J. Garcia-Rodriguez et al. / Bioorg. Med. Chem. Lett. 25 (2015) 4393–4398
4395
d. DHP, TsOH
e. allylMgBr, CuI
f. TBSCl
g. MgBr2
HO
THPO
HO
O
OH
OTBS
OTBS
55%
63%
7
8
9
h. Swern ox.
80%
a. TiCl4, (-)-sparteine
6b
OHC
S
OH
O
b.TBSCl
c. DIBAL
S
O
CHO
N
S
Me
N
S
R
R
4
i
Pr
iPr
R
60%, dr >15:1
75%, dr >10:1
3a: R = Me
3b: R = H
5a: R = Me
5b: R = H
70%
71%
6a: R = Me
6b: R = H
HO
Ph
NH2
Ph
i. In, 10,
propargylBr
10
I
I
k. MsCl
l. NaN3
m. Me3P
dr >20:1
j. Bu3SnH,
AIBN, I2
H2N
HO
HO
OTBS
OTBS
OTBS
R
R
63% (crude)
72%
70%
86%
85%
12c
12a: R = Me
12b: R = H
11a: R = Me
11b: R = H
Scheme 2. Reagents and conditions: (a) TiCl4 (6.0 equiv), (À)-sparteine (3.0 equiv), 4 (3.0 equiv), CH2Cl2, À78 °C, 30 min, (5a, 60%; 5b, 75%.); (b) TBSOTf (1.2 equiv), 2,6-
lutidine (1.5 equiv), CH2Cl2, 23 °C, 2 h; (c) DIBAL-H (1.8 equiv), toluene, À78 °C, 3 h, (6a, 70%; 6b, 71%; two steps); (d) p-TsOH (0.05 equiv), dihydropyran (DHP, 2.0 equiv),
THF, 23 °C, 4 h, 88%; (e) allylmagnesium bromide (3.0 equiv), CuI (0.1 equiv), THF, À20 °C, 1 h, 62%; (f) imidazole (5.0 equiv), TBSCl (1.2 equiv), DMF, 23 °C, 12 h, 87%; (g)
MgBr2ÁEt2O (3.0 equiv), Et2O, 23 °C, 2 h, 72%; (h) DMSO (2.0 equiv), oxalyl chloride (1.5 equiv), Et3N (5.0 equiv), CH2Cl2, À78 °C, 1 h, 80%; (i) propargyl bromide (4.5 equiv), 10
(3.0 equiv), In (3.0 equiv), pyridine (3.0 equiv), THF, À78 °C to 23 °C, 10 h, (11a, 86%; 11b, 85%); (j) Bu3SnH (1.2 equiv), AIBN (0.5 equiv) toluene, 110 °C, 2 h; then I2 (1.0 equiv),
THF, À78 °C, 30 min (12a, 72%; 12b, 70%); (k) MsCl (3.5 equiv), Et3N (3.5 equiv), Et2O, 23 °C, 1 h; (l) NaN3 (3.0 equiv), DMF, 60 °C, 2 h, (63%, two steps); (m) Me3P (3.0 equiv),
9:1 THF/H2O, 60 °C, 1 h, then MW, 100 °C, 30 min (used crude).
with substituted salicylic acid 1322 under Mitsunobu conditions
(DEAD/PPh3).10c For the esterification of 12a, a stoichiometric
amount of pyridine was required to prevent competing
elimination of the alcohol under the reaction conditions.23 The
N-acyl enamide side chain was installed by Cu-catalyzed amida-
tion of 14a,b with amide 15 under the conditions described by
Buchwald et al., providing enamides 16a,b in 60–78% yield after
re-installation of the acetate group.24,10f The ring-closing metathe-
sis of 16a,b with Grubbs I catalyst provided the benzolactones 2
and 17 in 31–36% yields after global deprotection.25,26 The E/Z iso-
mers could be separated by flash chromatography after removing
the acetate group. Of note is that the des-Me RCM-precursor 16b
cyclized with significant lower E/Z selectivity (4:1) compared to
the parent 16a (8:1), indicating a beneficial role for the Me-sub-
stituent in pre-organizing the substrate for an E-selective metathe-
sis.27 This lowered selectivity is observed for all des-methyl analogs
prepared (vide supra, Chart 1). Final desilylation with buffered
HFÁpyridine provided SaliPhe 2 and analog 17 in ten steps from
known aldehydes 3a,b.
When assayed for V-ATPase inhibitory activity, it was found
that des-Me analog 17 displayed potency comparable to SaliPhe
(IC50 = 5 nM, 17; 1 nM, 2), indicating only a marginal role for the
Me-group on V-ATPase inhibitory activity.28 Based on these results,
we decided to carry out the subsequent structural modifications in
the des-Me series, and survey the influence of various aromatic
substituents on biological activity. We envisioned exploring diflu-
oromethyl, fluoro, amino, nitro and hydroxyl substituents at the
ortho-, meta-, or para-position. Analogs 21–24 (see Chart 1) were
readily obtained using the chemistry outlined in Scheme 3, starting
from Mitsunobu reaction of alcohol 12b with various benzoic
acids.29,30 Interestingly, the RCM-product of an intermediate with
a meta-OH or final product with a meta-fluoro substituent were
unstable and could not be further processed or evaluated, whereas
the o-NO2 substituted analog 19 (Chart 1) could not be prepared
using this route due to failure of the corresponding bis-olefin inter-
mediate to undergo the ring closing metathesis.29
As discussed during the retrosynthetic analysis, we also investi-
gated whether we could intercept vinyl iodide intermediates in a
Pd-catalyzed carbonylation event to produce a,b-unsaturated ester
intermediates reminiscent of those utilized in our first-generation
synthetic approach (cf. IV ? II, Scheme 1). As shown in Scheme 4,
the palladium-catalyzed carbonylation of vinyl iodides 26a–c (pre-
pared via Mitsunobu reaction of alcohol 12b with the correspond-
ing benzoic acids)29 with catalytic PdCl2(PPh3)2 (Et3N, 1 atm of CO,
MeOH, rt) proceeded smoothly to yield, after ring-closing metathe-
sis as before, the desired unsaturated esters 27a–c in good overall
yields. Of note is that the ring-closing metathesis to o-NO2 substi-
tuted benzolactone 27c occurred with the best yield and E:Z-selec-
tivity (94%, 5.6:1) of all des-Me analogs evaluated in this study,
whereas the corresponding ring-closing metathesis with o-NO2
as implemented in the route depicted in Scheme 3 did not occur
at al. The remainder of the synthesis (enamide side chain forma-
tion) essentially followed our published sequence of ester hydrol-
ysis, acyl azide formation (28a–c) followed by Curtius
rearrangement and trapping of the resulting isocyanate with phe-
nylethynyl lithium, and final deprotection.8e,10c In addition to
providing alternate access to analog 17, this route also furnished
new analogs 18–20 (20 from reduction of 19 with Fe).
We tested the ability of all new analogs for their ability to inhi-
bit the activity of a purified reconstituted V-ATPase from bovine
brain (Chart 1, IC50-values shown in red).28 Analysis of the data
indicated that the des-Me analog 17, and the corresponding lactam
2531,32 remain potent V-ATPase inhibitors with IC50’s of 5 and
30 nM, respectively (versus 1 nM for SaliPhe 2). The modifications
around the phenol ring proved to be detrimental for activity.
Replacement of the o-OH with
a difluoromethyl bioisostere
resulted in a 100-fold reduction in activity of analog 21 (500 nM)
versus the corresponding phenol 17 (5 nM).33 A similar loss of