a/H2-11 supported the β-
orientation of OH-4a (Figure 3). On the basis of this
cumulative analysis, the structures of 2 and 3 were established
as depicted in Figures 1 and 3 and, as new members of the
frenolicin family, named frenolicins D and E, respectively.
Compound 4 was obtained as a yellow, amorphous powder.
The presence of a molecular ion at m/z = 371.1107 [M + Na]+
in the HRESIMS spectrum suggested a molecular formula of
C18H20O7, which, when compared to frenolicin (6), contained
two additional protons. The 13C NMR data of 4 was similar to
those of frenolicin (6), with both displaying similar C-4a−C-
10a epoxy signatures (Table 2). In contrast to 6, the 13C
spectrum suggested compound 4 to lack a C-5 carbonyl, and
the detection of two additional coupled proton signals at δH
6.09 (d, J = 8.6) and 4.78 (d, J = 8.6) also suggested the
presence of a secondary alcohol. Of these latter resonances, the
proton at δH 4.78 (d, J = 8.6) displayed HMBC correlations
with C-4a, C-5, C-4, C-5a, C-6, C-9a, and C-10a, consistent
with C-5 regiospecificity. The subsequent ROESY correlations
of OH-5/H-4β and H-4β/H-3 in compound 4 implicated a β-
orientation of the C-5 OH group. To determine the relative
configuration of the oxirane ring, compound 4 and the standard
6 were both treated with Dess-Martin periodinane at room
both eluted at the identical retention time and possessed
identical UV characteristics (Figure S46).
Compound 5 was obtained as a yellow, amorphous powder,
and its molecular formula was established as C36H38O14S by
HRESIMS at m/z 727.2107 [M + H]+ and m/z 725.1947 [M −
H]−. Only 19 hydrogen and 18 carbon atoms were observed in
the 1H NMR and 13C NMR, respectively. The tentative
molecular weight of the compound deduced from NMR
analysis appeared to be half of that observed in the HRESIMS
data, suggesting compound 5 to be a symmetrical dimer. The
NMR data of 5 were similar to those of frenolicin C (1). The
close chemical shift of C-4a of 5 to that of 1 suggested that
compound 5 contains a sulfur atom at the C-4a position.
Analysis of 2D NMR data including COSY and HMBC (Figure
4) led to the elucidation of the structure of 5. Compound 5
Figure 4. COSY, select HMBC, and ROESY correlations of frenolicin
G (5).
contained the 1 frenolicin core wherein a thioether bridge links
two symmetrical units at the 4a position. The relative
configuration of 5 was established through the analyses of
ROESY data. While the NOE correlations observed between
H-1 and OH-10a, H-1 and H-4α, and H-3 and H-4β helped
establish the configuration of most chiral centers, we were
unable to assign the relative configuration at the 4a position via
NMR experiments. A number of experiments to produce single
crystals for the X-ray diffraction analysis for compound 5 were
not successful. Hence, the structure of 5 is reported as depicted
in Figure 1 and, as a new family member, subsequently named
frenolicin G.
temperature following previously reported methods.14 The H
1
NMR data of the oxidized products of 4 and 6 were
indistinguishable, supporting an identical α-orientation of
oxirane in both compounds. On the basis of this cumulative
analysis, the structure of 4 was established as depicted in Figure
1 and, as a new member of the frenolicin family, named
frenolicin F.
Three previously reported compoundsfrenolicin (6),15−17
frenolicin B (7),17 and UCF76-A (8)18were also identified as
fermentation products of Streptomyces sp. RM-4-15 via
comparison to previously reported spectroscopic data. In light
of a recent report that indicated the positive effect of scandium
on the expression level of pathway-specific positive regulatory
genes,11 we tested the effect of this rare earth element on the
production of secondary metabolites of Streptomyces sp. RM-4-
15. Addition of 18 mg/L scandium chloride in a parallel
fermentation of the same strain revealed a new compound (5)
along with two additional known compounds, deoxyfrenolicin
(9)15−17 and UCF 13 (10).19 Interestingly, compounds 8 and 9
Frenolicins are typically characterized by the presence of a
pyranonaphthoquinone moiety with propyl substitution at the
C-1 position, and they are known for their antibacterial,
antifungal, and antiprotozoal activities.15,17,20−22 Frenolicin B
(7) was recently identified as a putative AKT-selective kinase
inhibitor, while UCF76-A (8) was reported to inhibit Ras p21
farnesylation.18,23−26 Inspired by the putative anticancer
mechanisms reflected within these recent studies, we compared
the new analogues (1−5) and the known frenolicins (6−10) in
a standard cancer cell line cytotoxicity assay using the human
non-small-cell lung carcinoma A549. While compounds 6−9
displayed moderate cytotoxicity in this assay (with IC50’s
1444
dx.doi.org/10.1021/np400231r | J. Nat. Prod. 2013, 76, 1441−1447

Journal of Natural Products
Article
performed using an Agilent 1260 system equipped with a DAD
detector and a Phenomenex C18 column (4.6 × 150 mm, 0.5 μm).
Semipreparative HPLC separation was performed using a Varian
Prostar 210 HPLC system equipped with a PDA detector 330 using a
DiscoveryBio wide pore C18 column (25 × 21.2 mm, 10 μm; flow rate,
8 mL/min). All solvents used were of ACS grade and purchased from
Pharmco-AAPER (Brookfield, CT, USA). Sephadex LH-20 (25−100
μm) was purchased from GE Healthcare (Little Chalfont, UK). C18-
functionalized silica gel (40−63 μm) was purchased from Material
Harvest Ltd. (Cambridge, UK). Silica gel (230−400 mesh) for column
chromatography was purchased from Silicycle (Quebec City, Canada).
TLC silica gel plates (60 F254) were purchased from EMD Chemicals
Inc. (Darmstadt, Germany).
ranging from 0.28 to 5.77 μM), the new isolates 1−5 were
inactive (IC50’s > 80 μM) in the parallel assays (Table 3).
Table 3. Cytotoxicity of Compounds 1−10 against the A549
Cell Line
compound
IC50 (μM)
95% confidence interval
1
2
3
4
5
6
7
8
9
10
>80
>80
>80
>80
>80
5.77
0.28
0.35
1.07
>80
Isolation of Streptomyces sp. RM-4-15 and Preliminary
Screening of the Metabolites. The soil sample was collected
from the Ruth Mullins underground mine fire, Perry County, KY
(coordinates: N 37°18.725′ and W 83°10.3335′). Streptomyces sp. RM-
4-15 was isolated from soil following minor modifications of prior
precedent.34 Specifically, a soil sample (0.5 g) was suspended in 1.0
mL of sterile H2O, and the suspension was heated to 75 °C for 10 min
to eliminate nonsporulating bacteria. Following serial dilution (10−1,
10−2, 10−3) of the suspension with sterile water, a 100 μL aliquot was
spread on oatmeal agar and on ISP4 agar plates supplemented with
nalidixic acid (75 μg/mL) and cycloheximide (50 μg/mL). A number
of sporulating bacterial colonies were observed after a week of
incubation at 28 °C, and each colony was subsequently streaked on an
M2 agar plate (glucose, 4.0 g/L; malt extract, 10.0 g/L; yeast extract,
4.0 g/L; and agar, 15.0 g/L; pH of the medium was adjusted to 7.2
with 2 M NaOH prior to sterilization). Individual bacterial colonies
were isolated from the second-generation plate and fermented in 50
mL of soytone−glucose medium (glucose, 20 g/L; soytone, 10 g/L;
CaCO3, 2 g/L; and CoCl2, 1 mg/L; pH 7.2) using 250 mL baffled
Erlenmeyer flasks. Cultures were allowed to grow for six days at 28 °C
with agitation (250 rpm). Amberlite XAD-16 (0.5 g, Sigma, St. Louis,
MO, USA) resin was added to the culture 24 h prior to harvesting.
Each individual culture was transferred to a 50 mL falcon tube, and the
mixture centrifuged at 300g for 15 min. The supernatant was
discarded, and the mycelia-XAD resin portion was washed twice
with 80 mL of distilled water prior to the addition of 15 mL of
methanol to generate the crude extract. Methanol extract was
separated from the resin using Whatman filter paper (150 mm
diameter, 11 μm pore size). Each extract was then subjected to HPLC/
HRESIMS analysis. Parental mass values were determined based on
consideration of all relevant adducts (e.g., [M + Na]+, [M + NH4]+,
and/or [M − H2O]+), and the corrected molecular masses used as a
batch query against the Natural Products Identifier AntiBase 201235 to
ascertain the potential for chemical novelty. On the basis of this
analysis, Streptomyces sp. RM-4-15 was identified as among the strains
capable of producing novel metabolites.
Identification of Streptomyces sp. RM-4-15. Genomic DNA
was isolated from streaked colonies and purified using Qiaquick PCR
purification kit (Qiagen). Universal primers (27F, 5′-AGAGTTT-
GATCMTGGCTCAG-3′; 1492R, 5′-GGTTACCTTGTTAC-
GACTT-3′) and HF Taq (New England Biolabs) were used for the
amplification of the 16S rRNA gene following prior precedent.36 The
amplified product was purified using Qiaquick gel extraction kit
(Qiagen) and sequenced. BLAST search of the amplified 1244 bp
fragment revealed 99% identity to the 16S rRNA gene of Streptomyces
coeruleorubidus strain ISP 5145. The sequence has been deposited in
GenBank under the accession number KC297232.
Fermentation. Streptomyces sp. RM-4-15 was cultivated in three
250 mL Erlenmeyer flasks, each containing 50 mL of medium A
(soluble starch, 20.0 g/L; glucose, 10.0 g/L; peptone, 5.0 g/L; yeast
extract, 5.0 g/L; NaCl, 4.0 g/L; K2HPO4, 0.5 g/L; MgSO4·7H2O, 0.5
g/L; CaCO3, 2.0 g/L). After three days of incubation at 28 °C with
200 rpm agitation, the cultures were used to inoculate 40 flasks (250
mL), each containing 100 mL of medium A. The fermentation was
continued for seven days at 28 °C with 200 rpm agitation. An identical
procedure was used to culture Streptomyces sp. RM-4-15 in 10 L of
medium A containing 18 mg/L ScCl3.
5.144−6.479
0.248−0.330
0.318−0.394
0.867−1.327
Although the molecular target that contributes to the cancer
cell cytotoxicity of the frenolicin family remains to be
determined, these results suggest an intact quinone as
indispensable for the cytotoxic activity. The similar potencies
of frenolicin B and UCF-76A suggest the exocyclic lactone to
have minimal influence on bioactivity, while the free C-14
carboxylate and/or the C-4 hydroxyl contribute to the differing
activities of deoxyfrenolicin and UCF76-A. This general trend is
consistent with the previous observation that frenolicin B was
both a more potent AKT inhibitor in vitro (∼2-fold) and a
more potent antiproliferative agent against the PTEN-deleted
breast tumor cell line MDA468 (∼5-fold) than deoxyfrenoli-
cin.24 Mechanistically, pyranonaphthoquinones are postulated
to inhibit AKT via irreversible alkylation of the allosteric T-loop
site.23 This mechanism is unique to classical AKT kinase
inhibitors, which act via competitive inhibition of the ATP-
binding site in a relatively nonselective manner.
The reported identification and characterization of five new
frenolicins support the contention that unique man-made
ecological environments may serve as reservoirs for new/novel
microbial natural products. Among a total of 138 reported
natural products containing a pyranonaphthoquinone core
known to date, only 21kalafungin,27 the nanomycins,28 the
lactoquinomycins,29 and arizonin30are considered to be
closely related structurally to the frenolicins. Therefore, the
addition of five new structures to this subset is a notable
expansion. In addition, amino acid conjugated naphthopyrones,
exemplified by frenolicin C, have not been previously identified,
and the thioether linkage sulfur source within the homodimeric
frenolicin G is intriguing from a biosynthetic perspective. This
study also highlights the first example, to the best of our
knowledge, wherein scandium chloride supplementation led to
the confirmed production of new bacterial secondary
metabolites. Given the intense focus upon the development
of technologies to induce metabolite production from cryptic
biosynthetic pathways within the past decade,31−33 this
observation might have a broader potential in the discovery
of previously inaccessible bacterial metabolites.
EXPERIMENTAL SECTION
■
General Experimental Procedures. UV spectra were recorded
on a GE Ultraspec 8000 spectrophotometer. IR spectra were measured
on a Nicolet FTIR 6700 spectrophotometer. All NMR data were
1
recorded at 500 MHz for H and at 100 MHz for 13C with Varian
Inova NMR spectrometers. HRESIMS experiments were carried out
using an AB SCIEX TripleTOF 5600 System. HPLC analyses were
1445
dx.doi.org/10.1021/np400231r | J. Nat. Prod. 2013, 76, 1441−1447

Journal of Natural Products
Article
Extraction, Isolation, and Purification. All 40 culture flasks were
combined, and 200 g of Celite was added into the culture to facilitate
filtration. The mycelial cake−Celite portion was extracted with acetone
(2 × 500 mL) and subjected to filtration. Evaporation of the effluent
afforded 0.5 g of a reddish-brown crude extract. The culture broth was
subjected to an XAD-16 resin column (800 g), washed with water until
the effluent became colorless, and then eluted with 5 L of methanol.
The MeOH extract was concentrated under reduced pressure to
obtain a crude extract (4.0 g). HPLC and TLC analysis indicated an
identical set of metabolites in both mycelial cake and culture extracts.
Therefore, the extracts were combined (4.5 g) for further processing.
Components of the crude extract (4.5 g) were separated by silica gel
column chromatography using a gradient of CHCl3−MeOH (100:0−
0:100) to yield nine fractions, I−IX. Fraction II (95 mg) was subjected
to a Sephadex LH-20 column, and methanol was used to elute
compounds at a flow rate of 2 mL/min. The major fraction obtained
was dried and then further chromatographed using a C18 column (10
× 1 cm) eluted with 60% MeOH−water to afford frenolicin B (7, 11
mg). Fraction III (80 mg) was first subjected to a Sephadex LH-20
(MeOH; flow rate, 2 mL/min), followed by preparative TLC
(CHCl3−10% MeOH) to yield UCF76-A (8, 2 mg). Similarly,
fraction IV (105 mg) was loaded onto a C18 column (10 × 1 cm) and
was separated with 50% aqueous MeOH to afford frenolicin (6, 9 mg).
Fraction VI (53 mg) was first subjected to a Sephadex LH-20 column
(MeOH), and the major fraction was further purified by using a
semipreparative HPLC to yield compound 4 (5 mg) following an
identical procedure described earlier in the General Experimental
Procedures section. Similarly, a gradient of acetonitrile and water was
used in HPLC to isolate 17, 10, and 12 mg of compounds 1−3,
respectively.
A separate fermentation batch (10 L), accomplished in medium A
containing scandium chloride (18 mg/L ScCl3), was centrifuged and
filtered over Celite. The supernatant was extracted with EtOAc (4 ×
800 mL), and the EtOAc extract subsequently evaporated in vacuo at
38 °C to afford 3.4 g of a yellow crude extract. The biomass
(mycelium) was extracted with EtOAc (3 × 600 mL), and then the
organic fraction was evaporated to yield 1.7 g of crude extract. Both
extracts were combined and separated by an HP-20 resin column (100
g) with a gradient of aqueous CH3CN (20%, 40%, 60%, 80%, and
100%) to yield five fractions. Five compounds were separated from
60% aqueous CH3CN fraction III (250 mg) by preparative HPLC,
frenolicin (6, 50 mg), frenolicin B (7, 42 mg), frenolicin G (5, 12 mg),
deoxyfrenolicin (9, 15 mg), and UCF 13 (10, 8 mg).
acetone (200 μL) was added, followed by 1 M NaHCO3 (50 μL). The
reaction was heated to 40 °C for 1 h, cooled to room temperature, and
acidified with 2 N HCl (25 μL). The reaction mixture was diluted with
MeOH (0.5 mL) and analyzed by HPLC using the following gradient:
0−40 min, linear gradient from 5% to 50% CH3CN in 50 mM TBDHS
buffer (a mixture of 40 mL of 0.5 M KH2PO4, 10 mL of 0.5 M
K2HPO4, 1.7 g of tetrabutylammonium bisulfate, and 20 mL of
acetonitrile, pH was adjusted to 6.0 with 0.5 M K2HPO4, then adjusted
with water to 1 L). Derivatized standards were prepared from
authentic D- and L-cysteine (50 μL of a 50 mM stock) following an
identical procedure. The retention times for Marfey’s derivatives were
as follows: derivative prepared from authentic L-cysteine, 38.6 min;
from authentic D-cysteine, 40.3 min; and from compound 1, 38.6 min
(Supporting Information Figure S9).
Oxidation of Frenolicin F (4). Oxidation of frenolicin F was
accomplished following prior precedent.14 Specifically, 1.0 mg of
compound 4 was dissolved in 0.5 mL of CDCl3, and 1.5 mg of Dess-
Martin periodinane was added in the solution. The mixture was stirred
for 10 min at room temperature and then subjected to NMR analyses.
1H NMR (500 MHz) data of the product was compared with the data
of the standard frenolicin (6). 1H NMR (500Mz, CDCl3): δH 4.62 (H-
1), 4.11 (H-3), 7.59 (H-6), 7.64 (H-7), 7.28 (H-8), 11.53 (OH-9),
0.98 (H3-13), 2.60, 2.92 (H2-14).
Cancer Cell Line Cytotoxicity. A resazurin-based cytotoxicity
assay, also known as the AlamarBlue assay, was used to assess the
cytotoxicity of agents against the human lung non-small-cell carcinoma
cell line A549 where the degree of cytotoxicity was based upon
residual metabolic activity as assessed via reduction of resazurin (7-
hydroxy-10-oxido-phenoxazin-10-ium-3-one) to its fluorescent prod-
uct resorufin. A549 (ATCC, Manassas, VA, USA) was grown in
DMEM/F-12 Kaighn’s modification and MEM/EBSS media,
respectively (Thermoscientific, Rockford, IL, USA), with 10% heat-
inactivated fetal bovine serum, 100 μg/mL penicillin, 100 mg/mL
streptomycin, and 2 mM L-glutamine. Cells were seeded at a density of
2 × 103 cells per well onto 96-well culture plates with a clear bottom
(Corning, NY, USA), incubated 24 h at 37 °C in a humidified
atmosphere containing 5% CO2, and exposed to standard toxin
(positive controls: 1.5 mM hydrogen peroxide, 10 mg/mL actino-
mycin D) and test compounds for two days. To assess residual
metabolic activity, 150 mM resazurin (Sigma, St. Louis, MO, USA)
was added to each well and the plate was shaken gently for 10 s and
incubated for another 3 h (A549 cells) in a 37 °C incubator to allow
viable cells to convert resazurin into resorufin. The fluorescence
intensity for resorufin was detected on a FLUOstar Omega scanning
microplate spectrofluorometer (BMG Labtech, Cary, NC, USA) using
an excitation wavelength of 560 nm and an emission wavelength of
590 nm. The assay was repeated in three independent experiment
replications. In each replication, the emission of fluorescence of
resorufin values in treated cells were normalized to, and expressed as a
percent of, the mean resorufin emission values of positive control
(untreated, metabolically active cells; 100%, all cells are viable).
Frenolicin C (1): yellow, amorphous powder; UV (MeOH) λmax (log
ε) 354 nm (3.60); IR (KBr) νmax 2930, 1717, 1690, 1653, 1603, 1453,
1257, 1220, 1002 cm−1; 13C and 1H NMR data, see Table 1;
HRESIMS m/z 510.1380 (calcd for C23H28O10NS, 510.1434), m/z
508.1340 (calcd for C23H26O10NS, m/z 508.1277).
Frenolicin D (2): yellow, amorphous powder; UV (MeOH) λmax
(log ε) 351 nm (4.80); IR (KBr) νmax 2935, 1702, 1644, 1602, 1454,
1164, 1265, 1220, 1123, 1047 cm−1; 13C and 1H NMR data, see Table
2; HRESIMS m/z 387.1016 (calcd for C18H20O8Na, m/z 387.1056).
Frenolicin E (3): yellow, amorphous powder; UV (MeOH) λmax (log
ε) 348 nm (5.21); IR (KBr) νmax 2959, 1705, 1651, 1604, 1454, 1261,
1223, 1127, 1020 cm−1; 13C and 1H NMR data, see Table 2;
HRESIMS m/z 365.1212 (calcd for C18H21O8, m/z 365.1236).
Frenolicin F (4): yellow, amorphous powder; UV (MeOH) λmax (log
ε) 341 nm (3.34); IR (KBr) νmax 3400, 2962, 1709, 1615, 1455, 1203,
1113, 1027 cm−1; 13C and 1H NMR data, see Table 2; HRESIMS m/z
371.1078 (calcd for C18H20O7Na, m/z 371.1107).
ASSOCIATED CONTENT
■
S
* Supporting Information
1H NMR, 13C NMR, HSQC, HMBC, ROESY, and HRESIMS
spectra of frenolicins C−G (1−5) and H and 13C NMR
1
spectra of compounds 6−10. This material is available free of
charge via the Internet at http://pubs.acs.org.
Frenolicin G (5): yellow, amorphous powder; UV (MeOH) λmax
(log ε) 359 nm (9.49); 13C and 1H NMR data, see Table 1; HRESIMS
m/z 727.2107 (calcd for C36H39O14S, m/z 727.2061), m/z 725.1947
(calcd for C36H37O14S, m/z 725.1904).
AUTHOR INFORMATION
■
Corresponding Author
Determination of Amino Acid Configuration. The absolute
configuration of the cysteine residue was determined following
Marfey’s method.13 Specifically, compound 1 (1.0 mg) was hydrolyzed
in 6 N HCl (1 mL) at 110 °C for 14 h. After drying under nitrogen,
the residue was dissolved in 2 mL of EtOAc−H2O (1:1). The aqueous
layer was dried in vacuo, to which a solution of 1% Marfey’s reagent in
*E-mail: jsthorson@uky.edu; madan.kharel@uky.edu.
Notes
The authors declare the following competing financial
interest(s): J.S.T. is a co-founder of Centrose (Madison, WI,
USA).
1446
dx.doi.org/10.1021/np400231r | J. Nat. Prod. 2013, 76, 1441−1447

Journal of Natural Products
Article
(26) Kraus, G. A.; Li, J.; Gordon, M. S.; Jensen, J. H. J. Am. Chem.
Soc. 1993, 115, 5859−5860.
ACKNOWLEDGMENTS
■
This work was supported, in part, by the University of
Kentucky College of Pharmacy, the University of Kentucky
Markey Cancer Center, and the National Center for Advancing
Translational Sciences (UL1TR000117).
(27) Johnson, L. E.; Dietz, A. Appl. Microbiol. 1968, 16, 1815−1821.
(28) Tanaka, H.; Nagai, T.; Marumo, H.; Omura, S. J. Antibiot. 1975,
XXVIII, 868−875.
(29) Leo, P.-M.; Morin, C.; Philouze, C. Org. Lett. 2002, 4, 2711−
2714.
(30) Hochloqski, J. E.; Brill, G. M.; Andres, W. W.; Spanton, S. G.;
McAlpine, J. B. J. Antibiot. 1987, XL, 401−407.
(31) Chiang, Y.-M.; Chang, S.-L.; Oakley, B. R.; Wang, C. C. C. Curr.
Opin. Chem. Biol. 2011, 15, 137−143.
REFERENCES
■
(1) Petranka, J. W. Salamanders of the United States and Canada;
Smithonian Institute Press: Washington, DC, 1998, p 587.
(2) Bernardo, J.; Spotila, J. R. Biol. Lett 2006, 2, 135−139.
(3) Neves, R. J.; Bogan, A. E.; Williams, J. D.; Ahlstedt, S. A.;
Hartfield, W. Aquatic Fauna in Peril: The Southeastern Perspective,
Special Publication No. 1; Southeastern Research Institute; Lenz
Design and Communications: Decatur, GA, 1997; pp 43−86.
(4) Stein, B. A.; Kutner, L. S.; Adams, J. S. Precious Heritage: The
Status of Biodiversity in the United States; Oxford University Press:
Oxford, UK, 2000.
(32) Ochi, K.; Hosaka, T. Appl. Microbiol. Biotechnol. 2013, 97, 87−
98.
(33) Corre, C.; Challis, G. L. Nat. Prod. Rep. 2009, 26, 977−986.
(34) Abdelfattah, M. S.; Kharel, M. K.; Hitron, J. A.; Baig, I.; Rohr, J.
J. Nat. Prod. 2008, 71, 1569−1573.
(35) Laatsch, H. AntiBase 2012, The Natural Compound Identifier;
Wiley-VCH: Hoboken, NJ, 2012.
(36) Lane, D. J. In Nucleic Acid Techniques in Bacterial Systematics;
Stackebrandt, E.; Goodfellow, M., Eds.; John Wiley and Sons Ltd.:
New York, 1991; pp 115−147.
(5) History of Coal in Kentucky. In Kentucky Coal Facts, 12th ed.;
Division of Fossil Energy Development: Frankfort, KY, 2012; pp 1−
101.
(6) O’Keefe, J. M. K.; Henke, K. R.; Hower, J. C.; Engle, M. A.;
Stracher, G. B.; Stucker, J. D.; Drew, J. W.; Staggs, W. D.; Murray, T.
M.; Hammond, M. L.; Adkins, K. D.; Mullins, B. J.; Lemley, E. W. Sci.
Total Environ. 2010, 408, 1628−1633.
(7) Zhao, Y.; Zhang, J.; Chou, C.; Li, Y.; Wang, Z.; Ge, Y. Int. J. Coal.
Geol. 2008, 73, 52−62.
(8) Carras, J. N.; Day, S. J.; Saghafi, A.; Williams, D. J. Int. J. Coal.
Geol. 2009, 78, 161−168.
(9) Hower, J. C.; Henke, K.; O’Keefe, J. M. K.; Engle, M. A.; Blake,
D. R.; Stracher, G. B. Int. J. Coal Geol. 2009, 80, 63−67.
(10) Silva, L. F. O.; Oliveira, M. L. S.; Philippi, V.; Serra, C.; Dai, S.
F.; Xue, W. F.; Chen, W. M.; O’Keefe, J. M. K.; Romanek, C. S.;
Hopps, S. G.; Hower, J. C. Int. J. Coal Geol. 2012, 94, 206−213.
(11) Kawai, K.; Wang, G.; Okamoto, S.; Ochi, K. FEMS Microbiol.
Lett. 2007, 274, 311−315.
(12) Maestrone, G.; Schildknecht, H.; Untawale, G. G. U.S. Patent
4,839, 382, 1989.
(13) Marfey, P. Carlsberg. Res. Commun. 1984, 49, 591−596.
(14) Dess, D. B.; Martin, J. C. J. Org. Chem. 1983, 48, 4155−4156.
(15) Ellestad, G. A.; Kunstmann, M. P.; Whaley, H. A.; Patterson, E.
L. J. Am. Chem. Soc. 1968, 90, 1325−1332.
(16) Ellestad, G. A.; Whaley, H. A.; Patterso, E. J. Am. Chem. Soc.
1966, 88, 4109−4110.
(17) Iwai, Y.; Kora, A.; Takahashi, Y.; Hayashi, T.; Awaya, J.;
Masuma, R.; Oiwa, R.; Omura, S. J. Antibiot. 1978, 31, 959−965.
(18) Hara, M.; Soga, S.; Shono, K.; Eishima, J.; Mizukami, T. J.
Antibiot. 2001, 54, 182−186.
(19) Nakano, H.; Hara, M.; Saito, Y.; Inata, Y.; Kawada, S.; Okabe,
M. Japan Patent 04154778, 1992.
(20) Armer, R. E.; Dutton, C. J.; Fenner, B. R.; Greenwood, S. D.;
Hall, K. T.; Rudge, A. J. Bioorg. Med. Chem. Lett. 1998, 8, 139−142.
(21) Fitzgerald, J. T.; Henrich, P. P.; O’Brien, C.; Krause, M.; Ekland,
E. H.; Mattheis, C.; Sa, J. M.; Fidock, D.; Khosla, C. J. Antibiot. 2011,
64, 799−801.
(22) Fitzgerald, J. T.; Ridley, C. P.; Khosla, C. J. Antibiot. 2011, 64,
759−762.
(23) Toral-Barza, L.; Zhang, W. G.; Huang, X.; McDonald, L. A.;
Salaski, E. J.; Barbieri, L. R.; Ding, W. D.; Krishnamurthy, G.; Hu, Y.
B.; Lucas, J.; Bernan, V. S.; Cai, P.; Levin, J. I.; Mansour, T. S.;
Gibbons, J. J.; Abraham, R. T.; Yu, K. Mol. Cancer Ther. 2007, 6,
3028−3038.
(24) Salaski, E. J.; Krishnamurthy, G.; Ding, W. D.; Yu, K.; Insaf, S.
S.; Eid, C.; Shim, J.; Levin, J. I.; Tabei, K.; Toral-Barza, L.; Zhang, W.
G.; McDonald, L. A.; Honores, E.; Hanna, C.; Yamashita, A.; Johnson,
B.; Li, Z.; Laakso, L.; Powell, D.; Mansour, T. S. J. Med. Chem. 2009,
52, 2181−2184.
(25) Fernandes, R. A.; Chavan, V. P.; Mulay, S. V.; Manchoju, A. J.
Org. Chem. 2012, 77, 10455−10460.
1447
dx.doi.org/10.1021/np400231r | J. Nat. Prod. 2013, 76, 1441−1447