Synthesis and structure-activity relationship of novel lactam-fused chroman derivatives having dual affinity at the 5-HT1A receptor and the serotonin transporter

Article abstract of DOI:10.1016/j.bmcl.2009.10.134

The structure-activity relationship (SAR) for three series of lactam-fused chroman derivatives possessing 3-amino substituents was evaluated. Many compounds exhibited affinities for both the 5-HT_1A receptor and the 5-HT transporter. Compounds 45 and 53 demonstrated 5-HT_1A antagonist activities in the in vitro cAMP turnover model.

Full text of DOI:10.1016/j.bmcl.2009.10.134

Bioorganic & Medicinal Chemistry Letters 20 (2010) 222–227
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Synthesis and structure–activity relationship of novel lactam-fused
chroman derivatives having dual affinity at the 5-HT1A receptor
and the serotonin transporter
Zhongqi Shen a, , P. Siva Ramamoorthy , Nicole T. Hatzenbuhler , Deborah A. Evrard , Wayne Childers ,
a
a
a
a
*
a
a
b
b
b
Boyd L. Harrison , Michael Chlenov , Geoffrey Hornby , Deborah L. Smith , Kelly M. Sullivan ,
b
b
Lee E. Schechter , Terrance H. Andree
a
Chemical and Screening Sciences, Wyeth Research, CN 8000, Princeton, NJ 08543, United States
Discovery Neuroscience, Wyeth Research, CN 8000, Princeton, NJ 08543, United States
b
a r t i c l e i n f o
a b s t r a c t
Article history:
The structure–activity relationship (SAR) for three series of lactam-fused chroman derivatives possessing
Received 20 August 2009
Revised 27 October 2009
Accepted 29 October 2009
Available online 31 October 2009
3
-amino substituents was evaluated. Many compounds exhibited affinities for both the 5-HT1A receptor
and the 5-HT transporter. Compounds 45 and 53 demonstrated 5-HT1A antagonist activities in the in vitro
cAMP turnover model.
Ó 2009 Elsevier Ltd. All rights reserved.
Keywords:
Selective serotonin reuptake inhibitors
Central nervous system
Antidepressant
5
-HT1A antagonist
Lactam-fused chroman amine derivatives
Selective serotonin reuptake inhibitors (SSRIs) have been impli-
cated in disorders of the central nervous system and have become
of particular interest for their roles in the autoregulation of the
brain 5-HT system and in the treatment of depression.1 Because
of the improved therapeutic index of SSRIs compared to traditional
effect in humans, although the clinical tool compounds used in
these studies were not optimal.
8
An alternative to administering two pharmaceutical agents is to
design molecules that possess both desired pharmacological
activities. To this end, significant amounts of effort have been
focused on a synthetic strategy developing molecules containing
a nitrogen-containing template and a 5-HT1A antagonist moiety
(1, Fig. 1) with a serotonin transporter group (2, Fig. 1). This strat-
egy has led to compounds that demonstrate desirable binding
affinities for both the 5-HT1A receptor and the 5-HT transporter.
As part of an ongoing effort to identify single molecules with dual
activity to inhibit both the 5-HT1A and 5-HT reuptake sites, we re-
ported on a series of analogs containing a chroman-based 5-HT1A
2
tricyclic antidepressants, attention has been increasingly focused
3
on overcoming their drawbacks. SSRIs display latency in the onset
of clinic antidepressant effects for more than three weeks as well
4
as non-consistent responses in some refractory patients. Other
notable side effects include sexual dysfunction, anxiety and gastro-
intestinal intolerance. The discovery of better antidepressants with
less adverse events continues to be an area of significant research.
One strategy to accelerate the onset of action of SSRIs involves co-
administering a 5-HT1A antagonist to block the initial reduced rate
9
,10
antagonist pharmacophore inherent in robalzotan (NAD-299)
5
11
of neuronal firing until neuronal accommodation occurs. Addi-
and indolealkylamine fragments such as 2. The SAR showed that
5-carboxamide-8-fluoro derivatives and 5-carboxamide-8-des-flu-
oro analogs with the proper N-alkyls displayed good affinities for
tionally, 5-HT1A antagonists may lessen the sexual dysfunction
6
liabilities seen with SSRIs. Support for this approach comes from
animal studies in which 5-HT1A antagonists decreased the amount
of time required to see antidepressant effects with SSRIs. Addi-
tionally, some clinical studies demonstrate the potential for similar
the 5-HT1A receptor and the 5-HT reuptake site in the GTPcS
7
12
assay. That work culminated in the identification of (3R)-3-
{cyclobutyl[3-(5-fluoro-1H-indol-3-yl) propyl] amino}-8-fluoroch-
roman-5-carboxamide 3, which possessed both potent 5-HT1A
receptor antagonist and SSRI activities (see data in Fig. 1). In a
micro dialysis study (30 mg/kg po), compound 3 induced acute
increases in serotonin levels in the rat frontal cortex that were
*
Corresponding author. Tel.: +1 732 274 4865; fax: +1 732 274 4505.
E-mail address: shenz@wyeth.com (Z. Shen).
0
960-894X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2009.10.134

Z. Shen et al. / Bioorg. Med. Chem. Lett. 20 (2010) 222–227
223
O
NH
2
NH
2
H N
N
O
F
F
2. 5-HT Transporter
1
. NAD-299
O
O
NH
2
HN
n = 1, 2
R
N
NH
NH
N
O
O
F
X
Y
F
3
. Lead Compound
Ki = 1.2 nM, IC50 = 27 nM in cAMP, Emax = 0%
5-HT1A receptor)
Ki= 1.5 nM, IC50 = 18.7 nM in functional assay
5-HT Transporter)
4A. n = 1, X = H, Y = F, CN
4B. n = 2, X = H, Y = F
4
C. n = 2, X = F, Y = F
(
(
Figure 1. Design strategy.
similar to those achieved only with chronic (14 days) dosing of an
SSRI alone. To continue the SAR effort around compound 3 in Fig-
ure 1, we designed and prepared a series of lactam-fused chroman
amine derivatives (4, Fig. 1). The present manuscript describes the
synthesis and pharmacological evaluation of those fused chroman
analogs.
Due to the difficulties in the synthesis of five-membered lac-
tam-fused chromans with 9-fluoro and 3-amino substituents, we
initially focused on the synthesis of key intermediate 13 (Scheme
indanone 7 as a single regioisomer, which was purified by chroma-
tography. The six-membered lactams 8 and 9 were obtained in the
respective ratio of 2:1 at room temperature by Schmidt reaction.
These two isomers were easily purified by chromatography. It
should be noted that higher reaction temperatures favored the for-
mation of the undesirable isomer 8. Compound 10 was synthe-
sized from 9 in two steps via a sequence that involved alkylation
with propargyl bromide followed by cyclization. We modified
the reported nitration procedure for the conversion of 10 to 11
by replacing the THF/pyridine solvent system with N-methylpyr-
rolidine for easy handling of the material during the work up of
the reaction. Reduction of the double bond in pyran derivative
11 followed by Raney nickel catalyzed reduction of the nitro group
in intermediate 12 yielded the desired six-membered lactam-
fused chroman amine 13. The enantiomers of 13 were not sepa-
rated at this stage of the synthesis.
1
). The formation of the lactam-fused chroman relies on the pres-
ence of a carboxylic acid moiety either in the para or meta position
of 2-fluorophenol. We began the synthesis by introducing an ,b
a
unsaturated propionate at the para position. This was achieved
in 56% yield by reacting 4-bromo-2-fluorophenol with methylac-
rylate under Heck conditions. Subsequent double bond reduction,
ester hydrolysis and intramolecular cyclization yielded substituted
O
O
O
MeO
HO
Br
a
b,c
d
e
OH
OH
OH
OH
F
F
F
F
5
6
7 (single regioisomer)
O
H
N
H
N
H
N
O
O
O
NH
F
i
f,g
h
+
NO2
OH
O
OH
O
F
F
F
11
8
(57 %)
9 (27 %)
10
H
H
N
N
O
O
NO2
j
NH2
O
O
F
F
12
13
Scheme 1. Synthesis of six-membered lactam chroman amine. Reagents and conditions: (a) methyl acrylate, Pd(OAc)
C, MeOH, 95%; (c) LiOH, MeOH/THF/H O, 90%; (d) AlCl , NaCl, 180 °C, 30 min; 78%; (e) MsOH, NaN , CHCl , 57% and 27%; (f) propargyl bromide, K
diethylaniline, 210 °C, 70%; (h) NaNO , I , NMP, 73%; (i) NaBH , SiO , CHCl , 60%; (j) NH NH , Raney Ni, EtOH, 50 °C, 72%.
2
, tri-o-tolylphosphine, Et
3
N, DMF, 80 °C, 84%; (b) H
2
, Pd/
2
3
3
3
2
CO , acetone, reflux, 90%; (g)
3
2
2
4
2
3
2
2

2
24
Z. Shen et al. / Bioorg. Med. Chem. Lett. 20 (2010) 222–227
H
N
H
N
H
N
O
O
O
O
d
a,b
c
O
OH
O
OMe
14
15
16
H
N
H
N
O
O
g
e,f
NO
2
NH2
O
O
17
18
Scheme 2. Synthesis of the des-fluoro six-membered lactam-fused chroman amine. Reagents and conditions: (a) MsOH, NaN
CH Cl , 77%; (c) propargyl bromide, K CO , acetone, reflux, 89%; (d) diethylaniline, 210 °C, 55%; (e) NaNO , I , NMP, 71%; (f) NaBH
EtOH, 50 °C, 61%.
3
, CHCl
3
, 0 °C, 87%; (b) BBr
3
(1 M in hexane),
NH , Raney Ni,
2
2
2
3
2
2
4
, SiO
2
, CHCl , 83%; (g) NH
3
2
2
We then turned our attention to the synthesis of the des-fluoro
six-membered lactam 18 (Scheme 2). Commercially available 6-
methoxy-1-indanone was used as starting material to obtain 14
in two steps. We were concerned that the cyclization of propargyl
ether 15 could yield two conceivable regioisomers. To our surprise,
compound 16 was the exclusive product obtained from the reac-
tion. Compound 18 was then prepared using analogous methods
to those shown in Scheme 1.
The synthesis of the des-fluoro five-membered lactam-fused
chroman amine 25 (Scheme 3), required an efficient route to inter-
mediate 22. The most significant reaction in this scheme was the
transformation of the aryl halide 19 to the toluene derivative 20.
Transition metal catalyzed cross coupling of an aryl bromide with
alkyl zinc is a known approach.13 The reaction of 19 with dimeth-
ylzinc in DMF at 50 °C was catalyzed by bis-(triphenylphosphine)
palladium(II) dichloride to produce 20. The formation of 20 pro-
vided the eventual synthesis of the five-membered lactam 22,14
which was elaborated to 25 as an exclusive regioselective isomer
using standard synthetic methodology.
Since we envisioned preparing the desired final compounds
using a reductive amination approach, the synthesis of indole alkyl
aldehyde intermediates was required. The preparation of these
compounds (27A and 27B) is shown in Scheme 4. Compounds 26
were synthesized by Fisher indole synthesis, and then converted
to aldehyde derivatives. The four-carbon homolog 27C was pre-
pared from 26A via a four-step sequence that involved bromina-
COOMe
COOMe
COOMe
COOH
Me
Br
c
Br
b
Br
a
OM
e
OMe
OMe
OMe
O
1
9
20
21
O
O
O
HN
HN
HN
HN
3
steps (42%)
d
e
f,g
NH
2
See Scheme 2
OMe
OH
O
O
2
2
23
24
25
2 2 3 2
Scheme 3. Synthesis of des-fluoro five-membered lactam-fused chroman amine. Reagents and conditions: (a) MeI, DBU, MeCN, reflux,73%; (b) Me Zn, Cl Ni(PPh ) , DMF,
5
5
5 °C, 82%; (c) NBS, AIBN, benzene, reflux, 88%; (d) NH
8%.
3
/MeOH, 130 °C, 92%; (e) BBr
3
2
, CH Cl
2
, 65%; (f) propargyl bromide, K
2
CO
3
, acetone, reflux, 92%; (g) diethylaniline, 210 °C,
CHO
OH
R1
R1
R1
a
b
+
N
H
NHNH
2
.HCl
O
N
H
1
1
27A R
1
= F
R = F, CN
26A R = F
2
1
1
6B R = CN
27B R = CN
CHO
F
N
H
27C
Scheme 4. Synthesis of indole-substituted alkyl aldehydes. Reagents and conditions: (a) dioxane/water, 90 °C, 95%; (b) PCC, Et
3
N, DMSO, 71–80%.

Z. Shen et al. / Bioorg. Med. Chem. Lett. 20 (2010) 222–227
225
H
N
O
CHO
a,b
O
HN
HN
R2
N
NH
2
R1
+
O
N
H
O
R1
X
X
1
1
2
2
1
1
5 n = 1, x = H
3 n = 2, x = F
8 n = 2, x = H
27 R = F, CN
n = 1, 2, X = H, F
R = F, CN, R = alkyl
3 3
Scheme 5. Target synthesis. Reagents and conditions: (a) NaBH CN, AcOH, MeOH, 80–90%; (b) aldehyde or cyclobutanone, NaBH CN, AcOH, MeOH.
tion, cyanation, hydrolysis and then reduction (not shown) in 45%
overall yield.
and the 5-HT transporter (e.g., compounds 28 and 29). Affinities
for both binding sites decreased when somewhat larger R groups
2
With intermediates in hand, the final products 28–56 were syn-
thesized through a two-step sequence (Scheme 5). Reductive ami-
nation of the three different lactam-fused chroman amines (13, 18
and 25) with indole-substituted aldehydes 27 generated penulti-
mate secondary amine derivatives. A second reductive amination
reaction with alkyl aldehydes, aryl aldehydes or cyclobutanone in-
were incorporated, such as the cyclopropylmethyl moiety in com-
pound 30. A similar trend was seen when the 10-fluoro substituent
2
was removed (compounds 31–40), with R = ethyl being the opti-
mal substitution pattern for both fluoro and des-fluoro derivatives.
Similar results were obtained when the 10-fluoro group was re-
placed with a cyano moiety (unpublished data). These results are
2
11
stalled the R group to give the final targets, as shown in Scheme 5.
similar to those seen in the 5-carboxamido analogs, although
2
We investigated the effect of three lactam-fused chroman
amines linked to a known serotonin transporter pharmacophore
as potential dual-acting 5-HT1A antagonist/SSRIs. The 5-HT1A
receptor affinity, serotonin transporter affinity, and in vitro
that system better tolerated somewhat larger groups in the R po-
sition at both the 5-HT1A receptor and the 5-HT transporter. Inter-
estingly, all of the racemic six-membered lactams tested displayed
agonist or partial agonist activity in the functional assay, although
previous evidence suggested that the two enantiomers might dis-
5
-HT1A intrinsic activity are summarized in Tables 1–3. 5-HT1A
1
1
functional activity was measured in CHO cells stably transfected
with the human 5-HT1A receptor by assessing test compound’s
ability to inhibit the agonist-induced decrease in adenylate cyclase
activity. Intrinsic activity is expressed as the percentage (Emax %) of
the full response of the agonist 8-OH-DPAT (Emax % = 100). Based
on early SAR studies for 8-fluoro-5-carboxamide-3-aminochroman
derivatives, we limited R2 to small alkyl groups (methyl, ethyl,
cyclopropylmethyl, cyclobutyl, etc).
play different functional characteristics.
We then turned our attention to five-membered ring lactam
derivatives (Table 2). Both the fluoro and the cyano substituent
at indole moiety were examined. Similar to the six-membered ring
lactams, substituents up to and including cyclopropylmethyl were
2
tolerated in the R position, with some loss in both 5-HT1A and
5-HT transporter affinity seen with larger groups such as cyclobu-
tyl (compound 47). Five-membered lactam derivatives displayed
somewhat greater affinity for both binding sites compared to
analogs six-membered ring lactams (e.g., compounds 43 and 44
compared to compounds 29 and 30, respectively). Because the
racemic mixtures 44 and 52 displayed antagonist profiles in the
in vitro functional assay, we separated and evaluated the enantio-
Table 1 summarizes data for six-membered lactam chroman
containing an indole-3-propylamine moiety. The resulting final
products were obtained and evaluated as racemic mixtures. When
1
the 10-fluoro substituent was present (R = F), small substituents
were tolerated on the basic amine by both the 5-HT1A receptor
Table 1
Data for substituted six-membered lactams
H
H
N
N
O
R2
N
O
F
R1
R1
R2
Stereo
5-HT1A
a
5-HT-T
cAMPc
EC50 (nM)
E
max (%)
d
Compound
b
K
i
(nM)
K
i
(nM)
28
29
30
31
32
33
34
35
36
47
38
39
40
F
F
F
H
Et
(CH
H
Me
Et
n-Pr
i-Pr
(CH
i-Bu
cBu
rac
rac
rac
rac
rac
rac
rac
rac
rac
rac
rac
rac
rac
6
4
50
44
23
6
13
129
19
927
95
966
1220
27
15
112
10
17
30
61
8
26
184
28
4
10
101
25
26
9
34
252
36
NT
54
NT
NT
98
92
62
87
99
97
100
94
91
NT
77
NT
NT
2
)cPr
H
H
H
H
H
H
H
H
H
H
2
)cPr
(CH
Bz
2 2
) CF
3
340
671
a
b
c
3
Binding affinity at 5-HT1A receptor in CHO cells labeled with [ H]-OH-DPAP (n >2).
Binding affinity for 5HT transporter labeled with paroxetine (n >2).
Maximal agonist effect relative to 5-HT in inhibiting forskolin-stimulated adenylate cyclase activity.
Maximal response of the compound as a result of 5-HT1A receptor-mediated stimulation of b( S)GTP
d
35
c
binding.

2
26
Z. Shen et al. / Bioorg. Med. Chem. Lett. 20 (2010) 222–227
Table 2
Data for des-fluoro five-membered lactamsa
H
O
N
R2
N
HN
O
R
1
Compound
R1
R2
Stereo
5-HT1A
5-HT-T
(nM)
cAMP
EC50 (nM)
Emax (%)
K
i
(nM)
K
i
41
42
43
44
45
46
47
48
49
50
51
52
53
54
F
F
F
F
F
F
F
F
F
F
CN
CN
CN
CN
H
Me
Et
(CH
(CH
(CH
cBu
i-Bu
(CH
Bz
H
(CH
(CH
rac
rac
rac
rac
24
22
2
6
8
8
9
8
15
6
5
62
75
28
223 (IC50
320 (IC50
24
350
NT
NT
436
100
80
78
0
2
2
2
)cPr
)cPr
)cPr
)
)
b
(À)
0
b
(+)
3
84
NT
NT
NT
96
91
0
rac
rac
rac
rac
rac
rac
101
NT
NT
85
36
4
63
617
1242
31
13
25
2
)cHex
22
63 (IC50)
203 (IC50
19
2
)cPr
)cPr
a
2
(À)
10
31
17
12
)
0
100
a
(CH2)cPr
(+)
a
See notes in Table 1.
As determined by optical rotation.
b
contained a propyl chain linker, the five-membered lactam chro-
man analogs displayed slightly greater affinity for both binding
sites compared to their six-membered ring lactam counterparts.
The rank order of potency for substituents on the basic amine
was similar for both systems. Like the previously reported car-
boxamides, one enantiomer appears to possess 5-HT1A antagonist
functional activity while the other profiles as an agonist, at least
in vitro. However, the functional potency of the two optimal com-
pounds from this series, 45 and 53, was significantly lower than
that seen with the corresponding carboxamides. We believe that
the constrained conformation of the lactam group has a subtle ef-
fect on the shift from antagonist to almost full agonist. It provides
a novel alternative to the freely rotating primary carboxamide
Table 3
Data for compounds possessing a 4-carbon spacera
O
HN
R
N
NH
O
F
Compound
R
Stereo 5-HT1A
(nM)
5-HT-T
cAMP
EC50 (nM)
E
max (%)
K
i
K
i
(nM)
5
5
5
6
H
(CH
rac
)cPr rac
22
19
20
139
284
49
96
87
moiety in compounds such as 3 and yields potent dual 5-HT1A
/
2
5
-HT transporter ligands that can display 5-HT1A antagonist activ-
a
See notes in Table 1.
ity. However, in these series, either the constrained amide confor-
mation inherent in the lactam group or the differences in
electronics or hydrogen bonding properties induced by moving
from a primary to a secondary amide result in less potent
5-HT1A antagonist activity.
mers. As was seen in the amide series,11 one enantiomer (com-
pounds 45 and 53) carried the antagonist activity while the other
(
compounds 46 and 54) profiled as a partial or full agonist.
However, the EC50 values of both compounds 45 and 53 in the
Acknowledgments
functional assay (IC50 values = 320 nM and 203 nM, respectively)
were much higher than that obtained with compound
IC50 = 27 nM).
Finally, we examined the effect of the length of the carbon
spacer’ between the chroman group and the indole moiety by pre-
paring compounds 55 and 56 (Table 3). The potency of compound
5 for both the 5-HT1A receptor and the 5-HT transporter was
comparable to that seen with the propyl analogs, but compound
6 displayed some loss in affinity relative to its propyl counter-
3
The authors thank Wyeth Discovery Analytical Chemistry for
their support in confirming the structures of the intermediates
and final products contained in this Letter and for their help with
chiral HPLC separations.
(
‘
5
References and notes
1.
2.
3.
Pineyro, G.; Biler, P. Pharmacol. Rev. 1999, 51, 533.
Blackwell, B. Drugs 1981, 21, 201.
Evrard, D. A.; Harrison, B. L. Annu. Rep. Med. Chem. 1999, 34, 1.
5
parts. Because the SAR of the present series correlated with that
of the previously reported carboxamides,1 the ethyl ‘spacer’ group
was not examined.
1
4. Annanth, J. Psychother. Psychosom. 1988, 67, 61.
5.
Zanardi, R.; Artigas, F.; Franchini, L.; Sfozini, L.; Gasperini, M.; Smeraldi, E.;
Perez, J. J. Clin. Psychopharmacol. 1997, 17, 446.
In the present Letter, we have expanded the SAR of a previ-
ously reported series of chroman indole alkylamine analogs that
possess both 5-HT1A and 5-HT transporter affinity. The results
suggested that cyclopropylmethyl played a fairly important role
in the antagonistic properties of the molecules along with the ste-
reochemistry at the 3-position of the chroman. When compounds
6
.
Popova, N. K.; Naumenko, V. S.; Plyunina, I. Z. Neurosci. Behav. Physiol. 2007, 37,
631.
Ballesteros, J.; Callado, L. F. J. Affect. Disord. 2004, 79, 137.
Sargent, P. A.; Kjaer, K. H.; Bench, C. J.; Rabiner, E. A.; Messa, C.; Meyer, J.; Gunn,
R. N.; Grasby, P. M.; Cowen, P. J. Arch. Gen. Psychiat. 2000, 57, 174.
7
8
.
.
9. Evenden, J.; Hammberg, E.; Hansson, H. S.; Hellberg, S. E.; Johansson, L. G.;
Lundkvist, J. R. M.; Ross, S. B.; Sohn, D. D.; Thorberg, S. O. (R)-5-Carbamoyl-8-

Z. Shen et al. / Bioorg. Med. Chem. Lett. 20 (2010) 222–227
227
fluoro-3-N,N-disubstituted amino-3,4 dihydro-2H-1 benzopyrans. U.S.
616610, April 1, 1997.
0. Ross, S. B.; Thorberg, S. O.; Jerning, E.; Mohell, N.; Stenfors, C.; Wallsten, C.;
Michert, I. G.; Ojteg, G. C. N. S. Drug Rev. 1999, 5, 213.
12. Hatzenbuhler, N. T.; Baudy, R.; Evrard, D. A.; Failli, A.; Harrison, B. L.; Lenicek,
S.; Mewshaw, R. E.; Saab, A.; Shah, U.; Sze, J.; Zhang, M.; Zhou, D.; Chlenov, M.;
Kagan, M.; Golembieski, J.; Hornby, G.; Lai, M.; Smith, D.; Sullivan, K. M.;
Schecheter, L. E.; Andree, T. H. J. Med. Chem. 2008, 51, 6980.
5
1
1
1. Hatzenbuhler, N. T.; Evrard, D. A.; Harrison, B. L.; Huryn, D.; Inghrim, J.; Kraml,
C.; Mattes, J. M.; Mewshaw, R. E.; Zhou, D.; Hornby, G.; Lin, Q.; Smith, D. L.;
Sullivan, K. M.; Schechter, L. E.; Beyer, C. E.; Andree, T. H. J. Med. Chem. 2006, 49,
13. Herbert, J. M. Tetrahedron Lett. 2004, 45, 817.
14. Jagtap, P. G.; Southan, G. J.; Baloglu, E.; Ram, S.; Mabley, J. G.; Marton, A.;
Salzman, A.; Szabo, C. Bioorg. Med. Chem. Lett. 2004, 14, 81.
4
785.