Bafilomycins produced in culture by streptomyces spp. isolated from marine habitats are potent inhibitors of autophagy

Article abstract of DOI:10.1021/np900632r

Five new bafilomycins, F (1) to J (5), have been isolated from laboratory cultures of two Streptomyces spp. obtained from marine sediments collected in British Columbia, and their structures have been elucidated by detailed analysis of spectroscopic data and the synthesis of model compounds. The new bafilomycins F (1), G (2), H (3), and J (5) along with several co-occurring known analogues showed potent inhibition of autophagy in microscopy and biochemical assays. The thiomorpholinone fragment present in bafilomycin F (1) has not previously been found in a natural product.

Full text of DOI:10.1021/np900632r

422
J. Nat. Prod. 2010, 73, 422–427
Bafilomycins Produced in Culture by Streptomyces spp. Isolated from Marine Habitats Are
Potent Inhibitors of Autophagy†
Gavin Carr,‡ David E. Williams,‡ Ana R. D´ıaz-Marrero,‡ Brian O. Patrick,§ Helen Bottriell,‡ Aruna D. Balgi,⊥
Elizabeth Donohue,⊥ Michel Roberge,⊥ and Raymond J. Andersen*,‡
Departments of Chemistry and Earth & Ocean Sciences, UniVersity of British Columbia, VancouVer, B.C., Canada V6T 1Z1, and Department
of Biochemistry and Molecular Biology, UniVersity of British Columbia, VancouVer, B.C., Canada V6T 1Z3
ReceiVed October 11, 2009
Five new bafilomycins, F (1) to J (5), have been isolated from laboratory cultures of two Streptomyces spp. obtained
from marine sediments collected in British Columbia, and their structures have been elucidated by detailed analysis of
spectroscopic data and the synthesis of model compounds. The new bafilomycins F (1), G (2), H (3), and J (5) along
with several co-occurring known analogues showed potent inhibition of autophagy in microscopy and biochemical
assays. The thiomorpholinone fragment present in bafilomycin F (1) has not previously been found in a natural product.
Autophagy is a catabolic process used by cells to degrade proteins
and organelles present in the cytoplasm into low molecular weight
products that are recycled.1 Autophagy is upregulated during periods
of nutrient starvation, and it is involved in mammalian develop-
mental processes, innate and adaptive immunity, degradation of
invading bacteria, and diseases such as neurodegeneration.1b Studies
in yeast have identified over 30 autophagy-related (Atg) genes.1a
There are currently no known direct inhibitors of any of the Atg
proteins. In an effort to discover new chemical genetics tools for
studying autophagy, a library of crude microbial extracts has been
screened in an automated microscopy assay designed to detect
modulators of autophagosome accumulation.2,3 The crude extract
library was prepared by extraction of solid agar laboratory cultures
of microorganisms obtained from both cold temperate and tropical
marine habitats. Two cultures that produced extracts with extremely
potent induction of autophagosome accumulation were identified
in the screen, and 16S RNA analysis identified both of the producing
organisms as Streptomyces sp. Assay-guided fractionation of the
extracts led to the isolation of the new compounds bafilomycin F
(1), bafilomycin G (2), bafilomycin H (3), bafilomycin I (4), and
bafilomycin J (5) along with the known compounds bafilomycin
A1 (6),4 bafilomycin B1 (7),4 and bafilomycin D (8).5 Details of
the isolation, structure elucidation, and autophagy-inhibiting activi-
ties of the new and known bafilomycins are presented below.
Results and Discussion
The Streptomyces sp. strains RJA71 and RJA635 were obtained
from marine sediments collected off the coast of British Columbia,
and production cultures were grown as lawns on solid agar. The
solid agar culture of strain RJA635 was extracted repeatedly with
EtOAc, and the combined EtOAc extracts were dried in Vacuo to
give a brown gum, which was fractionated by sequential application
of Si gel flash chromatography, reversed-phase flash chromatog-
raphy, and reversed-phase HPLC to give the new macrolide
bafilomycin F (1) along with the known compounds bafilomycin
A1 (6), bafilomycin B1 (7), and bafilomycin D (8). The structures
of the known compounds 6, 7, and 8 were confirmed by comparison
of their spectroscopic data with literature values.4,5
The cells and media from solid agar culture of strain RJA71
† Dedicated to the late Dr. John W. Daly of NIDDK, NIH, Bethesda,
Maryland, and to the late Dr. Richard E. Moore of the University of Hawaii
at Manoa for their pioneering work on bioactive natural products.
* To whom correspondence should be addressed. Tel: 604 822 4511.
Fax: 604 822 6091. E-mail: randersn@interchange.ubc.ca.
‡ Chemistry and Earth & Ocean Sciences.
were lyophilized together and then extracted with MeOH, which
was dried in Vacuo to give a brown gum. Partitioning the gum
between EtOAc and H2O gave an active EtOAc-soluble component
that was fractionated by sequential application of Sephadex LH-
20 chromatography and reversed-phase HPLC to give samples of
bafilomycin G (2), bafilomycin H (3), bafilomycin I (4), and
§ Chemistry.
⊥ Biochemistry and Molecular Biology.
bafilomycin J (5). ESIHRMS and 1D H NMR data showed that
1
10.1021/np900632r  2010 American Chemical Society and American Society of Pharmacognosy
Published on Web 12/22/2009

Bafilomycins Produced in Culture by Streptomyces spp.
Journal of Natural Products, 2010, Vol. 73, No. 3 423
Scheme 1
each of the bafilomycins G (2), H (3), and J (5) was contaminated
with small amounts of the corresponding analogues that simply have
a C-21 alcohol functionality in place of the C-21 methyl ether
assigned to the named structures. Significant effort was put into
attempts to separate the alcohol-containing contaminants from the
methyl ethers, but this proved to be impossible in our hands.
Therefore, only the major component methyl ethers 2, 3, and 5,
which are clearly defined in the NMR data (Supporting Informa-
tion), are described here.
Bafilomycin F (1) gave a [M - H]- ion at m/z 822.4096 in the
negative ion HRESIMS, consistent with a molecular formula of
C42H65NO13S, requiring 11 sites of unsaturation. Analysis of the
1D and 2D NMR data obtained for bafilomycin F showed that it
contained the C-1 to C-33 macrolide substructure present in
bafilomycin A1 (6), which was acylated at the C-21 alcohol. The
presence of the C-21 acyl substitution in 1 was identified by the
downfield shift of H-21 to δ 4.95 (δ 3.69 in 64) and the observation
of a HMBC correlation between the H-21 resonance and a carbonyl
resonance at δ 171.5 (C-1′). Subtraction of the atoms present in
the macrolide fragment (C35H57O9) of 1 from the molecular formula
(C42H65NO13S) showed that the C-21 acyl substituent had to account
for C7H8NO4S and four sites of unsaturation.
Two isolated spin systems that could be assigned to the C-21
acyl substituent were identified from the COSY data (CD3OD). The
first system consisted of a pair of geminal methylene proton
resonances at δ 3.20 (H-2a′) and 3.01 (H-2b′) that were coupled to
a methine resonance at δ 3.67 (H-3′), and the second consisted of
a two-proton methylene resonance at δ 3.12 (H2-7′) that was coupled
to a methine resonance at δ 4.25 (H-6′). HMBC correlations
observed between the H-2a′ (δ 3.20) and H-3′ (δ 3.67) resonances
and a resonance at δ 171.5 (C-1′) and between the H2a′, H2b′ (δ
3.01), and H-3′ resonances and a carbon resonance at δ 175.1 (C-
4′) indicated that the first spin system had carbonyl substituents on
each of the methylene and methine carbons. The H-3′ resonance at
δ 3.67 showed a HMBC correlation to a carbon resonance at δ
33.7 (C-7′), which was in turn correlated to the methylene resonance
at δ 3.12 (H2-7′) in the HSQC spectrum, and the H2-7′ methylene
resonance at δ 3.12 showed a HMBC correlation to a carbon
resonance at δ 39.6 (C-3′), which was correlated in the HSQC
spectrum to the H-3′ resonance (δ 3.67), demonstrating that C-3′
and C-7′ were linked via a nonprotonated atom. The chemical shifts
of the carbons indicated a possible linkage via the sulfur atom. A
HSQC correlation showed that the methine resonance at δ 4.25
(H-6′) in the second proton spin system was attached to a carbon
with a chemical shift of δ 60.6 (C-6′), typical of an amino acid R
methine carbon, and the possibility of a sulfur atom being attached
to its vicinal methylene neighbor (C-7′) suggested the presence of
a cysteine residue. The final carbonyl resonance observed at δ 178.1
in the 13C NMR spectrum (CD3OD) was assigned to C-9′, and a
COSY spectrum of 1 recorded in DMSO-d6 showed a correlation
between the H-6′ resonance at δ 3.76 and an exchangeable
resonance at δ 7.01, assigned to NH-5′, accounting for the remaining
pieces of the cysteine fragment. All of the above data were in
agreement with a biogenesis in which the fumaric acid residue found
resonances to each compound in the NMR data for the mixture.
NOESY correlations were observed between H-3 and H-6 in both
cis isomers 12 and 17 but not in the trans isomers 13 and 18. The
1H and 13C NMR shifts assigned to the thiomorpholinone rings in
the model compounds 12, 13, 17, and 18 all showed good agreement
with the shifts assigned to the corresponding atoms in 1 (Supporting
Information), confirming the nature of the C-21 acyl substituent.
The best fit between the chemical shifts for the thiomorpholinone
ring in 1 and the model compounds was with the trans isomers 13
and 18 (Supporting Information), especially at H-3′ (1: δ 3.67; 12:
δ 4.09; 13: δ 3.87; 17: δ 4.00; 18: δ 3.79), and, therefore, we
tentatively assigned the relative configuration of the thiomorpholi-
none ring in 1 as trans. The lack of NOESY correlations between
H-3′ and H-6′ in 1 was in agreement with the trans relative
configuration. The absolute configuration in the thiomorpholinone
ring of 1 was determined by chemical degradation to alanine
(Scheme 2), which was shown by Marfey’s analysis to be L.
Therefore, the absolute configuration of the C-21 acyl substituent
in 1 is assigned as 3′S,6′R. We have assumed that the remainder of
bafilomycin F (1) has the same absolute configuration reported for
bafilomycin A1 (6),8 as shown.
4
in bafilomycins B1 (7) and C1 had been elaborated by amide
formation with a cysteine R-amino nitrogen followed by an
intramolecular Michael addition of the sulfur atom to give a
thiomorpholinone ring as shown in 1.
In order to confirm the presence of the thiomorpholinone ring
and to determine the relative configurations at C-3′ and C-6′, model
compounds 12, 13, 17, and 18 were synthesized as shown in
Scheme 1.6 The diastereomers 12 and 13 were separated by Si gel
chromatography, and compound 12 was crystallized. Analysis of a
crystal of 12 by X-ray diffraction analysis showed that the
substituents on the thiomorpholinone ring were cis (Supporting
Information).7 We were not able to separate the mixture of
stereoisomers 17 and 18 by HPLC, but it was possible to assign
Bafilomycin G (2) gave a [M + Na]+ ion at m/z at 659.4142 in
the positive ion HRESIMS, consistent with a molecular formula
of C36H60O9, differing from the molecular formula of bafilomycin

424 Journal of Natural Products, 2010, Vol. 73, No. 3
Carr et al.
Scheme 2
at δ 80.2, assigned to C-21 in 2 by COSY and HSQC data,
confirmed that the C-21 OH present in bafilomycin A1 (6) had been
replaced by a OMe in bafilomycin G (2).
Bafilomycin H (3) gave a [M + Na]+ ion at m/z at 673.4301 in
the positive ion HRESIMS, consistent with a molecular formula
of C37H62O9, differing from the molecular formula of bafilomycin
G (2) simply by the addition of CH2. The 1D and 2D NMR data
obtained for 3 (Table 1) were nearly identical to the data obtained
for bafilomycin G (2), indicating that the compounds were closely
related. The major difference in the NMR data for 2 and 3 was
that the OH-19 proton at δ 5.84 in 2 was replaced by a methyl
singlet (19-OMe: δC 46.4; δH 3.12), suggesting that bafilomycin H
(3) was the methyl ketal of bafilomycin G (2). A correlation
observed between the methyl resonance at δ 3.12 and the C-19
ketal carbon (δ 103.8) in the HMBC spectrum obtained for 3
confirmed this assignment. The methyl ketal in 3 may be an
isolation artifact formed during MeOH extraction of the culture.9
Bafilomycin I (4) gave a [M + Na]+ ion at m/z at 623.3917 in
the positive ion HRESIMS, consistent with a molecular formula
of C36H56O7, which differed from the molecular formula of
bafilomycin A1 (6) by the loss of 2 × H2O. Comparison of the 1D
and 2D NMR data obtained for 4 with that of 2 (Table 1) showed
that these compounds were closely related. In particular, the NMR
data showed that the C-1 to C-15 macrocyclic rings were identical
in 2 and 4. Similarly, the NMR data confirmed that the substituents
at C-21 (OMe), C-22 (Me), and C-23 (isopropyl) were also the
same in 2 and 4 (Table 1). However, unlike in 2, the C-21 methine
proton resonance at δ 3.55 in 4 showed a COSY correlation to a
A1 (6) simply by the addition of CH2. Comparison of the 1D and
2D NMR data obtained for 2 (Table 1) with literature values and
the data that we re-collected for bafilomycin A1 (6) showed that 2
was closely related to 6. A HMBC correlation observed between a
methyl ether resonance at δ 3.25 (δC 56.3) and a carbon resonance
Table 1. 1H NMR Data for 1-5
1a
2b
3b
4b
5b
position
δH (J in Hz)
δH (J in Hz)
δH (J in Hz)
δH (J in Hz)
δH (J in Hz)
2-(O)Me/Me
3
26
5
3.61, s
6.69, s
1.97, s
5.88, d (8.9)
2.52, m
3.63, s
6.83, s
2.01, s
5.82, d (9.0)
2.29, m
3.48, s
6.83, s
1.97, s
5.83, d (8.9)
2.27, m
3.69, s
7.04, s
2.04, d (0.8)
5.99, d (9.1)
2.20, m
2.12, s
7.39, s
1.63, s
5.87, d (8.8)
2.23, m
6
27
7
7-OH
8
1.05, d (6.9)
3.24, m
0.88, d (7.0)
2.84, t (5.8)
0.83, d (5.8)
1.74, m
0.87, d (7.0)
2.86, m
0.87, m
0.89, d (7.0)
3.00, m
0.86, d (6.0)
1.66, m
0.83, d (7.0)
2.83, t (6.0)
0.83, d (6.0)
1.76, m
1.84, m
1.74, m
28
9a
9b
29
0.92, d (6.9)
2.02, m
2.02, m
0.73, d (6.2)
1.97, m
1.72, m
0.73, d (5.6)
1.95, m
1.74, m
0.81, d (7.0)
2.25, dd (14.4, 10.9)
1.84, d (14.4
1.66, s
0.72, d (6.1)
1.97, m
1.75, m
1.89, s
1.87, s
1.87, s
1.88, s
11
12
13
14
5.76, d (10.8)
6.59, dd (14.9, 10.8)
5.10, dd (14.9, 8.9)
3.98, t (8.4)
3.24, s
5.61, d (10.7)
6.47, dd (15.0, 10.7)
5.08, dd (15.0, 9.3)
3.90, t (9.2)
3.07, s
5.65, d (10.7)
6.50, dd (15.0, 10.7)
5.16, dd (15.0, 9.2)
3.93, t (9.0)
3.08, s
5.95, d (11.1)
6.69, dd (15.1, 11.1)
5.45, dd (15.1, 6.4)
3.94, t (5.4)
3.18, s
5.72, d (10.7)
6.53, dd (14.9, 10.7)
5.22, dd (14.9, 9.3)
3.96, t (8.9)
3.11, s
14-OMe
15
16
5.03, d (7.7)
2.07, m
5.38, d (8.6)
2.46, m
5.58, d (8.5)
2.39, m
5.42, dd (8.2, 4.3)
2.96, m
5.45, d (7.7)
2.43, m
30
17
17-OH
18
0.85, d (4.4)
4.14, dd (10.5, 1.4)
0.86, d (6.8)
4.54, ddd (10.7, 3.9, 1.7)
5.20, d (3.7)
1.91, m
0.99, d (6.9)
3.96, dd (10.2, 4.4)
4.51, d (4.4)
2.29, m
1.16, d (6.7)
6.35, d (10.1)
0.86, d (6.9)
4.54, d (10.6)
5.45, m
1.79, m
1.90, m
31
0.98, d (7.2)
1.21, d (7.1)
5.84, m
1.32, d (6.8)
3.12, s
1.75, s
1.21, d (7.5)
5.97, m
19-OH(OMe)
20a
20b
21
21-OMe
22
2.29, dd (12.0, 4.8)
1.30, m
4.95, dt (10.9, 4.8)
2.69, m
1.21, m
3.54, m
3.25, s
2.92, m
1.82, m
3.39, m
3.20, s
5.12, d (2.6)
2.70, m
1.22, m
3.55, m
3.25, s
3.55, dd (7.8, 2.6)
3.20, s
2.02, m
1.61, m
1.72, m
1.65, m
1.73, m
32
23
24
0.86, d (4.1)
3.56, dd (10.4, 1.5)
1.91, m
1.06, d (6.4)
3.85, d (10.4)
1.92, m
1.00, d (12.6)
3.19, d (9.3)
1.85, m
0.88, d (6.7)
3.37, dd (9.3, 3.4)
1.89, m
1.06, d (6.5)
3.88, d (10.1)
1.95, m
33
25
0.95, d (7.2)
0.80, d (6.9)
3.20, m
3.01, dd (14.8, 7.4)
3.67, dd (7.4, 2.5)
4.25, m
1.11, d (6.8)
0.94, d (6.8)
1.10, d (6.9)
0.94, d (6.7)
1.09, d (6.8)
0.92, d (6.8)
1.22, d (7.1)
0.95, d (6.7)
2′a
2′b
3′
6′
7′
3.12, m
a Spectra collected in MeOD at 600 MHz. b Spectra collected in C6D6 at 600 MHz.

Bafilomycins Produced in Culture by Streptomyces spp.
Journal of Natural Products, 2010, Vol. 73, No. 3 425
Figure 1. Stimulation of autophagosome accumulation by compounds 1-8. MCF-7 cells expressing EGFP-LC3 were exposed to the
compounds for 4 h. The formation of punctate EGFP-LC3 was determined quantitatively by automated microscopy and expressed relative
to control, untreated cells. The bars represent means ( SD (n ) 3).
vicinal olefinic methine resonance at δ 5.12 (H-20) and HMBC
correlations to olefinic carbon resonances at δ 97.6 (C-20) and 154.3
(C-19), indicating the presence of a ∆19,20 alkene. The H-16 methine
autolysosomes. The effect of various concentrations of 1 to 8 on
autophagosomal content is shown in Figure 1. Compound 4 was
the least potent, showing activity at 100 nM and above. Compounds
1, 3, and 8 showed good activity at 10 nM, and compounds 5, 6,
and 7 were active at 1 nM and above. Bafilomycin G (2) was the
most potent, being fully active at 0.3 nM, the lowest concentration
tested in this assay. For a comparison, 4 h exposure to well-
characterized autophagic stimuli such as amino acid and serum
starvation or 20 nM rapamycin caused ∼3-fold increase in
autophagosomal content, while some recently identified autophagy
modulators caused a 10-15-fold increase in this assay.2a
To determine whether the compounds induced autophagy or
inhibited this process at a late stage, they were subjected to a
biochemical test. Autophagy entails the recruitment of the protein
LC3 to autophagosomes and its degradation by lysosomal hydro-
lases upon fusion of autophagosomes with lysosomes. In the cell
line used, LC3 is tagged by fusion at its N-terminus with enhanced
green fluorescent protein (EGFP). The EGFP moiety of the EGFP-
LC3 fusion protein is much more resistant to degradation by
lysosomal hydrolases than the LC3 portion. Compounds that
stimulate autophagy, such as the mTOR inhibitor rapamycin, cause
increased accumulation of the EGFP moiety (Figure 2A). Com-
pounds 1-3 and 5-8 did not increase EGFP levels (Figure 2A),
indicating that they do not stimulate autophagy. Notably, compound
4 clearly and reproducibly increased EGFP levels, suggesting that
it induces autophagy. To determine whether the compounds
inhibited EGFP-LC3 degradation by autophagy, cells were co-
incubated with rapamycin to induce autophagy and with compounds
1-8. Compounds 1-3 and 5-8 inhibited EGFP production,
showing that they inhibit autophagic protein degradation. Compound
4 did not prevent EGFP formation by rapamycin, indicating that it
does not inhibit autophagy. Bafilomycins A1, B1, and D are potent
and selective inhibitors of the vacuolar type H+-ATPase responsible
for maintaining the acidic pH of lysosomes.10 The bafilomycins
thus block lysosomal acidification and the activity of the acid
hydrolases responsible for autophagic degradation. The observation
that 1-3 and 5-8 block autophagosomal degradation is consistent
with their known mechanism of action. However, the observation
1
resonance at δ 2.96 in the H NMR spectrum of 4 also showed a
COSY correlation to a vicinal olefinic methine resonance at δ 6.35
(H-17) and HMBC correlations to olefinic carbon resonances at δ
129.3 (C-18) and 129.8 (C-18), suggesting the presence of a ∆17,18
alkene. An olefinic methyl resonance at δ 1.75, assigned to Me-
31, showed HMBC correlations to the carbon resonances at δ 129.3
(C-17), 129.8 (C-18), and 154.3 (C-19), confirming that two alkenes
were conjugated. The observation of a ROESY correlation between
the Me-31 resonance at δ 1.75 and the H-16 methine resonance at
δ 2.96 showed that the ∆17,18 alkene had the E configuration. We
have assumed that the absolute configurations at C-21, C-22, and
C-23 and at the stereogenic centers in the C-1 to C-15 macrocyclic
ring in bafilomycin I (4) are identical to the corresponding absolute
configurations in bafilomycin A1 (6).8
Bafilomycin J (5) gave a [M + Na]+ ion at m/z at 643.4168 in
the positive ion HRESIMS, consistent with a molecular formula
of C36H60O8, which differed from the molecular formula of
bafilomycin G (2) simply by loss of one oxygen atom. Detailed
analysis of the 1D and 2D NMR data obtained for 5 showed that
it was nearly identical to 2 (Table 1). The only difference between
2 and 5 was that the methyl ether at C-2 in bafilomycin G (2) (OMe-
2: δC 59.7; δH 3.63) was replaced by a methyl group (δC 14.2; δH
2.12) in bafilomycin J (5), accounting for the one oxygen atom
difference in their molecular formulas. Consistent with this assign-
ment was the observation of a HMBC correlation between the
olefinic methyl resonance at δ 2.12 and the macrolide carbonyl
resonance at δ 172.5 (C-1). A methyl substituent at C-2 has
previously been encountered in the related macrolide oxohygroli-
din.5
The ability of bafilomycins 1 to 8 to induce autophagosome
accumulation was measured using an automated microscopy
screening assay.2 In the presence of 1 to 8 the fluorescence intensity
associated with autophagosomes increased, indicating that the
compounds induced autophagosome accumulation either by stimu-
lating the formation of autophagosomes or by inhibiting their
degradation by preventing the maturation of autophagosomes into

426 Journal of Natural Products, 2010, Vol. 73, No. 3
Carr et al.
The crude extract was subjected to Si gel flash chromatography (step
gradient: hexanes to EtOAc to MeOH). The fractions eluting with 3:1
EtOAc-hexanes were further purified by normal-phase HPLC (1:1
EtOAc-hexanes) and reversed-phase C18 flash chromatography (7:3
CH3CN-H2O) to give bafilomycin A1 (6, 1.2 mg), bafilomycin B1 (7,
3.8 mg), and bafilomycin D (8, 1.0 mg). The fractions eluting from
the Si gel flash chromatography fractionation with 100% MeOH were
subjected to reversed-phase C18 flash chromatography (step gradient:
H2O to MeOH). The fractions eluting with 3:1 MeOH-H2O were
purified by reversed-phase HPLC (3:1 MeOH-H2O) to give bafilo-
mycin F (1, 2.0 mg).
The solid agar culture of strain RJA71 was extracted twice with
MeOH. Concentration of the 2 L of combined MeOH extracts in Vacuo
gave a gummy brown residue, which was partitioned between EtOAc
(3 × 75 mL) and H2O (300 mL). The combined EtOAc extract was
evaporated to dryness and fractionated on Sephadex LH-20 using 4:1
MeOH-CH2Cl2 as eluent to generate an active fraction. Reversed-phase
HPLC, using a CSC-Inertsil 150A/ODS2, 5 µm 25 × 0.94 cm column
with 87:13 MeCN-H2O as eluent, gave pure samples of bafilomycin
G (2) (0.8 mg) and bafilomycin H (3) (4.4 mg) and a mixture of
bafilomycin I (4) and bafilomycin J (5). Pure samples of bafilomycin
I (4) (0.2 mg) and bafilomycin J (5) (0.4 mg) were obtained after an
additional reversed-phase HPLC purification using 17:3 MeOH-H2O
as eluent.
Bafilomycin F (1): clear glass; [R]20 +10.0 (c 0.3, MeOH); UV
D
Figure 2. Effect of compounds 1-8 on EGFP-LC3 degradation.
MCF-7 cells expressing EGFP-LC3 were exposed for 4 h to the
different compounds and combinations. The formation of the EGFP
band was monitored by Western blotting using antibodies to GFP.
Protein loading was monitored with an antibody to ꢀ-tubulin.
(MeOH) λmax (log ꢀ) 244 (4.4), 281 (4.0) nm; 1H NMR and 13C NMR,
see Tables
C42H64NO13S, 822.4098).
1 and 2; HRESIMS(-) m/z 822.4096 (calcd for
Table 2. 13C NMR (δC) Data for 1-5
position
1a
2b
3b
4b
5b
that the closely related bafilomycin I (4) appears to stimulate
autophagy instead of inhibiting it was very unexpected.
Bafilomycins F (1) to J (5) contain a number of structural variants
not previously encountered in this family of macrolides. The most
striking new variation is the thiomorpholinone fragment present in
bafilomycin F (1), which is an unknown substructure in any natural
product and does not appear to hinder activity. Also of note is the
conjugated ∆17,18, ∆19,20 diene in bafilomycin I (4), which results
in significant attenuation of the autophagy-inhibiting activity of the
bafilomycin scaffold.
1
2
168.1
142.3
60.6
134.9
133.3
14.1
145.6
38.5
17.9
81.2
41.7
22.5
42.5
145.0
20.1
125.6
134.7
127.0
84.4
55.9
77.7
39.5
10.5
72.0
43.6
7.2
167.6
142.1
59.7
133.1
133.1
14.3
142.9
37.1
17.3
80.7
40.5
21.8
41.5
142.9
20.3
125.5
133.4
127.8
82.5
55.2
77.3
37.5
10.1
71.2
42.9
7.6
167.0
142.3
59.4
132.4
133.0
14.2
142.5
37.1
17.3
80.7
40.4
22.0
41.6
142.7
20.2
125.6
133.2
127.9
83.0
55.3
77.6
39.2
11.0
70.2
39.3
8.0
103.8
46.4
35.2
79.8
38.9
12.6
77.9
28.8
21.0
14.4
56.0
164.1
142.7
59.7
131.8
132.1
14.1
140.8
38.4
18.6
80.8
38.1
23.9
41.8
140.7
18.0
125.8
131.1
126.1
83.5
55.8
78.6
35.8
17.4
172.5
122.8
14.2
146.5
134.4
15.0
144.9
37.0
17.4
80.7
40.4
21.6
41.5
142.7
20.3
125.4
133.0
128.0
82.9
55.3
76.7
37.9
10.1
71.0
42.6
7.4
2-(O)Me/Me
3
4
26
5
6
27
7
8
28
9
Experimental Section
10
29
11
12
13
14
14-OMe
15
16
30
17
18
31
19
19-OMe
20
21
22
32
23
24
33
25
1′
General Experimental Procedures. Optical rotations were recorded
with a JASCO P-1010 polarimeter equipped with a halogen lamp (589
nm) and a 10 mm microcell. UV spectra were recorded using a Waters
2487 dual λ absorbance detector. NMR spectra were recorded on a
Bruker Avance 600 equipped with a cryoprobe at 600 MHz in C6D6 or
MeOD and referenced to the solvent. ESIMS spectra were obtained
with Micromass LCT and Bruker Esquire-LC mass spectrometers. A
Waters 10 g silica Sep-pak and a Waters 2 g C18 Sep-pak were used
for normal-phase and reversed-phase flash chromatrography, respec-
tively. Sephadex LH-20 was used for column chromatography. A
Waters 1500 Series pump system equipped with a Waters 2487 dual λ
absorbance detector and a CSC-Inertsil 150A/ODS2 column or Alltech
Apollo silica semipreparative column (10 × 250 mm, 5 µm) was used
for HPLC. Single-crystal X-ray diffraction measurements were made
on a Bruker X8 APEX II diffractometer with graphite-monochromated
Mo KR radiation. The data were collected at a temperature of -100 (
0.1 °C to a maximum 2θ value of 55.0°. Data were collected in a series
of φ and ω scans in 0.50° oscillations with 4.0 s exposures. The crystal
to detector distance was 36.00 mm.
129.3
129.8
13.1
100.4
99.5
154.3
99.5
40.6
76.4
39.4
12.6
77.3
29.2
21.8
14.6
171.5
42.6
39.6
175.1
60.6
33.7
178.1
39.6
80.2
40.0
12.6
76.5
28.5
21.7
14.7
56.3
97.6
79.4
34.4
14.3
83.6
28.6
20.4
15.3
54.9
39.4
80.2
40.0
12.6
76.4
28.5
21.4
14.7
56.3
Microbial Isolates RJA71 and RJA635. Isolate RJA71 was
obtained from a marine sediment collected in the Queen Charlotte
Islands, British Columbia. It was identified as Streptomyces sp. YIM26
by 16S RNA analysis (GenBank accession number AF389343.1).
Production cultures were grown as lawns on solid tryptic soy agar at
25 °C. Isolate RJA635 was obtained from a marine sediment collected
in Indian Arm, British Columbia. It was identified as Streptomyces
halstediiby16SRNAanalysis(GenBankaccessionnumberAB184831.1).
Production cultures were grown as lawns on solid ISP2 media at 25
°C.
2′
3′
4′
6′
7′
9′
a Spectra collected in MeOD at 600 MHz. b Spectra collected in C6D6
at 600 MHz.
Isolation of Bafilomycins. The solid agar culture of strain RJA635
was extracted with EtOAc and dried in Vacuo to give the crude extract.

Bafilomycins Produced in Culture by Streptomyces spp.
Journal of Natural Products, 2010, Vol. 73, No. 3 427
Bafilomycin G (2): clear glass; [R]20D -29 (c 0.4, MeOH); UV (87%
MeCN-H2O) λmax (log ꢀ) 245 (4.6), 286 (4.0) nm; 1H NMR, see Table
1; 13C NMR, see Table 2; (+)-HRESIMS [M + Na]+ m/z 659.4142
(calcd for C36H60O9Na, 659.4135).
gel chromatography (step gradient elution: CH2Cl2 to EtOAc) to give
a mixture of 17 and 18 that could not be separated by chromatography
(191 mg, 15% over two steps).
1
Compound 17: H NMR (CD3OD, 400 MHz) δ 4.49 (1H, dd, J )
Bafilomycin H (3): clear glass; [R]20 -1.8 (c 2.1, MeOH); UV
6.5, 4.1 Hz, H-6), 4.00 (1H, t, J ) 6.8 Hz, H-3), 3.79 (3H, s, H-11),
3.69 (3H, s, H-10), 3.31 (1H, m, H-7a), 3.08 (1H, m, H-7b), 2.95 (1H,
m, H-2a), 2.60 (1H, dd, J ) 16.5, 7.3 Hz, H-2b); 13C NMR (CD3OD,
100 MHz) δ 172.8 (C, C-4), 171.6 (C, C-9), 170.8 (C, C-1), 57.5 (CH,
C-6), 53.4 (CH3, C-11), 52.5 (CH3, C-10), 38.9 (CH, C-3), 36.8 (CH2,
C-2), 28.8 (CH2, C-7); HRESIMS(+) m/z 270.0413 (calcd for
C9H13NO5SNa, 270.0412).
D
1
(87% MeCN-H2O) λmax (log ꢀ) 245 (4.6), 286 (4.0) nm; H NMR,
see Table 1; 13C NMR, see Table 2; (+)-HRESIMS [M + Na]+ m/z
673.4301 (calcd for C37H62O9Na, 673.4291).
Bafilomycin I (4): clear glass; [R]20 +140 (c 0.1, MeOH); UV
D
1
(85% MeOH-H2O) λmax (log ꢀ) 245 (4.6), 283 (4.0) nm; H NMR,
see Table 1; 13C NMR, see Table 2; (+)-HRESIMS [M + Na]+ m/z
623.3917 (calcd for C36H56O7Na, 623.3923).
Compound 18: 1H NMR (CD3OD, 400 MHz) δ 4.53 (1H, t, J ) 4.8
Hz, H-6), 3.79 (1H, m, H-3), 3.79 (3H, s, H-11), 3.69 (3H, s, H-10),
3.28 (1H, m, H-7a), 3.08 (1H, m, H-7b), 2.97 (1H, m, H-2a), 2.85
(1H, dd, J ) 16.6, 7.8 Hz, H-2b); 13C NMR (CD3OD, 100 MHz) δ
172.8 (C, C-4), 172.0 (C, C-9), 170.6 (C, C-1), 58.0 (CH, C-6), 53.4
(CH3, C-11), 52.4 (CH3, C-10), 38.5 (CH, C-3), 38.4 (CH2, C-2), 27.1
(CH2, C-7); HRESIMS(+) m/z 270.0413 (calcd for C9H13NO5SNa,
270.0412).
Bafilomycin J (5): clear glass; [R]20D +40 (c 0.2, MeOH); UV (85%
MeOH-H2O) λmax (log ꢀ) 245 (4.6), 283 (4.0) nm; 1H NMR, see Table
1; 13C NMR, see Table 2; (+)-HRESIMS [M + Na]+ m/z 643.4168
(calcd for C36H60O8Na, 643.4186).
Synthesis of Model Compounds 12 and 13. Compounds 12 and
13 were synthesized according to a literature protocol.6 A solution of
cysteine methyl ester hydrochloride (9) (343 mg, 2.0 mmol) in MeOH
was added to a solution of N-phenylmaleimide (10) (363 mg, 2.1 mmol)
in CH2Cl2. The reaction was stirred at room temperature for 2 h. The
mixture was concentrated in Vacuo and purified by Si gel chromatog-
raphy (step gradient elution: CH2Cl2 to MeOH) to give 11 (652 mg,
95%). To a solution of 11 (652 mg, 1.90 mmol) in MeOH was added
diisopropylethylamine, and the reaction mixture was stirred at room
temperature for 2 days. The reaction mixture was dried in Vacuo and
purified by Si gel chromatography (step gradient elution: CH2Cl2 to
EtOAc) to give a pure 12 (85 mg) and pure 13 (26 mg) along with a
mixture of 12 and 13 (320 mg, total yield ) 74%). Compound 12 was
recrystallized from MeOH.
Biological Assays. The automated microscopy cell-based assay used
to detect autophagosome accumulation and the biochemical EGFP-
LC3 degradation assay have been described in detail.2a
Acknowledgment. Financial support was provided by grants from
the Natural Sciences and Engineering Research Council of Canada
(R.J.A.), the National Cancer Institute of Canada (R.J.A.), the Canadian
Breast Cancer Foundation (M.R.), a UBC University Graduate Fel-
lowship (G.C.), and a UBC 4-Year Fellowship (E.D.).
Supporting Information Available: NMR spectra for bafilomycins
F (1) to J (5); experimental details for X-ray diffraction analysis of 12.
This material is available free of charge via the Internet at http://
pubs.acs.org.
1
Compound 12: colorless needles (MeOH); H NMR (CD3OD, 400
MHz) δ 7.52 (2H, d, J ) 7.8 Hz, H-12, H-16), 7.28 (2H, t, J ) 7.8
Hz, H-13, H-15), 7.07 (1H, t, J ) 7.8 Hz, H-14), 4.47 (1H, dd, J )
6.3, 4.2 Hz, H-6), 4.09 (1H, dd, J ) 8.2, 5.7 Hz, H-3), 3.78 (3H, s,
H-17), 3.28 (2H, m, H-2a, H-7a), 3.07 (1H, m, H-7b), 2.61 (1H, dd, J
) 15.4, 8.2 Hz, H-2b); 13C NMR (CD3OD, 100 MHz) δ 171.8 (C,
C-9), 171.4 (C, C-4), 170.8 (C, C-1), 139.8 (C, C-11), 129.8 (CH, C-13,
C-15), 125.2 (CH, C-14), 121.3 (CH, C-12, C-16), 57.7 (CH, C-6),
53.4 (CH3, C-17), 39.5 (CH2, C-2), 39.4 (CH, C-3), 28.8 (CH2, C-7);
HRESIMS(+) m/z 331.0734 (calcd for C14H16N2O4SNa, 331.0728).
References and Notes
(1) (a) Suzuki, K.; Ohsumi, Y. FEBS Lett. 2007, 581, 2156–2161. (b)
Martinez-Vicente, M.; Cuervo, A. M. Lancet Neurol. 2007, 6, 352–
361. (c) Yoshimori, T. Cell 2007, 128, 833–836. (d) Rubinstein, D. C.;
Gestwicki, J. E.; Murphy, L. O.; Klionsky, D. J. Nat. ReV. Drug
DiscoVery 2007, 6, 304–312. (e) Schmid, D.; Munz, C. Immunity 2007,
27, 11–21. (f) Xiao, G. Cytokine Growth Factor ReV. 2007, 18, 233–
243. (g) Ferraro, E.; Cecconi, F. Arch. Biochem. Biophys. 2007, 462,
210–219. (h) Mizushima, N. Proc. Jpn. Acad., Ser. B 2007, 83, 39–
46. (i) Yorimitsu, T.; Klionsky, D. J. Trends Cell Biol. 2007, 17, 279–
285.
(2) (a) Balgi, A. D.; Fonseca, B. D.; Donohue, E.; Tsang, T. C. F.; Lajoie,
P.; Proud, C. G.; Nabi, I. R.; Roberge, M. PLoS ONE 2009, 4, e7124.
(b) Klionsky, D. J.; Cuerva, A. M.; Seglen, P. O. Autophagy 2007, 3,
181–206.
1
Compound 13: H NMR (CD3OD, 400 MHz) δ 7.52 (2H, d, J )
7.8 Hz, H-12, H-16), 7.28 (2H, t, J ) 7.8 Hz, H-13, H-15), 7.07 (1H,
t, J ) 7.8 Hz, H-14), 4.50 (1H, m, H-6), 3.87 (1H, dd, J ) 8.6, 4.8 Hz,
H-3), 3.78 (3H, s, H-17), 3.26 (1H, m, H-7a), 3.07 (1H, m, H-7b),
3.04 (1H, m, H-2a) 2.85 (1H, dd, J ) 15.3, 8.6 Hz, H-2b); 13C NMR
(CD3OD, 100 MHz) δ 171.8 (C, C-9), 171.4 (C, C-4), 170.8 (C, C-1),
139.8 (C, C-11), 129.8 (CH, C-13, C-15), 125.2 (CH, C-14), 121.3
(CH, C-12, C-16), 58.1 (CH, C-6), 53.4 (CH3, C-17), 40.9 (CH2, C-2),
39.0 (CH, C-3), 27.0 (CH2, C-7); HRESIMS(+) m/z 331.0732 (calcd
for C14H16N2O4SNa, 331.0728).
(3) Keyzers, R. A.; Daoust, J.; Davies-Coleman, M. T.; Van Soest, R.;
Balgi, A.; Donohue, E.; Roberge, M.; Andersen, R. J. Org. Lett. 2008,
10, 2959–2962.
Synthesis of Model Compounds 17 and 18. To fumaric acid (14)
(1.45 g, 12.5 mmol) suspended in MeOH (20 mL) was added
concentrated H2SO4 (0.5 mL), and the mixture was heated to reflux
for 1 h. The mixture was then cooled in an ice bath and neutralized
with 10% Na2CO3. CH2Cl2 was added, and the layers were separated.
The organic layer was dried in Vacuo, redissolved in CH2Cl2, and
purified by Si gel chromatography (eluent: CH2Cl2) to give dimethyl-
fumarate (15) (1.63 g, 91%).
To a solution of L-cysteine methyl ester hydrochloride (9) (858 mg,
5.0 mmol) in MeOH (20 mL) was added diisopropylethylamine (1.8
mL, 10.3 mmol), followed by a solution of dimethylfumarate (15) (734
mg, 5.1 mmol) in CH2Cl2. The reaction was stirred for 15 min at room
temperature, at which point TLC analysis indicated that the reaction
was complete. The reaction mixture was purified by Si gel chroma-
tography (step gradient elution: CH2Cl2 to EtOAc) to give 16.
Compound 16 was dissolved in DMF and heated to 140 °C for 6 h.
The reaction mixture was cooled, dried in Vacuo, and purified by Si
(4) Werner, G.; Hagenmaier, H.; Albert, K.; Kohlshorn, H.; Drautz, H.
Tetrahedron Lett. 1983, 24, 5193–5196.
(5) Kretschmer, A.; Dorgerloh, M.; Deeg, M.; Hagenmaier, H. Agric. Biol.
Chem. 1985, 49, 2509–2511.
(6) Shih, H. C.; Rankin, G. O. J. Heterocycl. Chem. 1988, 25, 589–590.
(7) Crystallographic data for 12 have been deposited with the Cambridge
Crystallographic Data Centre (deposition number CCDC 757295).
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).
(8) Baker, G. H.; Brown, P. J.; Dorgan, R. J. J.; Everett, J. R.; Ley, S. V.;
Slawin, A. M. Z.; Williams, D. J. Tetrahedron Lett. 1987, 28, 5565–
5568.
(9) Hatfield, G. M.; Woodard, R. W.; Son, J.-K. J. Nat. Prod. 1992, 55,
753–759.
(10) Bowman, E. J.; Siebers, A.; Altendorf, K. Proc. Natl. Acad. Sci. U.S.A.
1988, 85, 7972–7976.
NP900632R