Synthesis and topoisomerase I inhibitory properties of luotonin A analogues

Article abstract of DOI:10.1016/j.bmc.2004.08.052

Luotonin A, a naturally occurring pyrroloquinazolinoquinoline alkaloid, has been previously demonstrated to be a topoisomerase I poison. A number of luotonin A derivatives have now been prepared through the condensation of anthranilic acid derivatives and 1,2-dihydropyrrolo[3,4-b]quinoline-3-one in the presence of phosphorus oxychloride. When dichloromethane was used as solvent the reaction proceeded to a single product. In contrast when the reaction was carried out in tetrahydrofuran or in phosphorus oxychloride, an additional isomeric product was obtained. The luotonin A analogues were evaluated for their ability to effect stabilization of the covalent binary complex formed between human topoisomerase I and DNA, and for cytotoxicity toward a yeast strain expressing the human topoisomerase I.

Full text of DOI:10.1016/j.bmc.2004.08.052

Bioorganic & Medicinal Chemistry 12 (2004) 6287–6299
Synthesis and topoisomerase I inhibitory properties of
luotonin A analogues
Ali Cagir, Brian M. Eisenhauer, Rong Gao, Shannon J. Thomas and Sidney M. Hecht*
Department of Chemistry and Department of Biology, University of Virginia, Charlottesville, VA 22901, USA
Received 28 June 2004; accepted 18 August 2004
Available online 30 September 2004
Abstract—Luotonin A, a naturally occurring pyrroloquinazolinoquinoline alkaloid, has been previously demonstrated to be a top-
oisomerase I poison. A number of luotonin A derivatives have now been prepared through the condensation of anthranilic acid
derivatives and 1,2-dihydropyrrolo[3,4-b]quinoline-3-one in the presence of phosphorus oxychloride. When dichloromethane was
used as solvent the reaction proceeded to a single product. In contrast when the reaction was carried out in tetrahydrofuran or
in phosphorus oxychloride, an additional isomeric product was obtained. The luotonin A analogues were evaluated for their ability
to effect stabilization of the covalent binary complex formed between human topoisomerase I and DNA, and for cytotoxicity toward
a yeast strain expressing the human topoisomerase I.
ꢀ
2004 Elsevier Ltd. All rights reserved.
9
2
7
1
. Introduction
5
O
10
8
3
6
4
N
N
O
1
1
3
21
20
Camptothecin (CPT, 1) is a cytotoxic alkaloid first iso-
lated from Camptotheca acuminata in 1966 by Wall
et al. (Fig. 1). Two semi-synthetic camptothecin ana-
logues, topotecan, and irinotecan, are now used clini-
cally for the treatment of ovarian small cell lung and
N
2
1
N
1
1
19
18
O
N
1
4
15
1
1
6
OH O
17
1
2
2
colon cancers. The success of these analogues in the
clinic has encouraged studies to identify additional topo-
Figure 1. Structures of CPT (1) and luotonin A (2).
3
isomerase I poisons. CPT and its analogues are believed
to function primarily at the locus of the topoisomerase
I–DNA covalent binary complex.
The nature of the CPT–topoisomerase I–DNA ternary
8
complex has been studied by molecular modeling and
9
X-ray crystallographic analysis. While the detailed
4
models resulting from these studies differ significantly,
two types of interactions appear to contribute to the
binding of CPT to the topoisomerase I–DNA covalent
binary complex. These include H-bonding interactions
with the (S)-a-hydroxylactone (E-ring) and stacking of
CPT between adjacent base pairs in the DNA substrate
at the site of reversible DNA cleavage. The importance
of the 20(S)-a-hydroxylactone has also been defined by
preparing E-ring modified analogues. While 20-chloro,
bromo, and amino derivatives of CPT could stabilize
the topoisomerase I–DNA covalent binary complex,
Topoisomerase I is an enzyme that catalyzes the relax-
ation of supercoiled chromosomal DNA during replica-
5
tion and transcription. This process is initiated by
reversible) covalent attachment of topoisomerase I to
the DNA backbone, creating a transient single strand
(
5
break in the DNA duplex. The formed topoisomerase
I–DNA covalent binary complex normally undergoes
religation to afford an intact duplex following DNA
relaxation, but can be stabilized by camptothecin via
formation of a ternary complex. Persistence of the sta-
7
bilized DNA break leads to cell death.
6
20(R)-OH CPT, and 20-deoxy CPT were essentially
incapable of stabilizing the binary complex.
1
0,11
While the relative contributions of hydrogen bonding
and p-stacking interactions to stabilization of the
CPT–topoisomerase I–DNA ternary complex are not
yet clear, the importance of p-stacking interactions
Keywords: Camptothecin; Luotonin A; Topoisomerase I–DNA inhi-
bition; Structure–activity relationships.
*
Corresponding author. Tel.: +1 434 924 3906; fax: +1 434 924
856; e-mail: sidhecht@virginia.edu
7
0
968-0896/$ - see front matter ꢀ 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2004.08.052

6288
A. Cagir et al. / Bioorg. Med. Chem. 12 (2004) 6287–6299
was reinforced by the finding that luotonin A (2) is a
1
O
1
topoisomerase I poison. Luotonin A (2) is a pyrrolo-
quinazolinoquinoline alkaloid, originally isolated from
OMe
NH2
O
1
2
N
Peganum nigellastrum (Fig. 1). The structure of 2 is
similar to that of CPT. The 20(S)-a-hydroxylactone ring
in CPT is replaced by a benzene ring in luotonin A and
the (14)-CH moiety of CPT is replaced by a nitrogen
atom in luotonin A. In addition to their structural sim-
ilarities, we have also reported that topoisomerase
I-mediated DNA cleavage exhibits the same pattern in
the presence of both 1 and 2, consistent with the possi-
bility that CPT and luotonin A interact with the topo-
isomerase I–DNA binary complex in a similar fashion.
One notable difference between the compounds is the
potency of luotonin A, which is significantly less efficient
than CPT in stabilizing the topoisomerase I–DNA cova-
lent binary complex and approximately 10-fold less
cytotoxic than camptothecin toward a yeast cell line
N
+
N
2
N
N
Cl
3
NH
N
O
4
Figure 2. Retrosynthetic analysis of luotonin A.
1
1,13
expressing human topoisomerase I.
This difference
might well be due to lack of the (S)-a-hydroxylactone
ring of CPT and consequent absence of hydrogen bonds
between the drug and the topoisomerase I–DNA binary
complex. We have shown previously that the potency of
ment of pyrroloquinoline 4 with phosphorus oxychlo-
ride (Figs. 2and 3 ). While Lee et al. reported that the
chlorination of 4 with phosphorus oxychloride followed
by treatment with methyl anthranilate gave either low
1
4
luotonin A may be altered by substitution in ring E.
1
5a
yields or side products, we found that the condensa-
tion of 4 with methyl anthranilate in the presence of
phosphorus oxychloride at 40ꢁC prevented the forma-
Presently, we have explored further alterations of the
E-ring of luotonin A to define the effects of potential
p-stacking and H-bonding interactions on the ability
of luotonin A derivatives to function as topoisomerase
I poisons.
14
tion of side products. As shown in Figure 3, eight
new luotonin A derivatives modified in ring E (5a–g)
were prepared by this method, as well as 16,17,18,19-
tetrahydroluotonin A (6) (Fig. 3). The yields of products
5 varied from 11% to 62%. Compound 7, prepared from
2. Results
5
d by demethylation with hydrogen bromide at reflux,
1
1
2
5
Since the initial report of the isolation of luotonin A,
eleven different synthetic routes have been reported.
has a phenolic OH substitutent oriented roughly in the
same position as the 20(S)-OH group of CPT (cf. Figs.
1 and 4). Luotonin B (8) (Fig. 4), another natural alka-
Five of these routes have utilized a coupling of the pyr-
1
6
12
roloquinoline precursor
derivatives as the final step.
4
and anthranilic acid
5a–e
We envisioned the
loid isolated from Peganum nigellastrum, was prepared
1
from luotonin A in two steps. First the methylene car-
bon was mono-brominated with N-bromosuccinimide,
and then the bromine was replaced with OH by treat-
preparation of luotonin A derivatives by the condensa-
tion of anthranilic acid methyl esters with imino chloride
1
5a
15j
3
,
the latter of which can be obtained in situ by treat-
ment with silver oxide at reflux.
O
N
POCl3,
o
CH Cl , 40 C
2
2
NH
N
N
N
R3
O
O
R3
R2
R1
R2
O
4
5a-g
H N
2
R1
R1
R2
R3
yields(%)
a H
b H
c H
Cl
F
F
H
H
F
22
31
22
17
11
62
12
O
N
N
d OMe
e H
H
H
N
OMe OMe
OMe OMe
f
OMe
6
g H
H
Me
yield (6 %)
Figure 3. Synthesis of luotonin A derivatives 5a–g and 6.

A. Cagir et al. / Bioorg. Med. Chem. 12 (2004) 6287–6299
6289
OH
N
by addition of methyl anthranilate reversed the ratio of
products, giving 4% luotonin A and 22% of the side
product (Fig. 6). Mass spectrometric analysis of the side
product indicated that it had the same molecular weight
O
N
O
N
N
N
1
N
as luotonin A. H NMR analysis showed that the pro-
7
HO
8
ton attached to C-7 of luotonin A (Fig. 1) was shifted
downfield approximately 0.5ppm, in principle consistent
with the interpretation that the adjacent methylene
group was substituted. However, the resonance for the
Figure 4. Structures of 16-hydroxyluotonin A (7) and luotonin B (8).
1
An interesting rearrangement has been reported previ-
ously for 4. Treatment of 4 with phosphorus pentasul-
fide gave 78% of 9 and 6% of isomeric 10. Slow
conversion of 9 to 10 was also observed in the presence
of phosphorus pentasulfide at 85ꢁC or hydrogen sulfide
at 120ꢁC in the presence of 1% aqueous pyridine
C-5 methylene group was still present in the H NMR
spectrum of the side product. Additionally, while the
NOESY spectrum of 2 showed a crosspeak between
the protons attached to C-7 and C-5, this crosspeak
was absent in the side product.
1
7
(
Fig. 5).
The structure of the side product was assigned as 11
(
Sugasawa et al. To support this assignment the iso-
Fig. 6) based on the rearrangement reported by
1
7
Similarly, the formation of a side product has been
reported for the condensation of 4 with methyl anthran-
1
8
meric pyrroloquinoline precursor 12 was prepared
and treated with methyl anthranilate under the same
conditions. Luotonin A was produced as the minor
product (3%) while isoluotonin A (11) was obtained in
somewhat greater (6%) yield (Fig. 6).
15a
ilate. We were able to separate this side product from
luotonin A. Treatment of a mixture of 4 and methyl
anthranilate with phosphorus oxychloride in tetra-
hydrofuran at 40ꢁC gave 12% luotonin A and 4% of
the side product. The same result was obtained when
the reaction was performed neat. In contrast, when the
sequence of addition was changed the ratio of the prod-
ucts also changed. Treatment of 4 with phosphorous
oxychloride at 40ꢁC in tetrahydrofuran for 1h followed
As shown in Table 1, a number of E-ring substituted
luotonin A derivatives (14a–20a) and isoluotonin A
derivatives (13, 14b–20b) were prepared and character-
ized. Additionally different luotonin A and isoluotonin
S
P2S
5
NH
NH
+
N
N
NH
O
S
N
4
9
10
Figure 5. Synthesis of thiolactam 9 and formation of rearranged product 10.
POCl , THF
3
O
N
O
o
40
C
N
N
N
NH
+
N
N
O
N
O
MeO
4
2
11
(4%)
H
2
N
(12%)
1
. POCl , THF
C, 1h
3
O
N
O
o
40
N
N
N
N
NH
+
N
O
N
2
.
O
MeO
H N
4
2
(4%)
11
(22%)
2
POCl , THF
O
N
O
O
3
o
4
0
C
N
N
+
NH
N
O
N
N
N
MeO
H N
12
2
(3%)
11
(6%)
2
Figure 6. Synthesis of luotonin A (2) and formation of rearranged product isoluotonin A (11).

6290
A. Cagir et al. / Bioorg. Med. Chem. 12 (2004) 6287–6299
Table 1. Synthesis of E-ring modified luotonin A derivatives
POCl
3
, THF
O
N
O
o
4
0
C
N
N
NH
N
+
N
O
N
N
R
3
N
R
3
O
R3
MeO
H
R1
5a, 14a-20a
R
R1
13, 14b-20b
R2
2
4
2
R
2
R
1
R
1
R
2
R
3
Yield (%)
Yield (%)
5
1
1
1
1
1
1
2
a, 13
H
H
H
H
H
H
H
CF
Cl
H
H
H
H
H
Br
H
H
37
13
24
19
15
14
9
9
<1
15
9
4a, 14b
5a, 15b
6a, 16b
7a, 17b
8a, 18b
9a, 19b
0a, 20b
CO
CO
Ph
2
Me
Bn
2
CN
H
12
2
a
a
NO
H
2
9
3
14
2
a
Prepared using the anthranilic acid derivative, rather than its methyl ester, at reflux for 1h.
1
1
A derivatives, having different aromatic ÔE-ringÕ systems
such as thiophene and naphthalene (21a–23a, and 21b–
In addition to luotonin A itself, we have recently de-
scribed the ability of several luotonin A analogues to
effect the stabilization of the human topoisomerase I–
2
3b), were prepared using the same method (Table 2).
1
4
DNA covalent binary. The derivatives tested included
5a–g, 6, and 7. Of these, compounds 5a,b, 6, and 7 were
found to exhibit reasonable stabilization of the binary
complex. In the present study, preliminary assays indi-
cated that many of the compounds stabilized the en-
zyme–DNA covalent binary complex weakly, if at all.
Accordingly, the assay was carried out initially at a high
concentration of the inhibitors (625lM) to facilitate the
detection of those compounds exhibiting any activity as
topoisomerase I poisons. The data are summarized in
Table 3, which quantified the extent of cleavage at each
of three sites relative to the cleavage observed in the
Several of the luotonin A and isoluotonin A derivatives
obtained were also converted to new analogues that may
either contribute to p-stacking interaction with DNA
bases or form hydrogen bonds with the topoisomerase
I–DNA covalent binary complex (Fig. 7). Compound
18a was transformed to 24 by the use of a Suzuki cou-
pling reaction in 7% yield. The nitro group of 19a was
reduced using tin(II)chloride in acidic medium to afford
2
5 in 40% yield. The methyl ester derivatives of luotonin
A (14a) and isoluotonin A (14b) were reduced to their
alcohols 26 and 27 in 21% and 32% yields, respectively.
Finally, the isomer of 17-fluoroluotonin A (28) was pre-
pared analogously starting from precursor 12.
1
9
presence of 50lM CPT. Those luotonin A derivatives
which exhibited reasonable activity in the initial assay
Table 2. Synthesis of luotonin A derivatives having different aromatic E-rings
POCl , THF
O
N
O
3
o
40
C
N
N
N
NH
N
+
N
N
O
O
R
1
O
4
21a-23a
21b-23b
H N
2
Entry
MeO
Yield (%)
Yield (%)
16
2
C
21a, 21b
22a, 22b
23a, 23b
22
33
2
S
S
2
H N
MeO C
2
3
2
H N
MeO C
2
0
H N
2

A. Cagir et al. / Bioorg. Med. Chem. 12 (2004) 6287–6299
6291
O
Pd(OAc)2
PhB(OH)
O
N
N
2
N
N
N
N
1
0% K
2
CO
3
Br
Ph
DMF
7%)
(
24
18a
O
O
N
N
SnCl
2
, HCl
N
N
N
N
MeOH, reflux
(40%)
NO
2
NH
2
1
9a
25
O
O
N
N
N
NaBH , THF
4
N
N
N
(21%)
CO
2
Me
CH
2
OH
1
4a
26
N
O
N
O
N
N
N
NaBH
32%)
4
, THF
N
(
CO
2
Me
CH
2
OH
14b
27
POCl , THF
O
N
O
O
3
o
40
C
N
N
+
NH
N
O
N
N
N
MeO
H N
F
F
12
5b
(2%)
28
(10%)
F
2
Figure 7. Conversion of luotonin A and isoluotonin A intermediates to new analogues.
were run again at 50lM concentration in direct compar-
ison with CPT and luotonin A (Fig. 8). As shown in the
figure, derivatives exhibiting reasonable activity as topo-
isomerase I poisons included isoluotonin A (11), 17-
cyanoluotonin A (17a), and 17-cyanoisoluotonin A
isomerase I gene under the control of a galactose pro-
1
3
moter. This yeast strain was also rendered permeable
to exogenous agents by virtue of a mutation in the gene
for an export pump. The topoisomerase I-dependent
cytotoxicity was determined by comparing the results
when this yeast strain was grown in raffinose or galac-
tose. In addition to those luotonin A derivatives previ-
ously shown to exhibit topoisomerase I-dependent
cytotoxicity (i.e., 5b,g, and 6), two additional derivatives
(11 and 25) were found to exhibit reasonably strong
topoisomerase I-dependent cytotoxic activity. Also
fairly strongly cytotoxic were thiophene derivatives
21a, 22a and 22b, although none was selectively cyto-
toxic to yeast expressing topoisomerase I.
(
17b), thiophene analogues 22a and 22b, as well as 17-
hydroxymethylluotonin A (26). Also found to be quite
active as a topoisomerase I poison was 17-aminoluoto-
nin A (25). It is interesting that the results obtained with
the corresponding luotonin A and isoluotonin A deriva-
tives were generally comparable. Thus both 17-chloro-
luotonin A (5a) and 17-chloroisoluotonin A (13)
afforded reasonably strong stabilization of the enzyme–
DNA covalent binary complex. This was also true
for the 17-fluoro derivatives of luotonin A (5b) and
isoluotonin A (28), which were both found to afford
reasonable stabilization of the covalent binary complex.
3. Discussion
Those compounds that mediated reasonable stabiliza-
tion of the topoisomerase I–DNA covalent binary com-
plex (Table 3 and Fig. 8) were tested for cytotoxicity in a
yeast strain that lacked the homologous topoisomerase
I, but harbored a plasmid containing the human topo-
The synthesis of luotonin A derivatives was accom-
plished using what has become a fairly common
strategy involving the condensation of pyrrolo-
quinoline 4 with anthranilic acid derivatives. While
the yields of products varied significantly, and were
1
5a–e

6292
A. Cagir et al. / Bioorg. Med. Chem. 12 (2004) 6287–6299
Table 3. Stabilization of human topoisomerase I mediated DNA
cleavage at three sites in the presence of luotonin A analogues at
a
625lM concentration
Compound
Site 1
Site 2Site 3
5
5
1
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
2
2
2
a
5.7
11.0
8.0
6.0
9.22
2.9
1.2
1.20.9
1.0
0.8
1.9
3.5
5.6
7.5
2.8
2
2.4
b
.5
1
2.6
0.9
3
2.6
1.0
4a
5a
5b
6a
6b
7a
7b
9a
9b
0a
0b
1a
1b
2a
2b
3a
3b
4
0.7
0.7
0.3
0.3
0.6
1.5
2.4
4.0
1.0
1.7
2.6
5.4
7.6
2.3
5.25.4.7
0.8
1.7
0.9
2.0
3.4
0
0.2
0.8
2.6
0
5.7
0.9
21.1
15.0
3.1
14.4
6.4
2.6
0
7.8
4.4
1.0
0.3
0.2
5.0
8.4
3.2
3.6
0.3
1.3
1.7
13.8
10.7
5.4
9.6
5
13.6
13.4
6.5
6
7
8
14.9
a
Relative to the intensity of the band at each site produced by 50lM
CPT.
sometimes rather low, it was possible to obtain sufficient
material in each case for the requisite biochemical and
biological assays. One interesting facet of the synthesis
was the formation of an isomer of luotonin A in which
the N atom in ring B was present at position-7 (luotonin
A numbering, Fig. 1), rather than position-1. Support
for this assignment was obtained in an experiment that
involved changing the order of addition of reagents,
and another that started with precursor 12, an isomer
of 4 that should lead unambiguously to the 7-aza, 1-dea-
za analogue of luotonin A (Fig. 6). An NOE experiment,
demonstrating the lack of a crosspeak between the pro-
tons at positions 5 and 7, also supported the assigned
structure 11.
Figure 8. Autoradiogram of a 10% denaturing gel showing the effect of
luotonin A analogues on human topoisomerase I-mediated cleavage of
0
32
a 222-base pair DNA restriction fragment 3 - P end labeled on the
scissile strand. The individual incubation mixtures contained 36ng of
topoisomerase I and 50lM luotonin A analogues as indicated.
Following incubation at 37ꢁC for 30min, the samples were digested
with proteinase K prior to polyacrylamide gel analysis. Lanes 1–10,
50lM compounds 11, 17a, 17b, 22a, 22b, 25, 26, 28, 2, and 1,
respectively. Lane 11, DNA + topoisomerase I.
Luotonin A analogues exhibiting good activity in the
stabilization of the enzyme–DNA covalent binary com-
plex included isoluotonin A (11), 17-aminoluotonin A
(25), 17-hydroxymethylluotonin A (26), and 17-fluoro-
isoluotonin A (28). Previously, luotonin A (2) and 17
fluoroluotonin A (5b) have been shown to exhibit good
The initial biochemical assays, involving stabilization of
topoisomerase I-mediated DNA cleavage of a 222-base
pair DNA restriction fragment, indicated rather modest
stabilization of cleavage for most of the luotonin A ana-
logues not studied previously. To assure that the lack of
stabilization was not due to unusual dose–response pro-
files, all of the new analogues were assayed at very high
concentration (Table 3). Those which afforded reasona-
ble stabilization of the topoisomerase I–DNA covalent
binary complex were tested at 50lM concentration in
direct comparison with luotonin A (Fig. 8). Surprisingly,
little difference in cleavage stabilization was observed at
the lower concentration, quite possibly reflecting the fact
that most of the analogues were poorly soluble in aque-
ous solution.
11,14
stabilization of the covalent binary complex,
indi-
cating that the position of the ring B N-atom does not
significantly affect stabilization of topoisomerase
I-mediated DNA cleavage.
Several luotonin A analogues were used to determine
their ability to mediate topoisomerase I-dependent cyto-
toxicity in a yeast strain that lacks the homologous
topoisomerase I, but expresses human topoisomerase I
when grown on galactose. As shown in Table 4, isoluoto-
nin A (11) exhibited topoisomerase I-dependent cyto-
toxicity and had an IC50 value (11.8lM) quite similar

A. Cagir et al. / Bioorg. Med. Chem. 12 (2004) 6287–6299
6293
Table 4. Human topoisomerase I-dependent cytotoxicity of CPT (1)
and luotonin A analogues toward S. cerevisiae
Table 4 (continued)
Com-
Concentration
% Inhibition on
growth medium
IC50 (lM)
for galactose
Com-
Concentration
% Inhibition on
growth medium
IC50 (lM)
for galactose
pound (lM)
pound (lM)
Raffinose Galactose
Raffinose Galactose
1
2.5
6.25
5
3
7
19
1
15
1
21
23
6
74
59
51
38
35
22
0.86
.5
c
2
2a
2b
100
63
81
81
74
—
0
.75
.39
.19
.03
5
2
1
0
0
26
19
40
5
94
86
0
2.5
6
97
97
90
93
0
.25
2
5
25
68
35
0
81
56
36
23
9.58
36.3
c
2
100
88
80
68
72
—
1
0
1
0
5
2
1
0
5
65
54
.5
0
2.5
6.25
31
49
59
36
b
100
25
14
0
71
73
40
0
5
2
1
0
5
0
5
1
2
2
5
6
100
0
4
7
0
63
39
16
0
15.7
5
2
1
0
5
5
2.5
4
0
0
0
c
—
100
44
28
5
18
0
a
5
c
d
g
50
2
44
0
—
5
2
1
0
5
41
0
0
0
0
0
1
0
5
1
53
10
2
0
0
2.5
0
0
c
—
6
0
27
28
100
14
0
0
0
5
2
1
0
a
5
50
2
16
0
—
5
0
5
47
2
2.5
0
0
3
1
0
5
1
36
35
41
0
c
—
0
0
100
50
2
35
20
5
0
b
5
10
1
39
8
49
14
—
1
2.5
0
0
a
b
c
Essentially no activity in TOP1-gal.
Not determined.
Cytotoxicity was not topoisomerase I-dependent.
6
100
18
3
51
46
91.4
3
5
2
1
0
5
2
0
0
1
34
32
0
0
to that of luotonin A (2) itself (9.58lM). Also found to
be cytotoxic in this assay was 17-aminoluotonin A (25)
(IC50 15.7lM), consistent with its strong activity in sta-
bilizing the topoisomerase I–DNA covalent binary com-
plex. In comparison, some analogues found to stabilize
the binary complex were found to be essentially non-
cytotoxic in the yeast strain grown on galactose; exam-
ples included 17-hydroxymethylluotonin A (26). Given
the poor aqueous solubility of many of the analogues,
the lack of cytotoxicity may simply reflect inadequate
concentrations of the analogues in solution, or poor
uptake by yeast cells.
7
100
7
0
13
9
>100
50
25
10
5
1
0
19
3
36
16
0
15
4
c
8
1
2
2
100
27
27
18
5
—
5
2
1
0
5
18
0
0
1
19
2
10
8
10
1
100
69
41
7
83
80
72
53
32
11.8
5
2
1
0
5
One facet of the yeast testing results worthy of mention
is illustrated by thiophene analogues 22a and 22b. Both
of these analogues effected quite reasonable stabilization
of the enzyme–DNA covalent binary complex (Table 3
and Fig. 8), and both were fairly strongly cytotoxic (Ta-
ble 4). However, the cytotoxicity for both was at least as
strong when the yeast strain was grown on raffinose as
when it was grown on galactose, indicating that the pre-
dominant cytotoxic response was not due to the activity
of the compounds as topoisomerase I poisons. Because
potency of covalent binary complex formation and
mammalian cell toxicity have been used to guide the
2.5
.25
7
6
0
c
1a
1b
100
31
51
62
60
57
57
62
66
64
67
—
5
2
1
0
5
2.5
6.25
100
65
19
5
57
28
177
5
5
2
0
4

6294
A. Cagir et al. / Bioorg. Med. Chem. 12 (2004) 6287–6299
2
0
structure modification of CPT lead compounds, it
seems essentially to also assure that the cytotoxicity is
mediated at the level of topoisomerase I.
temperature 0.50mL (820mg, 5.35mmol) of phosphorus
oxychloride. The resulting suspension was warmed to
40ꢁC and stirred overnight. The reaction mixture was
poured onto ice, and then concentrated ammonium
hydroxide was added to make the solution basic. The
reaction mixture was extracted with three 75mL por-
tions of chloroform. The combined organic phase was
washed with 75mL of water containing 10mL of con-
centrated ammonium hydroxide, dried over magnesium
sulfate, and concentrated under diminished pressure.
The crude product was dissolved in chloroform and
applied to a silica gel column. The column was washed
as described below.
Most of the luotonin A and isoluotonin A derivatives
that act as topoisomerase I poisons have substituents
in the 17-position, a position that is roughly spatially
analogous to the position of the 20(S)-OH group in
CPT. In common with CPT, in which the introduction
of 20-Cl, Br, and NH substituents in lieu of the 20-
2
OH group supported covalent complex stabilization
1
0b
and topoisomerase I-dependent cytotoxicity,
17-
chloro, 17-fluoro, and 17-aminoluotonin A all sup-
ported stabilization of the enzyme–DNA covalent
1
0b
binary complex.
However, while all of the 20-
4.1.2. 17-Chloroluotonin A (5a). Elution with 19:1 meth-
ylene chloride–methanol afforded 5a as a colorless solid:
substituted CPTs were cytotoxic in the yeast strain in
the presence of galactose, 17-chloroluotonin A (5a)
lacked cytotoxicity. Further, while 20-amino CPT was
much less cytotoxic than 20-chloro or 20-bromo
CPT, 17-aminoluotonin A (25) was the most cyto-
toxic of the 17-substituted derivatives, having an IC50
yield 11mg (22%); mp 263ꢁC; silica gel TLC R 0.63
f
1
(ethyl acetate); H NMR (CDCl ) d 5.34 (s, 2H), 7.53
3
(dd, 1H, J = 8.8 and 1.9Hz), 7.71 (dd, 1H, J = 8.2and
6.9Hz), 7.87 (dd, 1H, J = 8.2and 6.9Hz), 7.97 (d, 1H,
J = 8.2Hz), 8.09 (d, 1H, J = 1.9Hz), 8.36 (d, 1H,
J = 8.8Hz) and 8.40–8.55 (m, 2H); mass spectrum
(FAB), m/z 320.0592 (C H ClN O requires 320.0591).
1
0b
value (15.7lM) not much different than luotonin A itself
9.58lM). Other cytotoxic luotonin A derivatives (e.g.,
(
6
1
8
11
3
1
4
)
also differed in their behavior from the correspond-
ing CPT analogues (e.g., 20-deoxy CPT which does not
stabilize the covalent binary complex), underscoring our
4.1.3. 17-Fluoroluotonin A (5b). Elution with 19:1 ethyl
acetate–methanol gave 5b as a colorless solid: yield
11,14
1
16mg (31%); silica gel TLC R 0.43 (ethyl acetate); H
earlier conclusion
that the binding of luotonin A to
f
the topoisomerase I–DNA binary complex must differ
somewhat from the binding of CPT in spite of the obvi-
ous structural similarities between the two.
NMR (CDCl ) d 5.34 (s, 2H), 7.30–7.33 (m, 1H),
3
7.68–7.77 (m, 2H), 7.83–7.90 (m, 1H), 7.95–8.00 (m,
1H), and 8.42–8.50 (m, 3H); mass spectrum (FAB),
+
m/z 304.0887 (M + H)
3
(C H FN O requires
18 11 3
04.0886).
.1.4. 17,18-Difluoroluotonin A (5c). Elution with 1:2
ethyl acetate–methylene chloride afforded 5c as a color-
4. Experimental
4
4
.1. General methods
less solid: yield 14mg (22%); mp 229ꢁC (dec); silica gel
1
TLC R 0.58 (1:1 ethyl acetate–methylene chloride); H
High resolution mass spectra were obtained by the mass
spectrometry facility at Michigan State University. All
H and C NMR spectra were recorded on a General
f
NMR (CDCl ) d 5.35 (s, 2H), 7.68–7.76 (m, 1H),
3
1
13
7.82–8.00 (m, 3H), 8.13–8.23 (m, 1H), and 8.40–8.52
(m, 2H); mass spectrum (FAB), m/z 322.0791
Electric QE-300MHz or Varian Unity 300MHz NMR
spectrophotometer using residual solvent peaks at 2.50
ppm for DMSO-d or 3.31ppm for methanol-d for cal-
+
(M + H) (C H F N O requires 322.0792).
1
8
10
2
3
6
4
ibration. TLC separations were carried out on silica gel
0 F254 analytical TLC plates purchased from EM Sci-
ence. TLC plates were visualized under 254 and 365nm
4.1.5. 16-Methoxyluotonin A (5d). Elution with 1:1 ethyl
acetate–methylene chloride afforded 5d as a colorless
solid: yield 10mg (17%) silica gel TLC R 0.29 (1:3 ethyl
6
f
1
UV light. SiO (230–400mesh) was used for column
2
acetate–methylene chloride); H NMR (CDCl ) d 4.09
3
chromatography (Silicycle Chemicals). Dichlorometh-
ane was distilled from calcium hydride prior to use.
THF was distilled from sodium benzophenone ketal.
All reactions were carried out under a nitrogen or argon
atmosphere. Methyl 4-phenylanthranilate was prepared
as described. Methyl 2-cyanoanthranilate was pre-
pared from methyl 4-aminoterephthalate by sequential
treatment with thionyl chloride, concentrated ammonia,
and phosphorous oxychloride. Purchased anthranilic
acids were converted their esters by previously reported
procedures.
(s, 3H), 5.34 (s, 2H), 7.28 (dd, 1H, J = 8.0 and 1.2Hz),
7.51 (dd, 1H, J = 7.9 and 7.9Hz), 7.66 (ddd, 1H,
J = 7.5, 7.5, and 1.2Hz), 7.82 (ddd, 1H, J = 7.9, 7.7,
and 1.5Hz), 7.93 (dd, 1H, J = 8.1 and 1.3Hz), 7.98
(dd, 1H, J = 8.0 and 1.2Hz), and 8.38–8.46 (m, 2H);+
mass spectrum (FAB), m/z 316.1086 (M + H)
(C H N O requires 316.1086).
2
1
1
9
14
3
2
4.1.6. 17,18-Dimethoxyluotonin A (5e). Elution with
ethyl acetate afforded 5e as a colorless solid: yield 7mg
(11%); mp 282 ꢁC (dec); silica gel TLC R 0.30 (ethyl
f
1
acetate); H NMR (CDCl ) d 4.06 (s, 3H); 4.07
3
4
.1.1. General procedure for preparation of luotonin A
derivatives in dichloromethane (5a–g). To a suspension of
equiv of 1,2-dihydropyrrolo[3,4-b]quinoline-3-one (4)
(s, 3H), 5.36 (s, 2H), 7.56 (s, 1H), 7.70 (dd, 1H,
J = 8.1 and 7.0Hz), 7.76 (s, 1H), 7.86 (dd, 1H, J = 8.1
and 7.0Hz), 7.88 (d, 1H, J = 8.1Hz), and 8.42–8.45
(m, 2H); mass spectrum (FAB), m/z 346.1191
1
and 1–4equiv of methyl anthranilate in 5mL of anhy-
drous dichloromethane was added dropwise at room
+
(M + H) (C H N O requires 346.1191).
2
0
16
3
3

A. Cagir et al. / Bioorg. Med. Chem. 12 (2004) 6287–6299
6295
4.1.7. 17,18,19-Trimethoxyluotonin A (5f). Elution with
ethyl acetate afforded 5f as a colorless solid: yield
with three 100mL portions of ethyl acetate. The com-
bined organic phase was washed with two 50mL por-
tions of saturated sodium carbonate (or 1M sodium
hydroxide), dried over magnesium sulfate, and concen-
trated under diminished pressure. The crude product
was purified by silica gel column chromatography.
3
0
4
1
8
7mg (62%); mp 285 ꢁC (dec); silica gel TLC R
f
1
.42(ethyl acetate); H NMR (CDCl ) d 4.04 (s, 3H),
3
.09 (s, 3H), 4.27 (s, 3H), 5.33 (s, 2H), 7.54–7.63 (m,
H), 7.64–7.73 (m, 1H), 7.74–8.00 (m, 2H), and 8.40–
.46 (m, 2H); mass spectrum (FAB), m/z 376.1298
+
(
M + H) (C H N O requires 376.1297).
2
4.1.12. Luotonin A (2) and isoluotonin A (11). Elution of
the silica gel column with 1:4 ethyl acetate–methylene
chloride afforded 2 and 11 as colorless solids. For 2:
1
18
3
4
4.1.8. 18-Methylluotonin A (5g). Elution with 1:2ethyl
acetate–hexanes afforded 5g as a colorless solid: yield
yield 5mg (12%); silica gel TLC R 0.40 (1:4 ethyl ace-
f
1
tate–methylene chloride); H NMR was the same as
2
0
5
6
0mg (12%); mp 270 ꢁC (dec); silica gel TLC R
f
1
2,15
1
for the previously published data.
2
For 11: yield
mg (4%); silica gel TLC R 0.14 (1:4 ethyl acetate–
.60 (ethyl acetate); H NMR (CDCl ) d 2.57 (s, 3H),
3
.36 (s, 2H), 7.63–7.74 (m, 2H), 7.86 (ddd, 1H, J = 6.8,
.8, and 1.3Hz), 7.97 (d, 1H, J = 8.4Hz), 8.03 (d, 1H,
f
1
methylene chloride); H NMR (CDCl ) d 5.34 (s, 2H),
3
7.54 (ddd, 1H, J = 7.9, 6.3, and 1.9Hz), 7.68 (dd, 1H,
J = 8.4Hz), 8.23 (s, 1H), and 8.45–8.50 (m, 2H); mass
spectrum (FAB), m/z 300.1137 (M + H) (C H N O
requires 300.1137).
+
J = 8.1 and 6.9Hz), 7.78–7.92(m, 3H), 8.05 (d, 1H,
J = 8.1Hz), 8.21 (d, 1H, J = 8.5Hz), 8.42(d, 1H, J =
1
9
14
3
8
2
.1Hz), and 8.97 (s, 1H); mass spectrum (FAB), m/z
+
86.0979 (M + H) (C H N O requires 286.0980).
1
8
12
3
4
1
.1.9. 16,17,18,19-Tetrahydroluotonin A (6). Elution with
:1 ethyl acetate–methylene chloride afforded 6 as a
4.1.13. 17-Chloroluotonin A (5a) and 17-chloroisoluotonin
colorless solid: yield 4.4mg (6%); mp 275ꢁC (dec); silica
A (13). Elution with 1:4 ethyl acetate–dichloromethane
afforded 5a and 13 as colorless solids. For 5a: yield
gel TLC R 0.29 (1:1 ethyl acetate–methylene chlo-
f
1
ride); H NMR (CDCl ) d 1.76–1.97 (m, 4H), 2.60–2.74
3
130mg (37%); silica gel TLC R 0.50 (1:4 ethyl ace-
tate–dichloromethane); spectroscopic data were agree-
f
(
1
m, 2H), 2.84–3.00 (m, 2H), 5.22 (s, 2H), 7.69 (ddd,
H, J = 7.5, 7.5, and 1.3Hz), 7.84 (ddd, 1H, J = 8.2,
ment with those reported above. For 13: yield 30mg
(9%); silica gel TLC R 0.15 (1:4 ethyl acetate–dichloro-
7
.5, and 1.3Hz), 7.95 (d, 1H, J = 8.2Hz), and 8.38–8.51
m, 2H); mass spectrum (FAB), m/z 290.1293 (M + H)
+
f
(
(
1
methane); H NMR (CDCl ) d 5.35 (s, 2H), 7.50 (dd,
3
C H N O requires 290.1293).
3
1
8
16
1
H, J = 8.5 and 1.9Hz), 7.71 (dd, 1H, J = 8.2and
.5Hz), 7.86 (d, 1H, J = 1.9Hz), 7.92(ddd, 1H,
7
J = 8.5, 6.9, and 1.3Hz), 8.08 (d, 1H, J = 8.8Hz), 8.23
4
(
.1.10. 16-Hydroxyluotonin A (7). A solution of 14mg
0.04mmol) of 16-methoxyluotonin A (5d) in 10mL
(
(
(
d, 1H, J = 8.2Hz), 8.36 (d, 1H, J = 8.8Hz), and 8.98+
s, 1H); mass spectrum (FAB), m/z 320.0590 (M + H)
C H ClN O requires 320.0591).
of 48% aq HBr was heated at reflux overnight. The
resulting solution was neutralized with saturated sodium
bicarbonate and extracted with three 30mL portions of
ethyl acetate. The combined organic phase was washed
with water and dried over magnesium sulfate. After
removal of the solvent under diminished pressure,
chromatographic purification was performed on an
Alltech Alltima C18 reversed-phase HPLC column
1
8
11
3
4.1.14. Analogues 14a and 14b. Elution with 1:4 ethyl
acetate–methylene chloride afforded 14a and 14b as
colorless solids. For 14a: yield 6mg (13%); silica gel
TLC R 0.49 (1:4 ethyl acetate–methylene chloride);
f
1
H NMR (CDCl ) d 4.02(s, 3H), 5.36 (s, 2H ), 7.70
3
(
A
150 · 4.6mm) using a gradient of water and acetonitrile.
linear gradient was employed (90:10 H O–
(dd, 1H, J = 8.1 and 6.9Hz), 7.86 (dd, 1H, J = 8.4 and
6.7Hz), 7.96 (d, 1H, J = 8.4Hz), 8.17 (dd, 1H, J = 8.4
2
CH CN ! 10:90 H O–CH CN over a period of 40 min
3
2
3
and 1.6Hz), 8.44–8.52(m, 3H), and 8.78 (s, 1H); mass
spectrum (FAB), m/z 344.1037 (M + H) (C H N O
+
at a flow rate of 1mL/min). Fractions containing the
desired product were collected, frozen and lyophilized
to afford a colorless solid: yield 1.1mg (10%); H NMR
2
0
14
3
3
requires 344.1035). For 14b: yield <1mg; silica gel
TLC R 0.12(1:4 ethyl acetate–methylene chloride);mass
1
f
+
spectrum (FAB), m/z 344.1036 (M + H) (C H N O
requires 344.1035).
(
DMSO-d ) d 5.27 (s, 2H), 7.27 (d, 1H, J = 7.6Hz),
6
2
0
14
3
3
7
7
.40 (dd, 1H, J = 8.2and 7.6Hz), 7.66–7.75 (m, 2H ),
.87 (dd, 1H, J = 6.9 and 6.9Hz), 8.13 (d, 1H,
J = 8.2Hz), 8.22 (d, 1H, J = 8.2Hz), 8.72 (s, 1H), and
.95 (s, 1H); mass spectrum (FAB), m/z 302.0930
4
.1.15. Analogues of 15a and 15b. Elution with 1:2ethyl
9
acetate–methylene chloride afforded 15a and 15b as col-
orless solids. For 15a: yield 16mg (24%); mp 240ꢁC
+
(
M + H) (C H N O requires 302.0930).
18 12 3 2
(dec); silica gel TLC R 0.50 (1:2ethyl acetate–methylene
f
1
4
.1.11. General procedure for preparation of luotonin A
chloride); H NMR (CDCl ) d 5.37 (s, 2H), 5.44 (s, 2H),
3
and isoluotonin A derivatives in THF. To a suspension of
equiv of 1,2-dihydropyrrolo[3,4-b]quinoline-3-one (4)
and 1–5equiv of anthranilic acid derivative in 5mL of
7.37–7.55 (m, 5H), 7.71 (dd, 1H, J = 8.1 and 6.9Hz),
7.87 (dd, 1H, J = 8.4 and 6.9Hz), 7.97 (d, 1H,
J = 8.3Hz), 8.21 (dd, 1H, J = 8.3 and 1.5Hz), 8.43–
8.53 (m, 3H), 8.84 (d, 1H, J = 1.5Hz); mass spectrum
1
dry THF was added 1mL of POCl . The reaction mix-
3
+
ture was heated at 40ꢁC overnight. The resulting suspen-
sion was poured into ice and the solution was made
basic by addition of concentrated sodium carbonate
(FAB), m/z 420.1347 (M + H) (C H N O requires
2
6
18
3
3
420.1348). For 15b: yield 10mg (15%); mp 245ꢁC
(dec); silica gel TLC R 0.25 (1:2 ethyl acetate–methylene
f
1
(
or 1M sodium hydroxide). The mixture was extracted
chloride); H NMR (CDCl ) d 5.36 (s, 2H), 5.44 (s, 2H),
3

6296
A. Cagir et al. / Bioorg. Med. Chem. 12 (2004) 6287–6299
+
7
7
.36–7.54 (m, 5H), 7.70 (dd, 1H, J = 8.1 and 6.9Hz),
.91 (dd, 1H, J = 7.1 and 7.1Hz), 8.08 (d, 1H,
(FAB), m/z 331.0833 (M + H) (C H N O requires
18 11 4 3
331.0831). For 19b: yield 6mg (9%); silica gel TLC Rf
0.22 (1:4 ethyl acetate–methylene chloride); H NMR
(DMSO-d ) d 5.32(s, 2H ), 7.74 (dd, 1H, J = 8.4 and
7.7Hz), 7.95 (dd, 1H, J = 8.4 and 7.1Hz), 8.16 (d, 1H,
J = 8.4Hz), 8.24–8.34 (m, 2H), 8.45–8.54 (m, 2H), and
1
J = 8.3Hz), 8.14–8.25 (m, 2H), 8.48 (d, 1H,
J = 8.4Hz), 8.58 (s, 1H), and 9.00 (s, 1H); mass spec-
trum (FAB), m/z 420.1347 (M + H) (C H N O
6
+
2
6
18
3
3
requires 420.1348).
9
.24 (s, 1H); mass spectrum (FAB), m/z 331.0833
+
4.1.16. Analogues 16a and 16b. Elution with 1:2ethyl
acetate–methylene chloride afforded 16a and 16b as col-
(M + H) (C H N O requires 331.0831).
18 11 4 3
orless solids. For 16a: yield 30mg (19%); silica gel TLC
R 0.49 (1:2ethyl acetate–methylene chloride); H NMR
4.1.20. Analogues 20a and 20b. Elution with 1:4 ethyl
acetate–methylene chloride afforded 20a and 20b as col-
orless solids. For 20a: yield 8mg (14%); silica gel TLC Rf
1
f
(
CDCl ) d 5.31 (s, 2H), 7.39–7.55 (m, 3H), 7.63 (ddd,
3
1
1
H, J = 8.1, 6.9, and 1.2Hz), 7.69–7.90 (m, 5H), 8.32
d, 1H, J = 1.7Hz) and 8.37–8.46 (m, 3H); mass spec-
0.56 (1:4 ethyl acetate–methylene chloride); H NMR
(CDCl ) d 5.35 (s, 2H), 7.58–7.74 (m, 2H), 7.86 (m,
(
trum (FAB), m/z 362.1292 (M + H) (C H N O re-
3
+
1H), 7.95 (d, 1H, J = 8.3Hz), 8.20 (d, 1H, J = 7.3Hz),
8.46–8.54 (m, 2H), and 8.64 (d, 1H, J = 7.9Hz); mass
2
4
16
3
quires 362.1293). For 16b: yield 14mg (9%); mp 272ꢁC
dec); silica gel TLC R 0.23 (1:2 ethyl acetate–methylene
+
spectrum (FAB), m/z 354.0853 (M + H) (C H F N O
(
f
19 11
3
3
1
chloride); H NMR (CDCl ) d 5.36 (s, 2H), 7.41–7.56
requires 354.0854). For 20b: yield 1mg (2%); silica gel
TLC R 0.31 (1:4 ethyl acetate–methylene chloride);
3
(
m, 3H), 7.64–7.80 (m, 4H), 7.84–7.92(m, 1H), 8.01–
f
1
8.09 (m, 2H), 8.20 (d, 1H, J = 8.6Hz), 8.45 (d, 1H,
J = 8.3Hz), and 8.99 (s, 1H); mass spectrum (FAB),
H NMR (CDCl ) d 5.38 (s, 2H), 7.61 (dd, 1H, J = 8.2
3
and 7.9Hz), 7.72(dd, 1H, J = 7.1 and 6.9Hz), 7.93
(ddd, 1H, J = 8.2, 7.1, and 1.2Hz), 8.08–8.20 (m, 2H),
+
m/z 362.1294 (M + H) (C H N O requires 362.1293).
2
4
16
3
8.25 (d, 1H, J = 9.2Hz), 8.63 (d, 1H, J = 8.0Hz), and
9.05 (s, 1H); mass spectrum (FAB), m/z 354.0854
4
.1.17. 17-Cyanoluotonin A (17a) and 17-cyanoisoluoto-
+
(M + H) (C H F N O requires 354.0854).
nin A (17b). Elution with 1:4 ethyl acetate–dichloro-
methane afforded 17a and 17b as light yellow solids.
For 17a: yield 26mg (15%); silica gel TLC R 0.40 (1:4
ethyl acetate–dichloromethane); H NMR (CDCl ) d
5.39 (s, 2H), 7.72–7.80 (m, 2H), 7.90 (dd, 1H, J = 8.4
and 7.1Hz), 8.00 (d, 1H, J = 8.4Hz), 8.41 (s, 1H), and
1
9
11
3
3
4.1.21. Analogues 21a and 21b. Elution with 1:3 ethyl
acetate–dichloromethane afforded luotonin A deriva-
tives 21a and 22b as light yellow solids. For 21a: yield
f
1
3
7mg (22%); silica gel TLC R 0.30 (1:3 ethyl acetate–
f
1
8
3
.45–8.56 (m, 3H); mass spectrum (FAB), m/z
11.0934 (M + H) (C H N O requires 311.0933).
dichloromethane); H NMR (CDCl ) d 5.33 (s, 2 H), 7.39
3
+
(d, 1H, J = 5.7Hz), 7.63 (d, 1H, J = 5.7Hz), 7.70
(ddd, 1H, J = 8.1, 6.9, and 1.3Hz), 7.86 (ddd, 1H,
J = 8.1, 6.9, and 1.3Hz), 7.96 (d, 1H, J = 8.2Hz), 8.43
(d, 1H, J = 9.4Hz), and 8.46 (s, 1H); mass spectrum
1
9
11
4
For 17b: yield 20mg (12%); silica gel TLC R 0.10 (1:4
f
ethyl acetate–dichloromethane); H NMR (CDCl ) d
1
3
5
.37 (s, 2H), 7.70–7.77 (m, 2H), 7.94 (dd, 1H, J = 7.1
+
and 7.1Hz), 8.11 (d, 1H, J = 7.9Hz), 8.17 (d, 1H,
J = 1.5Hz), 8.24 (d, 1H, J = 8.3Hz), 8.51 (d, 1H,
J = 8.3Hz), and 9.02(s, 1H); mass spectrum (FAB),
(FAB), m/z 292.0543 (M + H) (C H N OS requires
1
6
10
3
292.0545). For 21b: yield 5mg (16%); silica gel TLC Rf
0.16 (1:3 ethyl acetate–dichloromethane); H NMR
1
+
m/z 311.0934 (M + H) (C H N O requires 311.0933).
1
(CDCl ) d 5.37 (s, 2H), 7.34 (d, 1H, J = 5.7Hz), 7.63
9
11
4
3
(d, 1H, J = 5.7Hz), 7.72(m, 1H), 7.92(m, 1H), 8.08
4.1.18. Analogues 18a and 18b. Elution with 1:2ethyl
acetate–methylene chloride afforded 18a and 18b as yel-
(d, 1H, J = 7.6Hz), 8.27 (d, 1H, J = 8.2Hz), and 8.94+
(s, 1H); mass spectrum (FAB), m/z 292.0543 (M + H)
(C H N OS requires 292.0545).
low solids. For 18a: yield 97mg (14%); silica gel TLC Rf
0
16
10
3
1
.57 (1:2ethyl acetate–methylene chloride); H NMR
(
(
CDCl ) d 5.35 (s, 2H), 7.66–7.74 (m, 1H), 7.82–8.00
4.1.22. Analogues 22a and 22b. Elution with 1:4 ethyl
acetate–dichloromethane afforded luotonin A deriva-
tives 22a and 22b as light yellow solids. For 22a: yield
3
m, 4H), 8.44–8.49 (m, 2H), 8.55 (d, 1H, J = 2.3Hz);
+
mass spectrum (FAB), m/z 364.0091 (M + H)
(
C H BrN O requires 364.0085). For 18b: yield
1
20mg (33%), silica gel TLC R 0.34 (1:4 ethyl acetate–
f
8
11
3
1
1
5mg (2%); silica gel TLC R 0.30 (1:2ethyl acetate–
f
dichloromethane); H NMR (CDCl ) d 5.34 (s, 2 H), 7.63
3
1
methylene chloride); H NMR (CDCl ) d 5.36 (s, 2H),
7
J = 7.7Hz), 8.22 (d, 1H, J = 8.3Hz), 8.55 (d, 1H,
J = 2.3Hz), and 9.00 (s, 1H); mass spectrum (ESI), m/z
3
(d, 1H, J = 5.7Hz), 7.69 (dd, 1H, J = 8.2and
7.6Hz), 7.82–7.90 (m, 2H), 7.95 (d, 1H, J = 8.2Hz),
8.43–8.46 (m, 2H); mass spectrum (FAB), m/z
3
.65–7.76 (m, 2H), 7.91–7.95 (m, 2H), 8.07 (d, 1H,
+
292.0543 (M + H) (C H N OS requires 292.0545).
3
1
6
10
+
64.4 (M + H) (theoretical 364.0).
For 22b: yield 2mg (3%); mp 280 ꢁC (dec); silica gel TLC
1
R 0.17 (1:4 ethyl acetate–dichloromethane); H NMR
f
4.1.19. Analogues 19a and 19b. Elution with 1:4 ethyl
acetate–methylene chloride afforded 19a and 19b as col-
(CDCl ) d 5.37 (s, 2H), 7.49 (d, 1H, J = 5.7Hz), 7.70
3
(dd, 1H, J = 8.2, and 7.6Hz), 7.88–7.93 (m, 2H), 8.07
(d, 1H, J = 7.6Hz), 8.23 (d, 1H, J = 8.8Hz), and 8.96+
(s, 1H); mass spectrum (FAB), m/z 292.0543 (M + H)
(C H N OS requires 292.0545).
orless solids. For 19a: yield 6mg (9%); silica gel TLC Rf
0
1
.50 (1:4 ethyl acetate–methylene chloride). H NMR
(
DMSO-d ) d 5.35 (s, 2H), 7.79 (dd, 1H, J = 7.8 and
6
16 10
3
7
.1Hz), 7.93 (dd, 1H, J = 8.5 and 6.7Hz), 8.20 (d, 1H,
J = 7.8Hz), 8.24–8.39 (m, 2H), 8.51 (d, 1H, J = 8.5Hz),
.67 (d, 1H, J = 2.3Hz), and 8.81 (s, 1H); mass spectrum
4.1.23. Analogues 23a and 23b. Elution with 1:3 ethyl
acetate–methylene chloride afforded 23a and 23b as yel-
8

A. Cagir et al. / Bioorg. Med. Chem. 12 (2004) 6287–6299
6297
low solids. For 23a: yield 23mg (20%); silica gel TLC Rf
0
tion mixture was heated to reflux for 8h. Then 5mL
of methanol was added and the reaction mixture was
heated to reflux for an additional 10 min. The resulting
yellow solution was poured into ice and extracted with
three 75mL portions of ethyl acetate. The combined or-
ganic phase was extracted with brine, dried over magne-
sium sulfate, and concentrated under diminished
pressure. The product was purified by silica gel column
chromatography. Elution with 9:9:1 ethyl acetate–
dichloromethane–methanol afforded 26 as a yellow
1
.49 (1:3 ethyl acetate–methylene chloride); H NMR
(
(
8
CDCl ) d 5.42(s, 2H ), 7.56–7.74 (m, 3H), 7.83–7.90
m, 1H), 7.97 (d, 1H, J = 8.1Hz), 8.07–8.14 (m, 2H),
.47 (s, 1H), 8.51 (d, 1H, J = 8.6Hz), 8.65 (s, 1H), and
3
9
.04 (s, 1H); mass spectrum (FAB), m/z 336.1138
M + H) (C H N O requires 336.1137). For 23b:
+
(
yield 5mg (4%); silica gel TLC R 0.20 (1:3 ethyl ace-
tate–methylene chloride); H NMR (CDCl ) d 5.39 (s,
2
2
14
3
f
1
3
2H), 7.53–7.72 (m, 3H), 7.89 (ddd, 1H, J = 8.4, 6.8,
and 1.5Hz), 8.00–8.12(m, 3H), 81 .2 (d, 1H,
solid: yield 9mg (21%); silica gel TLC R 0.13 (9:9:1
f
1
ethyl acetate–dichloromethane–methanol); H NMR
J = 8.2Hz), 8.34 (s, 1H), 9.01 (s, 1H), and 9.02 (s, 1H);
mass spectrum (FAB), m/z 336.1138 (M + H)
+
4:1 CDCl –CD OD d 4.71 (s, 2H), 5.22 (s, 2H), 7.48
3
3
(
C H N O requires 336.1137).
2
(d, 1H, J = 8.1Hz), 7.56–7.62(m, 1H), 7.69–7.75 (m,
2
14
3
1
H), 7.84–7.88 (m, 2H), 8.20–8.26 (m, 2H), and 8.40+
4
.1.24. Analogue 24. Nitrogen gas was passed through a
(s, 1H); mass spectrum (FAB), m/z 316.1086 (M + H)
(C H N O requires 316.1086).
suspension of 75mg (0.21mmol) of 18-bromoluotonin A
18a) and 28mg (0.23mmol) of phenyl boronic acid in a
:1 mixture of 10% potassium carbonate–DMF for 40
1
9
14
3
2
(
1
4.1.27. Analogue 27. This compound was prepared as de-
scribed for 26 starting from 14b. Elution with 9:9:1 ethyl
acetate–dichloromethane–methanol afforded 27 as a yel-
min. Then 5mg (0.02mmol) of palladium(II)acetate
was added and the mixture was heated to reflux for 4h.
The resulting solution was poured into 50mL of water
and extracted with three 50mL portions of ethyl acetate.
The combined organic phase was washed with 50mL
portions of water and brine. After drying over magne-
sium sulfate, the solution was concentrated under dimin-
ished pressure and purified by silica gel column
chromatography. Elution with 1:9 ethyl acetate–methyl-
ene chloride afforded 24 as a colorless solid: yield 4mg
low solid: yield 3mg (32%); silica gel TLC R 0.13 (9:9:1
f
1
ethyl acetate–dichloromethane–methanol); H NMR
(4:1 CDCl –CD OD) d 4.78 (s, 2H), 5.35 (s, 2H),
3
3
7.54–7.71 (m, 2H), 7.88–7.93 (m, 1H), 7.96 (s, 1H),
8.11–8.16 (m, 2H), 8.30 (d, 1H, J = 8.1Hz), and 9.41+
(s, 1H); mass spectrum (ESI), m/z 316.1086 (M + H)
(C H N O requires 316.1086).
1
9
14
3
2
(
7%); silica gel TLC R 0.30 (1:9 ethyl acetate–methylene
f
4.1.28. 17-Fluoroluotonin A (5b) and 17-fluoroisoluotonin
A (28). To a suspension of 50mg (0.27mmol) of 2,3-dihy-
dro-1H-pyrrolo[3,4-b]quinolin-1-one (12) and 50mg
(0.30 mmol) of methyl 4-fluoroanthranilate in 6mL of
dry THF was added 1mL of phosphorus oxychloride.
The reaction mixture was stirred overnight at 40ꢁC.
The resulting suspension was poured onto ice and the
resulting solution was made basic by the addition of
1M NaOH. The resulting solution was extracted with
three 75mL portions of ethyl acetate. The combined or-
ganic phase was washed with 50mL of 1N NaOH, dried
over magnesium sulfate, and concentrated under dimin-
ished pressure. The crude product was purified by silica
gel column chromatography. Elution with 1:2ethyl ace-
tate–dichloromethane afforded 17-fluoroluotonin A (5b)
and 17-fluoroisoluotonin A (28) as light yellow solids.
1
chloride); H NMR (CDCl ) d 5.40 (s, 2H), 7.39–7.55 (m,
3
1
and 2.2Hz), 8.21 (d, 1H, J = 8.4Hz), 8.47–8.53 (m, 2H),
and 8.67 (d, 1H, J = 2.3Hz); mass spectrum (FAB), m/z
3
3
H), 7.68–7.79 (m, 3H), 7.87 (ddd, 1H, J = 8.5, 7.1, and
.4Hz), 7.98 (d, 1H, J = 8.3Hz), 8.12(dd, 1H, J = 8.5
+
62.1294 (M + H) (C H N O requires 362.1293).
2
4
16
3
4
(
.1.25. Analogue 25. To a suspension of 6mg
0.018mmol) of 17-nitroluotonin A (19a) in 10mL of
methanol, was added 34mg (0.18mmol) of tin(II)chlo-
ride. The reaction mixture was heated and stirred at
reflux for 2h. Then 1mL of concentrated aq HCl was
added dropwise and stirring was continued for an addi-
tional 2h. The resulting solution was poured onto ice
and the pH of the solution was adjusted to 12by addition
of 1M sodium hydroxide. The solution was extracted
with three 50mL portions of chloroform and three
For 5b: yield 1.4mg (2%); silica gel TLC R 0.41 (1:2ethyl
f
acetate–dichloromethane); For 28: yield 8.1mg (10%);
silica gel TLC Rf 0.19 (1:2ethyl acetate–dichloro-
5
0mL portions of ethyl acetate. The combined organic
1
phase was dried over magnesium sulfate and concen-
trated under diminished pressure. The product was puri-
fied by silica gel column chromatography. Elution with
5
solid: yield 2mg (40%); silica gel TLC R 0.38 (5%
methane); H NMR (CDCl ) d 5.35 (s, 2H), 7.24–7.30
3
(m, 1H), 7.53 (dd, 1H, J = 9.7 and 2.4Hz), 7.71 (m,
1H), 7.92(ddd, 1H, J = 8.4, 7.1, and 1.4Hz), 8.09 (d,
1H, J = 8.3Hz), 8.23 (d, 1H, J = 8.4Hz), 8.44 (dd, 1H,
J = 8.8 and 6.1Hz), and 9.02(s, 1H); mass spectrum
% methanol in ethyl acetate afforded 25 as a yellow
f
1
+
methanol in ethyl acetate); H NMR (DMSO-d ) d
(FAB), m/z 304.0885 (M + H) (C H FN O requires
18 11
6
3
5
2
7
.24 (s, 2H), 6.22 (br s,2H), 6.85 (dd, 1H, J = 8.5 and
.2Hz), 6.90 (s, 1H), 7.71 (dd, 1H, J = 7.8 and 7.1Hz),
.86-7.97 (m, 2H), 8.18 (d, 1H, J = 8.3Hz), 8.27 (d, 1H,
304.0885).
4.1.29. Topoisomerase I-mediated DNA cleavage. The
topoisomerase I-mediated DNA cleavage reaction was
carried out (37ꢁC, 30min) in a 40lL (total volume) reac-
tion mixture containing 20mM Tris–HCl (pH7.6),
J = 8.1Hz), and 8.74 (s, 1H); mass spectrum (FAB),
m/z 301.1088 (M + H) (C H N O requires 301.1089).
+
1
8
13
4
4
(
1
.1.26. Analogue 26. To a suspension of 30mg
0.09mmol) of ester 14a in 30mL of dry THF was added
50mg (3.97mmol) of sodium borohydride. The reac-
10mM KCl, 5mM MgCl , 0.5mM EDTA, 0.5mM
2
DTT, 30 lg/mL BSA, 12fmol of labeled DNA restriction
fragment (HindIII-PvuII fragment from pSP64 DNA),

6298
A. Cagir et al. / Bioorg. Med. Chem. 12 (2004) 6287–6299
and 36 ng of human DNA topoisomerase I. CPT ana-
logues were employed at 50lM concentration. The reac-
tions were terminated by SDS-proteinase K treatment.
After extraction with phenol and chloroform, the DNA
was recovered by ethanol precipitation. The DNA was
dissolved in 80% formamide loading buffer (10mM
NaOH, 1mM EDTA, 0.1% xylene cyanol, and bromo-
phenol blue), and analyzed on a 10% denaturing gel.
62, 2039; (g) Garcia-Carbenero, R.; Supko, J. G. Clin.
Cancer Res. 2002, 8, 641.
. (a) Wall, M. E.; Wani, M. C.; Nicholas, A. W.; Maniku-
mar, G.; Tele, C.; Moore, L.; Truesdale, A.; Leitner, P.;
Besterman, J. M. J. Med. Chem. 1993, 36, 2689;
3
(
b) Yaegashi, T.; Sawada, S.; Nagata, H.; Furuta, T.;
Yokokura, T.; Miyasaka, T. Chem. Pharm. Bull. 1994,
2, 2518; (c) Lackey, K.; Besterman, J. M.; Fletcher, W.;
Leitner, P.; Morton, B.; Sternbach, D. D. J. Med. Chem.
995, 38, 906; (d) Luzzio, M. J.; Besterman, J. M.;
4
1
4
.1.30. Yeast strain growth. A transformed strain of S.
Emerson, D. L.; Evans, M. G.; Lackey, K.; Leitner, P. L.;
McIntyre, G.; Morton, B.; Myers, P. L.; Peel, M.; Sisco, J.
M.; Sternbach, D. D.; Tong, W.-Q.; Truesdale, A.;
Uehling, D. E.; Vuong, A.; Yates, J. J. Med. Chem.
cerevisiae, RS321Nph-TOP1, had the genotype
RS321Nph-TOP1 (Mat a ade2-1 his3-1 leu3,112 trp1-1
ura3-1 can1-100 erg6 rad52::TRP1 top1-8::LEU2 phTO-
P1::URA). This strain was grown from a 15% glycerol
stock to log phase (OD595 1–3) at 30ꢁC in minimal med-
ia (0.9% Yeast Nitrogen Base without amino acids,
1
995, 38, 395; (e) Lackey, K.; Sternbach, D. D.; Croom,
D. K.; Emerson, D. L.; Evans, M. G.; Leitner, P. L.;
Luzzio, M. J.; McIntyre, G.; Vuong, A.; Yates, J.;
Besterman, J. M. J. Med. Chem. 1996, 39, 713; (f)
Lavergne, O.; Lesueur-Ginot, L.; PlaRodas, F.; Bigg, D.
C. H. Bioorg. Med. Chem. Lett. 1997, 7, 2235; (g)
Lavergne, O.; Lesueur-Ginot, L.; PlaRodas, F.; Kasprzyk,
P. G.; Pommier, J.; Demarquay, D.; Prevost, G.; Ulibarri,
G.; Rolland, A.; Schiano-Liberatore, A.-M.; Harnett, J.;
Pons, D.; Camara, J.; Bigg, D. C. H. J. Med. Chem. 1998,
0
.025mg/mL each of adenine and histidine), which con-
tained 3% glucose as the carbon source. The yeast was
then transferred to the same minimal medium contain-
ing 3% raffinose, a neutral carbon source, instead of glu-
cose. Cultures were then grown to log phase (OD595 1–3).
4
1, 5410; (h) Garbarda, A. E.; Du, W.; Isarno, T.;
Tangirala, R. S.; Curran, D. P. Tetrahedron 2002, 58,
329.
4
.1.31. Yeast cytotoxicity assay. The exponentially
growing yeast strain was diluted to OD595 0.015 with
minimal media having 3% galactose or raffinose. Sam-
ples to be assayed were dissolved in DMSO to a concen-
trations of 50mM, then diluted in the yeast incubation
to the appropriate concentrations; the final DMSO con-
centration were less than 1% and the assays were carried
out in 96-well microtiter plates. Camptothecin was used
as a positive control. The plates were incubated in a high
humidity chamber (30ꢁC, 48h) and the optical density
of cells in each well was measured (595 nm) by using a
microplate reader. Data was plotted as OD595 versus
natural log of concentration. This provided a linear plot
for easy analysis of IC50 values, which was defined as the
concentration of a compound at which yeast growth was
inhibited by 50%.
6
4
. (a) Wang, J. C. Annu. Rev. Biochem. 1996, 65, 635; (b)
Wang, J. C. Q. Rev. Biophys. 1998, 31, 107; (c) Pommier,
Y. Biochimie 1998, 80, 255; (d) Champoux, J. J. Annu.
Rev. Biochem. 2002, 70, 369.
5. (a) Pommier, Y.; Pourquier, P.; Fan, Y.; Strumberg, D.
Biochim. Biophys. Acta 1998, 1400, 83; (b) Redinbo, M.
R.; Champoux, J. J.; Hol, W. G. J. Biochemistry 2000, 39,
6
832.
6
. (a) Hsiang, Y.-H.; Hertzberg, R.; Hecht, S.; Liu, L. F. J.
Biol. Chem. 1985, 260, 14873; (b) Hsiang, Y.-H.; Liu, L.
F.; Wall, M. E.; Wani, M. C.; Nicholas, A. W.; Maniku-
mar, G.; Kirschenbaum, S.; Silber, R.; Potmesil, M.
Cancer Res. 1989, 49, 4385; (c) Covey, J. M.; Jaxel, C.;
Kohn, K. W.; Pommier, Y. Cancer Res. 1989, 49,
5
016.
. Hsiang, Y.-H.; Lihou, M. G.; Liu, L. F. Cancer Res. 1989,
9, 5077.
7
4
Acknowledgements
8. (a) Fan, Y.; Weinstein, J. N.; Kohn, K. W.; Shi, L. M.;
Pommmier, Y. J. Med. Chem. 1998, 41, 2216; (b)
Kerrigan, J. E.; Pilch, D. S. Biochemistry 2001, 40, 9792;
We thank Dr. Mary-Ann Bjornsti, St. Jude ChildrenÕs
Hospital, for the yeast strain used for evaluation of luot-
onin derivatives. This work was supported by NIH
Research Grant CA78415, awarded by the National
Cancer Institute.
(
c) Laco, G. S.; Collins, J. R.; Luke, B. T.; Kroth, H.;
Sayer, J. M.; Jerina, D. M.; Pommier, Y. Biochemistry
002, 41, 1428.
2
9
. (a) Redinbo, M. R.; Stewart, L.; Kuhn, P.; Champoux, J.
J.; Hol, W. G. J. Science 1998, 279, 1504; (b) Staker, B. L.;
Hjerrild, K.; Feese, M. D.; Behnke, C. A.; Burgin, A. B.,
Jr.; Stewart, L. Proc. Natl. Acad. Sci. U.S.A. 2002, 99,
15387; (c) Stewart, L.; Redinbo, M. R.; Qui, X.; Hol, W.
G.; Champoux, J. J. Science 1998, 279, 1534.
References and notes
1
. Wall, M. E.; Wani, M. C.; Cook, C. E.; Palmer, K. H.;
McPhail, A. T.; Sim, G. A. J. Am. Chem. Soc. 1966, 88,
10. (a) Jaxel, C.; Kohn, K. W.; Wani, M. C.; Wall, M. E.;
Pommier, Y. Cancer Res. 1989, 49, 1465; (b) Wang, X.;
Zhou, X.; Hecht, S. M. Biochemistry 1999, 38, 4374; (c)
Hertzberg, R. P.; Caranfa, M. J.; Hecht, S. M. Biochem-
istry 1989, 28, 4629.
11. Cagir, A.; Jones, S. H.; Gao, R.; Eisenhauer, B. M.;
Hecht, S. M. J. Am. Chem. Soc. 2003, 125, 13628.
12. (a) Xiao, X.-H.; Qou, G.-L.; Wang, H.-L.; Lui, L.-S.;
Zheng, Y.-L.; Jia, Z.-J.; Deng, Z.-B. Chin. J. Pharmacol.
Toxicol. 1988, 232; (b) Ma, Z. Z.; Hano, Y.; Nomura, T.;
Chen, Y. J. Heterocycles 1997, 46, 541.
3888.
2
. (a) Takimoto, C. H.; Wright, J.; Arbuck, S. G. Biochim.
Biophys. Acta 1998, 1400, 107; (b) OÕLeary, J.; Muggia, F.
M. Eur. J. Cancer 1998, 34, 1500; (c) Saltz, L. B.; Cox, J.
V.; Blanke, C.; Rosen, L. S.; Fechrenbacher, L.; Moore,
M. J.; Maroun, J. A.; Ackland, S. P.; Locker, P. K.;
Pirotta, N.; Elfring, G. L.; Miller, L. L. New Engl. J. Med.
2
2
000, 343, 905; (d) Ozols, R. F. Int. J. Gynecol. Cancer
000, 10, 33; (e) Vanhoefer, U.; Harstrick, A.; Achterrath,
W.; Cao, S.; Seeber, S.; Rustum, Y. M. J. Clin. Oncol.
001, 19, 1501; (f) Ulukan, H.; Swaan, P. W. Drugs 2002,
13. Bjornsti, M. A.; Benedetti, P.; Viglianti, G. A.; Wang, J.
C. Cancer Res. 1989, 49, 6318.
2

A. Cagir et al. / Bioorg. Med. Chem. 12 (2004) 6287–6299
6299
1
4. Cagir, A.; Jones, S. H.; Eisenhauer, B. M.; Gao, R.;
Hecht, S. M. Bioorg. Med. Chem. Lett. 2004, 14, 2051.
5. (a) Lee, E. S.; Park, J.-G.; Jahng, Y. Tetrahedron Lett.
17. Sugasawa, T.; Toyota, T.; Sasakura, K.; Hidaka, T. Chem.
Pharm. Bull. 1971, 19, 1971.
18. Tanaka, T.; Mashimo, K.; Wagatsuma, M. Tetrahedron
Lett. 1971, 30, 2803.
1
2
003, 44, 1883; (b) Osborne, D.; Stevenson, P. J. Tetra-
hedron Lett. 2002, 43, 5469; (c) Dallavalle, S.; Merlini, L.
Tetrahedron Lett. 2002, 43, 1835; (d) Yadav, J. S.; Reddy,
B. V. S. Tetrahedron Lett. 2002, 43, 1905; (e) Wang, H.;
Ganesan, A. Tetrahedron Lett. 1998, 39, 9097; (f) Ma,
Z.-Z.; Hano, Y.; Nomura, T.; Chen, Y.-J. Heterocycles
19. Repetition of these experiments at 50lM concentration
indicated surprisingly little difference, presumably reflect-
ing the limited aqueous solubility of the luotonin A
analogues.
20. (a) Wall, M. E.; Wani, M. C.; Nicholas, A. W.; Mani-
kumar, G.; Tele, G.; Moore, L.; Truesdale, A.; Leitner, P.;
Besterman, J. M. J. Med. Chem. 1993, 36, 2689; (b)
Kingsbury, W. D.; Boehm, J. C.; Jakas, D. R.; Holden, K.
G.; Hecht, S. M.; Gallagher, G.; Caranfa, M. J.; McCabe,
F. L.; Faucette, L. F.; Johnson, R. K.; Hertzberg, R. P.
J. Med. Chem. 1991, 34, 98.
1
999, 51, 1593; (g) Kelly, T. R.; Chamberland, S. Silva,
R. A. Tetrahedron Lett. 1999, 40, 2723; (h) Molina, P.;
Tarraga, A.; Gonzalez-Tejero, A. Synthesis-Stuttgart
2
000, 11, 1523; (i) Toyota, M.; Komori, C.; Ihara, M.
Heterocycles 2002, 56, 101; (j) Harayama, T.; Morikami,
Y.; Shigeta, Y.; Abe, H.; Takeuchi, Y. Synlett 2003, 6, 847;
(
1
k) Mhaske, S. B.; Argade, N. P. J. Org. Chem. 2004, 69,
4563.
21. Quan, M. L.; Dominguez, W. C.; Duncia, J. V.; Kettner,
C. A.; Ellis, C. D.; Liauw, A. Y.; Park, J. M.; Santella, J.
B.; Knabb, R. M.; Thoolen, M. J.; Weber, P. C.; Wexler,
R. R. Bioorg. Med. Chem. Lett. 1997, 13, 1595.
1
6. Danishefsky, S.; Bryson, T. A.; Puthenpurayil, J. J. Org.
Chem. 1975, 40, 796.