Synthesis and SAR of tetracyclic pyrroloquinolones as phosphodiesterase 5 inhibitors

Article abstract of DOI:10.1016/j.bmc.2003.12.044

The synthesis of the fused tetracyclic pyrroloquinolones 9a-i in four steps is described. The PDE5 inhibitory activities of these compounds, their selectivities against PDE1, PDE2, PDE3, PDE4 and PDE6, the preclinical pharmacokinetic assessments and the in vivo efficacy in increasing intracavernosal pressure are presented and discussed.

Full text of DOI:10.1016/j.bmc.2003.12.044

Bioorganic & Medicinal Chemistry 12 (2004) 1505–1515
Synthesis and SAR of tetracyclic pyrroloquinolones as
phosphodiesterase 5 inhibitors
Weiqin Jiang,* Vernon C. Alford, Yuhong Qiu, Sheela Bhattacharjee, T. Matthew John,
Donna Haynes-Johnson, Patricia J. Kraft, Scott G. Lundeen and Zhihua Sui*
Drug Discovery, Johnson & Johnson Pharmaceutical Research & Development, L.L.C., 1000 Route 202 South; POBox 300,
Raritan, NJ 08869, USA
Received 12 September 2003; accepted 18 December 2003
Abstract—The synthesis of the fused tetracyclic pyrroloquinolones 9a–i in four steps is described. The PDE5 inhibitory activities of
these compounds, their selectivities against PDE1, PDE2, PDE3, PDE4 and PDE6, the preclinical pharmacokinetic assessments
and the in vivo efficacy in increasing intracavernosal pressure are presented and discussed.
# 2004 Elsevier Ltd. All rights reserved.
1. Introduction
MEDis defined as the inability to achieve and maintain
an erection sufficient to permit satisfactory sexual inter-
course. The incidence of MED2a increases dramatically
with age. The large patient population, estimated to be
30 million men in the United States alone,2b combined
with the success of Viagra1 (sildenafil)3 have provided
strong stimuli for the discovery and development of
additional PDE5 inhibitors,6 such as recently launched
Levitra1 (vardenafil)5 and Cialis1 (tadalafil).4 One
important issue facing a new PDE5 inhibitor is its
selectivity over other PDEs. Accumulation of clinical data
suggested that some of the adverse side effects observed in
sildenafil therapy, such as visual disturbances, might be
the results of noticeable inhibition of sildenafil toward
PDE6. So both potencies toward PDE5 and selectivities
against other PDEs are important for the successful
development of new PDE5 inhibitors. During the SAR
studies on pyrroloquinolones in this group,8 we dis-
covered that a wide variety of substituents are tolerated
on the pyrrole nitrogen. These included acyl groups8j
and various heterocycles.8f In addition, the b-carboline
precursors such as 1a were less potent than 1b.8e We
decided to tie back the acyl group on the pyrrole nitrogen
to the 3-position to generate conformationally con-
strained pyrroloquinolones such as 2b (Fig. 1). We were
also interested in the comparison of biological activities of
these heterocyclic pyrroloquinolones with their b-carbo-
line precursors such as 2a.7 SAR study of pyrimidine
pyrroloquinolones8e showed that both potencies and
PDE6/5 selectivities are very sensitive to the C-1 aromatic
substituents. More importantly, pyrimidinyl pyrrolo-
quinolones bearing methylenedioxyphenyl or dihydro-
benzofuranyl groups are active and selective PDE5
Cyclic nucleotides, that is, cyclic adenosine 30,50-mono-
phosphate (cAMP) and cyclic guanosine 30,50-mono-
phosphate (cGMP) are important second messengers
that control many physiological processes.1 The levels
of intracellular cyclic nucleotides are determined by the
activities of cyclases that synthesize them and phospho-
diesterases (PDEs) that degrade them. To date, 21
mammalian PDE genes have been cloned and are classi-
fied into 11 families (PDE1–PDE11) according to the
sequence homology and their biochemical properties.1c
Phosphodiesterase type 5 (PDE5), a cGMP specific PDE,
is abundant in smooth muscles, lung and platelets. In
human corpus carvernosum, PDE5 is the major enzyme
to hydrolyze cGMP to GMP. Upon sexual stimulation,
release of nitric oxide from nonadrenergic, non-
cholinergic neurons activates soluble guanylyl cyclase,
which cyclizes guanosine triphosphate (GTP) to generate
cGMP. Increased cGMP levels eventually cause a
decrease in intracellular calcium concentration. Lower
calcium concentration leads to relaxation of smooth
muscle in the corpus cavernosum, resulting in increased
arterial blood flow to the penis and ultimately erection.
PDE5 inhibition blocks cGMP degradation, facilitates
cGMP accumulation, induces adequate relaxation of
corpus cavernosal smooth muscle, and thus erection in
patients with male erectile dysfunction (MED).
Keywords: Pyrroloquinolone; b-Carboline; PDE5 inhibitor; MED
(male erectile dysfunction).
* Correspondence author. Tel.: +1-908-704-4351; fax: +1-908-526-
6469; e-mail: wjiang1@prdus.jnj.com
0968-0896/$ - see front matter # 2004 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmc.2003.12.044

1506
W. Jiang et al. / Bioorg. Med. Chem. 12 (2004) 1505–1515
eomers 4a and 4b in 25% and 24% yields, respectively.7g
The X-ray of 4b was obtained to exclusively assign the
absolute stereochemistry of 4a and 4b. Condensation of
4a with chloroacetyl chloride provided acylated inter-
mediate 6a in almost quantitative yield. Subsequent
cyclization of 6a with N-methyl amine in methanol at
50 ꢀC for 16 h provided diastereomers 7a in 54% yield
along with small amount of 7b in 1.1% yield.7g Com-
pound 7a is in full accordance with the literature data
{[a]2D0 +71.4 (c 1.00, CHCl3); lit.7g [a]2D0 +71.2 (c 1.00,
CHCl3)}. It should be noted that the amount of C-12a
epimerized product 7b generated from the cyclization
reaction varied according to the length of reaction
times. Thus, under the elongated reaction time, 48 h,
compound 7c was obtained from precursor 6b with
decreased yield of 21%, while C-12a epimerized product
7d increased significantly to 18% yield. NOE studies
were implemented on 7c/7d for confirmation of the
structural assignments. An NOE was observed between
the two hydrogens on C-6 and C-12a for 7d, while there
is no NOE observed for the same pair of hydrogens in
7c. When starting from 2,3-dihydro-benzofuran-5-carb-
aldehyde 3b, compounds 5a and 5b were obtained in
17% and 9.0% yields, respectively. The stereochemistry
of compound 5a and 5b were assigned by 13C NMR
spectroscopy according to the method reported by
James Cook.10a The signals for C-1 and C-3 in the trans
diastereomer 5b (55.9, 53.2 ppm) is more upfield than
those of corresponding cis isomer 5a (59.3, 57.8 ppm).
Following the similar cyclization conditions (50 ꢀC, 48 h)
for compounds 7c and 7d, b-carbolines 7e and 7f were
obtained from 6c in 25% and 38% yields, respectively.
An NOE was observed for 7f between C-12a hydrogen
and Ha and Hb on the 2,3-dihydro-benzofuran group,
Figure 1. Derivatization of tri-cyclic pyrroloquinolones to tetra-cyclic
pyrroloquinolones.
inhibitors.8e We thus incorporated these two components
in our initial synthetic plan.
2. Chemistry
The functional quinolone skeleton can be assembled by
the application of the Winterfeldt oxidation reaction of
b-carbolines, generated from the Pictet–Spengler reaction
of d-tryptophan methyl ester with selected aldehydes.
The implementation of our synthetic protocol to access
these target pyrroloquinolone molecules is shown in
Scheme 1. Thus, d-tryptophan methyl ester reacted
with piperonal (3a) under Pictet–Spengler reaction con-
dition (TFA/CH2Cl2/MeOH) to furnish two diaster-
Scheme 1. Synthesis of compound 7a–f.

W. Jiang et al. / Bioorg. Med. Chem. 12 (2004) 1505–1515
1507
but not for 7e. By carefully monitoring the reaction
progress by HPLC-MS, we were able to stop the cycli-
zation reaction right before significant C-12a epimeriza-
tion had occurred. Thus, as shown in Scheme 2, when
benzyl amine and 2-(aminomethyl)-pyridine was used in
place of N-methyl amine in cyclization reaction, tetra-
cyclic b-carbolines 7g and 7h were obtained (50 ꢀC, 24 h)
as a single diastereomer in 42% and 26% yields,
respectively.
After the tetracyclic b-carboline precursors 7a–7i were
successfully obtained, they were subjected to Winter-
feldt oxidation conditions.9 As shown in Scheme 4,
when b-carbolines 7a was treated with KOtBu/O2, a
very polar intermediate 8a formed quickly with 32 mass
units higher than compound 7a. Attempts to isolate
intermediate 8a failed due to its high water solubility.
We tentatively assigned intermediate 8a as an oxidized
product with concomitant D-ring hydrolysis when
quenched with water. So, we stopped the reaction by
neutralizing the excess amount of KOtBu with HCl, and
then subjected the reaction mixture to standard coupling
condition (PyBrOP/DIEA). The desired pyrroloquino-
lone 9a was isolated in 43% yield. This one-pot proce-
dure has been applied to compounds 7b through 7h to
provide compounds 9b–9h in 8–63% yield.
In order to expand the substrate types, we turned our
attention to 7-membered Dring. As shown in Scheme 3,
we first obtained acylated compound 11a from com-
pound 5a in 89% yield when treated with acroyl chloride.
Michael adduct 12a was formed quickly in 90% yield
after the addition of N-methyl amine to a,b-unsaturated
ketone 11a. However, compound 12a was reluctant to
cyclize after heated at 50 ꢀC for 48 h. Longer reaction
times or higher temperature only led to slow decom-
position of starting material. This was presumably due
to the 1,3-interaction of C-3-CO2Me group and C-1
benzofuranyl group in cis compound 12a, which forces
both C-1 and C-3 substituents to pseudo-equatorial
positions. Molecular modeling showed that it is difficult
for -NHMe in the N-2 side chain to approach the C-3
carbonyl group when the C-3 substituent is pseudo-
equatorial. On the contrary, the trans relationship of
substituents on C-1 and C-3 in compound 12b results in the
most populated conformation of C-1 pseudo-equatorial
and C-3 pseudo-axial since C-1 substituent is bigger than
C-3. The pseudo-axial-CO2Me of C-3 is more accessible
when NHMe on N-2 side chain approaches in an anti-
periplanar version, so that 7-membered Dring was
formed to provide compound 7i in 96% yield.
Due to the less strain in a 7-membered ring than in a 6-
membered ring, we observed the direct formation of
pyrroloquinolone 9i from 7i with D-ring intact, under
oxidation condition (KOtBu/O2). This was shown in
Scheme 5. All of these reactions were carefully mon-
itored by HPLC-MS, and no epimerization at C-3 was
observed. Table 1 summarizes the experimental results
for obtaining compounds 9a through 9i.
After considerable work had been carried out in this
project, we discovered the potassium superoxide as an
alternative reagent for Winterfeldt oxidation of b-car-
bolines.8h Compound 7a was converted to compound
9a in 38% yield upon treatment with KO2/18-crown-6
(Scheme 6). This is in consistence with our findings that
KO2/18-crown-6 oxidation condition is complementary
to Winterfeldt condition for base sensitive substrates.8h
Scheme 2. Synthesis of compounds 7g and 7h.
Scheme 3. Synthesis of compound 7i.

1508
W. Jiang et al. / Bioorg. Med. Chem. 12 (2004) 1505–1515
Scheme 4. Construction of compound 9a.
than sildenafil against all tested PDEs though its
potency was 4-fold less.
Preliminary pharmacokinetics studies have been con-
ducted with selected analogues. In particular, com-
pound 9a had an oral bioavailibility of 19% in male
rats, compared with 11% for sildenafil in a parallel
study.
Scheme 5. Construction of compound 9i.
Selected pyrroloquinolones from this series were studied
in vivo in an anesthetized canine model for erectile dys-
function. In these studies, drugs were dosed via the
intravenous route using sildenafil citrate as a positive
control. Compound 9a was found to be efficacious at
300 mg/kg in the canine model (Fig. 2). Compound 9a is
8-fold less potent than sildenafil with slightly higher oral
bioavailability. Even at higher dosage (300 mg/kg),
compound 9a did not reach 50% of sildenfil’s efficacy.
This might be due to protein binding in the systemic
circulation, leading to low concentration of unbound
drug. This could also be explained by unfavorable local
tissue distribution.
3. Results and discussion
We have observed that all pyrimidinyl pyrroloquino-
lones in our previous studies are much more potent than
their corresponding b-carboline precursors.8 Yet, for the
tetracyclic compounds in this paper, the SAR was dif-
ferent (Table 2). Firstly, for compounds 7c/9c, 7f/9f and
7i/9i, pyrroloquinolones are more potent than b-carbo-
lines. For 7b/9b, pyrroloquinolone has the same potency
as that of b-carboline. For compounds 7a/9a, 7d/9d, 7e/
9e, 7g/9g and 7h/9h, pyrroloquinololones are less potent
than b-carbolines. Secondly, this study demonstrated
that, for pyrroloquinolones pairs 9a/9b, 9d/9c and 9e/9f,
cis isomers (C-4a and C-11) are more potent than trans
isomers. When the substituent on the nitrogen in D-ring
was changed from methyl to benzyl in 9a/9g, potency
decreased 9-fold. When benzyl group in 9g was replaced
by pyridinylmethyl in 9h, potency slightly increased.
Finally, 7-membered D-ring pyrroloquinolone 9i was
less potent than any 6-membered D-ring pyrroloquino-
lones 9a–h. The potency of the best compound (9e) is
approaching that of sildenafil. In addition, the tetra-
cyclic analogues reported in this paper are generally less
potent than pyrimidinyl pyrroloquinolones8e and acyl
pyrroloquinolones.8j
4. Conclusion
In summary, we have discovered a series of novel pyr-
roloquinolone analogues as potent and selective PDE5
inhibitors. Selected compounds are orally bioavailable
in preliminary pharmacokinetics studies. Representative
compound demonstrated in vivo efficacy in a canine
model for erectile dysfunction.
5. Experimental
One of the most important issues for the clinical utility
of PDE5 inhibitors is that these inhibitors should
demonstrate strong inhibitory potencies toward PDE5
while maintaining little or no inhibition toward other
PDEs. The inhibitory activities of PDE1, PDE2, PDE3,
PDE4 and PDE6 of all compounds in this series, toge-
ther with sildenafil were tested. We controlled the sub-
strate cGMP concentration relatively low (30 nM) to
meet the condition of [S]< <Km, so that IC50 values
approximate the Ki values. The values of Ki of PDE1-4
and PDE6 divided by Ki of PDE5 were calculated. All
the data were shown in Table 3. Compounds 9d and 9h
showed better PDE1/5 selectivity than sildenafil. Com-
pounds 9a, e and f had better PDE6/5 selectivity than
sildenafil. Weaker inhibition of PDE6 will probably
decrease the visual side effects observed in current Via-
gra1 therapy. Compound 9e showed better selectivities
5.1. PDE Isolation
PDE1, 2, 3, 4, 5, and 6 were isolated from human heart,
corpus cavernosum, platelet, skeletal muscle, corpus
cavernosum, and retina respectively, as described in ref
8k.
5.2. PDE Assay and Ki Determination
The PDE assays were carried out as described in refer-
ence 8k. Stock solutions of compounds were prepared in
100% DMSO, diluted in 100% DMSO to the appropriate
concentrations, and added to the assay buffer to give a
final concentration of 2% DMSO. The amount of enzyme
used in each reaction was such that the hydrolysis of
substrates did not exceeded 15% so that the amount
of product increased linearly with time. Duplicates were

W. Jiang et al. / Bioorg. Med. Chem. 12 (2004) 1505–1515
Table 1. Transformation of compound 7 to compound 9a
Entry Substrate
1509
Product
Timec (h)
Yield (%)
1
4
43
2
3
50
3
6
63
4
5
6
7
5
8
5
3
20
30
8
35
(continued on next page)

1510
W. Jiang et al. / Bioorg. Med. Chem. 12 (2004) 1505–1515
Table 1 (continued)
Entry
Substrate
Product
Timec (h)
Yield (%)
8
7
8
9b
3
10
a Reaction condition for entry 1–8: (1) KOtBu/O2, (2) HCl, (3) PyBrOP, DIEA.
b Reaction condition for entry 9: KOtBu/O2.
c Reaction time for the oxidation step.
model was used and the animals were prepared as
described in reference 8k. After period of stabilization,
the control ICP response curves were generated at an
appropriate current output setting (ICP increase to 20–
30% of systolic pressure). When similar ICP increases
were obtained from at least two control stimulations
applied in 15-min interval, the baseline was established.
Area under the curve was computed and designated as
baseline. Ascending doses of compound were admini-
strated via bolus injection to the right femoral vein at 40
min interval. The effects of compound on ICP increase
were evaluated 15 min after each dosing by electrical
stimulation with the frequency setting the baseline. To
determine the effect of the compound, the area under
the curve (AUC) for ICP at each stimulation was com-
puted and subtracted from the baseline AUC. At the
end of each experiment, 300 mg/kg sildenafil was given
intravenously. This dose was showed to induce a max-
imal response in our hands. For each animal, the highest
ICP increase was designated as 100%.
Scheme 6. An alternative route to compound 9a.
Table 2. In vitro PDE5 inhibitory data for compounds 7 and 9
b-Carboline
sildenafil
Ki (nM)
1.8
Pyrroloquinolone
Ki (nM)
7a
7b
7c
7d
7e
7f
7g
7h
7i
1.6
47
100
5.5
2.3
299
4.2
12
9a
9b
9c
9d
9e
9f
9g
9h
9i
14
48
80
17
7.1
69
123
63
586
148
5.4. Synthesis of compounds reported
5.4.1. General. NMR spectra were obtained at 400 MHz
or 300 MHz on a Bruker AVANCE300 or AVANCE400
spectrometer. Chemical shifts are reported in ppm
downfield from TMS as an internal standard. Thin-
layer chromatography was carried out using 2.5ꢁ7.5 cm
silica gel 60 (250 mM layer) plates with UV detection.
Magnesium sulfate was employed to dry organic
extracts prior to concentration by rotary evaporation.
Flash chromatography was performed using EM science
silica gel 60 (230–400 mesh). Standard solvents from J.
T. Baker were used as received. Anhydrous solvents
from J. T. Baker or Aldrich and all other commercially
available reagents were used without further purifica-
tion. Melting points were taken using a Thomas-Hoover
run in each assay. IC50 values were obtained from a
nonlinear regression curve fitting program. At very low
substrate concentrations ([S]< <Km), IC50 values
approximate the Ki values.
5.3. In vivo efficacy study
All animals were handled in accordance with the NRC
Guide for Care and Use of Laboratory Animals and the
protocol was approved by Internal Animal Care and
Usage Committee of Johnson & Johnson Pharmaceutical
Research & Development, LLC. An anesthetized dog

W. Jiang et al. / Bioorg. Med. Chem. 12 (2004) 1505–1515
Table 3. Selectivities of compound 9a–i on PDE isoforms
1511
Compd
sildenafil
Ki (nM)
1.80
PDE 1/5
170
PDE 2/5
6000
PDE 3/5
7110
PDE 4/5
2,610
PDE 6/5 (rod)
5.4
PDE 6/5 (cone)
9.6
9a
9b
9c
9d
9e
9f
9g
9h
9i
14
48
80
17
7.1
69
123
63
860
430
310
2160
3660
540
810
1580
680
4500
2100
840
5810
11,410
1450
70
7000
2100
1250
5810
12,440
1440
810
5,000
2,100
940
3,150
9,820
1,450
70
142
12
37
39
385
76
15
41
147
22
35
64
554
125
43
42
24
1580
680
1580
680
530
250
148
20
5.4.2. (1R, 3R)-1-(Benzo[1,3]dioxol-5-yl)-2,3,4,9-tetrahy-
dro-1H-ꢀ-carboline-3-carboxylic acid methyl ester (4a)
and (1S, 3R)-1-(Benzo[1,3]dioxol-5-yl)-2,3,4,9-tetrahydro-
1H-ꢀ-carboline-3-carboxylic acid methyl ester (4b).
d-Trytophan methyl ester hydrochloride (7.2 g, 28.3
mmol) was dissolved in methanol (74 mL) and di-
chloromethane (274 mL), followed by the addition of
piperonal 3a (4.7 g, 31.1 mmol) and TFA (5.0 mL). The
reaction mixture was stirred at room temperature for 60
h. After cooled to 0 ꢀC and carefully neutralized with
saturated sodium bicarbonate aqueous solution, the
organic layer was separated, dried and concentrated.
The residue was purified by silica gel column (800 g)
eluted with ethyl acetate/hexane (30/100) to provide cis
isomer 4a as a white solid (2.5 g, 25%). Continuous
elution with ethyl acetate/hexane (40/60) yielded the
Figure 2. In vivo efficacy of 9a in canine model.
1
trans isomer 4b as a white solid (2.4 g, 24%). 4a: H
NMR (CDCl3) d 3.11 (m, 1H), 3.23 (m, 1H), 3.69 (s,
3H), 3.98 (t, 1H, J=6.1 Hz), 5.31 (s, 1H), 6.78–7.62 (m,
1
7H). MS (m/z): 351 [M+H]+. 4b: H NMR (CDCl3) d
2.95 (m, 1H), 3.20 (m, 1H), 3.81 (s, 3H), 3.92 (m, 1H),
5.18 (s, 1H), 5.89 (s, 2H), 6.78-7.52 (m, 7H); MS (m/z):
351 [M+H]+. Anal. calcd for 4a C20H18N2O4: C 68.56,
H 5.18, N 8.00; found, C 68.27, H 4.91, N 7.81.
Recrystalization of 4b from dichloromethane provided
crystals suitable for X-ray analysis. X-ray ORTEP pic-
ture of 4b was shown in Figure 3.
5.4.3. (1R,3R)-1-(2,3-Dihydro-benzofuran-5-yl)-2,3,4,9-
tetrahydro-1H-ꢀ-carboline-3-carboxylic acid methyl ester
(5a) and (1S,3R)-1-(2,3-Dihydro-benzofuran-5-yl)-2,3,4,9-
tetrahydro-1H-ꢀ-carboline-3-carboxylic acid methyl ester
(5b). Following the same procedure as 4a and 4b, start-
ing from 2,3-dihydro-benzofuran-5-carbaldehyde 3b,
compound 5a was obtained in 17% yield and com-
pound 5b was obtained in 9% yield. 5a: 1H NMR
(CDCl3) d 2.90–3.28 (m, 4H), 3.78 (s, 3H), 3.91 (m, 1H),
4.51 (t, 2H, J=8.2 Hz), 5.12 (s, 1H), 6.71 (d, 1H, J=6.7
Hz), 7.12 (m, 5H), 7.52 (m, 1H); 13C NMR (CD3OD
with two drops of DMSO-d6) d 26.4, 30.4, 52.6, 57.8,
59.3, 72.4, 108.7, 109.9, 112.1, 118.6, 119.8, 122.2, 126.5,
128.1, 128.9, 129.8, 134.1, 136.0, 138.0, 161.5, 174.4; MS
Figure 3. ORTEP picture of compound 4b.
MelTemp apparatus without any correction. Micro-
analysis was done by Quantitative Technologies Inc.,
Whitehouse, NJ. Mass spectra were obtained on a
Hewlett-Packard 5989A quadruple mass spectrometer.
Silica gel (E. Merck, 230–400 mesh) was used for all
flash chromatography. Thin-layer chromatography was
performed on Analtech silica gel (250 um). HPLC ana-
lysis was carried on Agilent 1100 Series LC/MSD. High
resolution mass spectra were obtained on M-Scan’s VG
Analytical ZAB 2SE high field mass spectrometer.
1
(m/z): 349 [M+H]+. 5b H NMR (CDCl3) d 3.02–3.28
(m, 4H), 3.71 (s, 3H), 3.92 (m, 1H), 5.38 (s, 1H), 6.71 (d,
1H, J=6.7 Hz), 6.92–7.23 (m, 5H), 7.51 (m, 1H), 7.61
(s, 1H); 13C NMR (CD3ODwith two drops of DMSO-
d6) d 25.8, 30.6, 52.7, 53.2, 55.9, 72.5, 108.5, 110.1, 112.2,
118.9, 120.0, 122.5, 126.5, 128.2, 128.9, 129.7, 135.1,
135.6, 138.1, 161.3, 175.2; MS (m/z): 349 [M+H]+.

1512
W. Jiang et al. / Bioorg. Med. Chem. 12 (2004) 1505–1515
5.4.4. (6R,12aR)-6-(Benzo[1,3]dioxol-5-yl)-2-methyl-2,3,
6,7,12,12a hexahydropyrazino [10,20:1,6]pyrido[3,4-b]indole-
1,4-dione (7a) and (6R,12aS)-6-(Benzo[1,3]dioxol-5-yl)-
2-methyl-2,3,6,7,12,12a hexahydropyrazino[10,20:1,6]pyr-
ido[3,4-b]indole-1,4-dione (7b). To a mixture of 4a (1.01
g, 2.59 mmol) and sodium bicarbonate (0.30 g) in
methylene chloride (30 mL) was added chloroacetyl
chloride (0.54 mL, 6.78 mmol) in dichloromethane (5
mL) at 0 ꢀC under nitrogen. After 1 h at 0 ꢀC, the reac-
tion mixture was poured into ice-cold saturated sodium
bicarbonate aqueous solution (20 mL). The organic
layer was separated and the aqueous layer was extracted
with dichloromethane (2ꢁ50 mL). The combined
organic layers were dried and concentrated to provide
acylated compound 6a as a yellow solid (1.12 g, 91%).
The product was used for the next step without further
purification.
quartet, 2H, J=60.2, 13.9 Hz), 4.31 (m, 1H), 5.91 (d,
2H, J=7.6 Hz), 6.18 (s, 1H), 6.68-7.85 (m, 7H); MS (m/
z): 390 [M+H]+; HRMS calcd M+ for C22H19N3O4
389.1376, found 389.1387.
5.4.6. (6R,12aR)-6-(2,3-Dihydro-benzofuran-5-yl)-2-methyl-
2,3,6,7,12,12a-hexahydro-pyrazino[10,20:1,6]pyrido [3,4-
b]indole-1,4-dione (7e) and (6R,12aS)-6-(2,3-Dihydro-
benzofuran-5-yl)-2-methyl-2,3,6,7,12,12a-hexahydro-pyr-
azino[10,20:1,6]pyrido[3,4-b]indole-1,4-dione (7f). Follow-
ing the same procedure as 7a and 7b, starting from 5a, in
the reaction with methyl amine step, reaction was
allowed to stir at 50 ꢀC for 48 h. 7e was obtained in 25%
1
yield and 7f was obtained in 38% yield. 7e: H NMR
(CDCl3) d 2.96 (s, 3H), 3.12 (t, 2H, J=8.6 Hz), 3.21 (m,
1H), 3.79 (m, 1H), 4.01 (AB quartet, 2H, J=52.7, 17.0
Hz), 4.32 (m, 1H), 4.51 (t, 2H, J=8.6 Hz), 6.17 (s, 1H),
6.65 (d, 1H, J=5.7 Hz), 7.09–7.82 (m, 7H); MS (m/z):
388 [M+H]+; HRMS calcd M+ for C23H21N3O3
To a solution of 1-benzo[1,3]dioxol-5-yl-2-(2-chloro-
acetyl)-2,3,4,9-tetrahydro-1H-b-carboline-3-carboxylic
acid ethyl ester 6a (1.12 g, 2.29 mmol) in methanol (100
mL) was added methylamine (2.0 M in methanol, 4 mL,
8.0 mmol) at room temperature under nitrogen and
stirred for 1 h. The resulting mixture was heated at
50 ꢀC for 16 h before cooling down to room temper-
ature. After most of the solvent was removed under
reduced pressure, the reaction mixture was then diluted
with water (100 mL) and extracted with methylene
chloride (1ꢁ100 mL, 2ꢁ30 mL). The combined organic
layers were dried and concentrated. The residue was
purified on silica gel column (50 g) eluted with di-
chloromethane/methanol (99:1) to provide 7a (542 mg,
54%) as colorless crystals and 7b (11.1 mg, 1.1%) as a
white powder. 7a: mp 303–305 ꢀC; [a]D20 +71.4 (c 1.00,
CHCl3); 1H NMR (CDCl3) d 3.08 (s, 3H), 3.19 (m, 1H),
3.79 (m, 1H), 4.05 (AB quartet, 2H, J=61.8, 18.5 Hz),
4.31 (m, 1H), 5.92 (d, 2H, J=7.7 Hz), 6.18 (s, 1H), 6.70-
7.82 (m, 7H); MS (m/z): 390 [M+H]+; HRMS calcd
M+ for C22H19N3O4 389.1376; found 389.1395. Anal.
calcd for C22H19N3O4: C, 67.86; H, 4.92; N, 10.79; O,
16.4ꢀ3; found: C, 67.64; H, 4.89; N, 10.64. 7b: mp 288–
1
387.1583, found 387.1590. 7f: H NMR (CDCl3) d 2.82
(m, 1H), 2.95 (s, 3H), 3.15 (t, 2H, J=8.6 Hz), 3.52 (m,
1H), 4.08 (AB quartet, 2H, J=45.6, 14.0 Hz), 4.38 (m,
1H), 4.52 (t, 2H, J=8.6 Hz), 6.75-7.95 (m, 8H); MS (m/
z): 388 [M+H]+; HRMS calcd M+ for C23H21N3O3
387.1583, found 387.1590.387.1590.
5.4.7. (6R,12aR)-6-(Benzo[1,3]dioxol-5-yl)-2-benzyl-2,3,6,7,
12,12a-hexahydro-pyrazino[10,20:1,6]pyrido[3,4- b]indole-
1,4-dione (7g). Following the same procedure as 7a
using benzyl amine instead of methyl amine in the
cyclization step (50 ꢀC, 24 h) the product was obtained
in 42% yield as a white solid. 7g: mp 180–182 ꢀC; H
1
NMR (CDCl3) d 3.28 (m, 1H), 3.82 (m, 1H), 3.91 (d,
2H, J=4.6 Hz), 4.32 (m, 1H), 4.61 (AB quartet, 2H,
J=150.0, 13.8 Hz), 5.81 (d, 2H, J=4.0 Hz), 6.15 (s,
1H), 6.60–8.24 (m, 12H); MS (m/z): 466 [M+H]+, 464
[MꢂH]ꢂ; HRMS calcd M+ for C28H23N3O4 465.1688;
found 465.1686.
5.4.8. (6R,12aR)-6-(Benzo[1,3]dioxol-5-yl)-2-pyridin-4-yl-
methyl-2,3,6,7,12,12a-hexahydro-pyrazino[10,20:1,6]pyrido
0:1,6]pyrido [3,4-b]indole-1,4-dione (7h). Following the
same procedure as 7a, using C-pyridin-2-yl-methyl-
amine instead of methyl amine in the cyclization step
(50 ꢀC, 24 h), the product was obtained as a yellow solid
1
290 C; H NMR (CDCl3) d 2.91 (m, 1H), 3.02 (s, 3H),
3.51 (m, 1H), 4.07 (AB quartet, 2H, J=58.0, 18.0 Hz),
4.35 (broad d, 1H, J=7.6 Hz), 5.91 (s, 2H), 6.68-7.96
(m, 8H); MS (m/z): 390 [M+H]+; HRMS calcd M+ for
C22H19N3O4 389.1376; found 389.1386.
ꢀ
1
in 26% yield. 7h: mp 238–240 C; H NMR (CDCl3) d
2.48 (m, 1H), 2.75 (m, 2H), 3.02 (s, 3H), 3.09 (t, 2H,
J=8.5 Hz), 3.42 (m, 1H), 4.73 (AB quartet, 2H,
J=129.0, 15.0 Hz), 6.60 (d, 1H, J=7.6 Hz), 7.29–6.96
(m, 6H), 7.52 (d, 1H, J=7.6 Hz), 8.98 (s, 1H); MS
(m/z): 467 [M+H]+; 465 [MꢂH]ꢂ; HRMS calcd
(M+H)+ for C27H22N4O4 467.1719; found 467.1721.
5.4.5. (6S,12aR)-6-(Benzo[1,3]dioxol-5-yl)-2-methyl-2,3,6,
hexahydropyrazino[10,20,5:1,6]pyrido[3,4-b]-
7,12,12a
indole-1,4-dione (7c) and (6S,12aS)-6-(Benzo[1,3]dioxol-
5-yl)-2-methyl-2,3,6,7,12,12a hexahydropyrazino [10,20:1,6]
pyrido[3,4-b]indole-1,4-dione (7d). Following the same
procedure as 7a and 7b, starting from 4b, and in the
reaction with methyl amine step, reaction was allowed
to stir at 50 ꢀC for 48 h. 7c was obtained as a white solid
in 21% yield; 7d was obtained as an off-white solid in
18% yield. 7c: [a]D20 +250 (c 1.00, CHCl3); mp 286–
5.4.9. Preparation of compound 7i:. (1S,3R)-2-Acryloyl-
1-(2,3-dihydro-benzofuran-5-yl)-2,3,4,9-tetrahydro-1H-ꢀ-
carboline-3-carboxylic acid methyl ester (11b). To a
solution of 5b (1.80 g, 5.17 mmol) in dichloromethane
(165 mL) was added acryloyl chloride (0.85 mL, 10.34
mmol) and sodium bicarbonate (0.504 g, 6.00 mmol).
The reaction mixture was stirred at room temperature for
4 h before quenched with water (100 mL). The aqueous
layer was further extracted with dichloromethane (100
mL). After washing with brine (2ꢁ150 mL), the
288 ꢀC; H NMR (CDCl3) d 2.95 (m, 1H), 3.02 (s, 3H),
1
3.56 (m, 1H), 4.10 (AB quartet, 2H, J=55.6, 14.0 Hz),
4.39 (m, 1H), 5.96 (s, 2H), 6.72-7.88 (m, 8H); HRMS
calcd M+ for Cꢀ22H19N3O4 389.1376, found 389.1367.
1
7d: mp 302–304 C decomposed; H NMR (CDCl3) d
3.03 (s, 3H), 3.21 (m, 1H), 3.76 (m, 1H), 4.03 (AB

W. Jiang et al. / Bioorg. Med. Chem. 12 (2004) 1505–1515
1513
combined organic layers were dried and concentrated.
The crude product was purified by silica gel column
eluted with 10–50% ethyl acetate/hexane to provide the
6.98–7.16 (m, 2H), 7.25 (d, 1H, J=10.0 Hz), 7.48 (d,
1H, J=10.0 Hz); MS (m/z): 404 [M+H]+, 426
[M+Na]+, 829 [2M+Na]+; 402 [MꢂH]ꢂ; HRMS
calcd (M+H)+ for C22H17N3O5 404.1246; found
404.1245.
1
product as a white solid (1.84 g, 89%). 11b: H NMR
(CDCl3) d 3.11 (m, 5H), 3.65 (m, 1H), 4.49 (t, 2H,
J=8.2 Hz), 5.72 (d, 1H, J=10.2 Hz), 6.61 (m, 2H), 6.81
(m, 1H), 6.98 (m, 1H), 7.31–7.11 (m, 5H), 7.57 (d, 1H,
J=6.2 Hz), 8.03 (s, 1H); MS (m/z): 403 [M+H]+, 425
[M+Na]+, 401 [MꢂH]ꢂ.
5.4.13. (4aS,11R)-11-(Benzo[1,3]dioxol-5-yl)-3-methyl-
2,3,4a,11-tetrahydro-10H-3,10,11a-triaza-benzo[b]fluorene-
1,4,5-trione (9b). Compound 9b was prepared following
the same procedure as compound 9a in 50% yield. 9b:
mp 185–187 ꢀC; 1H NMR (CD3OD) d 3.02 (s, 3H), 3.33
(m, 1H), 3.53 (m, 1H), 4.73 (m, 1H), 5.92 (m, b, 2H),
6.76 (s, 1H), 6.87 (s, 1H), 6.94 (s, 1H), 7.12 (m, 2H), 7.29
(d, 1H, J=8.7 Hz), 7.52 (d, 1H, J=8.7 Hz), 7.98 (s,
1H). MS (m/z): 404 [M+H]+, 426 [M+Na]+, 829
[2M+Na]+, 402 [MꢂH]ꢂ; HRMS calcd (M+H)+ for
C22H17N3O5 404.1246, found 404.1236.
5.4.10. (1S,3R)-1-(2,3-Dihydro-benzofuran-5-yl)-2-(3-
methylamino-propionyl)-2,3,4,9-tetrahydro-1H-ꢀ-carbo-
line-3-carboxylic acid methyl ester (12b). To a solution
of 11b (1.84 g, 4.58 mmol) in methanol (40 mL) was
added methylamine (2.0 M in THF, 38.9 mL, 77.8
mmol). The reaction mixture was stirred at 50 ꢀC for 2 h
before concentrated to yield a yellow oil. The crude
product was purified by silica gel column eluted with
5% methanol/dichloromethane to provide the product
as a white solid (1.78 g, 90%). 12b: 1H NMR (CDCl3) d
2.38 (s, 3H), 2.51 (m, 1H), 2.75 (m, 2H), 3.05 (s, 3H),
3.10 (m, 2H), 3.48 (m, 2H), 4.51 (t, 2H, J=6.8 Hz,),
4.71 (d, 1H, J=4.0 Hz), 6.61 (d, 1H, J=7.6 Hz), 6.91 (s,
2H), 7.09–7.29 (m, 4H), 7.52 (d, 1H, J=4.0 Hz), 9.0 (s,
1H),; MS (m/z): 434 [M+H]+, 432 [MꢂH]ꢂ.
5.4.14. (4aR,11S)-11-(Benzo[1,3]dioxol-5-yl)-3-methyl-
2,3,4a,11-tetrahydro-10H-3,10,11a-triaza-benzo[b]fluor-
ene-1,4,5-trione (9c). Compound 9c was prepared fol-
lowing the same procedure as compound 9a in 63%
ꢀ
1
yield. 9c: mp 232–234 C; H NMR (CD3OD) d 3.01–
3.27 (m, 4H), 3.22 (s, 3H), 3.58 (m, 1H), 4.75 (m, 1H),
5.92 (m, 2H), 6.74–7.19 (m, 5H), 7.34 (d, 1H, J=10.3
Hz), 7.52 (d, 1H, J=10.3 Hz); MS (m/z): 426 [M+Na]+,
404 [M+H]+, 829 [2M+Na]+, 402 [MꢂH]ꢂ; HRMS
calcd (M+H)+ for C22H17N3O5 404.1246, found
404.1261.
5.4.11. (11S,5aR)-11-(2,3-Dihydro-benzofuran-5-yl)-7-
methyl-5,5a,8,9,11,12-hexahydro-7H-7,10a,12-triaza-cy-
clohepta[b]fluorene-6,10-dione (7i). A solution of 12b
(1.78 g, 4.13 mmol) in methanol (80 mL) was stirred at
50 ꢀC for 48 h until the complete consumption of start-
ing material monitored by HPLC. After concentration,
the crude product was purified by silica gel column
eluted with 5-10% methanol/dichloromethane to pro-
vide the ꢀproduct as a white solid (1.60 g, 96%). 7i: mp
5.4.15. (4aS,11S)-11-(Benzo[1,3]dioxol-5-yl)-3-methyl-
2,3,4a,11-tetrahydro-10H-3,10,11a-triaza-benzo[b]fluorene-
1,4,5-trione (9d). Compound 9d was prepared following
the same procedure as compound 9a in 20% yield. 9d:
mp 180–182 ꢀC; 1H NMR (CD3OD) d 3.03 (s, 3H), 3.54
(m, 1H), 4.03 (m, 1H), 4.24 (m, 1H), 4.74 (m, 1H), 5.88
(m, 2H), 6.69-7.21 (m, 5H), 7.32 (d, 1H, J=10.8 Hz),
7.51 (d, 1H, J=10.8 Hz); MS (m/z): 426 [M+Na]+, 404
[M+H]+, 829 [2M+Na]+, 402 [MꢂH]ꢂ; HRMS calcd
(M+H)+ for C22H17N3O5 404.1246, found 404.1236.
1
190–192 C; H NMR (CDCl3) d 2.89 (m, 2H), 3.08 (s,
3H), 8.01 (s, 1H), 3.15 (t, 2H, J=6.2 Hz), 3.22 (m, 1H),
3.48–3.38 (m, 3H), 3.75 (m, 1H), 4.61–4.48 (m, 5H), 6.65
(d, 1H, J=6.0 Hz), 7.31–6.92 (m, 6H), 7.48 (d, 1H,
J=5.2 Hz); MS (m/z): 402 [M+H]+, 424 [M+Na]+,
825 [2M+Na]+, 400 [MꢂH]ꢂ; HRMS calcd (M+H)+
for C24H23N3O3 402.1818, found 402.2047.
5.4.16. (4aR,11R)-11-(2,3-Dihydro-benzofuran-6-yl)-3-
methyl-2,3,4a,11-tetrahydro-10H-3,10,11a-triaza-ben-
zo[b]fluorene-1,4,5-trione (9e). Compound 9e was pre-
pared following the same procedure as compound 9a in
5.4.12. Procedure for converting 7a to 9a:. (4aR,11R)-
11-(Benzo[1,3]dioxol-5-yl)-3-methyl-2,3,4a,11-tetrahydro-
10H-3,10,11a-triaza-benzo[b]fluorene-1,4,5-trione (9a).
To a solution of compound 7a (0.106 g, 0.272 mmol) in
N,N-dimethylformamide (2.0 mL) was added KOtBu
(0.46 mL, 1.0 M in THF, 0.46 mmol). Dry air was bub-
bled through the reaction mixture under stirring for 4 h.
Excess amount of KOtBu was neutralized by HCl (0.23
mL, 2.0 M in ether, 0.46 mmol). To the reaction mix-
ture, PyBrOP (0.14 g, 0.299 mmol, 1.1 equiv) and di-iso-
propyl ethylamine (0.095 mL, 0.544 mmol) were then
added. After 16 h at 25 ꢀC, the reaction mixture was
partitioned in ethyl acetate/water (30 mL/30 mL). The
organic layer was washed with brine (3ꢀ30 mL), dried
and concentrated. The crude product was purified by
silica gel column eluted with 1% methanol/dichloro-
methane to provide the product as a yellow solid (16.2
1
30% yield. 9e: H NMR (CD3OD) d 3.05 (s, 3H), 3.15
(t, 2H, J=9.3 Hz), 3.54 (m, 1H), 4.05 (d, 1H, J=18.0
Hz), 4.28 (d, 1H, J=18.0 Hz), 4.57 (t, 2H, J=9.3 Hz),
4.72 (m, 1H), 6.65 (m, 1H), 6.92-7.32 (m, 5H), 7.51 (m,
1H), 7.98 (broad s, 1H;-NH); MS (m/z): 424 [M+Na]+,
402 [M+H]+, 825 [2M+Na]+, 400 [MꢂH]ꢂ; HRMS
calcd (M+H)+ for C23H19N3O4 402.1454, found
402.1450.
5.4.17. (4aS,11R)-11-(2,3-Dihydro-benzofuran-6-yl)-3-
methyl-2,3,4a,11 -tetrahydro-10H-3,10,11a-triaza-ben-
zo[b]fluorene-1,4,5-trione (9f). Compound 9f was pre-
pared following the same procedure as compound 9a in
8% yield. 9f: mp 190–192 ꢀC; 1H NMR (CD3OD) d
3.01-3.15 (m, 4H), 3.23 (s, 3H), 3.14 (t, 2H, J=9.3 Hz),
3.55 (m, 1H), 4.56 (t, 2H, J=9.3 Hz), 4.71 (m, 1H), 6.61
(m, 1H), 6.91-7.28 (m, 5H), 7.51 (m, 1H); MS (m/z): 424
[M+Na]+, 402 [M+H]+, 825 [2M+Na]+, 400
mg, 44%). 9a: mp 208–210 ꢀC; H NMR (CD3OD) d
1
3.12 (m, 2H), 3.20 (s, 3H), 3.52 (m, 2H), 4.68 (m, 1H),
5.88 (m, 2H), 6.74 (s, 1H), 6.84 (s, 1H), 6.94 (s, 1H),

1514
W. Jiang et al. / Bioorg. Med. Chem. 12 (2004) 1505–1515
[MꢂH]ꢂ; HRMS calcd (M+H)+ for C23H19N3O4
Center as supplementary publication number CCDC
225080. Copies of the data can be obtained, free of charge,
on application by e-mail to deposit@ccdc.cam.ac.uk.
402.1454, found 402.1472.
5.4.18. (4aR,11R)-11-(Benzo[1,3]dioxol-5-yl)-3-benzyl-
2,3,4a,11-tetrahydro-10H-3,10,11a-triaza-benzo[b]fluorene-
1,4,5-trione (9g). Compound 9g was prepared following
the same procedure as compound 9a in 35% yield. 9g:
1H NMR (CD3OD) d 3.03 (m, 1H), 3.54 (m, 2H), 3.95
(m, 1H), 4.12 (m, 1H), 4.78 (m, 1H), 5.88 (m, 2H), 6.69–
7.51 (m, 12H); MS (m/z): 502 [M+Na]+, 981
[2M+Na]+, 478 [MꢂH]ꢂ; HRMS calcd (M+H)+ for
C28H21N3O5 480.1559, found 480.1558.
Acknowledgements
The authors would like to express gratitude to Ms. Amy
Maden, Dr. Alexandra Shedlow for NOE and X-ray
data, Dr. Naresh Jain for assistance in HPLC and Dr.
Raymond Ng for assistance in manuscript preparation.
References and notes
5.4.19. (4aR,11R)-11-(Benzo[1,3]dioxol-5-yl)-3-pyridin-2-
ylmethyl-2,3,4a,11-tetrahydro-10H-3,10,11a-triaza-ben-
zo[b]fluorene-1,4,5-trione (9h). Compound 9h was pre-
pared following the same procedure as compound 9a in
1. (a) Corbin, J. D.; Francis, S. H. J. Biol. Chem. 1999, 274,
13729. (b) Soderling, S. H.; Beavo, J. A. Curr. Opin. Cell
Biol. 2000, 12, 174. (c) Francis, S. H.; Turko, I. V.; Cor-
bin, J. D. Prog. Nucleic Acid Res. Mol. Biol. 2001, 65, 1.
2. (a) For MED, see: Lue, T. F. Engl. J. Med. 2000, 342,
1802. (b) Feldman, H. A.; Goldstein, I.; Hatzichristou,
D. G.; Krane, R. J.; McKinlay, J. B. J. Urol. 1994, 151,
54.
1
8% yield. 9c: H NMR (CD3OD) d 3.54 (m, 2H), 4.81
(m, 2H), 5.18 (m, 2H), 5.92 (s, 2H), 6.71–8.71 (m, 11H);
MS (m/z): 481 [M+H+], 503 [M+Na]+, 983
[2M+Na]+; 479 [MꢂH]ꢂ; HRMS calcd (M+H)+ for
C27H20N4O5 481.1512, found 481.1554.
3. For Viagra1 (sildenafil), see: Brook, G. Drugs Today
2000, 26, 125 and references therein..
5.4.20. (5bR,11S)-11-(2,3-Dihydro-benzofuran-5-yl)-7-
methyl-5,5a,8,9,11,12-hexahydro-7H-7,10a,12-triaza-cyclo-
hepta[b]fluorene-6,10-dione (9i). To a solution of com-
pound 7i (0.095 g, 0.24 mmol) in N,N-dimethylformamide
(2.5 mL) was added KOtBu (0.47 mL, 1.0 M in THF, 0.47
mmol). The reaction mixture was stirred with dry air bub-
bling for 1 h. Then the reaction mixture was partitioned
in ethyl acetate/water (30 mL/30 mL). The organic layer
was washed with brine (3ꢁ15 mL), dried and con-
centrated. The crude product was purified on silica gel
eluted with 1% methanol/methylene chloride to provide
the product as a yellow solid 9i (10.0 mg, 10%) with
recovered starting material and dehydrogenated starting
material (ꢃ10 mg, ꢃ10%). 9i: mp 193–195 ꢀC; 1H
NMR (CD3OD) d 2.68 (t, 2H, J=8.2 Hz), 3.21 (s, 3H),
3.23 (s, 1H ), 3.39 (t, 2H, J=8.2 Hz), 3.85 (m, 2H), 4.68
(m, 2H), 6.98 (d, 1H, J=9.6 Hz), 7.38–7.86 (m, 6H); MS
(m/z): 416 [M+H]+, 414 [MꢂH]ꢂ; HRMS calcd
(M+H)+ for C23H19N3O5 416.1610, found 416.1612.
4. (a) For Cialis1 (tadalafil), see:. Drugs Future 2001, 26 (2),
141 and references therein. (b) For discovery of tadalafil,
see: Daugan, A.; Grondin, P.; Ruault, C.; Gouville, A.-
C. L. M.; Coste, H.; Kirilovsky, J.; Hyafil, F.; Labaudi-
niere, R. J. Med. Chem. 2003, 46, 4525. (c) Daugan, A.;
Grondin, P.; Ruault, C.; Gouville, A.-C. L. M.; Coste, H.;
Linget, J. M.; Kirilovsky, J.; Hyafil, F.; Labaudiniere, R.
J. Med. Chem. 2003, 46, 4533.
5. For Levitra1 (vardenafil), see:. Drugs Future 2001, 26 (1),
15 and references therein..
6. (a) For comparison of the three launched PDE5 inhibi-
tors, see: Gresser, U.; Gleiter, C. H. Eur. J. Med. Res.
2002, 7, 435. (b) Sui, Z. Expert Opin. Therap. Pat. 2003,
13, 1373.
7. (a) Sorbera, L. A.; Martin, L.; Leeson, P. A.; Castaner, J.
Drugs Future 2001, 26, 15. (b) Wang, H.; Ganesan, A. J.
Org. Chem. 2000, 65, 4685. (c) Wang, H.; Ganesan, A.
Org. Lett. 1999, 1, 1647. (d) Wang, H.; Usui, T.; Osada,
H.; Ganesan, A. J. Med. Chem. 2000, 43, 1577. (e)
Nussbaum, F.; Danishefsky, S. J. Angew, Chem. Int. Eng.
Ed. 2000, 39, 2175. (f) Nakagawa, M.; Fukushima, H.;
Kawate, T.; Hongu, M.; Une, T.; Kodato, S.; Tani-
guchi, M.; Hino, T. Chem. Pharm. Bull. 1989, 37, 23. (g)
Daugan, A. C.-M. JP 199509221, US 6140329, WO
9703675.
8. (a) Sui, Z.; Macielag, M. J.; Guan, J.; Jiang, W.; Lanter,
J. C. PCT Int. Appl. 2001, WO0187882. (b) Sui, Z.;
Macielag, M. J. PCT Appl. 2001, WO 0187038. (c) Jiang,
W.; Sui, Z. PCT Int. Appl. 2003, WO 0342213. (d) Sui, Z.;
Guan, J.; Macielag, M. J.; Jiang, W.; Qiu, Y.; Kraft, P.;
Bhattacharjee, S.; John, T. M.; Craig, E.; Haynes-John-
son, D.; Clancy, J. Bioorg. Med. Chem. Lett. 2003, 13,
761. (e) Sui, Z.; Guan, J.; Macielag, M. J.; Jiang, W.;
Zhang, S.; Qiu, Y.; Kraft, P.; Bhattacharjee, S.; John,
T. M.; Craig, E.; Haynes-Johnson, D.; Clancy, J. J. Med.
Chem. 2002, 45, 4094. (f) Jiang, W.; Sui, Z.; Macielag,
M. J.; Walsh, S. P.; Fiordeliso, J. J.; Lanter, J. C.; Guan,
J.; Qiu, Y.; Kraft, P.; Bhattacharjee, S.; Craig, E.;
Haynes-Johnson, D.; John, T. M.; Clancy, J. J. Med.
Chem. 2003, 46, 441. (g) Jiang, W.; Sui, Z.; Chen, X. Tet.
Lett. 2002, 43, 8941. (h) Jiang, W.; Zhang, X.; Sui, Z. Org.
Lett. 2003, 5, 43. (i) Zhang, X.; Jiang, W.; Sui, Z. J. Org.
Chem. 2003, 68, 4523. (j) Lanter, J. C.; Walsh, S. P.; Sui,
KO2 oxidation of compound 7a to compound 9a
To a solution of b-carboline 7a (38.9 mg, 0.10 mmol)
and 18-crown-6 (26.4 mg, 0.10 mmol) in N,N-dimethyl-
formamide (1.0 mL) was added KO2 (35.5 mg, 0.50
mmol). The reaction mixture was allowed to stir at
25 ꢀC for 6 h before quenched with a few drops of water.
The crude mixture was then partitioned between ethyl
acetate/brine (50 mL/50 mL). The organic layer was
separated and washed with brine (3ꢁ50 mL). After
dried and concentrated, the oily crude product was
purified by silica gel column eluted with 1% methanol/
methylene chloride to provide compound 9a as a yellow
solid (15.3 mg, 38%).
6. Supplementary material
Crystallographic data for the structure 4b have been
deposited with the Cambridge Crystallographic Data

W. Jiang et al. / Bioorg. Med. Chem. 12 (2004) 1505–1515
1515
Z.; Macielag, M. J.; Jiang, W.; Fiordeliso, J. J.; Guan, J.;
Qiu, Y.; Kraft, P.; Lombardi, E.; John, T. M.; Clancy, J.
J. Med. Chem. in press. (k) Qiu, Y.; Bhattacharjee, S.;
Kraft, P.; John, T. M.; Craig, E.; Haynes-Johnson, D.;
Guan, J.; Jiang, W.; Macielag, M.; Sui, Z.; Clancy, J.;
Lundeen, S. Euro. J. Pharmacol. 2003, 472, 73.
(c) Boch, M.; Korth, T.; Nelke, J. M.; Pike, D.; Radunz,
H.; Winterfeldt, E. Chem. Ber. 1972, 105, 2126. (d) Car-
niaux, J.-F.; Kan-Fan, C.; Royer, J.; Husson, H.-P. Tet-
rahedron Lett. 1997, 38, 2997.
10. (a) Ungemach, F.; Soerens, D.; Weber, R.; DiPierro, M.;
Campos, O.; Mokry, P.; Cook, J. M.; Silverton, J. V. J.
Am. Chem. Soc. 1980, 102, 6979. (b) Bailey, P. D.; Hol-
linshead, S. P. Heterocycles 1987, 26, 389.
9. (a) Winterfeldt, E. Liebigs Ann. Chem. 1971, 745, 23. (b)
Warneke, J.; Winterfeldt, E. Chem. Ber. 1972, 105, 2120.