Arylsulfonamidothiazoles as a new class of potential antidiabetic drugs. Discovery of potent and selective inhibitors of the 11β-hydroxysteroid dehydrogenase type 1

Article abstract of DOI:10.1021/jm025530f

Novel antidiabetic arylsulfonamidothiazoles are presented that exert action through selective inhibition of the 11β-hydroxysteroid dehydrogenase type 1 (11β-HSD1) enzyme, thereby attenuating hepatic gluconeogenesis. The diethylamide derivative 2a was shown to potently inhibit human 11β-HSD1 (IC₅₀ = 52 nM), whereas the N-methylpiperazinamide analogue 2b only inhibited murine 11β-HSD1 (IC₅₀ = 96 nM). Both compounds showed >200-fold selectivity over human and murine 11β-HSD2. 2b was subsequently shown to reduce glucose levels in diabetic KKA^y mice, substantiating the 11β-HSD1 enzyme as a target for the treatment of type 2 diabetes.

Full text of DOI:10.1021/jm025530f

© Copyright 2002 by the American Chemical Society
Volume 45, Number 18
August 29, 2002
Letters
Ar ylsu lfon a m id oth ia zoles a s a New Cla ss
of P oten tia l An tid ia betic Dr u gs.
Discover y of P oten t a n d Selective
In h ibitor s of th e 11â-Hyd r oxyster oid
Deh yd r ogen a se Typ e 1
Tjeerd Barf,*,† J erk Vallgårda,‡ Rikard Emond,†
Charlotta Ha¨ggstro¨m,† Guido Kurz,† Alf Nygren,|
Vivienne Larwood, Erifili Mosialou,# Kent Axelsson,#
Rolf Olsson,# Lars Engblom,∇ Naimie Edling,∇
Yuko Ro¨nquist-Nii,∞ Birgitta O¨ hman,∇
F igu r e 1. Interconversion of cortisone and cortisol by 11â-
HSD types 1 and 2 enzymes.
generation by key enzymes. Cortisol and the inactive
counterpart cortisone may be interconverted by 11â-
hydroxysteroid dehydrogenases (in rodents corticoster-
one and 11-dehydrocorticosterone) (Figure 1).1 The type
1 isoform of 11â-hydroxysteroid dehydrogenase (11â-
HSD1) is a reductase and highly expressed in the liver.
Gluconeogenesis in this organ is reduced when the 11â-
HSD1 gene is knocked out, resulting in lower fasting
blood glucose levels.2 Recent data suggest that improved
insulin release might contribute to resistance to hyper-
glycemia in 11â-HSD1 -/- mice.3
Peteris Alberts,∇ and Lars Abrahmse´n∇
Departments of Medicinal Chemistry,
Chemical Technologies, Structural Chemistry,
Assay Development and Screening, Preclinical R&D,
and Biology, Biovitrum, Box 6443, SE-751 37,
Uppsala, Sweden, and Computational Chemistry, BioFocus,
130 Abbott Drive, Sittingbourne Research Centre,
Sittingbourne, Kent ME9 8AZ, U.K.
Received April 16, 2002
Abstr a ct: Novel antidiabetic arylsulfonamidothiazoles are
presented that exert action through selective inhibition of the
11â-hydroxysteroid dehydrogenase type 1 (11â-HSD1) enzyme,
thereby attenuating hepatic gluconeogenesis. The diethyla-
mide derivative 2a was shown to potently inhibit human 11â-
HSD1 (IC50 ) 52 nM), whereas the N-methylpiperazinamide
analogue 2b only inhibited murine 11â-HSD1 (IC50 ) 96 nM).
Both compounds showed >200-fold selectivity over human and
murine 11â-HSD2. 2b was subsequently shown to reduce
glucose levels in diabetic KKAy mice, substantiating the 11â-
HSD1 enzyme as a target for the treatment of type 2 diabetes.
The second isoform, 11â-HSD2, inactivates cortisol by
converting it to cortisone. Disruption or mutations in
the 11â-HSD2 gene result in sodium retention, hy-
pokalemia, and hypertension because of inappropriate
glucocorticoid occupation of the mineralocorticoid recep-
tor in the kidney.4
Various compounds have previously been shown to
inhibit 11â-HSD1, but none of these were reported to
be 11â-HSD1 selective, potent, and druglike.5,6 For
example, the endogenous steroid chenodeoxycholic acid
is selective but not potent (Ki in the micromolar range)
and might not be considered druglike.5 Glycyrrhetinic
acid and the hemisuccinyl derivative carbenoxolone, the
only known potent inhibitors to date, display poor
selectivity between the two isoforms of 11â-HSD.5,6
Carbenoxolone has earlier been demonstrated to im-
prove insulin sensitivity and to decrease glucose produc-
tion in healthy human volunteers.7 In that study, the
11â-HSD2 inhibitory activity of carbenoxolone was a
limiting factor because it induces renal mineralocorti-
coid excess at higher doses. Thus, while inhibition of
Excessive levels of glucocorticoids can cause metabolic
complications. Glucocorticoid action is dependent on
circulating hormone levels and local activation or re-
* To whom correspondence should be addressed. Phone: +46 8
6973885. Fax: +46 8 6973914. E-mail: tjeerd.barf@biovitrum.com.
† Department of Medicinal Chemistry, Biovitrum.
‡ Department of Structural Chemistry, Biovitrum.
| Department of Chemical Technology, Biovitrum.
BioFocus.
#
Department of Assay Development & Screening, Biovitrum.
∇ Department of Biology, Biovitrum.
∞ Department of Preclinical R&D, Biovitrum.
10.1021/jm025530f CCC: $22.00 © 2002 American Chemical Society
Published on Web 08/03/2002

3814 J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 18
Sch em e 1. Synthesis of Arylsulfonamidothiazolesa
Letters
Ta ble 1. In Vitro Inhibitory Activities for the 11â-HSD1 and
11â-HSD2 Enzymes
IC50 (nM) ( SEMa
mouse
human
human
compound
11â-HSD1
11â-HSD1
11â-HSD2
a
(a) 3-Chloro-2-methylphenylsulfonyl chloride, pyridine; (b)
carbenoxolone
2a
2b
108 ( 6
284 ( 24
96 ( 14
330 ( 16
52 ( 3
3341 ( 412
83 ( 5b
>10000
>10000
AlCl3, N,N-diethylamine, DCM; (c) (i) KOH, EtOH, (ii) N-meth-
ylpiperazine, EDCI, HOBT, TEA, DCM.
the 11â-HSD1 enzyme could provide a very attractive
and novel treatment of type 2 diabetes, it is obvious that
an inhibitor against 11â-HSD1 must not affect 11â-
HSD2. This communication describes the discovery of
the first potent and selective druglike inhibitors of 11â-
HSD1.
a
IC50 values ((SEM) for inhibition of the 11â-HSD1 catalyzed
formation of [3H]-cortisol from cortisone or vice versa for 11â-
HSD2. Values were obtained with 11 inhibitor concentrations from
three to seven independent determinations, each performed in
b
triplicate. Value was obtained with 11 inhibitor concentrations
from two independent determinations, each performed in tripli-
cate.
A high-throughput screen of our compound collection
followed by a lead optimization process provided us with
arylsulfonamidothiazole 2a and 2b as examples of this
structural class (Scheme 1). Compounds of interest were
prepared as follows. Ethyl (2-aminothiazol-4-yl)acetate
was sulfonylated with 3-chloro-2-methylbenzenesulfonyl
chloride in pyridine in 79% yield. The resulting ester
intermediate can directly be treated with N,N-diethy-
lamine in the presence of aluminum chloride in dichlo-
romethane, giving compound 2a in 68% yield. Hydrol-
ysis of the ethyl ester 1 in ethanol using potassium
hydroxide gave the carboxylic acid in a high yield.
Treatment of the intermediate acid with N-methylpip-
erazine using 1-hydroxybenzotriazole and 1-[3-(dim-
ethylamino)propyl]-3-ethylcarbodiimide hydrochloride
as amide coupling reagents afforded compound 2b as
the hydrochloride salt in a moderate 48% yield.
To evaluate the ability of the compounds to inhibit
human or mouse 11â-HSD1, they were assayed in a
scintillation proximity assay (SPA). Recombinant hu-
man and murine 11â-HSD1 were produced in Pichia
pastoris6,8 and treated with a substrate/cofactor mixture
of [3H]-cortisone/NADPH and different concentrations
of inhibitor. The conversion of cortisone to cortisol was
stopped by addition of glycyrrhetinic acid. The labeled
cortisol product was captured by mouse monoclonal
anticortisol antibodies (6D6.7) and was subsequently
bound to SPA beads coated with antimouse antibodies.
The amount of [3H]-cortisol bound to the beads was
determined in a scintillation counter.
Assessment of human 11â-HSD2 activity was based
on the conversion of [3H]-cortisol to the tritium-labeled
product (cortisone) in the presence of inhibitor. The
enzymatic reaction was performed with a surplus of the
cofactor NAD+ and stopped with perchloric acid. Sub-
strate and product were separated by HPLC and
monitored using a flow scintillation counter. Enzyme
activity was quantified as the percentage area of the
product peak compared to the total area.
Since the mouse and human 11â-HSD1 enzymes
display only 79% amino acid identity, we were not
surprised that the compounds did not inhibit 11â-HSD1
from both species to a similar degree (Table 1). Appar-
ently, the arylsulfonamidothiazole moiety can be ac-
commodated in both 11â-HSD1 orthologues while the
right-hand side of the molecule determines the extent
of inhibition of either murine or human 11â-HSD1. The
more flexible diethylamide group of 2a is well accom-
modated in the human enzyme, whereas the N-meth-
ylpiperazinamide group of 2b is not at all tolerated
because of steric reasons or maybe its basic character.
In contrast, this group is well accepted in the mouse
11â-HSD1 enzyme. Inhibition of both enzymes was
evaluated with cortisone as the substrate, which is the
natural substrate for the human enzyme.
Importantly, the arylsulfonamidothiazole derivatives
exhibited no appreciable inhibitory activity against
human 11â-HSD2, and 2b showed no inhibition of the
mouse 11â-HSD2 enzyme at 0.2 mM concentration (data
not shown). Thus, 2a and 2b are more than 200-fold
selective for the human and mouse 11â-HSD1 enzymes,
respectively. Carbenoxolone was less effective than 2a
in inhibiting human 11â-HSD1 and was shown to
potently inhibit the human 11â-HSD2 enzyme with an
IC50 value of 83 nM. A calculation using the published
Km value of 25.5 nM yields a Ki of 11 nM, well in line
with earlier reported Ki values.5,9 Unlike the arylsul-
fonamidothiazole derivatives, carbenoxolone displays
negligible ability to discriminate between the mouse and
human 11â-HSD1 enzymes.
The favorable pharmacokinetic and in vitro selectivity
profile of compound 2b allowed us to assess the proposed
importance of the 11â-HSD1 enzyme for the treatment
of diabetes in the hyperglycemic KKAy mouse model.
Pharmacokinetic experiments with compound 2b at
doses of 10-100 mg/kg yielded a serum terminal half-
life in the range 2.5-3.5 h and an oral bioavailability
of approximately 21% in the mouse.
Indeed, we found that compound 2b (BVT.2733; 25,
50, 100 mg/kg po) administered twice daily significantly
lowered blood glucose levels in a dose-dependent man-
ner in KKAy mice (n ) 12-13), as measured on days
3-11 (Figure 2). The maximal reduction in glucose was
53% of the control after 11 days of treatment at the
highest dose. Serum compound levels 12 h after the last
administration were 0.06 ( 0.02, 0.13 ( 0.02, and 0.19
( 0.04 µM, respectively. The plasma-free fraction of 2b
in serum amounted to 5%, indicating that at the highest
doses used, enough unbound compound is available to

Letters
J ournal of Medicinal Chemistry, 2002, Vol. 45, No. 18 3815
the in vitro inhibition assays and in vivo pharmacology. This
material is available free of charge via the Internet at http://
pubs.acs.org.
Refer en ces
(1) For a review, see the following. (a) Seckl, J . R.; Walker, B. R.
Minireview: 11â-Hydroxysteroid dehydrogenase type 1sA tis-
sue-specific amplifier of glucocorticoid action. Endocrinology
2001, 142, 1371-1376. (b) Kotelevtsev, Y.; Seckl, J . R.; Mullins,
J . J . 11â-Hydroxysteroid dehydrogenases: key modulators of
glucocorticoid action in vivo. Curr. Opin. Endocrinol. Diabetes
1999, 6, 191-198. (c) Stewart, P. M.; Krozowski, Z. S. 11â-
Hydroxysteroid dehydrogenase. Vitam. Horm. 1999, 57, 249-
324.
(2) Kotelevtsev, Y.; Holmes, M. C.; Burchill, A.; Houston, P. M.;
Schmoll, D.; J amieson, P.; Best, R.; Brown, R.; Edwards, C. R.
W.; Seckl, J . R.; Mullins, J . J . 11-â-Hydroxysteroid dehydroge-
nase type 1 knockout mice show attenuated glucocorticoid-
inducible responses and resist hyperglycemia on obesity or
stress. Proc. Natl. Acad. Sci. U.S.A. 1997, 94, 14924-14929.
(3) Davani, B.; Khan, A.; Mårtensson, E.; Efendic, S.; J o¨rnvall, H.;
Oppermann, U. C. T. Type I 11â-hydroxysteroid dehydrogenase
mediates glucocorticoid activation and insulin release in pan-
creatic islets. J . Biol. Chem. 2000, 275, 34841-34844.
(4) (a) Kotelevtsev, Y. V.; Brown, R. W.; Fleming, S.; Kenyon, C.;
Edwards, C. R. W.; Seckl, J . R.; Mullins, J . J . Hypertension in
mice caused by inactivation of 11â-hydroxysteroid dehydroge-
nase type 2. J . Clin. Invest. 1999, 103, 683-689. (b) Dave-
Sharma, S.; Wilson, R. C.; Harbison, M. D.; Newfield, R.; Azar,
M. R.; Krozowski, Z. S.; Funder, J . W.; Shackleton, C. L.;
Bradlow, H. L.; Wei, J .-Q.; Hertecant, J .; Moran, A.; Neiberger,
R. E.; Balfe, J . W.; Fattah, A.; Daneman, D.; Akkurt, H. I.; de
Santis, C.; New, M. I. Examination of genotype and phenotype
relationships in 14 patients with apparent mineralcorticoid
excess. J . Clin. Endocrinol. Metab. 1998, 83, 2244-2254.
(5) Diederich, S.; Grossmann, C.; Hanke, B.; Quinkler, M.; Her-
rmann, M.; Ba¨hr, V.; Oelkers, W. In the search for specific
inhibitors of human 11â-hydroxysteroid dehydrogenases (11â-
HSDs): chenodeoxycholic acid selectively inhibits 11â-HSD-I.
Eur. J . Endocrinol. 2000, 142, 200-207.
(6) Hult, M.; J o¨rnvall, H.; Oppermann, U. C. T. Selective inhibition
of human type 1 11â-hydroxysteroid dehydrogenase by synthetic
steroids and xenobiotics. FEBS Lett. 1998, 441, 25-28.
(7) Walker, B. R.; Connacher, A. A.; Lindsay, M.; Webb, D. J .;
Edwards, C. R. W. Carbenoxolone increases hepatic insulin
sensitivity in man: a novel role for 11-oxosteroid reductase in
enhancing glucocorticoid receptor activation. J . Clin. Endocrinol.
Metab. 1995, 80, 3155-3159.
(8) Nobel, C. S. I.; Dunås, F.; Abrahmse´n, L. A. Purification of full-
length recombinant human an rat type 1 11â-hydroxysteroid
dehydrogenases with retained oxidoreductase activity. Protein
Expression Purif., in press.
F igu r e 2. Glucose-lowering effect of 2b in KKAy mice.
exert the pharmacological effect observed. Compound
2b did not alter liver or heart marker enzymes (AST,
ALT, ALP), suggesting that liver impairment can be
excluded in the present experiment. From evidence
generated in a different experiment, we inferred that
reduced hepatic glucose output contributes to the glucose-
lowering effect obtained here. A 1-week treatment of
KKAy mice with 2b (167 mg/kg/day) using osmotic
minipumps lowered mRNA levels encoding for two key
enzymes in hepatic glucose production: phosphoe-
nolpyruvate carboxykinase (75% of control; P < 0.05)
and glucose-6-phosphatase (55% of control; P ) 0.089).10
Our data support the literature finding that 11â-
HSD1 is a key mediator of gluconeogenesis.2 The results
presented in this paper strengthen 11â-HSD1 inhibition
as an interesting approach for type 2 diabetes treat-
ment, and pharmacological tools have been developed
to explore the potential of this concept further. Develop-
ment of compound 2a , a selective inhibitor of the human
11â-HSD1 enzyme, is the first step toward the clinical
evaluation of this novel compound class aimed to benefit
type 2 diabetic patients without the risk of hypertension.
(9) Diederich, S.; Hanke, B.; Ba¨hr, V.; Oelkers, W. The metabolism
of 9R-fluorinated steroids in the human kidney. Endocr. Res.
1996, 22, 803-810.
(10) Alberts, P.; Engblom, L.; Edling, N.; Forsgren, M.; Klingstro¨m,
G.; Larsson, C.; Ro¨nquist-Nii, Y.; O¨ hman, B.; Abrahmse´n, L. The
selective 11â-hydroxysteroid dehydrogenase type 1 (11â-HSD1)
inhibitor BVT.2733 lowers blood glucose levels in hyperglycemic
KKAy mice. ADA 62nd Scientific Sessions, San Francisco, CA,
J une 14-18, 2002; Diabetes 51: Abstract 172-OR Suppl. 2 J UN
2002.
Ack n ow led gm en t. The authors thank Hanna Ax-
elsson for producing inhibition data, and Bjo¨rn Norrlind
and Andreas Hugerth for the skillful formulation of
compounds for the in vivo experiment.
Su p p or tin g In for m a tion Ava ila ble: Experimental sec-
tion containing preparation of 2a and 2b and description of
J M025530F