Antihyperglycemic activities of cryptolepine analogues: An ethnobotanical lead structure isolated from Cryptolepis sanguinolenta

Article abstract of DOI:10.1021/jm970735n

Cryptolepine (1) is a rare example of a natural product whose synthesis was reported prior to its isolation from nature. In the previous paper we reported the discovery of cryptolepine's antihyperglycemic properties. As part of a medicinal chemistry progr

Full text of DOI:10.1021/jm970735n

2754
J . Med. Chem. 1998, 41, 2754-2764
An tih yp er glycem ic Activities of Cr yp tolep in e An a logu es: An Eth n obota n ica l
Lea d Str u ctu r e Isola ted fr om Cr yptolepis sa n gu in olen ta †
Donald E. Bierer,* Larisa G. Dubenko, Pingsheng Zhang, Qing Lu, Patricia A. Imbach, Albert W. Garofalo,
Puay-Wah Phuan, Diana M. Fort, J oane Litvak, R. Eric Gerber, Barbara Sloan, J ian Luo, Raymond Cooper, and
Gerald M. Reaven
Shaman Pharmaceuticals, Inc., 213 East Grand Avenue, South San Francisco, California 94080
Received October 29, 1997
Cryptolepine (1) is a rare example of a natural product whose synthesis was reported prior to
its isolation from nature. In the previous paper we reported the discovery of cryptolepine’s
antihyperglycemic properties. As part of a medicinal chemistry program designed to optimize
natural product lead structures originating from our ethnobotanical and ethnomedical field
research, a series of substituted and heterosubstituted cryptolepine analogues was synthesized.
Antihyperglycemic activity was measured in vitro and in an NIDDM mouse model to generate
the first structure-bioactivity study about the cryptolepine nucleus.
In tr od u ction
try program designed to optimize ethnobotanically
generated leads,13,14 we considered cryptolepine as a
template for synthetic modification, recognizing that
such optimization should address the reduced food
consumption and body weight issues. A portion of this
structure optimization study on the cryptolepine nucleus
is reported below.
Cryptolepine (1), a member of the indoloquinoline
alkaloid family, is a rare example of a natural product
whose synthesis was reported prior to its isolation from
nature. Cryptolepine was first synthesized in 1906 by
Fichter and Boehringer1 for use as a possible dye while
its isolation from Cryptolepis triangularis N. E. Br was
first reported by Clinquart2 23 years later.3-8 Cryptol-
epine-containing plants have been used by indigenous
peoples of Africa as a dye,9,10 and extracts obtained from
various Cryptolepis sp. have a number of reported and
current ethnomedical uses.11 Cryptolepine and its
hydrochloride salt (2) possess numerous biological prop-
Ch em istr y
Our earlier exploration into the synthesis of cryptol-
epine led to three synthetic routes being established.8
Since these routes allowed for the preparation of
structurally diverse cryptolepine analogues, all three
synthetic strategies were pursued further. The first
route began with the condensation reaction between a
substituted indolyl acetate 3 and an isatin derivative
4, affording quindoline carboxylic acids 5. Decarboxy-
lation in diphenyl ether followed by alkylation with
methyl iodide using the Fichter conditions1 or with
methyl triflate afforded cryptolepine derivatives 7 and
8, respectively. In most cases the hydroiodide and
hydrotrifluoromethanesulfonate (hydrotriflate) salts were
converted to their hydrochloride salts prior to biological
testing using procedures described earlier8 (Scheme 1).
erties, including antimalarial activity.11 In previous
papers we reported that cryptolepine (1) and its hydro-
chloride salt (2) possess antihyperglycemic properties8
and that extracts of Cryptolepis sanguinolenta possess
antihyperglycemic activity in non-insulin-dependent
diabetes mellitus (NIDDM) rodent models and in hu-
mans.12 Yet while the aqueous extract of C. sanguino-
lenta demonstrated a remarkable ability to lower fasting
blood glucose levels in a patient study conducted in
Ghana12 and the glucose-lowering properties of cryp-
tolepine hydrochloride were shown to be separate from
an apparent anorexigenic effect,8,12 we were faced with
the prospect of separating these effects with future
cryptolepine analogues, which would require extensive
experimentation. Thus, as part of a medicinal chemis-
When alkylation of quindoline 6 with ethyl iodide in
methanol was attempted using the Fichter conditions,1
a near 1:1 mixture of the ethyl and methyl derivatives
7 (R5 ) Et and Me, respectively) were obtained, pre-
sumably due to the formation of ethyl methyl ether and
possible subsequent formation of methyl iodide, either
of which could serve as the alkylating agent. This
problem was eliminated when the solvent was changed
from methanol to chloroform. Similarly using ethanol-
free chloroform, the butyl, benzyl, and p-fluorobenzyl
derivatives 11 were obtained.
4-Methoxycryptolepine hydrochloride 11j was pre-
pared according to the route shown in Scheme 2.
Condensation of 1-acetyl-3-oxoindole8 12 with 2-nitro-
3-methoxybenzaldehyde 13 using a catalytic amount of
piperidine gave 14 as a mixture of E/Z isomers in 78-
85% yield. Hydrogenation of 14 gave quindoline 15 in
† Dedicated to Professor Henry Rapoport on the occasion of his 79th
birthday.
* Address correspondence to this author at Shaman Pharmaceuti-
cals. E-mail: dbierer@shaman.com.
S0022-2623(97)00735-8 CCC: $15.00 © 1998 American Chemical Society
Published on Web 06/25/1998

Antihyperglycemic Activities of Cryptolepine Analogues
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 15 2755
Sch em e 1a ,b
Sch em e 3a ,b
a
(a) Bromoacetyl bromide, DMF/dioxane; (b) DMF; (c) PPA; (d)
POCl3; (e) MeI, MeOH, bomb; (f) MeOTf; (g) Na2CO3; (h) basic
b
alumina, CHCl3; (i) MeOH-CHCl3; (j) HCl-Et2O. See Table 1
a
(a) KOH, H2O, N2; (b) Ph2O, 255 °C, 6 h; (c) R5X, bomb, 120
for definitions of Ra and Rb.
°C; (d) MeOTf; (e) Na2CO3; (f) basic alumina, CHCl3; (g) MeOH-
b
CHCl3; (h) HCl-Et2O. See Table 1 for definitions of Ra, Rb, and
approach to 11j involved removal of the acetate protect-
ing group prior to alkylation. Oftentimes quindoline 15
was difficult to purify, as it was contaminated with a
reduction byproduct (14 with double bond reduced). In
these instances, it became necessary to resort to HPLC
to obtain 15 free of this impurity. This problem was
solved by removing the acetate group with KOH,
providing 7j, which was easily separated from the
reduction byproduct. Alkylation with methyl triflate
and counterion exchange provided 4-methoxycryptol-
epine hydrochloride 11j.
R5.
Sch em e 2a
A series of 11-chlorocryptolepine analogues was pre-
pared according to the route depicted in Scheme 3. A
substituted anthranilic acid 16 was treated with bro-
moacetyl bromide to afford bromoacetyl derivative 17
(50-95%), which was treated with a substituted aniline
18 to provide anthranilic acid derivative 19 (50-90%).
Acid-promoted cyclization of 19 with PPA gave quin-
dolone 20, which often was isolated only as the crude
product and subsequently treated with POCl3 to provide
11-chloroquindoline 21.15-19 The yield for this two-step
transformation varied from 10 to 50% depending on the
ring substituents. Alkylation of 21 with methyl iodide
gave an inseparable mixture of the desired salt 22 and
the product from halogen exchange 23.8 For this reason,
the preferred procedure utilized methyl triflate as the
alkylation agent which gave >95% yields of the hydro-
triflate salt 24. Conversion to the free base and
subsequent treatment with HCl provided the target
cryptolepine hydrochloride salts 25.
a
(a) Piperidine (cat.), toluene/CHCl3, 4 Å MS, room tempera-
ture; (b) H2, Pd/C, MeOH; (c) KOH; (d) MeOTf, toluene, room
temperature; (e) K2CO3; (f) basic alumina, CHCl3; (g) CHCl3-
MeOH; (h) HCl.
53% yield after chromatography. Alkylation with meth-
yl iodide and subsequent chromatography over basic
alumina followed by acidification with HCl afforded the
target methoxycryptolepine analogue. A more desirable
Heterocyclic variation at position 10 was pursued,
with the indolo nitrogen atom being replaced with
oxygen, sulfur, and carbon. The requisite 11-chloroben-
zofuro-20,21 and 11-chlorobenzothienoquinolines21 31 and

2756 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 15
Bierer et al.
Sch em e 4a ,b
Ta ble 1. Cryptolepine Analogues
compd
Ra
R5
Rb
R11
Z
X
2
7a
H
H
2-F
H
H
H
H
H
H
CH3
CH3
CH3
CH3
CH3
Et
Bu
Bn
4-F-Bn
H
H
H
H
H
H
H
H
H
H
H
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
NH
O
Cl
I
11b
11c
11d
11e
11f
11g
11h
11j
25k
25l
25m
25n
25o
25p
25r
25sa
25ta
33
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
OTf
OTf
OTf
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
OTf
OTf
I
7-Br, 8-Cl
6-Cl, 7-Br
H
H
H
H
H
H
H
H
4-OMe CH3
H
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
CH3
Cl
Cl
Cl
Cl
Cl
Cl
Cl
I
2-F
H
2-Cl
1-Cl
H
H
H
H
H
H
H
H
2-F
H
2-F
2-F
H
H
2-F
H
H
6-F
H
a
b
(a) PPA; (b) POCl3; (c) MeOTf. See Table 1 for definitions of
Ra and Rb.
H
Sch em e 5a
8-F
7-Ph
7-F
9-F
H
H
H
H
H
H
H
H
H
H
H
H
H
I
Cl
Cl
Cl
NHPh
NHPh
OPh
OPh
4-Cl-PhS NH
SPh
Ph
Ph
4-Cl-PhS
4-Cl-PhS CH2
COOMe
COOMe
34
40
S
CH2
NH
NH
NH
NH
41a
41b
42a
42b
43b
43c
44a
44b
45a
46a
49
NH
NH
NH
O
H
H
H
H
H
NH
NMe
50
I
a
(a) PPA; (b) POCl3; (c) MeOTf.
a
Contains 11-chloro impurity.
32 were prepared according to literature or modified
literature procedures (Scheme 4). Alkylation of 31 and
32 with methyl triflate gave the 11-chlorotriflate salts
33 and 34.
Indenoquinoline derivative 40 (Scheme 5) required
the synthesis of 11-oxo-5,11-dihydroindeno[1,2-b]quino-
line 38.22 This was obtained by the Thiele-Falk reac-
tion of o-phthalaldehyde 35 with 2-(N-tosylamino)-
acetophenone23 36 and subsequent cyclization of the
intermediate indanone 37 with PPA. Treatment of 38
with POCl3 followed by alkylation with methyl triflate
gave 11-chlorotriflate salt 40.
The availability of the 11-chloro functionality in 24,
25, 33, 34, and 40 made it attractive to prepare a series
of 11-substituted cryptolepine analogues. The 11-chlo-
roquindoline and 2-fluoro-11-chloroquindoline systems
(Z ) NH, O, S, and CH2) were chosen as representative
substrates. Treatment of 24 or 25 (or the corresponding
free base) with aniline, phenol, and 4-chlorothiophenol
or thiophenol gave 41, 42, and 43 respectively (Scheme
6). All of these were converted into their chloride salts
using the sodium carbonate/basic alumina procedure.
In a similar manner, 4-chlorothiophenol was added to
benzofuroquinoline 33 and indenoquinoline 40 to pro-
vide the 11-thiophenol derivatives 45 and 46, respec-
tively. 11-Phenyl derivatives 44 were prepared via
phenylmagnesium bromide addition to the free base of
24. The intermediate free bases were acidified with HCl
to provide 44a ,b as their HCl salts.
In Vitr o Stu d ies
The activity demonstrated by cryptolepine in the 3T3-
L1 glucose transport assay8 provided us with a useful
method to screen analogues as they were prepared and
to study the effect that substitution about and within
the cryptolepine nucleus had on bioactivity. Most of the
analogues shown in Table 1 were screened at three
concentrations in the absence of added insulin, and the
results are shown in Table 2. Our earlier study noted
the importance of the N-5 methyl group for bioactivity,
as the N-10 CH3 regioisomer was inactive.8 Increasing
the lipophilicity of the substituent at N-5 increased the
glucose transport activity 11f (Bu) > 11e (Et) > 2 (CH3).
Among the two N-5 benzyl derivatives prepared, benzyl
(11g) was superior to the 4-fluorobenzyl derivative (11h )
and equipotent with 11f. In the absence of other factors,
halogen or phenyl substitution on the ring system had
a minimal effect on efficacy (11b vs 2, 25n vs 25k ) or
decreased activity (11c, 25p , 25r ). Likewise, the 11-
chloro substituent reduced the activity of the cryptol-
epine analogue, possibly due to a deactivating effect
(25k vs 2, 25l vs 11b). Substitution at the 4-position
with OMe (11j) or at the 6-position with fluorine (25m )

Antihyperglycemic Activities of Cryptolepine Analogues
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 15 2757
Sch em e 6a ,b
a
Reagents: (a) aniline (80%); (b) phenol (60%); (c) 4-chlorothiophenol (66-75%); (d) thiophenol (63%); (e) PhMgBr (45%). For conversion
b
to the chloride salts when Z ) NH, the procedure in Scheme 1 was used. See Table 1 for the definition of Ra and X.
improved activity, possibly due to an interaction be-
tween the methoxy or fluoro group and the N-5 methyl
group altering the planarity of the molecule. A similar
argument can be made for the simultaneous substitu-
tion at the 1- and 11-positions (25o), as this analogue
displayed similar activity with the parent 2 and was
more active than 25k or 25n .
Replacement of the nitrogen at the 10-position with
oxygen (33), sulfur (34), and CH2 (40) led to a reduction
in the glucose transport activity, with the indenoquino-
line derivative (40) showing activity only at the 30 µM
concentration. Substitution at the 11-position with
aniline (41b) maintained activity while substution with
phenol (42b), 4-chlorothiophenol (43b), or phenyl (44b)
appeared to reduce activity.
Some of the cryptolepine analogues which displayed
significant glucose transport activity in the initial screen
were investigated further at additional concentrations.
As shown in Figure 1, an increase in the lipophilicity of
the substituent at R5 led to an increase in the glucose
transport activity. With the butyl (11f) and benzyl (11g)
derivatives, cellular toxicity was seen at the higher
concentrations. The 4-methoxy derivative (11j) showed
a marked improvement in glucose transport activity
over cryptolepine hydrochloride (2), with cytotoxicity
being observed at the higher concentration (30 µM). In
contrast, the 2-fluoro derivative (11b) showed improved
glucose transport across the concentration range with-
out signs of cytotoxicity.
altered obese diabetic mice (designated C57BL/KS-db/
db or db/db). The results are shown in Table 3. The
4-methoxy derivative 11j was the most efficacious
derivative tested, lowering the plasma glucose level
58.3% 24 h postdose while showing improved food intake
as compared with cryptolepine hydrochloride (2). The
2-fluoro derivative 11b was also efficacious, lowering
the plasma glucose level 42.5% 3 h postdose, with
activity still being seen at 24 h. However, as with
cryptolepine hydrochloride (2), this effect was accom-
panied by a large reduction in both mean body weight
and food intake. The 7- and 9-fluoro, 11-iodo derivatives
(25s and 25t, respectively) were less active than 11b,
with 25s showing less of an effect on the food consump-
tion and mean body weight than did 11b and 25t.
Substitution at the 11-position decreased the ability of
the compound to lower plasma glucose based on the
three analogues tested (41b, 42b, and 44b), with 41b
showing improvement in food consumption and body
weight as well as demonstrating some activity. Het-
eroquindoline 33 was statistically inactive while show-
ing no effect on food intake or body weight.
Substitution at the N-5 position dramatically altered
the glucose-lowering ability and the in vivo toxicity of
the cryptolepine analogue, possibly due to increased
lipophilicity. In comparison with 2, the ethyl derivative
11e showed a faster onset of action at 3 h postdose,
possibly due to pharmacokinetic differences, yet resulted
in the death of two animals. The butyl derivative 11f
was acutely toxic at 100 mg/kg, resulting in the death
of six out of eight animals. When 11e and 11f were
tested at a lower dose (30 mg/kg), no activity and a trend
toward activity was observed, respectively. With both
In Vivo Stu d ies
Select cryptolepine analogues were tested under
single dose conditions at 100 mg/kg in genetically

2758 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 15
Bierer et al.
Ta ble 2. In Vitro Glucose Transport Activity in 3T3-LI
Adipocytes of Cryptolepine Analoguesa
% glucose transportb,c
compd
3 µM
10 µM
30 µM
2
7a
148 (*)
305 (**)
163 (**)
360 (***)
127 (*)
359 (**)
287 (**)
443 (****)
0.1 (ns)
404 (****)
490 (****)
161 (*)
134 (**)
253 (***)
104 (ns)
197 (***)
561 (****)
611 (***)
168 (**)
237 (****)
107 (ns)
142 (*)
11b
11c
11e
11f
11g
11h
11j
25k
25l
25m
25n
25o
25p
25r
25s
25t
33
228 (****)
868 (****)
781 (***)
248 (**)
486 (****)
101 (ns)
224 (***)
457 (****)
138 (**)
282 (****)
159 (*)
190 (****)
117 (ns)
255 (***)
114 (ns)
100 (ns)
119 (ns)
208 (***)
207 (***)
154 (**)
215 (*)
323 (****)
543 (****)
0.33 (-)
3.3 (-)
143 (*)
6.4 (-)
108 (ns)
120 (ns)
107 (ns)
122 (*)
115 (ns)
129 (*)
215 (**)
352 (****)
6 (-)
16 (-)
0.92 (-)
0.1 (-)
107 (ns)
110 (ns)
297 (****)
0.9 (-)
0.39 (-)
13 (-)
116 (ns)
141 (*)
122 (ns)
371 (****)
260 (****)
119 (ns)
96 (ns)
F igu r e 1. In vitro effect of chronic treatment of cryptolepine
hydrochloride (2), its 2-fluoro derivative 11b, its N-R5 deriva-
tives 11e-g, and its 4-methoxy derivative 11j on glucose
transport in 3T3-L1 adipocytes (absence of insulin).
34
40
125 (ns)
325 (****)
140 (**)
128 (**)
281 (****)
118 (ns)
108 (ns)
63 (ns)
41b
42b
43b
44b
45a
46a
49
lowered plasma glucose 36.3% with improved food
intake (3.1 vs 0.9 g for 2).
142 (*)
110 (ns)
159 (**)
207 (***)
In an attempt to differentiate a glucose lowering effect
with a cryptolepine analogue from an anorexigenic
effect, a pair-fed experiment was conducted on a rep-
resentative analogue. The 4-methoxy derivative 11j
was chosen on the basis of its ability to lower plasma
glucose significantly and because concomitant effects in
food intake were observed with this derivative. In this
experiment, db/db mice were divided into a control
group given free access to food and two pair-fed groups
treated with either vehicle or 100 mg/kg of 11j. Plasma
glucose concentrations were measured 3 and 24 h
postdose. While glucose concentrations were lower than
control in both pair-fed groups, mice receiving 11j
treatment had significantly lower plasma glucose con-
centrations than did the vehicle-treated group (Table
4).
50
155 (**)
a
3T3-L1 glucose transport assay in the absence of added
insulin. See ref 8 and procedure cited therein. Data is the average
of triplicate runs.
following codes: **** p < 0.0001; *** p < 0.001; ** p < 0.01; * p
< 0.05; ns denotes p > 0.05. c p as determined by Student’s t-test,
one tailed, independent.
b
p value in parentheses according to the
derivatives, no effect on body weight or food intake was
observed at the 30 mg/kg dose. Increasing the steric
bulk at N-5 as in 11g and 11h resulted in two com-
pounds which demonstrated an ability to lower plasma
glucose with an improvement in the food intake and
body weight parameters. Benzyl derivative 11g lowered
plasma glucose 15% 3 h postdose with no effect on food
intake or body weight. 4-Fluorobenzyl derivative 11h
Ta ble 3. In Vivo Results in db/ db Mice of Selected Cryptolepine Analoguesa
plasma glucose, % change predose
ANOVAb
mean body weight, g/mouse
dose,
mg/kg
food intake,
g/mouse/day
compd
3 h
24 h
3 h
24 h
0 h
24 h
2
100
100
100
100
30
-16.4
-42.5
-33.0
-23.2
-7.7
-43.1
-34.5
-18.9
-26.5
-4.1
0.059
0.007
0.011
ns
ns
0.090
ns
0.0223
<0.0001
ns
0.026
ns
0.0002
0.006
0.037
0.021
ns
ns
0.1006
ns
41.0 ( 0.5
43.8 ( 0.7
45.5 ( 0.8
44.6 ( 0.6
41.7 ( 0.5
45.8 ( 1.3
38.7 ( 0.8
39.3 ( 0.6
40.2 ( 1.2
38.4 ( 0.7
42.2 ( 1.4
43.4 ( 0.7
40.6 ( 0.6
44.8 ( 0.6
46.6 ( 0.8
40.5 ( 0.9
41.9 ( 0.8
39.8 ( 0.5
41.9 ( 0.9
44.6 ( 0.6
43.3 ( 1.0
41.2 ( 0.6
45.8 ( 1.2
38.6 ( 0.8
38.9 ( 0.7
39.0 ( 1.2
37.1 ( 0.7
41.8 ( 1.3
41.4 ( 0.8
41.1 ( 0.6
44.2 ( 0.6
45.2 ( 1.0
40.4 ( 0.9
41.8 ( 0.8
0.9
0.5
2.8
1.0c
5.1
2.6d
6.0
5.1
3.1
2.0
2.7
0.7
4.9
3.2
1.6
5.1
5.0
11b
11c
11e
11e
11f
11f
11g
11h
11j
25s
25t
33
100
30
-33.8
4.0
-2.2
-22.0
-20.0
-9.7
100
100
100
100
100
100
100
100
100
100
-15.0
-36.3
-8.3
-33.1
19.9
-17.9
-10.5
-11.8
-11.2
-3.2
0.091
<0.0001
ns
-58.3
-13.2
-32.9
-20.8
-10.7
-7.5
0.15e
nse
ns
41b
42b
44b
0.085
0.049
ns
0.017
0.044
ns
-4.9
-1.3
mean vehiclef
a
Single dose experiments at t ) 0 (n ) 8, except 2, where n ) 5) in genetically altered obese diabetic mice (designated C57BL/KS-
b
db/db). Plasma glucose levels measured at 3 and 24 h postdose. Analysis of variance (one factor), ns ) p > 0.15. Vehicle used as a
negative control. Metformin used as a positive control. c Two animals died. Six animals died. e Vehicle control dropped at 24 h in this
d
experiment. f Mean values of seven treatment groups shown for comparison.

Antihyperglycemic Activities of Cryptolepine Analogues
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 15 2759
Ta ble 4. Pair-Fed Experiment on 4-Methoxycryptolepine
Hydrochloride (11j) in db/ db Mice
1H), 7.66 (ddd, J ) 8.4, J ) 7.6, J ) 1.2, 1H), 7.62 (ddd, J )
9.2, J ) 7.6, J ) 2.8, 1H), 7.34 (ddd, J ) 8.0, J ) 7.2, J ) 0.8,
1H); 13C NMR (DMSO-d6) δ 167.66, 160.45 (J ) 245), 147.27
(d, J ) 2.3), 144.44, 140.48, 132.87, 132.22 (d, J ) 10), 130.51,
124.70 (d, J ) 11), 121.22, 120.57, 120.48, 115.85 (d, J ) 27),
112.70, 112.66, 109.80 (d, J ) 9), 108.53 (J ) 26); MS (EI,
m/z) 280 (M+). Anal. (C16H9N2O2F‚0.5H2O) C, H, N.
% change predoseb
ANOVA
treatmenta
control
pair-fed control
pair-fed 11j, 100 mg/kg
3 h
24 h
3 h
24 h
-9
-14
-15
2
-43
-55*
ns
ns
<0.001
<0.001
Gen er a l P r oced u r e B. 2-F lu or oqu in d olin e (6b).
A
mixture of 5b (4.50 g, 16.1 mmol) and diphenyl ether (40 mL)
was heated at 250 °C for 4 h. The mixture was cooled,
petroleum ether (40 mL) was added, and the mixture was
filtered. The filter cake was washed with petroleum ether (100
mL) and dried. The crude material was taken up into
methanol (250 mL), and the insoluble portion was filtered and
washed with additional methanol (2 × 75 mL). Concentration
of the filtrate gave 2.47 g (65%) of 6b as a brown solid, mp
>246 °C: 1H NMR (DMSO-d6) δ 11.52 (s, 1H), 8.34 (d, J )
8.0, 1H), 8.28 (s, 1H), 8.24 (dd, J ) 9.2, J ) 5.6, 1H), 7.90 (dd,
J ) 10, J ) 2.8, 1H), 7.65-7.52 (m, 3H), 7.29 (ddd, J ) 7.6, J
) 6.8, J ) 0.8, 1H); 13C NMR (DMSO-d6) δ 158.96 (d, J ) 241),
145.53, 143.97, 140.42, 132.86, 131.22 (d, J ) 10), 129.83,
127.25 (d, J ) 10.6), 121.31, 120.84, 119.54, 116.23 (d, J )
27), 112.47 (d, J ) 5.3), 111.58, 110.01 (d, J ) 22); MS (EI,
m/z) 236 (M+). Anal. (C15H9N2F) C, H, N.
Gen er a l P r oced u r e C. 5-Eth ylqu in d olin iu m Hyd r o-
iod id e (7e). A suspension of quindoline8 (1.0 g, 4.6 mmol)
and ethyl iodide (3 mL, 37.7 mmol) was heated in a bomb at
135 °C for 15 h. The excess ethyl iodide was removed in vacuo,
and the brown precipitate was collected, washed with ether,
and dried, yielding 1.5 g (88%) of 7e. Recrystallization from
water gave bright yellow crystals, mp 256 °C (lit.1 mp 222-
223 °C): 1H NMR (DMSO-d6) δ 12.93 (s, 1H), 9.32 (s, 1H), 8.79
(d, J ) 9.2, 1H), 8.66 (d, J ) 8.4, 1H), 8.60 (d, J ) 8.0, 1H),
8.18 (dd, J ) 7.6, J ) 7.6, 1H), 7.95 (t, J ) 8.0, 2H), 7.87 (d,
J ) 8.4, 1H), 7.55 (dd, J ) 7.6, J ) 7.6, 1H), 5.55 (q, J ) 6.8,
2H), 1.76 (t, J ) 7.2, 3H); 13C NMR (DMSO-d6) δ 145.78,
136.92, 134.34, 133.94, 133.52, 132.71, 130.06, 127.09, 126.42,
125.50, 125.14, 121.81, 117.31, 113.37, 112.80, 47.33, 13.30;
MS (EI, m/z) 247 (M+). Anal. (C17H15N2I) C, H, N.
5-Bu tylqu in d olin iu m Hyd r oiod id e (7f). Reaction of
quindoline (1.0 g, 4.6 mmol) and butyl iodide (3 mL, 26.4 mmol)
according to general procedure C gave, after washing with 10%
ethanol in ethyl ether and drying, 1.82 g (100%) of 7f as a
brown solid. Recrystallization from water gave 7f as bright
yellow crystals, mp 252-253 °C: 1H NMR (DMSO-d6) δ 12.96
(s, 1H), 9.34 (s, 1H), 8.77 (d, J ) 9.2, 1H), 8.61 (dd, J ) 8.4, J
) 1.2, 1H), 8.54 (d, J ) 8.4, 1H), 8.19 (ddd, J ) 8.4, J ) 6.8,
J ) 1.6, 1H), 7.96 (t, J ) 8.4, 2H), 7.88 (d, J ) 8.4, 1H), 7.58
(ddd, J ) 8.0, J ) 6.8, J ) 0.8, 1H), 5.51 (t, J ) 7.2, 2H), 2.09
(m, 2H), 1.69 (m, 2H), 1.00 (t, J ) 7.2, 3H); 13C NMR (DMSO-
d6) δ 145.78, 137.07, 134.64, 133.93, 133.52, 132.71, 130.07,
127.11, 126.40, 125.38, 125.29, 121.88, 117.55, 113.47, 112.89,
51.29, 30.11, 19.22, 13.83; MS (EI, m/z) 275 (M+). Anal.
(C19H19N2I‚0.86 H2O) C, H, N.
Gen er a l P r oced u r e D. 2-F lu or o-5-m eth ylqu in d olin i-
u m Hyd r och lor id e (11b). A suspension of 6b (0.50 g, 2.12
mmol) and methyl iodide (3.95 mL, 63.5 mmol) was heated at
150 °C for 24 h in a Parr bomb. The bomb was cooled, and
the reaction mixture was concentrated. The orange residue
was suspended in chloroform, adsorbed onto Na2CO3, and
purified by chromatography over basic alumina. Elution with
chloroform to remove unreacted starting material, and then
elution with 2% methanol in chloroform gave violet fractions,
which were collected and concentrated to a small volume (50
mL). The violet solution was acidified with a 1 M solution of
HCl in ether to a bright yellow endpoint. The mixture was
chilled, filtered, and then dried to give 0.41 g (68%) of 11b as
a bright orange solid, mp >246 °C: 1H NMR (DMSO-d6) δ
13.40 (s, 1H), 9.27 (s, 1H), 8.88 (dd, J ) 10, J ) 4.4, 1H), 8.80
(d, J ) 8.4, 1H), 8.44 (dd, J ) 8.8, J ) 2.8, 1H), 8.11 (dd, J )
10.4, J ) 8.0, 1H), 7.94 (dd, J ) 8.0, J ) 8.0, 1H), 7.85 (d, J )
8.4, 1H), 7.51 (dd, J ) 8.0, J ) 8.0, 1H), 5.05 (s, 3H); 13C NMR
(DMSO-d6) δ 159.35 (J ) 246), 145.89, 138.28, 134.13, 134.03,
132.39, 127.39 (d, J ) 10.6), 126.37, 123.62 (d, J ) 5.3), 121.86
a
b
n ) 8, single dose. p < 0.01 (ANOVA) as compared to pair-
fed control group.
Con clu sion
On the basis of the ethnobotanically generated lead
structure of cryptolepine, a series of substituted and
heterosubstitued cryptolepine analogues have been
synthesized, and their bioactivities have been measured
in a 3T3-L1 glucose transport assay. While cryptolepine
itself has been the subject of numerous biological
studies, this report to our knowledge is the first struc-
ture-bioactivity study reported on the cryptolepine
nucleus.24 In this study, 33 analogues are reported. A
select group of these synthesized analogues were further
evaluated in genetically altered obese diabetic C57BL/
KS-db/db mice. Most of the cryptolepine analogues that
demonstrated in vitro activity, which were tested in
vivo, also demonstrated in vivo activity. With most of
the analogues that were tested in vivo, there appears
to be a correlation between the ability of the compound
to lower plasma glucose levels and the effect seen in the
body weight and food intake parameters. Yet while it
is possible that some of the analogues tested might
manifest their ability to lower plasma glucose levels
based on reduced food intake or based on a general toxic
phenomenon (as with 11e and 11f), we have shown that
two derivatives in this series, cryptolepine hydrochloride
(2) and 4-methoxycryptolepine hydrochloride (11j), dis-
play true antihyperglycemic properties. One derivative,
the N-5 benzyl analogue 11g, exhibited a significant
glucose lowering effect with no effect on food intake or
body weight. Further results derived from this study
which successfully address the food intake and body
weight issues will be published in due course.
Exp er im en ta l Section
Gen er a l. General experimental conditions have been
described.8 Low-pressure liquid chromatography (LPLC) was
performed on E. Merck 230-400 mesh silica gel using nitrogen
pressure unless otherwise noted. Chromatography on cryp-
tolepine salt derivatives was done on Fisher Activity I basic
alumina using ethanol-free chloroform8 or on Fisher neutral
alumina.
Gen er a l P r oced u r e A. 2-F lu or oqu in d olin e-11-ca r box-
ylic Acid (5b). A nitrogen-purged solution of 3-fluoroisatin
(4.71 g, 28.5 mmol) in 4 M KOH (135 mL) at 5 °C was added
to a nitrogen-purged flask containing 3-indolyl acetate (5.00
g, 28.5 mmol). The mixture was mechanically stirred for 5
days at room temperature. Water (80 mL) was added, and
the solution was heated at 70 °C for 20 min with air being
drawn through the solution. The solution was filtered while
hot through Celite, and the bed was rinsed with warm water
(75 mL). The yellow filtrate was combined with an equal
volume of ethanol (300 mL). The mixture was acidified to pH
2 with dilute HCl, chilled, and filtered. The filter cake was
washed with water and ethanol, and then dried under high
vacuum for 2 days to afford 4.91 g (61%) of 5b as a bright
yellow solid, mp > 243 °C: 1H NMR (DMSO-d6) δ 14.3 (broad
s, 1H), 11.49 (s, 1H), 8.88 (dd, J ) 12.8, J ) 2.8, 1H), 8.34
(broad d, J ) 8.8, 1H), 8.32 (d, J ) 9.2, 1H), 7.79 (d, J ) 8.4,

2760 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 15
Bierer et al.
(d, J ) 27), 121.41, 121.19 (d, J ) 9.0), 113.70, 113.20, 112.65
(d, J ) 23.0), 40.74; MS (LSIMS, m/z) 251 (M+). HRMS (FAB+,
m/z) calcd for C16H12N2F+ 251.0984, found 251.0998. Anal.
(C16H12N2ClF‚1.25H2O) C, H, N.
113.9, 113.3, 56.6, 24.7; MS (FAB, m/z) 339 (M + H+). Minor
isomer 14b: 1H NMR (CDCl3) δ 8.00 (d, J ) 8.4, 1H), 7.74-
7.70 (m, 1H), 7.68-7.63 (m, 2H), 7.47 (t, J ) 8.0, 1H), 7.25
(dt, J ) 7.6, J ) 0.8, 1H), 7.19 (dt, J ) 8.0, J ) 0.8, 1H), 7.14-
7.06 (m, 1H), 3.94 (s, 3H), 2.65 (s, 3H); 13C NMR (CDCl3) δ
182.9, 168.9, 151.3, 148.7, 136.5, 131.7, 130.8, 128.2, 124.8,
124.6, 123.9, 122.1, 119.8, 117.1, 112.8, 56.5, 26.7; MS (FAB,
m/z) 339 (M + H+). HRMS (FAB) calcd for C18H14N2O5+H+
339.0981, found 339.1010. Anal. (C18H14N2O5) C, H, N.
Meth od B. A solution of 1-acetyl-3-oxoindole8 (3.0 g, 17.1
mmol), 2-nitro-3-methoxybenzaldehyde (4.65 g, 25.7 mmol),
and piperidine (20 drops) in benzene (200 mL) was refluxed
in a Dean-Stark apparatus for 4 h. During the first 3 h, the
solvent in the trap was removed four times and replaced with
freshly distilled benzene. After cooling, the solvent was
removed and the residue was purified by LPLC, eluting with
dichloromethane, to give 4.5 g (78%) of 14 as a 1:2 mixture of
Z and E isomers.
4-Meth oxyqu in d olin e (7j). A mixture of 14 (6.0 g, 17.75
mmol, obtained via Method A) and 5% Pd/CaCO3 (poisoned
with lead, 1.5 g) in methanol (600 mL) was hydrogenated
(balloon, 1 atm) at room temperature overnight. The catalyst
was filtered over Celite, and the solvent was removed on a
rotary evaporator to afford crude 15. A solution of KOH (12.0
g) in methanol (180 mL) was added to crude 15, and the
mixture was stirred at room temperature for 30 min, during
which the mixture turned dark brown. The mixture was
poured into a saturated NH4Cl solution and extracted several
times with dichloromethane. The organic solutions were
combined, washed with brine, dried over MgSO4, and then
concentrated. The crude product was purified by LPLC,
eluting with dichloromethane and ethyl acetate 20:1, to give
1.82 g (41% yield over two steps) of 7j, mp 204.5-205.2 °C:
1H NMR (DMSO-d6) δ 11.40 (s, 1H), 8.34 (d, J ) 8.0, 1H), 8.22
(s, 1H), 7.66-7.54 (m, 3H), 7.44 (dd, J ) 8.0, J ) 8.0, 1H),
7.29 (dd, J ) 7.2, J ) 7.2, 1H), 7.06 (d, J ) 7.2, 1H), 4.05 (s,
3H); 13C NMR (DMSO-d6) δ 155.3, 144.3, 143.8, 135.5, 132.6,
129.3, 127.8, 125.0, 121.4, 121.2, 119.31, 119.27, 112.8, 111.4,
104.7, 55.4; MS (FAB, m/z) 249 (M + 1+). HRMS (FAB) calcd
for C16H12N2O+H+ 249.1028, found 249.1035.
Gen er a l P r oced u r e F . 4-Meth oxy-5-m eth ylqu in d o-
lin iu m Hyd r otr iflu or om eth a n esu lfon a te (11i). A suspen-
sion of 7j (570 mg, 2.3 mmol) in anhydrous toluene (30 mL)
was treated with methyl triflate (1.88 g, 11.48 mmol) and the
mixture was stirred at room temperature for 3 days. The
reaction mixture was diluted with diethyl ether (100 mL). The
solid was filtered, washed several times with ether, and then
dried to give 742 mg (78%) of 11i as a yellow solid. The
product was recrystallized from methanol, mp 234.0-235.0
°C: 1H NMR (DMSO-d6) δ 12.81 (s, 1H), 9.19 (s, 1H), 8.68 (d,
J ) 8.0, 1H), 8.09 (dd, J ) 8.4, J ) 0.8, 1H), 7.94 (ddd, J )
7.2, J ) 7.2, J ) 0.8, 1H), 7.86-7.80 (m, 2H), 7.66 (dd, J )
7.6, J ) 1.2, 1H), 7.51 (ddd, J ) 7.6, J ) 7.6, J ) 1.2, 1H),
5.13 (s, 3H), 4.14 (s, 3H); 13C NMR (DMSO-d6) δ 150.8, 145.8,
139.9, 133.9, 133.4, 128.5, 128.1, 127.6, 126.5, 124.7, 121.6,
121.3, 114.1, 113.3, 113.1, 57.0, 46.4; MS (FAB, m/z) 263 (M+).
HRMS (FAB) calcd for C17H15N2O+ 263.1184, found 263.1188.
Anal. (C18H15N2F3O4S) C, H, N.
Gen er a l P r oced u r e E. 5-Eth ylqu in d olin iu m Hyd r o-
ch lor id e (11e). 5-Ethylquindolinium hydroiodide 7e (0.75 g,
2.0 mmol) was treated with a 5% solution of Na2CO3 (100 mL),
extracted with chloroform (2 × 250 mL), and then the
combined chloroform extract was concentrated to a small
volume. Purification over basic alumina, first eluting with
chloroform, and then eluting with 1-2% methanol in chloro-
form, gave a solution of the free base. The purple extract was
evaporated to a small volume and acidified to a bright yellow
endpoint with a 1.0 M solution of HCl in ether to give, after
filtration and drying, 300 mg (53%) of 11e as yellow crystals,
mp 268.5-269 °C (dec): 1H NMR (DMSO-d6) δ 13.29 (s, 1H),
9.34 (s, 1H), 8.79 (d, J ) 9.2, 1H), 8.66 (d, J ) 8.4, 1H), 8.61
(dd, J ) 8.4, J ) 1.6, 1H), 8.19 (ddd, J ) 8.4, J ) 6.8, J ) 1.6,
1H), 7.96 (t, J ) 7.2, 2H), 7.90 (d, J ) 8.0, 1H), 7.56 (ddd, J )
8.4, J ) 6.8, J ) 1.2, 1H), 5.56 (q, J ) 7.2, 2H), 1.76 (t, J )
7.6, 3H); 13C NMR (DMSO-d6) δ 145.88, 136.93, 134.34, 133.92,
133.60, 132.68, 130.11, 127.07, 126.44, 125.50, 125.18, 121.75,
117.30, 113.43, 112.79, 47.32, 13.29. Anal. (C17H15N2Cl‚
0.86H2O) C, H, N.
5-Bu tylqu in dolin iu m Hydr och lor ide (11f). 5-Butylquin-
dolinium hydroiodide (0.8 g, 2.0 mmol) was converted to its
HCl salt according to general procedure E, affording 400 mg
(63.5%) of 11f as yellow crystals, mp 253-254 °C: 1H NMR
(DMSO-d6) δ 13.55 (s, 1H), 9.35 (s, 1H), 8.76 (d, J ) 9.2, 1H),
8.61 (d, J ) 8.4, 1H), 8.52 (d, J ) 8.4, 1H), 8.17 (dd, J ) 8.4,
J ) 8.4, 1H), 7.94 (t, J ) 8.0, 2H), 7.88 (d, J ) 8.4, 1H), 7.55
(dd, J ) 8.0, J ) 8.0, 1H), 5.50 (t, J ) 7.6, 2H), 2.08 (m, 2H),
1.69 (m, 2H), 1.00 (t, J ) 7.2, 3H); 13C NMR (DMSO-d6) δ
145.85, 136.94, 134.58, 133.81, 133.53, 132.63, 130.09, 127.02,
126.36, 125.34, 125.22, 121.76, 117.51, 113.48, 112.75, 51.27,
30.11, 19.22, 13.83; MS (EI, m/z) 275 (M+). HRMS (EI) calcd
for C19H19N2+ 275.1548, found 275.1545. Anal. (C19H19N2Cl‚
1.5H2O) C, H, N.
5-Ben zylqu in d olin iu m Hyd r och lor id e (11g). Reaction
of quindoline (1.0 g, 4.6 mmol) and benzyl bromide (3 mL, 8.5
mmol) in chloroform (5 mL) according to general procedure C
for 48 h gave 1.78 g of crude hydrobromide salt. Purification
on neutral alumina and conversion to the hydrochloride salt
according to general procedure E gave 0.40 g (25%) of 11g as
yellow crystals, mp 241-242 °C: 1H NMR (DMSO-d6) δ 13.40
(s, 1H), 9.49 (s, 1H), 8.68 (d, J ) 8.4, 1H), 8.53 (d, J ) 8.8,
1H), 8.31 (d, J ) 8.4, 1H), 8.12 (dd, J ) 7.6, J ) 7.6, 1H), 7.96
(dd, J ) 7.6, J ) 7.6, 1H), 7.90 (d, J ) 5.4, 2H), 7.42-7.32 (m,
4H), 7.22 (m, 2H), 6.84 (s, 2H); 13C NMR (DMSO-d6) δ 146.14,
138.11, 135.21, 134.16, 133.84, 133.28, 133.00, 130.23, 129.19,
128.14, 127.21, 126.41, 126.10, 126.02, 125.24, 121.65, 117.59,
113.53, 112.86, 54.89; MS (EI, m/z) 308 (M+, free base). HRMS
(EI) calcd for C22H16N2 308.1313, found 308.1309. HRMS
+
(FAB) calcd for C22H17N2 309.1392, found 309.1387. Anal.
(C22H17N2Cl‚1.75H2O) C, H, N.
1-Acetyl-2-(3-m eth oxy-2-n itr op h en ylm eth ylen e)-3-oxo-
2,3-dih ydr oin dole (14). Meth od A. To a mixture of 1-acetyl-
3-oxoindole8 (3.0 g, 17.14 mmol), 2-nitro-3-methoxybenzalde-
hyde (9.31 g, 51.43 mmol), and 4 Å molecular sieves (60 g) in
toluene (250 mL) and chloroform (10 mL) was added 15 drops
of piperidine. The mixture was stirred for 1 week during
which time a yellow precipitate formed. The reaction mixture
was diluted with dichloromethane to dissolve the precipitate.
The molecular sieves were filtered off and washed several
times with dichloromethane. The organic solution was con-
centrated and the crude product was purified by LPLC, eluting
with dichloromethane, to give 5.86 g of unreacted 2-nitro-3-
methoxybenzaldehyde and 4.9 g (85%) of 14 as a 2:1 mixture
of Z and E isomers. Major isomer 14a : 1H NMR (CDCl3) δ
8.22 (d, J ) 8.0, 1H), 7.86 (ddd, J ) 7.6, J ) 1.2, J ) 0.4, 1H),
7.71-7.68 (m, 1H), 7.48 (dd, J ) 8.0, J ) 8.0, 1H), 7.32 (ddd,
J ) 7.6, J ) 7.6, J ) 0.8, 1H), 7.14-7.06 (m, 3H), 3.97 (s, 3H),
1.98 (s, 3H); 13C NMR (CDCl3) δ 184.7, 169.9, 151.7, 150.2,
137.5, 136.8, 136.1, 129.1, 125.2, 124.6, 123.3, 120.8, 117.5,
Gen er a l P r oced u r e G. 4-Meth oxy-5-m eth ylqu in d o-
lin iu m Hyd r och lor id e (11j). A suspension of 11i (700 mg,
1.70 mmol) in chloroform (200 mL) was shaken vigorously with
a saturated K2CO3 solution (100 mL) in a separatory funnel.
The purple chloroform extract was washed with aqueous K2-
CO3 solution (3×) and brine, dried over anhydrous K2CO3, and
then concentrated. The residue was dissolved in chloroform
(20 mL) and treated with a solution of HCl in ether (1.0 M, 20
mL). The yellow precipitate was filtered, washed several times
with ether, and then dried. Recrystallization from methanol
and acetone gave 430 mg (85%) of 11j, mp 212.4-212.9 °C;
1H NMR (DMSO-d6) δ 13.24 (s, 1H), 9.22 (s, 1H), 8.68 (d, J )
8.4, 1H), 8.10 (dd, J ) 8.4, J ) 0.8, 1H), 7.93 (ddd, J ) 7.2, J
) 7.2, J ) 1.2, 1H), 7.86-7.80 (m, 2H), 7.66 (dd, J ) 8.0, J )
0.8, 1H), 7.50 (ddd, J ) 7.2, J ) 7.2, J ) 1.2, 1H), 5.14 (s, 3H),
4.15 (s, 3H); 13C NMR (DMSO-d6) 150.7, 145.9, 139.8, 133.8,

Antihyperglycemic Activities of Cryptolepine Analogues
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 15 2761
133.4, 128.4, 128.0, 127.4, 126.4, 124.6, 121.6, 121.2, 113.9,
113.3, 113.1, 57.1, 46.6; MS (FAB, m/z) 263 (M+). HRMS
(FAB) calcd for C17H15N2O+ 263.1184, found 263.1169.
2-F lu or o-11-ch lor oqu in d olin e (21l). A solution of 20l
(6.0 g, 23.8 mmol) in POCl3 (60 mL) was refluxed for 2 h. The
reaction mixture was cooled, poured into ice, neutralized with
a cold saturated solution of KOH while keeping the internal
temperature below 45 °C, and then extracted with EtOAc (3
× 500 mL). The combined EtOAc extract was washed with
water and brine, dried (Na2SO4), and then concentrated.
Purification by LPLC, eluting with EtOAc-hexane (1:6), gave
3.5 g (54.5%) of 21l as a yellow solid, mp 208.5-210.0 °C: 1H
NMR (DMSO-d6) δ 11.89 (s, NH), 8.32 (s, J ) 9.6, 1H), 8.30
(d, J ) 8.4, 1H), 7.90 (dd, J ) 10.4, J ) 2.8, 1H), 7.69-7.60
(m, 3H), 7.33 (ddd, J ) 8.0, J ) 6.8, J ) 1.2, 1H); 13C NMR
(DMSO-d6) δ 159.98 (d, J ) 243.4), 145.83, 144.00, 140.83,
132.25 (d, J ) 9.8), 130.46, 130.42, 124.45 (d, J ) 9.9), 121.62,
121.10, 120.39, 116.93 (d, J ) 25.8), 112.06, 105.52 (d, J )
24.3); MS (EI, m/z) 270 (M+). Anal. (C15H8N2ClF‚0.25H2O)
C, H, N.
Gen er a l P r oced u r e H. 2-[(2-Br om oa cetyl)a m in o]-5-
flu or oben zoic Acid (17l). A solution of 2-amino-5-fluoroben-
zoic acid (15.0 g, 96.6 mmol) in DMF (35 mL) and dioxane (35
mL) was cooled to 0 °C. Bromoacetyl bromide (8.43 mL, 96.6
mmol) was added dropwise over a 20 min period, keeping the
internal temperature between 0 and 1 °C. After the addition
was completed, the ice bath was removed and stirring was
continued overnight at room temperature. Water (300 mL)
was added, and the light yellow precipitate which formed was
filtered, washed with water until neutral, and then dried to
give 25.4 g (95%) of 17l as a white solid, mp 191.2-191.8 °C:
1H NMR (acetone-d6) δ 11.60 (s, 1H), 8.71 (dd, J ) 9.2, J )
5.2, 1H), 7.80 (dd, J ) 9.6, J ) 3.2, 1H), 7.48-7.43 (m, 1H),
4.18 (s, 2H); 13C NMR (acetone-d6) δ 168.80, 165.71, 158.37
(d, J ) 241.9), 138.56, 122.78 (d, J ) 7.8), 122.70, 122.06 (d, J
) 21.9), 117.98 (d, J ) 24.0), 30.45; MS (EI, m/z) 275 (M+).
HRMS (EI) calcd for C9H7NO3BrF 274.9593, found 274.9594.
Anal. (C9H7NO3BrF) C, H, N.
Gen er a l P r oced u r e J . 6-F lu or o-11-ch lor oqu in d olin e
(21m ) a n d 8-F lu or o-11-ch lor oqu in d olin e (21p ). A suspen-
sion of 19m (6.75 g, 23.4 mmol) in PPA (420 g) was heated at
140 °C for 5 h with mechanical stirring. The mixture was
poured over ice (1 L) and placed in an ice bath where the
internal temperature was kept below 45 °C while the mixture
was neutralized to pH 7 with a saturated KOH solution (total
volume 2 L). The resulting precipitate was filtered, washed
with water (250 mL), and then dried under high vacuum at
45 °C for several days to afford 7.37 g of a mixture of 6-fluoro-
11-quindolone and 8-fluoro-11-quindolone as a green solid: MS
(FAB, m/z) 253 (M + H+).
The crude mixture obtained above (7.29 g) was refluxed in
POCl3 (60 mL) for 3 h, cooled, and then slowly poured over ice
(1 L). The mixture was neutralized to pH 7 with a cold slurry
of aqueous KOH while maintaining an internal temperature
below 45 °C with an external ice bath. The aqueous portion
(1.3 L) was extracted with chloroform (5 × 400 mL). The
combined organic extracts were washed with water (600 mL)
and brine (600 mL), dried (Na2SO4), filtered, and then con-
centrated. The crude product was adsorbed onto neutral
alumina with acetone. Chromatography over neutral alumina,
eluting with EtOAc-hexane (1:5 gradiated to 1:4), and then
rechromatography of mixed fractions in the same manner
afforded 0.21 g (4% overall) of 21m and 1.39 g (22% overall)
of 21p as yellow solids.
Gen er a l P r oced u r e I. 2-[2-[(P h en yla m in o)a cetyl]a m i-
n o]-5-flu or oben zoic Acid (19l). A solution of 17l (22.4 g,
91 mmol) and aniline (26 mL, 285 mmol) in DMF (150 mL)
was heated at 120 °C for 30 h. After cooling, the reaction
mixture was poured into ice-water (800 mL), aqueous 5%
KOH (100 mL) was added to solubilize the solid product and
adjust the pH to 10-11, and then the mixture was extracted
with dichloromethane (3 × 400 mL). The combined dichlo-
romethane extracts were set aside and the aqueous layer was
acidified to pH 3 with a solution of 5% HBr. The precipitate
which formed was collected, washed with water, and then
dried, yielding 19.6 g (75%) of 19l as white crystals, mp 194-
195 °C: 1H NMR (acetone-d6) δ 11.87 (s, 1H), 8.88 (dd, J )
9.6, J ) 5.2, 1H), 7.71 (dd, J ) 9.2, J ) 3.2, 1H), 7.45-7.39
(m, 1H), 7.14-7.10 (m, 2H), 6.68-6.64 (m, 3H), 3.92 (s, 2H);
13C NMR (acetone-d6) δ 171.30, 168.14, 163.22, 157.97 (d, J )
241.3), 148.97, 138.66, 129.82, 122.60 (d, J ) 7.1), 121.90 (d,
J ) 21.9), 118.63, 117.78 (d, J ) 24.0), 113.67, 50.07; MS (EI,
m/z) 288 (M+). HRMS (EI) calcd for C15H13N2O3F 288.0910,
found 288.0914.
2-[2-[[(3-F lu or op h en yl)a m in o]a cet yl]a m in o]b en zoic
Acid (19m ). Reaction of 2-[(2-bromoacetyl)amino]benzoic
acid16 (19k ) (8.00 g, 31.0 mmol) and 3-fluoroaniline (7.45 mL,
77.5 mmol) in DMF (80 mL) at 80 °C for 8 h according to
general procedure I gave 7.00 g (78%) of 19m as a gray solid,
mp 209 °C: 1H NMR (DMSO-d6) δ 13.5 (bs, 1H), 11.94 (s, 1H),
8.70 (d, J ) 8.4, 1H), 7.94 (dd, J ) 7.8, J ) 1.4, 1H), 7.60 (ddd,
J ) 8.0, J ) 8.0, J ) 1.4, 1H), 7.15-7.07 (m, 2H), 6.83 (bs,
1H), 6.43-6.35 (m, 3H), 3.88 (s, 2H); 13C NMR (DMSO-d6) δ
170.29, 169.12, 163.25 (d, J ) 240.3), 150.17 (d, J ) 10.6),
140.52, 134.18, 131.13, 130.43 (d, J ) 10.6), 122.73, 119.40,
115.98, 108.42, 103.09 (d, J ) 21.2), 98.97 (d, J ) 25), 48.48;
MS (EI, m/z) 388 (M+). HRMS (EI) calcd for C15H13N2O3F
288.0910, found 288.0915.
Characterization of 21m : mp 241.8 °C; TLC Rf 0.39 (ethyl
acetate-hexane (1:2)); 1H NMR (DMSO-d6) δ 12.09 (s, 1H),
8.28-8.25 (m, 2H), 7.79-7.71 (m, 2H), 7.65 (ddd, J ) 8.0, J )
8.0, J ) 5.6, 1H), 7.44 (d, J ) 8.0, 1H), 7.10 (dd, J ) 10.4, J )
8.0, 1H); 13C NMR (DMSO-d6) δ 157.70 (d, J ) 254.0), 145.86
(d, J ) 8.5), 144.19, 143.68, 131.54 (d, J ) 9.2), 129.81, 129.41,
127.07, 126.71, 123.37, 122.08, 118.15, 109.26 (d, J ) 18.1),
108.20 (d, J ) 3.0), 106.10 (d, J ) 18.3); MS (FAB, m/z) 271
(M + H+). Anal. (C15H8N2FCl‚0.75H2O) C, H, N.
Characterization of 21p : mp 240.5 °C; TLC Rf 0.47 (ethyl
acetate-hexane (1:2)); 1H NMR (DMSO-d6) δ 11.95 (s, 1H),
8.34 (dd, J ) 8.4, J ) 5.6, 1H), 8.24 (m, 2H), 7.74 (ddd, J )
17.6, J ) 6.8, J ) 1.6, 1H), 7.72 (ddd, J ) 17.6, J ) 6.8, J )
1.6, 1H), 7.32 (dd, J ) 9.6, J ) 2.4, 1H), 7.14 (ddd, J ) 9.6, J
) 8.8, J ) 2.4, 1H); 13C NMR (DMSO-d6) δ 163.80 (d, J )
244.9), 145.35 and 145.21 and 145.08 (two carbons), 144.03,
130.44, 129.19, 127.05, 126.37, 123.51 (d, J ) 10.7), 123.35,
122.17, 118.06, 117.91, 108.41 (d, J ) 24.1), 98.57 (d, J ) 26.9);
MS (FAB, m/z) 271 (M + H+). Anal. (C15H8N2FCl‚0.88H2O)
C, H, N.
5,11-Dih yd r o-2-flu or o-11-oxo-10H-in d olo[3,2-b]qu in o-
lin e (2-F lu or o-11-qu in d olon e) (20l). A mixture of 19l (4.5
g, 15.8 mmol) and polyphosphoric acid (PPA, 150 g) was heated
with mechanical stirring at 130 °C for 2 h. The reaction
mixture was poured into ice-water (1.2 L), neutralized with
saturated KOH solution, and then extracted with EtOAc (2 ×
750 mL). The extract was washed with water and brine, dried
(Na2SO4), and then concentrated to give 2.25 g (56.3%) of 20l
as a yellow solid, mp >270 °C. A portion of this material was
purified by LPLC, eluting with EtOAc-MeOH (5:1): 1H NMR
(DMSO-d6) δ 12.65 (s, NH), 11.75 (s, NH), 8.18 (d, J ) 8.0,
1H), 7.99 (dd, J ) 9.6, J ) 2.4, 1H), 7.79 (dd, J ) 8.8, J ) 4.4,
1H), 7.60 (ddd, J ) 10.8, J ) 8.0, J ) 2.4, 1H), 7.53-7.45 (m,
2H), 7.21 (dd, J ) 7.6, J ) 7.6, 1H); 13C NMR (DMSO-d6) δ
166.37, 156.77 (d, J ) 237), 138.85, 135.75, 129.34, 127.71,
123.46 (d, J ) 5.7), 122.67, 120.91, 120.27 (d, J ) 7.7), 119.65
(d, J ) 25.3), 118.99, 115.74, 112.66, 108.75 (d, J ) 21.8); MS
(EI, m/z) 252 (M+).
2-F lu or o-5-m eth yl-11-ch lor oqu in d olin iu m Hyd r otr i-
flu or om eth a n esu lfon a te (24l). Reaction of 21l (1.56 g, 5.76
mmol) and methyl triflate (1.89 g, 11.52 mmol) in anhydrous
toluene (50 mL) for 1 day according to general procedure F
gave 2.40 g (96%) of 24l as an orange solid, mp 297.5-299.5
°C: 1H NMR (DMSO-d6) δ 13.29 (s, 1H), 8.97 (dd, J ) 10.0, J
) 4.4, 1H), 8.82 (d, J ) 8.4, 1H), 8.41 (dd, J ) 9.6, J ) 3.2,
1H), 8.26-8.18 (m, 1H), 7.99 (ddd, J ) 8.4, J ) 8.4, J ) 0.8,
1H), 7.84 (d, J ) 8.4, 1H), 7.57 (ddd, J ) 8.4, J ) 8.4, J ) 0.8,
1H), 5.04 (s, 3H); 13C NMR (DMSO-d6) δ 160.50 (d, J ) 250.5),

2762 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 15
Bierer et al.
145.92, 138.99, 134.85, 132.88, 132.54, 127.60 (d, J ) 5.0),
126.74, 124.96 (d, J ) 9.9), 122.48 (d, J ) 10.0), 122.39 (d, J
) 26.9), 122.26, 114.47, 113.49, 108.46 (d, J ) 25.4), 40.99;
MS (EI, m/z) 285 (M+), 287 (M + 2+). HRMS (FAB, m/z) calcd
for C16H11N2ClF+ 285.0595, found 285.0597. Anal. (C17H11N2O3-
ClF4S) C, H, N.
122.81, 119.88, 117.13, 42.79; MS (FAB, m/z) 284 (M+) MS
(-FAB, m/z) 149 (OTf-). HRMS (FAB) calcd for C16H11NClS+
284.0301, found 284.0313.
11-Ch lor o-5-m eth ylin d en o[1,2-b]qu in olin iu m Tr iflu o-
r om eth a n esu lfon a te (40). A solution of 39 (0.09 g, 0.35
mmol) in anhydrous toluene (8 mL) was treated with methyl
triflate (0.05 mL, 0.42 mmol). The reaction mixture was
stirred for 20 h at room temperature. Ether (50 mL) was
added, and the mixture was stirred for 1 h. The product was
filtered, dried, refluxed for 1 h with benzene, filtered while
hot, and then dried again to afford 0.11 g (77%) of 40, mp 193-
194 °C: 1H NMR (DMSO-d6) δ 8.82 (d, J ) 8.8, 1H), 8.72 (d,
J ) 8.0, 1H), 8.60 (dd, J ) 8.0, J ) 1.2, 1H), 8.32 (ddd, J )
8.8, J ) 6.8, J ) 1.2, 1H), 8.13 (ddd, J ) 8.0, J ) 6.8, J ) 0.8,
1H), 8.01 (d, J ) 7.6, 1H), 7.95 (dd, J ) 7.6, J ) 7.6, 1H), 7.79
(dd, J ) 7.2, J ) 7.2, 1H) 4.88 (s, 3H), 4.51 (s, 2H); 13C NMR
(DMSO-d6) δ 157.70, 148.30, 144.16, 139.51, 136.34, 134.85,
134.71, 133.09, 130.04, 128.83, 126.80, 125.39, 124.86, 119.81,
40.97, 34.97; MS (FAB, m/z) 266 (M+); MS (FAB, m/z) 149
(OTf-).
Gen er a l P r oced u r e K. 2-F lu or o-5-m eth yl-11-ch lor o-
qu in d olin iu m Hyd r och lor id e (25l). A suspension of 24l
(2.34 g, 5.38 mmol), anhydrous Na2CO3 (100 g), and chloroform
(∼100 mL) was sonicated and then concentrated to dryness.
The adsorbate was loaded onto a basic alumina column and
eluted with chloroform to remove the minor quindoline impu-
rity. Elution with 2-4% methanol in chloroform gave 1.48 g
(97%) of 11-chloro-5-methylquindoline as a purple solid.
The purple base was dissolved in chloroform (20 mL) and
acidified to a yellow endpoint with a 1 M solution of HCl in
ether. The product was filtered, washed with ether, dried, and
then recrystallized from chloroform-methanol (5:1), using a
minimum amount of ether to facilitate crystallization, provid-
ing 1.42 g (90%) of 25l as yellow crystals, mp 211.8-212.5 °C:
1H NMR (DMSO-d6) δ 13.55 (s, 1H), 8.99 (dd, J ) 10.0, J )
4.4, 1H), 8.82 (d, J ) 8.4, 1H), 8.40 (dd, J ) 9.2, J ) 2.8, 1H),
8.25-8.18 (m, 1H), 7.99 (ddd, J ) 8.4, J ) 8.4, J ) 0.8, 1H),
7.88 (d, J ) 8.0, 1H), 7.56 (ddd, J ) 8.4, J ) 8.4, J ) 0.8, 1H),
5.04 (s, 3H); 13C NMR (DMSO-d6) δ 160.48 (d, J ) 250.5),
146.01, 138.95, 134.75, 132.85, 132.53, 127.58 (d, J ) 5.6),
126.71, 124.94 (d, J ) 10.6), 122.51 (d, J ) 9.3), 122.33 (d, J
) 27.5), 122.20, 114.42, 113.56, 108.44 (d, J ) 25.4), 40.02;
MS (EI, m/z) 285 (M+), 287 (M + 2+). Anal. (C16H12N2ClF‚
2.6H2O) C, H, N.
Gen er a l P r oced u r e L. 2-F lu or o-5-m eth yl-11-(p h en yl-
a m in o)qu in d olin iu m Hyd r och lor id e (41b). A solution of
2-fluoro-5-methyl-11-chloroquindoline (freebase of 25l) (530
mg, 1.86 mmol) and aniline (0.75 g, 8.06 mmol) in 2-ethoxy-
ethanol (50 mL) was heated at 120 °C for 10 min. During this
time the color of the reaction mixture changed from purple to
yellow, and a yellow precipitate formed. After cooling, the
reaction mixture was diluted with ether (100 mL). The
precipitate was collected, washed thoroughly with ether, and
then dried to give 555 mg (79%) of 41b, mp 280 °C (sub-
limed): 1H NMR (DMSO-d6) δ 11.31 (s, NH), 10.62 (s, NH),
8.64 (d, J ) 8.0, 2H), 8.47 (d, J ) 10.8, 1H), 8.04 (dd, J ) 8.0,
J ) 8.0, 1H), 7.77-7.68 (m, 2H), 7.47-7.39 (m, 3H), 7.29-
7.22 (m, 3H), 4.83 (s, 3H); 13C NMR (acetone-d6) δ 152.50 (d,
J ) 245), 136.01, 132.03 (d, J ) 3.0), 130.95, 130.52, 127.76,
125.68, 123.12, 119.52, 117.72, 115.57, 115.32, 115.08, 113.86
(d, J ) 9.6), 113.37, 112.07 (d, J ) 9.7), 107.81, 106.57, 101.60
(d, J ) 24.7), 32.13; MS (EI, m/z) 342 (M+). HRMS (FAB) calcd
for C22H17N3F+ 342.1406, found 342.1433. Anal. (C22H17N3-
ClF) C, H, N.
6-F lu or o-5-m eth yl-11-ch lor oqu in d olin iu m Hyd r och lo-
r id e (25m ). Hydrotriflate salt 24m was prepared from 21m
(0.11 g, 3.9 mmol) and methyl triflate (0.11 mL, 0.97 mmol)
according to general procedure F. The hydrotriflate salt was
converted to its hydrochloride salt according to general pro-
cedure G. Purification on basic alumina (chloroform-EtOAc
(1:0 to 4:1), then 2% methanol in chloroform) and reconversion
to the hydrochloride salt according to general procedure K gave
0.11 g (98%) of 25m as a bright orange solid, mp 232-233.2
°C (dec): 1H NMR (DMSO-d6/CD3OD 1:1 v/v) δ 8.64 (d, J )
9.2, 1H), 8.63 (dd, J ) 8.0, J ) 1.2, 1H), 8.18 (ddd, J ) 9.2, J
) 6.6, J ) 1.2, 1H), 8.00 (ddd, J ) 8.4, J ) 6.2, J ) 1.2, 1H),
7.89-7.84 (m, 1H), 7.59 (dd, J ) 8.4, J ) 0.4, 1H), 7.22 (ddd,
J ) 13.6, J ) 8.0, J ) 0.8, 1H), 4.98 (s, 3H). HRMS (EI) calcd
for C16H10N2Cl2F 284.0516, found 284.0534. Anal. (C16H11N2-
Cl2F‚1.5H2O) C, H, N.
11-Ch lor o-5-m eth ylben zofu r o[3,2-b]qu in olin e Tr iflu o-
r om eth a n esu lfon a te (33). A suspension of 31 (0.40 g, 1.58
mmol) in anhydrous toluene (15 mL) was treated with methyl
triflate (0.52 g, 0.36 mL, 200 mol %). The reaction mixture
was stirred for 20 h at room temperature, diluted with ether,
and then filtered and dried to give 0.41 g (62%) of 33, mp 221-
223 °C: 1H NMR (DMSO-d6) δ 8.90 (d, J ) 8.0, 1H), 8.88 (d,
J ) 8.0, 1H), 8.73 (dd, J ) 8.0, J ) 1.2, 1H), 8.38 (ddd, J )
8.2, J ) 7.2, J ) 1.2, 1H), 8.22 (dd, J ) 8.2, J ) 0.8, 1H),
8.19-8.14 (m, 2H), 7.83 (ddd, J ) 7.6, J ) 4.4, J ) 1.2, 1H),
5.01 (s, 3H); 13C NMR (DMSO-d6) δ 159.53, 145.04, 142.89,
137.44, 136.69, 134.57, 130.32, 129.93, 127.05, 126.02, 125.43,
124.73, 119.23, 116.62, 113.80, 40.00; MS (FAB, m/z) 268 (M+).
HRMS (FAB) calcd for C16H11NOCl+ 268.0529, found 268.0545.
2-F lu or o-5-m eth yl-11-(p h en oxy)qu in d olin iu m Hyd r o-
ch lor id e (42b). Reaction of 2-fluoro-5-methyl-11-chloroquin-
doline (510 mg, 1.79 mmol) and phenol (1.5 g, 15.96 mmol) in
2-ethoxyethanol (50 mL) at 100 °C for 10 min according to
general procedure L gave 410 mg (60.5%) of 42b, mp 249.5-
249.8 °C: 1H NMR (DMSO-d6) δ 12.94 (s, 1H), 8.98 (dd, J )
10.0, J ) 4.4, 1H), 8.84 (d, J ) 8.4, 1H), 8.16 (ddd, J ) 10.0,
J ) 8.0, J ) 2.8, 1H), 8.01 (dd, J ) 8.8, J ) 2.8, 1H), 7.94
(ddd, J ) 8.4, J ) 6.8, J ) 0.8, 1H), 7.77 (d, J ) 8.4, 1H), 7.55
(ddd, J ) 8.4, J ) 6.8, J ) 0.8, 1H), 7.44-7.40 (m, 2H), 7.22
(dd, J ) 8.4, J ) 8.4, 1H), 7.15-7.12 (m, 2H), 5.06 (s, 3H); 13
C
NMR (DMSO-d6) δ 159.65 (d, J ) 248), 156.97, 145.58, 144.49
(d, J ) 5.2), 141.74, 134.57, 134.30, 130.29, 127.14, 126.18,
124.21, 122.41 (d, J ) 7.7), 122.23 (d, J ) 9.0), 122.02 (d, J )
9.8), 121.85, 116.10, 114.57, 113.51, 106.46 (d, J ) 26.2), 40.60;
MS (EI, m/z) 342 (M - H+). HRMS (FAB) calcd for C22H16N2-
FO+ 343.1247, found 343.1222. HRMS (EI) calcd for C22H15N2F
342.1168, found 342.1183. Anal. (C22H16N2OClF‚2H2O) C, H,
N.
2-F lu or o-5-m et h yl-11-[(4-ch lor op h en yl)t h io]q u in d o-
lin iu m Hyd r och lor id e (43b). Reaction of 2-fluoro-5-methyl-
11-chloroquindoline (56 mg, 0.21 mmol) and 4-chlorothiophenol
(36 mg, 120 mol %) in 2-ethoxyethanol (5 mL) according to
general procedure L gave 60 mg (73%) of 43b, mp 254-256
°C: 1H NMR (DMSO-d6) δ 13.16 (s, 1H), 9.00 (dd, J ) 10.0, J
) 4.4, 1H), 8.85 (d, J ) 8.4, 1H), 8.37 (dd, J ) 9.6, J ) 2.8,
1H), 8.15 (ddd, J ) 10.0, J ) 7.6, J ) 2.8, 1H), 7.98 (dd, J )
8.0, J ) 8.0, 1H), 7.87 (d, J ) 8.4, 1H), 7.57 (dd, J ) 7.6, J )
7.6, 1H), 7.38-7.33 (m, 4H), 5.10 (s, 3H); 13C NMR (DMSO-
d6) δ 160.35 (d, J ) 249), 146.23 (146.04), 138.84, 138.10
(137.95), 134.86, 132.62, 132.32, 131.99, 130.40, 129.60, 128.51
(d, J ) 5.3), 126.91, 125.68 (d, J ) 2.0), 122.64 (d, J ) 9.8),
122.08, 121.49 (d, J ) 25.6), 114.51, 113.59, 109.66 (d, J )
11-Ch lor o-5-m eth ylben zoth ien o[3,2-b]qu in olin iu m Tr i-
flu or om eth a n e-su lfon a te (34). A solution of 32 (0.16 g, 0.59
mmol) in anhydrous toluene (12 mL) was treated with methyl
triflate (0.081 mL, 120 mol %), and the reaction mixture was
stirred for 20 h at room temperature. The yellow product
which formed was filtered, triturated with EtOAc for 24 h,
filtered, and then dried to give 97 mg (38%) of 34, mp 187-
189 °C: 1H NMR (DMSO-d6) δ 9.01 (d, J ) 8.0, 1H), 8.95 (d,
J ) 9.2, 1H), 8.69 (d, J ) 8.4, 1H), 8.53 (d, J ) 8.0, 1H), 8.43
(dd, J ) 8.8, J ) 8.8, 1H), 8.21 (dd, J ) 8.0, J ) 8.0, 1H), 8.09
(dd, J ) 7.6, J ) 7.6, 1H), 7.89 (dd, J ) 8.0, J ) 8.0, 1H), 5.07
(s, 3H); 13C NMR (DMSO-d6) δ 148.34, 143.68, 143.11, 139.34,
135.72, 133.95, 130.59, 130.08, 128.35, 127.10, 125.12, 124.89,

Antihyperglycemic Activities of Cryptolepine Analogues
J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 15 2763
25.1), 41.24; MS (FAB, m/z) 393 (M+). HRMS (FAB) calcd for
C
Cl2FS‚1.25H2O) C, H, N.
(3) It is possible that the accepted structure for cryptolepine was
actually isolated 45 years after its initial synthesis by Fichter
and Boehringer. Clinquart isolated a purple compound in 1929
(mp 160 °C), assumed to possess a primary NH2 group, from
the roots of C. triangularis N. E. Br. collected in Kisantu (Belgian
Congo). He named this compound cryptolepine and proposed the
formula C14H17N2O4 (see ref 2). Delvaux isolated a purple base
in 1931 (mp 193-194 °C) from the same plant material which
he named cryptolepine and proposed the formula C17H16N2O (see
refs 4 and 5). The accepted formula for cryptolepine, C16H12N2,
was first isolated as a purple solid (mp 175-178 °C) by Gelle´rt
in 1951 from C. sanguinolenta (see ref 6). Gelle´rt assigned the
structure of (1) to cryptolepine. The structure of cryptolepine
was recently verified by NMR spectroscopy in 1991 by Tackie
et al. (see ref 7) and by unambiguous synthesis in 1998 by Bierer
et al. (see ref 8).
(4) Delvaux, E. Sur la Cryptole´pine. (On Cryptolepine.) J . Pharm.
Belg. 1931, 13, 955-959.
(5) Delvaux, E. Sur la Cryptole´pine. (On Cryptolepine.) J . Pharm.
Belg. 1931, 13, 973-977.
(6) Gelle´rt, E.; Raymond-Hamet; Schlittler, E. Die Konstitution des
Alkaloids Cryptolepin. (The Constitution of the Alkaloid Cryp-
tolepine.) Helv. Chim. Acta 1951, 34, 642-651.
(7) Tackie, A. N.; Sharaf, M. H. M.; Schiff, P. L., J r.; Boye, G. L.;
Crouch, R. C.; Martin, G. E. Assignment of the Proton and
Carbon NMR Spectra of the Indoloquinoline Alkaloid Cryptol-
epine. J . Heterocycl. Chem. 1991, 28, 1429-1435.
(8) Bierer, D. E.; Fort, D. M.; Mendez, C. D.; Luo, J .; Imbach, P. A.;
Dubenko, L. G.; Gerber, R. E.; Litvak, J .; Lu, Q. Zhang, P.; Reed,
M. J .; Waldeck, N.; Bruening, R. C.; Noamesi, B. K.; Hector, R.
F.; Carlson, T. J .; King, S. R. Ethnobotanical-Directed Discovery
of the Antihyperglycemic Properties of Cryptolepine: Its Isola-
tion from Cryptolepis sanguinolenta, Synthesis, and In Vitro and
In Vivo Activities. J . Med. Chem. 1998, 41, 894-901.
(9) Dalziel, J . M. African Leather Dyes. Kew Bull. 1926, 225-238.
(10) Dalziel, J . M. The Useful Plants of West Tropical Africa; Crown
Agents for Overseas Governments and Administrations: Lon-
don, 1955; 2nd reprint, p 387.
22H15N2ClFS+ 393.0629, found 393.0618. Anal. (C22H15N2-
2-F lu or o-5-m eth yl-11-p h en ylqu in d olin iu m Hyd r och lo-
r id e (44b). A suspension of 2-fluoro-5-methyl-11-chloroquin-
doline (300 mg, 0.35 mmol) in dry dioxane (30 mL) was added
to a 1.0 M solution of phenylmagnesium bromide in ether (6
mL, 6.0 mmol) at room temperature. The reaction mixture
was stirred for 1 h at room temperature and then heated for
1 h at 50 °C. The reaction mixture was cooled, poured into
ice-water (300 mL), left to stand overnight, and then extracted
with EtOAc (4 × 50 mL) and chloroform (4 × 50 mL). The
combined chloroform and ethyl acetate extract was washed
with water, dried, and then concentrated. The residue was
purified on a basic alumina column, eluting 0.5-1.5% metha-
nol in chloroform. The product containing fractions were
combined, acidified with 1.0 M HCl solution in ether, and then
filtered to give 175 mg (45.8%) of 44b as a yellow solid, mp
251.4-251.8 °C: 1H NMR (DMSO-d6) δ 12.44 (s, NH), 9.00
(dd, J ) 9.2, J ) 4.4, 1H), 8.85 (d, J ) 8.4, 1H), 8.17 (dd, J )
7.6, J ) 7.6, 1H), 7.93 (dd, J ) 7.8, J ) 7.8, 1H), 7.81-7.70
(m, 6H), 7.62 (dd, J ) 9.6, J ) 2.8, 1H), 7.54 (dd, J ) 7.2, J )
7.2, 1H), 5.12 (s, 3H); 13C NMR (DMSO-d6) δ 159.66 (d, J )
249.1), 146.35, 138.27, 136.45, 134.13, 132.80 (d, J ) 4.2),
131.15, 130.34, 129.94, 129.60, 129.39, 126.20 (d, J ) 9.9),
121.88 (d, J ) 9.1), 121.72, 121.66, 121.46, 114.01, 113.49,
109.94 (d, J ) 24.1), 49.94; (LSIMS, m/z) 327 (M+). HRMS
(FAB) calcd for C22H16N2F+ 327.1297, found 327.1281.
Biologica l Assa ys. In Vitr o Glu cose Tr a n sp or t Assa y
(w ith ou t Exogen ou sly Ad d ed In su lin ): [3H]-2-Deoxy-D-
glu cose Up ta k e in Differ en tia ted 3T3-L1 Ad ip ocytes.
Compounds were initially screened at three concentrations:
3, 10, and 30 µM. Compounds of interest were evaluated
further at 0.1, 0.3, 1, 3, 10, 30 µM final concentrations or at
0.3, 1, 3, 10, 30, and 100 µM final concentrations. Adipocytes
were treated (in triplicate) for 18 h with a test compound at
the appropriate concentration. Details of this assay have been
previously described.8
(11) See ref 8 and references therein.
(12) Luo, J .; Fort, D. M.; Carlson, T. J .; Noamesi, B.; nii-Amon-Kotei,
D.; King, S. R.; Tsai, J .; Quan, J .; Hobensack, C.; Lapresca, P.;
Waldeck, N.; Mendez, C. D.; J olad, S. D.; Bierer, D. E.; Reaven,
G. M. Cryptolepine, A Potentially Useful New Antihyperglycae-
mic Agent Isolated from Cryptolepis sanguinolenta: An Example
of the Ethnobotanical Approach to Drug Discovery. Diabetic Med.
1998, 15, 367-374.
(13) Bierer, D. E.; Carlson, T. J .; King, S. R. Integrating Indigenous
Knowledge, Tropical Medicinal Plants, Medicine, Modern Sci-
ence and Reciprocity into a Novel Drug Discovery Approach.
Network Sci. 1996, 2 (No. 5) (http://www.awod.com/ netsci/Issues/
May96/feature2.html).
(14) Bierer, D. E.; Dener, J . M.; Dubenko, L. G.; Gerber, R. E.; Litvak,
J .; Peterli, S.; Peterli-Roth, P.; Truong, T. V.; Mao, G.; Bauer,
B. E. Novel 1,2-Dithiins: Synthesis, Molecular Modeling Studies,
and Antifungal Activity. J . Med. Chem. 1995, 38, 2628-2648.
(15) Yamato, M.; Takeuchi, Y.; Chang, M.; Hashigaki, K.; Tsuruo,
T.; Tashiro, T.; Tsukagoshi, S. Synthesis and Antitumor Activity
of Fused Quinoline Derivatives. Chem. Pharm. Bull. 1990, 38,
3048-3052.
(16) Yamato, M.; Takeuchi, Y.; Chang, M.; Hashigaki, K. Synthesis
and Antitumor Activity of Fused Quinoline Derivatives. II. Novel
4- and 7-Hydroxyindolo-[3,2-b]quinolines. Chem. Pharm. Bull.
1992, 40, 528-530.
(17) Go¨rlitzer, K.; Stockmann, R.; Walter, R. D. Gegen Malaria
wirksame 10H- Indolo[3,2-b]chinolin-11-yl-amine. Part 1: Phenol-
Mannich-Basen vom Amodiaquin- und Cycloquin-Typ. (Antima-
larial 10H-Indolo[3,2-b]quinolin-11-ylamines. Part I: Phenol-
Mannich Bases of the Amodiaquine and Cycloquine Type.)
Pharmazie 1994, 49, 231-235.
(18) Go¨rlitzer, K.; Weber, J . Anellierte Chinoline, 6. Mitt. 10H-Indolo-
[3,2-b]chinoline (Fused Quinolines VI: 10H-Indolo[3,2-b]quino-
line). Arch. Pharm. (Weinheim) 1981, 314, 852-861.
(19) Chang, M.; Takeuchi, Y.; Hashigaki, K.; Yamato, M. Synthesis
of 7-Substituted Indolo[3,2-b]quinoline Derivatives. Heterocycles
1992, 33, 147-152.
(20) Sunder, S.; Peet, N. P. Synthesis of Benzofuro[3,2-b]quinoline-
6(11H)one and Derivatives. J . Heterocycl. Chem. 1978, 15, 1379-
1382.
In Vivo Stu d ies Usin g d b/d b Mice. The testing protocol
in db/db mice has been previously described.8
Ack n ow led gm en t. The authors wish to thank Zhi
J un Ye, J ohn Kuo, J ian Lu Chen, and Nigel Parkinson
for their assistance in obtaining NMR and mass spectral
data; Steven R. King, Thomas J . Carlson, and Richard
Hector for valuable scientific discussions; Michael Ce-
niceros for editorial assistance; Dominic Brignetti for
assistance with biological figures; Dr. D. nii-Amon-Kotei
and Dr. Ben K. Noamesi from the University of Science
and Technology, Kumasi, Ghana, for their scientific
collaboration; and Professor Henry Rapoport at the
University of California, Berkeley for valuable consulta-
tion. We also thank the traditional healers, communi-
ties, botanists, physicians, governmental agencies, and
nongovernmental agencies from Nigeria, Guinea, Cam-
eroon, Zaire, Tanzania, Ghana, and Congo that have
contributed to diabetes research at Shaman Pharma-
ceuticals.26
Su p p or tin g In for m a tion Ava ila ble: Experimental pro-
cedures and characterization data for all compounds not
described above (26 pages). Ordering information is given on
any current masthead page.
(21) Go¨rlitzer, K.; Weber, J . Anellierte Chinoline 4. Mitt. 11-Oxo-5,
11-dihydrobenzothieno[3,2-b][1]chinoline, S, S.-Dioxide und Thio-
nierungsprodukte. (Fused Quinolines. Part 4. 5,11-Dihydroben-
zothieno[3,2-b][1]quinolin-11-ones, S,S-Dioxides and Thionation
Products.) Arch. Pharm. (Weinheim) 1980, 314, 76-84.
(22) Go¨rlitzer, K. Untersuchungen an 1,3-Dicarbonylverbindungen,
11 Mitt. Anellierte Chinolone II, N- und O-Alkylierungsprodukte.
(Studies on 1,3-Dicarbonyl Compounds, XI: Fused Quinolines
II, N- and O-Alkylated Products.) Arch. Pharm. (Weinheim)
1979, 312, 254-261.
Refer en ces
(1) Fichter, F.; Boehringer, R. Ueber Chindolin. (Over Quindoline.)
Chem. Ber. 1906, 39, 3932-3942.
(2) Clinquart, E. Sur la composition chimique de Cryptolepis trian-
gularis, plante congolaise. (On the Chemical Composition of
Cryptolepis triangularis, Plants from the (Belgian) Congo.) Bull.
Acad. R. Med. Belg. 1929, 9, 627-635.

2764 J ournal of Medicinal Chemistry, 1998, Vol. 41, No. 15
Bierer et al.
(23) Kempter, G.; Noack, H.; Russ, G. Heterocyclen aus Aminoke-
tonen; Reaktionen monoalkylierter Amine mit aktiven Halo-
geniden. (Heterocycles from Amino Ketones: Reaction of Mono-
alkylated Amines with Active Halides.) Z. Chem. 1963, 3, 352-
353.
3:4 Benz-∆-carboline is a Requirement for Renal Vasodilation
and Inhibition of Platelet Aggregation in the Rat. Med. Chem.
Res. 1996, 6, 602-610.
(25) Bierer, D. E. Cryptolepine Analogs with Hypoglycemic Activity.
U.S. Patent 5,681,958, October 28, 1997, filed J une 7, 1995.
(26) Moran, K. In Biodiversity and Human Health; Grifo, F., Rosenthal,
J ., Eds.; Island Press: Washington, D.C., 1997; pp 243-263.
(24) After this study was completed,25
a report appeared which
studied the effect that N-5 substitution of quindoline had on
renal vasodilation and inhibition of platelet aggregation in a rat
model. See: Oyekan, A. O.; Ablordeppy, S. Y. N-5 Alkylation of
J M970735N