Bafilomycin analogue site-specifically fluorinated at the pharmacophore macrolactone ring has potent vacuolar-type ATPase inhibitory activity

Article abstract of DOI:10.1016/j.tetlet.2016.04.075

Based on previously reported structure-activity relationships, an analogue of bafilomycin A₁ with a site-specific fluorine label at the C2 position was designed and efficiently synthesized. The fluorinated compound exhibited potent vacuolar-typ

Full text of DOI:10.1016/j.tetlet.2016.04.075

Tetrahedron Letters 57 (2016) 2426–2429
Contents lists available at ScienceDirect
Tetrahedron Letters
journal homepage: www.elsevier.com/locate/tetlet
Bafilomycin analogue site-specifically fluorinated
at the pharmacophore macrolactone ring has potent
vacuolar-type ATPase inhibitory activity
c
Hiroshi Tsuchikawa a, Tatsuru Hayashi a, Hajime Shibata a,b, Michio Murata a,b, , Yoko Nagumo ,
⇑
Takeo Usui c
a Graduate School of Science, Osaka University, 1-1 Machikaneyama, Toyonaka, Osaka 560-0043, Japan
b ERATO, Lipid Active Structure Project, Japan Science and Technology Agency, 1-1 Machikaneyama, Toyonaka, Osaka 560-0043, Japan
c Faculty of Life and Environmental Sciences, University of Tsukuba, Tennodai, Tsukuba, Ibaraki 305-8572, Japan
a r t i c l e i n f o
a b s t r a c t
Article history:
Based on previously reported structure–activity relationships, an analogue of bafilomycin A1 with a site-
specific fluorine label at the C2 position was designed and efficiently synthesized. The fluorinated com-
pound exhibited potent vacuolar-type ATPase (V-ATPase) inhibitory activity, comparable to that of the
natural product, representing the first example of highly bioactive analogues with a modified macrolac-
tone core from the plecomacrolide family compounds.
Received 15 March 2016
Revised 13 April 2016
Accepted 20 April 2016
Available online 21 April 2016
Ó 2016 Elsevier Ltd. All rights reserved.
Keywords:
Fluorine-label
Chemical synthesis
V-ATPase
Bafilomycin
Conformational analysis
A vacuolar-type ATPase (V-ATPase) is an ATP-driven proton
pump occurring in eukaryotic cells. V-ATPase acts upon acidifica-
tion of endomembrane organelles such as lysosomes, endosomes,
and Golgi apparatus to regulate protein degradation and trans-
port.1 Subtypes of V-ATPase are also distributed in the plasma
membranes of various cells, including osteoclasts, renal interca-
lated cells, and cancer cells, and are correspondingly deeply
involved in bone resorption, pH homeostasis and cell growth.1–4
Consequently, specific inhibitors of V-ATPase have received signif-
icant attention as useful tools for functional analysis, and also as
drug candidates. For example, it has been suggested that suppres-
sion of V-ATPase activity induces the apoptotic death of cancer
cells.5,6
One of the most potent and specific V-ATPase inhibitors, bafilo-
mycin A1 (Baf, 1, Scheme 1), was first isolated from Streptomyces
griseus sp. sulphurus.7 Although it has been widely utilized as a
diagnostic agent for V-ATPase-implicated biological events, the
molecular mechanism of action is not fully elucidated; the binding
site has been deduced to be the subunit c ring of the transmembra-
nous Vo domain.8,9
To elucidate the exact binding site for attachment of Baf to V-
ATPase, we utilize structural analysis using 13C{19F}REDOR,10
which is a distance measurement solid-state NMR technique often
used for complicated non-crystalline systems such as membrane
proteins.11 Preparation of a bioactive fluorine-labeled derivative
is essential as the first step of this strategy; we recently success-
fully synthesized 24,24-didesmethyl-24-F3-Baf (24-F-Baf) possess-
ing a CF3 group at the C23 position instead of the i-propyl group.
This structure was carefully designed based on consideration of
previously reported structure–activity relationship (SAR) data.12
As anticipated we confirmed the potent V-ATPase inhibitory activ-
ity of 24-F-Baf.13 Another objective for the synthesis of fluorine-
labeled Baf is to elucidate the exact conformation of Baf in the V-
ATPase bound form since knowledge of the active structure of a
specific inhibitor upon binding to a potential drug target protein
generally provides important information for drug development.
The significance of the macrolide conformation of Baf to the overall
binding structure could be determined using a site-specifically flu-
orinated Baf analogue at this moiety. However the potential per-
turbation induced by fluorine substitution14 was of particular
concern since the macrolactone ring was thought to be the phar-
macophore for V-ATPase binding. Previous SAR experiments
demonstrated that any modifications on the macrolactone resulted
in loss of activity. For example, opening of the lactone ring,
⇑
Corresponding author. Tel./fax: +81 6 6850 5774.
E-mail address: murata@chem.sci.osaka-u.ac.jp (M. Murata).
http://dx.doi.org/10.1016/j.tetlet.2016.04.075
0040-4039/Ó 2016 Elsevier Ltd. All rights reserved.

H. Tsuchikawa et al. / Tetrahedron Letters 57 (2016) 2426–2429
2427
O
OH
O
OPMB
Macrolactonization
a,b
c
MeO
F
N
OBn
R
OBn
24
EtO
1
O
OH
TMS
2
8
7
OTMS
O
OH
O
11
OH OPMB
OH OPMB
O
OH
17
18
d,e
HO
12
HO
23
OBn
23
+
OTBS
OMe
3
Stille
coupling
Aldol reaction
11
O
9
23R-isomer
10
23S-isomer
1 : 4
11
separable
1
2
: R= OMe (bafilomycin A1, Baf)
: R= F (2-desmethoxy-2-F-Baf, 2-F-Baf)
I
6
t-Bu
O
t-Bu
O
OH OH
Si
g
f
23
OBn
OBn
F
t-Bu
O
t-Bu
O
1
Si
12
23R-isomer
OH
13
t-Bu
O
O
2
EtO2C
OH
t-Bu
Si
24
17
18
11
I
t-Bu
O
t-Bu
SnBu3
O
O
Si
MMTrO
j,k
h,i
12
O
O
OMe
18
23
24
3
C18-C24 segment
4
C12-C17 segment
5
C1-C11 segment
3
C18-C24 segment
14
Scheme 1. Retrosynthesis of 2-F-Baf.
Scheme 2. Synthesis of the C18–C24 segment 3. Reagents and conditions: (a)
PMBOCNHCCl3, Sc(OTf)3, toluene, rt, 85%; (b) DIBAL, CH2Cl2, À78 °C, 92%; (c) i-
PrMgBr, THF, 0 °C, 81%; (d) AZADOL, TBAB, NaClOÁ5H2O, KBr, CH2Cl2, NaHCO3 aq,
0 °C, 99%; (e) DDQ, CH2Cl2, PBS buffer (pH 7.0), rt, 74%; (f) Me4NBH(OAc)3, AcOH, rt,
79% and 10% 23S-isomer; (g) (t-Bu)2Si(OTf)2, 2,6-lutidine, DMF, rt, 85%; (h) H2, Pd
black, EtOH, rt, 68%; (i) Dess-Martin periodinane, NaHCO3, CH2Cl2, rt, 86%; (j)
EtMgBr, THF, rt, 99%; (k) Dess-Martin periodinane, NaHCO3, CH2Cl2, rt, 92%.
oxidation of the 7-hydroxy group to the ketone, and even
demethylation of the two methyl groups at the C6 and C8 positions
remarkably diminished the activity.12c,13b
Considering the previous results, we designed and synthesized
a novel fluorine-labeled Baf analogue modified at the macrolactone
backbone; this analogue exhibited V-ATPase inhibitory activity
comparable to that of the native structure. Compound 2 (2-F-Baf,
2) is the first analogue with the modified macrolactone that pos-
sesses potent bioactivity.
In selecting the position for fluorine substitution, we first chose
the hydroxy or methoxy group to minimize perturbation of the
polarity. Among the three possible positions, the 7-hydroxy group
and 14-methoxy group were ruled out because of the known bio-
logical importance and the expected instability, respectively.
Hence, the 2-F-Baf derivative (2-F-Baf, 2) was selected as a synthe-
sis target; this moiety was expected to be easier to synthesize
despite concerns about the effect of fluorine in destabilizing the
surrounding dienoate moiety (Scheme 1). The strategy involved
the convergent synthetic method established for synthesis of 24-
F-Baf from three key segments via the Stille coupling, macrolac-
tonization, and diastereoselective aldol reactions.13 For synthesis
of the C18–C24 segment, 3, which was already reported by two
groups,15 we examined a new route by modifying our scheme for
synthesis of the CF3-C18–C24 segment to improve the efficiency.13
The 2-F-labeled C1–C11 segment, 5, could be synthesized from the
known aldehyde 6 via Z-selective fluoroolefination.
formed as shown in Scheme 3. At first, the fluoroolefination of alde-
hyde 6, which was synthesized as outlined in our previous report,13
was examined using fluoro(trimethylsilyl)ketene ethyl trimethylsi-
lyl acetal 15 (Z:E = 1:1) in the presence of a catalytic amount of n-
Bu4NOAc as a Lewis base according to Michida’s procedure.20 The
reaction proceeded smoothly to afford the desired (Z)-a-fluoroacry-
late product selectively and the subsequent deprotection of t-
butyldimethylsilyl (TBS) ether gave the C1–C11 segment, 5 in 48%
yield in two steps. The Stille coupling reaction with the separately
prepared C12–C17 segment, 413b and hydrolysis of the ethyl ester
furnished the carboxylic acid 17; this process was followed by
macrolactonization under Yamaguchi conditions21 to successfully
furnish the fluorine-containing macrolactone in a similar yield to
that obtained with the non-fluorinated seco acid.13 Protection of
the secondary alcohol with a diethylisopropylsilyl (DEIPS) group
via a previously optimized process15a and selective removal of the
monomethoxytrityl (MMTr) group with pyridinium p-
toluenesulfonate (PPTS) furnished the primary alcohol 18 in 83%
yield despite the expected lability of the fluorinated dienoate
moiety under acidic conditions. After oxidation to aldehyde 19, the
(E)-boron enolate generated from ketone 3 was added and stirred
at À78 °C for 3 h to give the desired hydroxyketone 20 with
excellent stereoselectivity in 60% yield,22 which was comparable
to the data for natural Baf synthesis.15a,23 Finally, deprotection of
the two silyl groups was performed by using carefully tuned
conditions, including the use of 18% HFÁpyridine and TBAF with
AcOH, leading to completion of the synthesis of 2-F-Baf.24
Initially, the C18–C24 segment, 3, was prepared by starting
from the known Weinreb amide 716 as shown in Scheme 2. After
several attempts at direct alkylation to 7,17 it was found that trans-
formation to the aldehyde 8 and the subsequent addition of iso-
propyl magnesium bromide afforded the isopropylated alcohols 9
and 10 as 1:4 diastereomer mixtures. This undesired stereoselec-
tivity could be explained by Felkin-Ahn model applied to a-chiral
To evaluate the biological activity of 2-F-Baf, 2, the inhibitory
effect of V-ATPase on the acidification of acidic organelles was
examined using HeLa cells (Fig. 1).13 The control cells exhibited
the red fluorescence of acidic vesicles, which were acidified by
properly functioning V-ATPase. In contrast, treatment with 2-F-
Baf, 2 markedly reduced the red fluorescence to a similar extent
to that achieved with Baf, 1, which implied that 2-F-Baf, 2 pos-
sessed potent V-ATPase inhibitory activity, comparable to that of
the natural product. To evaluate the inhibition potency of 2, the
inhibitory effect on V-ATPase activity was evaluated based on
quantification of the inorganic phosphate produced by ATP hydrol-
ysis (Table 1).13,25 The results demonstrated that 2-F-Baf, 2, inhib-
aldehydes. The undesired 23S-epimer 10 was separately oxidized
to the ketone,18 and removal of the p-methoxybenzyl (PMB) group
yielded the b-hydroxyketone 11, which was followed by 1,3-anti-
selective reduction reported by Evans to furnish the desired 23R-
isomer 12 with good selectivity.19 After cyclic silyl protection of
the diols and silica gel purification, the desired silyl ether 13 was
obtained in 85% yield, followed by a conventional sequence in four
steps to afford the C18–C24 segment, 3. The overall steps were
shortened and modification of the C23 substituent was more facile
than in previous reports.15
Subsequently, synthesis of the 2-fluorinated C1–C11 segment, 5,
and the ensuing coupling reactions leading to 2-F-Baf, 2, were per-

2428
H. Tsuchikawa et al. / Tetrahedron Letters 57 (2016) 2426–2429
F
2
F
F
1
OH
OTBS
b
OTBS
EtO
3
EtO2C
OH
EtO2C
TMS
17
c,d
SnBu3
O
+
OTMS
11
I
MMTrO
15
11
12
a
OMe
I
I
16
6
4
C12-17 segment
C1-C11 segment 5
F
F
F
t-Bu
O
t-Bu
O
O
ODEIPS
OH
O
ODEIPS
Si
HO2C
OH
e-g
i
O
O
h
O
O
+
HO
MMTrO
OMe
OMe
OMe
C18-C24 segment 3
18
17
19
F
F
F
O
ODEIPS
j
O
O
OH
t-Bu
O
t-Bu
O
Si
k
O
O
OH
O
OH
O
O
OH
O
18
17
HO
HO
OH
OH
OMe
OMe
OMe
21
20
2-F-Baf 2
Scheme 3. Synthesis of 2-F-Baf 2. Reagents and conditions: (a) 15, n-Bu4NOAc, CH2Cl2, rt; (b) TBAF, THF, rt, 48% (2 steps); (c) [Pd2(dba)3]ÁCHCl3, Ph3As, LiCl, NMP, rt, 63%; (d)
KOH, dioxane, 80 °C; (e) 2,4,6-trichlorobenzoyl chloride, i-Pr2NEt, toluene, rt, then diluted with toluene, DMAP, 34% (2 steps); (f) DEIPSOTf, 2,6-lutidine, CH2Cl2, 0 °C, 90%; (g)
PPTS, THF, MeOH, rt, 83%; (h) (COCl)2, DMSO, Et3N, CH2Cl2, À78 °C to 0 °C; (i) 3, PhBCl2, i-Pr2NEt, CH2Cl2, À78 °C, 60% (2 steps); (j) 18% HFÁPy, THF, rt, 67%; (k) TBAF, AcOH, THF,
rt, 48%.
Figure 1. Effect of 2-F-Baf on acridine orange stain of HeLa cells.
ited V-ATPase function as potently as natural Baf, 1, (IC50 = 5.3 and
4.3 nM, respectively), indicating that fluorine substitution at the C2
position of Baf induced no perturbation effect.
After confirming the potent activity of 2-F-Baf, we next exam-
ined whether the analogue maintained enough of the structural
features of Baf to be used for accurate conformational analysis in
solid-state NMR by comparing their structures in solution. The
1H NMR data of 2 were compared with those of 1 as shown in Fig-
ure 2. Most of the 1H chemical shifts of these two species were sim-
fluorine atom and susceptible hydroxy protons (H3, H5,17-OH,
19-OH) (see Supporting information for detailed data). Moreover,
all coupling constants of the protons on the ring shown in Figure 2A
were consistent with those of Baf, and NOE correlations attributed
to H3/H5 and H8/H11, characteristic to Baf, were observed
(Fig. 2B). These observations suggested that the entire conforma-
tion of the macrolactone ring is virtually the same as that of Baf.
To confirm the conformation, a detailed structural analysis was
performed using Macromodel with some restrictions derived from
NMR experiments (Fig. 2B) (see Supporting information for
detailed data). The superimposed structures clearly showed that
2-F-Baf and Baf have a similar conformation in the macrolactone
moiety, especially in the direction of the 7-hydroxy group; this
conformation was largely altered in non-active analogues such as
the 6,8-desmethylated derivative.13b These results suggest that 2-
F-Baf retains not only the V-ATPase activity but also the
conformation of natural Baf in CDCl3 solution.
ilar (D
d < 0.05 ppm), except for those of the protons close to the
Table 1
V-ATPase inhibition of budding yeast.
IC50 (nM)
Baf 1
2-F-Baf 2
4.3
5.3

H. Tsuchikawa et al. / Tetrahedron Letters 57 (2016) 2426–2429
2429
Graduate Schools, MEXT, Japan. H.S. expresses his special thanks
for the supports from Global COE Programs of Osaka University.
Supplementary data
Supplementary data associated with this article can be found, in
the online version, at http://dx.doi.org/10.1016/j.tetlet.2016.04.
075.
References and notes
1. Forgac, M. Nat. Rev. Mol. Cell Biol. 2007, 8, 917–929.
2. Cotter, K.; Stransky, L.; McGuire, C.; Forgac, M. Trends Biochem. Sci. 2015, 40,
611–622.
3. Marshansky, V.; Rubinstein, J. L.; Grüber, G. Biochim. Biophys. Acta 2014, 1837,
857–879.
4. Sun-Wada, G.-H.; Wada, Y. Biochim. Biophys. Acta 2015, 1847, 1166–1172.
5. Graham, R. M.; Thompson, J. W.; Webster, K. A. Oncotarget 2013, 5, 1162–1173.
6. Schneider, L. S.; Schwarzenberg, K. V.; Lehr, T.; Ulrich, M.; Kubisch-Dohmen, R.;
Liebl, J.; Trauner, D.; Menche, D.; Vollmar, A. M. Cancer Res. 2015, 75, 2863–
2874.
7. (a) Werner, G.; Hagenmaier, H.; Albert, K.; Kohlshorn, H. Tetrahedron Lett. 1983,
24, 5193–5196; (b) Werner, G.; Hagenmaier, H.; Drautz, H.; Baumgartner, A.;
Zähner, H. J. Antibiot. 1984, 37, 110–117.
8. (a) Bowman, B. J.; Bowman, E. J. J. Biol. Chem. 2002, 277, 3965–3972; (b)
Bowman, E. J.; Graham, L. A.; Stevens, T. H.; Bowman, B. J. J. Biol. Chem. 2004,
279, 33131–33138; (c) Bowman, B. J.; McCall, M. E.; Baertsch, R.; Bowman, E. J.
J. Biol. Chem. 2006, 281, 31885–31893.
9. (a) Huss, M.; Ingenhorst, G.; Kçnig, S.; Gaßel, M.; Drçse, S.; Zeeck, A.; Altendorf,
K.; Wieczorek, H. J. Biol. Chem. 2002, 277, 40544–40548; (b) Osteresch, C.;
Bender, T.; Grond, S.; Zezschwitz, P. V.; Kunze, B.; Jansen, R.; Huss, M.;
Wieczorek, H. J. Biol. Chem. 2012, 287, 31866–31876.
10. Gullion, T.; Schaefer, J. J. Magn. Reson. 1989, 81, 196–200.
11. Gullion, T. Annu. Rep. NMR Spectrosc. 2009, 65, 111–137.
12. (a) Lu, C.; Shen, Y. J. Antibiot. 2003, 56, 415–418; (b) Ohta, E.; Kubota, N. K.;
Ohta, S.; Suzuki, M.; Ogawa, T.; Yamasaki, A.; Ikegami, S. Tetrahedron 2001, 57,
8463–8467; (c) Gagliardi, S.; Gatti, P. A.; Belfiore, P.; Zocchetti, A.; Clarke, G. D.;
Farina, C. J. Med. Chem. 1998, 41, 1883–1893.
13. (a) Shibata, H.; Tsuchikawa, H.; Matsumori, N.; Murata, M.; Usui, T. Chem. Lett.
2014, 43, 474–476; (b) Shibata, H.; Tsuchikawa, H.; Hayashi, T.; Matsumori, N.;
Murata, M.; Usui, T. Chem. Asian J. 2015, 10, 915–924.
Figure 2. (A) Differences of 1H NMR chemical shifts between 2-F-Baf 2 and Baf 1
and observed coupling constants in CDCl3. (B) Lowest-energy conformation of Baf
(orange) and 2-F-Baf (green) and experimentally observed NOE correlations of
them.
14. Smart, B. E. J. Fluorine Chem. 2001, 109, 3–11.
It is recognized that there must be some differences between
the conformation in solution and the active 3D structure upon
binding to V-ATPase in the membrane, because the former is lar-
gely influenced by intramolecular hydrogen bonding, whereas,
the latter is potentially affected by interaction with protein. Never-
theless, we deduce that there may be some correlations since Baf is
15. (a) Toshima, K.; Jyojima, T.; Yamaguti, H.; Noguchi, Y.; Yoshida, T.; Murase, H.;
Nakata, M.; Matsumura, M. J. Org. Chem. 1997, 62, 3271–3284; (b) Yadav, J. S.;
Reddy, K. B.; Sabitha, G. Tetrahedron 2008, 64, 1971–1982.
16. Rauhala, V.; Nevalainen, M.; Koskinen, A. M. P. Tetrahedron 2004, 60, 9199–
9204.
17. The direct alkylation of 7 or its PMB ether derivative with i-PrMgBr resulted in
no reaction or only demethoxylation. On the other hand addition of the
isopropenyl magnesium bromide to 7 proceeded in 82% yield, however, any of
reduction conditions did not give the desired b-alcohol 7b as a main product.
thought to bind to the hydrophobic interface in the
a-helix
domains in subunit c of V-ATPase, where the formation of multiple
intermolecular hydrogen bonds is less likely.8
MgBr
O
OH
O
OH
In summary, we successfully synthesized site-specifically fluo-
rinated bafiromycin A1 at the C2 position of the macrolactone ring.
This analogue exhibits potent V-ATPase inhibitory activity, and is
the first bioactive agent of the plecomacrolide family compounds
in which the macrolactone core is modified as far as we know. In
addition, fluorine-labeling of Baf backbone with no structural per-
turbation would pave the way to a better understanding of the
mechanism underlying V-ATPase activity and to the development
of a specific inhibitor of the enzyme. Solid-state NMR experiments
are now underway in our laboratory.
MeO
OBn
OH OR
N
OBn
O
11
7
OH
MgBr
OBn
, 82%
7a
7b
18. Shibuya, M.; Sasano, Y.; Tomizawa, M.; Hamada, T.; Kozawa, M.; Nagahana, N.;
Iwabuchi, Y. Synthesis 2011, 3418–3425.
19. Evans, D. A.; Chapman, K. T.; Carreira, E. M. J. Am. Chem. Soc. 1988, 110, 3560–
3578.
20. Michida, M.; Mukaiyama, T. Chem. Lett. 2008, 37, 890–891.
21. Inanaga, J.; Hirata, K.; Saeki, H.; Katsuki, T.; Yamaguchi, M. Bull. Chem. Soc. Jpn.
1979, 52, 1989–1993.
Acknowledgments
22. The configuration at C17 and C18 of the major product 20 was determined to
be the same as reported in Ref. 13, see the Supporting information for details.
23. Evans, D. A.; Calter, M. A. Tetrahedron Lett. 1993, 34, 6871–6874.
24. When a strong acid such as TsOH was used for removal of the 7-O-silyl group
by following our previous synthesis,13 no deprotected product was obtained.
Under acidic conditions, the lactone ring was easily opened probably due to
strong inductive effect of a fluorine atom.
We thank Prof. N. Matsumori (Kyushu University) and Dr. S.
Hanashima (Osaka University) for discussions and Dr. N. Inazumi
(Osaka University) for his help in NMR measurements. This work
was partly supported by JST ERATO, Lipid Active Structure Project,
Grant-In-Aids for Scientific Research (C) (No. 15K01821) and for
Scientific Research (A) (No. 25242073) and Grants for Excellent
25. Uchida, E.; Ohsumi, Y.; Anraku, Y. J. Biol. Chem. 1985, 260, 1090–1095.