C. Constant-Urban et al. / Bioorg. Med. Chem. Lett. 23 (2013) 5878–5881
5879
openers like diazoxide, a reference mitoKATP channel opener, have
been proposed to improve organ and cells survival before
transplantation.7
In spite of numerous studies performed on the mitoKATP chan-
nel, the molecular structure of the latter channel remains the ob-
ject of scientific controversy. State of the art indicates that,
among potassium channel openers already studied on cardiomyo-
cytes, diazoxide (interacting with SUR1/Kir6.2 and SUR2B/Kir6.1/
Kir6.2-type channels) was more efficient at opening mitoKATP
channels than sarcolemmal KATP channels (cell membrane
SUR2A/Kir6.2-type channels), positioning diazoxide as a pharma-
cological tool able to discriminate between sarcolemmal and mito-
chondrial KATP channels in cardiomyocytes.6a Moreover, a recent
study demonstrated the existence of a mitoKATP channel composed
of Kir6.1 and SUR1 in rat hepatocytes,7c supporting the view that
KATP channel openers interacting with the SUR1-type channel
could become interesting tools to target mitochondria.
Our laboratories have been involved for a long time in the
development of KATP channel openers structurally related to diaz-
oxide (a benzothiadiazine dioxide) and cromakalim (a dihydro-
benzopyran).8 Several original benzothiadiazines were found to
be highly active and selective for the SUR1-type KATP channel (ex-
pressed i.e. on insulin-secreting cells and neurons), while others
strongly and specifically activated the SUR2B-type KATP channel
(expressed i.e. on smooth muscle cells).8,9
Recently, we found that the presence of an (R)-1-hydroxy-2-
propylamino chain at the 3-position of 4H-1,2,4-benzothiadiazine
1,1-dioxides KATP channel openers considerably increased the
selectivity for the pancreatic-type (SUR1/Kir6.2) versus the
vascular smooth muscle-type (SUR2B/Kir6.1) channels.8l Thus,
the SUR1-selective diazoxide analogue (R)-7-chloro-3-(1-hydro-
xy-2-propyl)amino-4H-1,2,4-benzothiadiazine 1,1-dioxide8l was
chosen as the starting point for the design of new compounds tar-
geting mitochondria and mitoKATP channels of cardiomyocytes,
supposing that the SUR1 subunit would be present on the cardiac
mitoKATP channels. We decided to introduce a triphenylphospho-
nium group through an ester link on (R)-7-chloro-3-(1-hydroxy-
2-propyl)amino-4H-1,2,4-benzothiadiazine 1,1-dioxide so as to
obtain a new generation of mitoKATP channel activators or to
design prodrugs of the SUR1-selective opener (R)-7-chloro-3-(1-
hydroxy-2-propyl)amino-4H-1,2,4-benzothiadiazine 1,1-dioxide
which, after accumulation into the mitochondria, could be con-
verted, by hydrolysis of the ester link, to the active compound.
Lipophilic triphenylphosphonium cations are known to pass di-
rectly through phospholipid bilayers due to their large hydropho-
bic surface area lowering the activation energy for uptake.10 The
inherent positive charge causes these cations to selectively accu-
mulate several hundred folds within mitochondria, driven by the
plasma and mitochondrial membrane potentials.10a,b
The synthetic pathway giving access to (R)-7-chloro-3-(1-hy-
droxy-2-propyl)amino-4H-1,2,4-benzothiadiazine 1,1-dioxide (5)
is described in Scheme 1. The key intermediate for the synthesis
of this compound was the previously reported 7-chloro-3-meth-
ylsulfanyl-4H-1,2,4-benzothiadiazine 1,1-dioxide 4.8c This inter-
mediate was obtained from the corresponding aniline 1 in three
steps (Scheme 1). The first step allowed ring closure of a chloro-
sulfonylurea intermediate through FriedelꢀCrafts conditions. The
ring-closed sulfonylurea 2 obtained was then converted into its
sulfonylthiourea analogue 3 by the action of phosphorus penta-
sulfide in pyridine. In the next step, 7-chloro-3-thioxo-4H-1,2,
Scheme 1. Reagents and conditions: (a) ClSO2NCO, AlCl3, CH3NO2 (52%); (b) P2S5,
pyridine (85%); (c) CH3I, NaHCO3, CH3OH/H2O (80%); (d) (R)-1-hydroxy-2-propyl-
amine, 150 °C, 5 h (71%).
The introduction of a phosphonium group on compound 5 was
realized by esterification between the hydroxy function of 5 and
the acid function of triphenylphosphonium fragments. The differ-
ent phosphonium salts were easily synthesized via alkylation
(quaternization) of triphenylphosphine with the corresponding
commercially available haloalkylcarboxylic acids in toluene,
dichloromethane or without solvent, except for compound 9
(Scheme 2).11 Conditions were dependent on starting material.
The esterification reaction between the precursor 5 and tri-
phenylphosphonium fragments was not very easy (Scheme 3). In
classical acid catalysis conditions, the expected products were
formed in very few yields. Then, various coupling agents such as
DCC/NHS or DCC/HOBt were tested but without better results.
Esterification was finally realized with the coupling agent
N,N0-dicyclohexylcarbodiimide (DCC) alone to activate the carbox-
ylic acid group of compounds 12–16. Five different compounds
from the five phosphonium salts, with variable carbon chains
(n = 1, 2, 3), were obtained (compounds 17ꢀ21). Three of them
(the chloride salts 17, 18 and 19) were selected for biological
evaluation. The final compounds were characterized by RMN 1H,
13C and elemental analysis.
The first in vitro model used to characterize the phosphonium
salts determined the ability of the compounds at inhibiting the glu-
cose-induced insulin secretion from rat pancreatic islets, a SUR1-
expressing tissue. Data collected with this model were expressed
as the percentage of residual insulin release recorded at a 10 lM
drug concentration. Results obtained with the new compounds
were compared to previously reported data obtained with
(R)-7-chloro-3-(1-hydroxy-2-propyl)amino-4H-1,2,4-benzothiadi-
azine 1,1-dioxide 5 and diazoxide (Table 1).
4-benzothiadiazine 1,1-dioxide
3 was alkylated with methyl
iodide to give the corresponding 3-methylsulfanyl-substituted
key intermediate 4. The product 4 was then heated during five
hours with optically pure (R)-1-hydroxy-2-propylamine at 150 °C,
leading to the expected R stereoisomer 5 with a global yield of
25% on five steps.
Scheme 2. Reagents and conditions: (a) PPh3, CH2Cl2, rt, 18 h (97%); (b) PPh3,
toluene, 110 °C, 18 h (93%); (c) PPh3, toluene, 180 °C, 3 h (78%); (d) various
conditions tested but without success; (e) PPh3, toluene, 120 °C, 4 h (93%); (f) PPh3,
toluene, 110 °C, 48 h (64%).