Thiourea-based gemfibrozil analogues as HDL-elevating agents.

Article abstract of DOI:10.1016/S0960-894X(02)00435-3

A series of gemfibrozil analogues with a thiourea moiety embedded in the side chain was prepared and evaluated as HDL-elevating agents. Derivatives 8b, 9b, 9c, and 9d were found to be approximately as effective as gemfibrozil (1) for HDL cholesterol eleva

Full text of DOI:10.1016/S0960-894X(02)00435-3

Bioorganic & Medicinal Chemistry Letters 12 (2002) 2439–2442
Thiourea-Based Gemfibrozil Analogues as HDL-Elevating Agents
Gary M. Coppola,a,* Robert E. Damon,a J. Bruce Eskesen,a
Dennis S. Francea and James R. Paterniti, Jr.b
aDepartment of Metabolic and Cardiovascular Diseases, Novartis Institute for Biomedical Research, 556 Morris Avenue,
Summit, NJ 07901, USA
bAmylin Pharmaceuticals, 9373 Towne Centre Drive, San Diego, CA 92121, USA
Received 14 January 2002; accepted 13 May 2002
Abstract—A series of gemfibrozil analogues with a thiourea moiety embedded in the side chain was prepared and evaluated
as HDL-elevating agents. Derivatives 8b, 9b, 9c, and 9d were found to be approximately as effective as gemfibrozil (1) for HDL
cholesterol elevation. # 2002 Elsevier Science Ltd. All rights reserved.
Cardiovascular disease remains the leading cause of
death in the United States and many other indus-
trialized countries.1 Atherosclerosis, a progressive dis-
ease involving metabolic abnormalities in both
lipoprotein and artery wall metabolism, underlies much
of the total cardiovascular disease burden.2 Major risk
factors for atherosclerotic cardiovascular disease are
well-known and include such dyslipidemias as elevated
low-density lipoprotein (LDL) cholesterol, low levels of
high-density lipoprotein (HDL) cholesterol, and high
levels of triglycerides. The Scandinavian Simvastatin
Survival Study3 and the CARE trial4 clearly established
that lowering LDL cholesterol is associated with a
reduction in total and cardiovascular mortality and
nonfatal cardiac events.
to treatment with gemfibrozil (1), a member of the
fibrate class of drugs. The beneficial effect was asso-
ciated with a drug-related increase in HDL levels.
Additionally, the Veterans Administration HDL Inter-
vention Trial (VA-HIT) demonstrated that gemfibrozil
therapy in men with coronary heart disease and low
HDL cholesterol levels raised HDL modestly and
reduced coronary events by 22%, despite not lowering
LDL cholesterol levels.8 Triglyceride levels were also
lowered in this study.
In contrast to the well-established pathways for LDL
production and LDL receptor-mediated clearance, the
molecular mechanisms involved in overall HDL meta-
bolism have not been completely elucidated, and phar-
macological approaches aimed at selectively modifying
HDL levels in patients are not currently available. In
response to this clear medical need, we embarked on a
program to identify agents that would elevate HDL
cholesterol.
Extensive epidemiological evidence supports the con-
cept that HDL is a protective factor against athero-
sclerosis and that a low plasma HDL level is associated
with a high risk of disease.5 Several recent studies sug-
gest the potential importance of pharmacological inter-
vention to raise HDL levels. A proportional hazards
model analysis of 12-year death rates of Framingham
men indicated that a 10 mg/dL increase in HDL cho-
lesterol was associated with a 19% decrease in coronary
artery disease death and a 12% decrease in all-cause
mortality.6 The Helsinki heart study,7 a primary pre-
vention trial in dyslipidemic men, demonstrated a
reduced incidence of cardiovascular events in response
A systematic analysis of our in-house chemical database
resulted in the identification of thiourea 2 as a com-
pound that was found to raise serum levels of HDL
cholesterol and apolipoprotein A-I (Apo A-I) in rats fed
a cholic acid and cholesterol-supplemented diet.9
Upon investigation of structures 1 and 2 one can envi-
sion certain similarities between the two. Two such
similarities are the substitution pattern of the aromatic
ring and the protonated oxygen atom at the terminus of
both chains. We reasoned that combining the features
common to both gemfibrozil and 2 into a new molecule
*Corresponding author. Fax: +1-908-277-2405; e-mail: gary.
coppola@pharma.novartis.com
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00435-3

2440
G. M. Coppola et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2439–2442
Retrosynthetic analysis of generalized target structure 8
revealed that the skeletal framework of the molecule can
be assembled convergently at the thiourea functionality.
This can readily be accomplished by reacting an appro-
priately substituted aryl isothiocyanate 7 with an a-
substituted-b-amino acid derivative. The flexibility of
this synthesis would also allow the preparation of the
corresponding urea derivatives by simply using an iso-
cyanate in place of the isothiocyanate (Scheme 1).
The required 2,5-disubstituted phenyl isothiocyanates
were commercially available. However, the b-amino
acid reactants needed to be synthesized. These were
readily prepared by amino methylation of ketene silyl
acetals 5 with N,N-bis(trimethylsilyl)methoxymethyl-
amine (4)12 catalyzed with trimethylsilyltrifluoro-
methanesulfonate.13 The resulting N,N-bis(trimethyl-
silyl)-b-amino acid esters 6 were allowed to react
directly with the aryl isothiocyanate (or isocyanate) 7 in
methanol at room temperature. Under these conditions,
desilylation of 6 occurs slowly and the resulting free
amine reacts with 7 to provide thiourea 8 (or urea) in
good yields (Table 1). Hydrolysis of the ester with 1 N
sodium hydroxide gives the desired acid 9.
would produce a ‘hybrid’ structure 3. Incorporation of
the thiourea function into the chain portion of the
molecule would also partially rigidify it thereby redu-
cing the number of possible conformations the side
chain can adopt. Since thioureas are known to raise
HDL cholesterol10,11 the inclusion of this function into
1 should be compatible.
This paper describes the synthesis of a series of analo-
gues based on structure 3 and reports on the biological
activity of these derivatives as HDL-elevating agents.
The primary amide 13 was prepared by an alternate
route (Scheme 2). Reaction of the thiourea 10 with 3-
chloropivaloyl chloride (11) under phase transfer con-
ditions14 produces the b-lactam 12 in 45% yield. Ring
opening of the b-lactam heterocycle with lithium amide
affords the desired amide 13 in 35% yield along with
24% of the corresponding acid 9b.
In an effort to further conformationally restrict the
chain in compound 9, we designed a molecule where one
of the methyl groups adjacent to the carboxylate is
tethered to the closest thiourea nitrogen atom. The
resulting cyclic analogue 19 would have the carboxylic
acid group in a fixed location while still retaining the a-
dialkylated motif.
The synthesis of 19 was accomplished in a convergent
manner (Scheme 3). The requisite a-methyl nipecotate
17 was prepared from commercially available ethyl
nipecotate (14) by protecting the nitrogen with a Boc
group (98% yield), generating the enolate of ester 15 with
LDA in THF at ꢀ78 ꢁC, alkylation of the enolate with
1.1 equivof methyl iodide (90% yield), then deprotection
Scheme 1. General preparation of arylthiourea carboxylic acid
derivatives.
Table 1. Physical data for esters 8 and acids 9
Compd
R1
R2
R
X
8
Yield (%)
8
9
Yield (%)
9
mp (ꢁC)b
mp (ꢁC)b
a
b
c
d
e
f
Me
Cl
Cl
Cl
Cl
Cl
Cl
Me
Me
Me
Me
Me
OMe
Me
Me
Me
–(CH2)3–
–(CH2)4–
–(CH2)5–
Me
S
S
S
S
S
S
O
66
82
62a
44
70
95
51
80–83
95–98
74–77
108–111
150–152
81–84
75
85
89
90
94
86
79
174–176
164–167
156–159
151–154
164–167
138–141
190–191
g
Me
138–140
aEthyl ester.
bSatisfactory elemental analysis obtained for all compounds.

G. M. Coppola et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2439–2442
2441
of 16 with trifluoroacetic acid (85% yield). Reaction of
17 with 7 in methylene chloride furnishes 18 in 77%
yield. Hydrolysis of the ester with 1 N sodium hydroxide
in ethanol gives the desired product 19 in 51% yield.
for 7 days. The compounds were then administered in
the diet to groups of six rats for 8 days. At termination
the rats were anesthetized with carbon dioxide and
sacrificed by exsanguination from the corotid artery.
The blood was centrifuged and the serum separated.
Male Sprague–Dawley rats, weighing 200–225 g, were
housed two per cage and fed Purina rodent chow (spe-
cial mix 5001-S) which was supplemented with 0.25%
cholic acid and 0.75% cholesterol, and water ad libitum
The lipoproteins were separated by classical ultra-
centrifugation techniques.9,15 Serum was adjusted to a
density of 1.060 g/mL with sodium chloride solution
and 175 mL of adjusted sera were centrifuged at 42,000
rpm in a Beckman Ti42.2 rotor at 20 ꢁC for 2.5 h. Lipid
and apoprotein analyses were performed on fractions
removed from the top of each tube and on the fraction
remaining on the bottom. Cholesterol was quantitated
by enzymatic colorimetric methods. Serum Apo A-I was
determined by SDS polyacrylamide gel electrophor-
esis16,17 or a rat competition ELISA.18
Within the series of analogues in Table 2, 9a, which has
the same aromatic substitution pattern as gemfibrozil,
raised HDL only about half that of 1, retained Apo A-I
activity, but significantly raised triglycerides. Com-
pound 9b was nearly as potent as 1 with regard to HDL,
raised Apo A-I almost 1.5 times that of 1, and modestly
lowered triglycerides. Ester 8b exhibited roughly the
same profile as the acid 9b, which suggests that the ester
is metabolized to the active analogue. Amide derivative
13 led to reduced HDL elevation levels while retaining
Apo A-I levels, whereas cyclobutyl and cyclopentyl
analogues 9c and 9d retained HDL levels while reducing
Apo A-I levels. Conformationally restrained derivative
19, where one methyl is tethered to the thiourea nitro-
gen, was essentially inactive. The importance of sulfur is
illustrated by the lack of activity of the urea analogue 9g
in comparison to the corresponding thiourea 9b. Of the
most active compounds, 8b and 9b showed no elevation
of total cholesterol levels.
Scheme 2. Preparation of carboxamide derivative.
The liver enzymes ALT, AST, and ALP were measured
for all thiourea and urea analogues. Both AST and ALP
remained at normal levels relative to untreated controls.
ALT levels were elevated 60–127% for all compounds
except 2, 9a, 9b, and 9g which remained normal. None
Scheme 3. Preparation of conformationally restrained derivative.
Table 2. In vivo biological results
Compd
Relative to gemfibrozila
Apo A-I
Total Cholesterol
(% change)
Triglycerides
(% change)
Dose
(mg/kg/day)
HDL
1
2
1
1
1.06c
1.81
ꢀ23
40
ꢀ43b
ꢀ56c
46b
52c
74
78
75b
76
67
88
77
73
41
81
0.61c
0.93
0.54
0.85b
1.19
1.09
0.73
0.62
0.36
0.57
0.33
8b
9a
9b
9c
9d
9e
9f
9g
13
19
ꢀ12 (NS)
ꢀ12 (NS)
69
0.93
73
1.44b
0.49
15(NS)
36
41
21
9 (NS)
74
ꢀ19b (NS)
ꢀ39 (NS)
ꢀ61
0.55
0.32
0.27 (NS)
0.32 (NS)
1.15
ꢀ54
ꢀ28 (NS)
ꢀ21 (NS)
ꢀ49
ꢀ3 (NS)
ꢀ0.16 (NS)
14 (NS)
15 (NS)
aGemfibrozil=1 (typical results for gemfibrozil: HDL cholesterol "113%; apo A-I "49%; triglycerides #43% relative to an untreated control group
at about 46 mg/kg/day which is the maximally effective dose for this compound in this model).
bMean value of three studies.
cMean value of 26 studies. Values are significantly different (P<0.05) relative to untreated controls unless otherwise indicated (NS).

2442
G. M. Coppola et al. / Bioorg. Med. Chem. Lett. 12 (2002) 2439–2442
of the gemfibrozil analogues had any statistically sig-
nificant effect on body weight gain.
Norola, S.; Pasternack, A.; Pikkarainen, J.; Romo, M.; Sjo-
blom, T.; Nikkila, E. A. N. Engl. J. Med. 1987, 317, 1237.
8. Rubins, H. B.; Robins, S. J.; Collins, D.; Fye, C. L.;
Anderson, J. W.; Elam, M. B.; Faas, F. H.; Linares, E.;
Schaefer, E. J.; Schectman, G.; Wilt, T. J.; Wittes, J. N. Engl.
J. Med. 1999, 341, 410.
9. France, D. S.; Quinby, R. E.; Babiak, J.; Lapen, D. C.;
Paterniti, J. R.; Weinstein, D. B. Lab. Rob. Autom. 1990, 2, 155.
10. Coppola, G. M.; Damon, R. E. Canadian Patent Appli-
cation CA 2072704 1993. Chem. Abstr. 1993, 119, 95134.
11. Failli, A. A.; Shumsky, J. S. Internat. Patent WO 9911621.
Chem. Abstr. 1999, 130, 209605.
In conclusion, we have shown that a thiourea moiety
can be incorporated into the gemfibrozil skeleton to
produce derivatives which still exhibit varying degrees
of HDL-elevating activity. Compounds 8b, 9b, 9c, and
9d were as potent as 1 while 8b and 9b raised Apo A-1
levels significantly higher than 1.
References and Notes
12. Morimoto, T.; Takahashi, T.; Sekiya, M. J. Chem. Soc.,
Chem. Commun. 1984, 794.
13. Okano, K.; Morimoto, T.; Sekiya, M. J. Chem. Soc.,
Chem. Commun. 1984, 883.
14. Okawara, T.; Nakayama, K.; Furukawa, M. Chem. Lett.
1982, 1791.
15. Bronzert, T. J.; Brewer, Jr., H. B. Clin. Chem. 1977, 23,
2089.
16. Hughes, T. E.; Sasak, W. V.; Ordovas, J. M.; Forte, T. M.;
Lamon-Fava, S.; Schaefer, E. J. J. Biol. Chem. 1987, 262,
3762.
17. France, D. S.; Hughes, T. E.; Miserendino, R.; Spirito,
J. A.; Babiak, J.; Eskesen, J. B.; Tapparelli, C.; Paterniti, Jr.,
J. R. J. Lipid Res. 1989, 30, 1997.
18. Burkey, B.; France, D. S.; Wang, H.; Ma, X.; Brand, B.;
Abuhani, C.; Diffederfer, M.; Marsh, J.; Paterniti, Jr., J. R.;
Fisher, E. J. Lipid Res. 1995, 36, 1463.
1. Lacko, A. G. Trends Cardiovascular Med. 1994, 4, 84.
2. Gordon, T.; Castelli, W. P.; Hjortland, M. C.; Kannel,
W. B.; Dawber, T. R. Am. J. Med. 1977, 62, 704.
3. Scandinavian Simvastatin Survival Study Group. Lancet
1994, 344, 1383.
4. Sacks, F. M.; Moye, L. A.; Davis, B. R.; Cole, T. G.;
Roleau, J. L.; Nash, D.T; Pfeffer, M. A.; Braunwald, E. Cir-
culation 1998, 97, 1446.
5. Gordon, D. J.; Rifkind, B. M. N. Engl. J. Med. 1989, 321,
1311.
6. Gordon, T.; Castelli, W. P.; Hjortland, M. C.; Kannel,
W. B.; Dawber, T. R. Am. J. Med. 1977, 62, 704.
7. Frick, M. H.; Elo, O.; Kaapa, K.; Heinonen, O. P.; Hein-
salmi, P.; Helo, P.; Huttunen, J. K.; Kaitaniemi, P.; Koskinen,
P.; Manninen, V.; Maenpaa, H.; Malkonen, M.; Manttari, M.;