Structure-activity relationships in a series of NPY Y5 antagonists: 3-Amido-9-ethylcarbazoles, core-modified analogues and amide isosteres

Article abstract of DOI:10.1016/S0960-894X(03)00329-9

Beginning with carbazole 1a, the amide and alkyl substituents were optimized to maintain potency while adding solubilizing groups. Efforts to replace the 3-amino-9-ethylcarbazole core, a known carcinogen, used the SAR generated in the carbazole series for guidance and led to the synthesis of a number of core-modified analogues. In addition, an isosteric series, in which the amide was replaced with an imidazole, was prepared. Two potent new series lacking the putative toxicophore were identified from these endeavors.

Full text of DOI:10.1016/S0960-894X(03)00329-9

Bioorganic & Medicinal Chemistry Letters 13 (2003) 1989–1992
Structure–Activity Relationships in a Series of NPY Y5
Antagonists: 3-Amido-9-ethylcarbazoles, Core-Modified Analogues
and Amide Isosteres
Marlys Hammond,a,* Richard L. Elliott,a Melissa L. Gillaspy,a David C. Hager,a
Richard F. Hank,a Janet A. LaFlamme,a Robert M. Oliver,a Paul A. DaSilva-Jardine,a
Ralph W. Stevenson,a Christine M. Mackb and James V. Cassellab
aDepartment of Cardiovascular and Metabolic Diseases, Pfizer Global Research and Development, Groton, CT 06340, USA
bNeurogen Corporation, Branford, CT 06405, USA
Received 24 January 2003; revised 24 March 2003; accepted 25 March 2003
Abstract—Beginning with carbazole 1a, the amide and alkyl substituents were optimized to maintain potency while adding solubi-
lizing groups. Efforts to replace the 3–amino-9-ethylcarbazole core, a known carcinogen, used the SAR generated in the carbazole
series for guidance and led to the synthesis of a number of core-modified analogues. In addition, an isosteric series, in which the
amide was replaced with an imidazole, was prepared. Two potent new series lacking the putative toxicophore were identified from
these endeavors.
# 2003 Elsevier Science Ltd. All rights reserved.
Neuropeptide Y (NPY), a member of the PYY family of
peptides, exerts its physiological effects through at least
6 receptor subtypes, the Y1 through Y5 and y6 recep-
tors.1 Of interest to us has been the central orexigenic
effect observed when NPY is given ICV to rodents.2 It is
believed that NPY exerts these effects primarily through
the Y1 and Y5 receptor subtypes. Peptide agonist analo-
gues of NPY having selectivity for the Y5 receptor have
been found to be orexigenic, with the food intake effect
directly correlated to the affinity for the receptor.3 In
addition, it has been postulated that blockade of either or
both of Y1 or Y5 receptors might result in an anorectic
effect. In fact, a number of studies have shown that both
Y1 and Y5 antagonists are capable of blocking the orexi-
genic effects of selective Y1 and Y5 peptide agonists.4
analogues of 1a were prepared as shown in Scheme 1.
Those compounds having R1=Et were synthesized begin-
ning with the commercially available 3-amino-9-ethylcar-
bazole (4, R1=Et) by acylation at the 3-amino position.
Intermediate 4 for analogues in which R1 was modified
from ethyl were prepared by nitration and reduction of
intermediate 3, which was either commercially available
(R1=i-Pr) or prepared by alkylation of carbazole (2). The
aminocarbazoles 4 were then acylated following the same
protocol as for the commercial aminocarbazole.
As part of a program directed at discovering novel,
potent antagonists of the Y5 receptor, we (and others)5
have identified aminocarbazole-based Y5 antagonists.
The initial lead was 3-trifluoroacetamido-9-ethylcarba-
zole 1a, a high-throughput screening hit with an affinity
for the Y5 receptor of 14 nM. In an effort to improve
the potency and solubility within the series, several
Analogues of 1a were screened against human Y5 recep-
tor using 125[I]-PYY as the radioligand.6 Selected binding
data for this series is given in Table 1, and shows that
solubilizing amines present on the amide substituents are
well tolerated. For example, the 4-pyridylacetamide-
containing aminocarbazole 1b was found to have a Y5 Ki
of 4.7 nM. More striking was the effect seen when the
alkyl group at the 9-position was modified. Replacement
of the ethyl group with a methyl group resulted in a
substantial loss in binding potency (1e), while adding
*Corresponding author. Tel.:+1-860-441-0712; fax:+1-860-715-8095;
e-mail: marlys_hammond@groton.pfizer.com
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00329-9

1990
M. Hammond et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1989–1992
Scheme 1. Reagents and conditions: (a) R1Br, NaH, DMF, 23 ꢀC; (b)
HNO3, HOAC, 60 ꢀC; (c) H2, Pd/C, EtOH, 23 ꢀC; (d) trifluoroacetic
anhydride, Et3N, CH2Cl2, 23 ꢀC; (e) N,N-dimethylglycine or 4-pyr-
idylacetic acid, EDC, Et3N, DMAP, CH2Cl2, 23 ꢀC; (f) Me3Al, ethyl
3-N-pyrrolidinopropionate, 1,2-dichloroethane, 50 ꢀC.
Scheme 2. Reagents and conditions: (a) 1-bromo-2-methylpropane,
potassium tert-butoxide, THF, 60 ꢀC; (b) H2NNH2, CH3OH, reflux;
(c) NaNO2, HCl, 0 ꢀC; (d) HOAc, H2O, reflux; (e) Me3Al, ethyl 3-N-
pyrrolidinopropionate, 1,2-dichloroethane, 50 ꢀC.
Table 1. Y5 binding data for 1a–1 h
Compd
R1
R2
Avg. Y5 Ki, nM (n)
1a
Et
14 (2)
1b
1c
1d
1e
1f
Et
Et
4.7 (5)
21 (2)
11 (7)
93 (2)
15 (2)
3.4 (2)
0.3 (2)
0.84 (2)
2.6 (7)
Scheme 3. Reagents and conditions: (a) KMnO4, H2O/acetone, 70 ꢀC;
(b) CDI, CH2Cl2, 23 ꢀC; 1-(3-aminopropyl)morpholine, DMAP,
23 ꢀC.
Et
Me
i-Pr
i-Pr
i-Bu
Et
Scheme 4. Reagents and conditions: (a) 3-morpholino-1-propan-
amine, NaBH(OAc)3, 1,2-dichloroethane, 23 ꢀC.
1g
1h
1i
1j
Et
the more bulky and lipophilic substituents isopropyl
and isobutyl resulted in 3- and 36-fold increases in
potency, respectively (cf. 1d, 1g, and 1h).
Scheme 5. Reagents and conditions: (a) (CO)2Cl2, CH2Cl2, DMF
23 ꢀC, then PhNHR1, Et3N, DMAP, CH2Cl2, 0 ꢀC; (b) Pd(OAc)2,
Ag2CO3, PPh3, CH3CN, reflux; (c) H2, Pd/C, EtOH, 23 ꢀC; (d) LAH,
THF, reflux; (e) trifluoroacetic anhydride, pyridine, CH2Cl2, 23 ꢀC or
R2CO2H, EDC, Et3N, CH2Cl2, 23 ꢀC.
Although potent antagonists of the Y5 receptor having a
core derived from 3-aminocarbazole were readily identi-
fied, we were acutely aware of the risks associated with
this series, as 3-amino-9-ethylcarbazole is a known car-
cinogen.7 Although the structural origin of this toxicity
(the ‘toxicophore’) has not been definitively identified,
we targeted the 1,4-phenylenediamine substructure as a
likely culprit. Thus, guided by the SAR we had devel-
oped in this original series, we began to search for a
non-aminocarbazole core. We investigated a variety of
modified cores, as shown in Schemes 2–6. The b-carbo-
line-derived analogue 8 (Scheme 2) had nearly 250-fold
reduced potency relative to the analogous carbazole 1h.
Both the ‘reverse amide’ 10 (Scheme 3) and the amino-

M. Hammond et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1989–1992
1991
Another potent series arose from replacement of the
amide functionality in the 3-position of the carbazole
core with an isosteric imidazole. Compounds were pre-
pared as depicted in Scheme 6, starting with aldehyde 9.
Heating of 9 and a 1,2-diketone in acetic acid with
ammonium acetate provided imidazolyl carbazoles
17a–c, while heating of 9 and the appropriate aryl 1,2-
diamine in nitromethane provided analogues 17d–f
(Table 3). With respect to binding potency, a range of
imidazole substitution was well-tolerated within these
analogues, including alkyl (17a) and aryl (17b). Incor-
poration of a more polar, potentially solubilizing group
(17c) initially resulted in a drop in potency, but by
incorporating the basicity in a fused system, potency
returned (17d–f). Interestingly, compounds from this
series examined in the Ca2+ functional assay demon-
strated a 10- to 20-fold reduction in potency (17a and
17c) relative to binding affinity.
Scheme 6. Reagents and conditions: (a) ii. R1(CO)2R2, NH4OAc,
AcOH, 100 ꢀC; (b) aryl 1,2-diamine; CH3NO2, 145 ꢀC.
methylcarbazole 11 (Scheme 4) also demonstrated a
significant loss in binding affinity relative to the close-in
carbazole analogue 1i.
Binding potency was ultimately restored to the range of the
original aminocarbazole series when we prepared a series
of phenanthridinone-based analogues. The phenan-
thridinones were prepared as shown in Scheme 5, starting
with the appropriately substituted nitro 2-iodobenzoic acid
12. Amide formation under standard conditions afforded
intermediate 13, which underwent facile Pd-catalyzed ring-
closure to afford phenanthridinones 14.8 Reduction of
both the nitro group and the amide with LAH followed by
acylation afforded phenanthridine 15, which showed shar-
ply decreased potency over the parent carbazoles. Chemo-
selective reduction of the nitro group followed by acylation
provided the phenanthridinones 16a–f (Table 2). Com-
pounds were examined for binding activity and were
found to be of comparable potency to the aminocarba-
zoles. Compounds from this series were also submitted to
a functional assay (Ca2+ mobilization),6 where they
showed 3-fold or more reduction in potency relative to the
binding assay. The most potent of these analogues were
16b and 16e, each containing a 4-pyridylacetamide side-
chain. The R1 alkyl group appeared to have less influence
over binding potency, as evidenced by the similarities in
potency between pairs 16b, 16e and 16c, 16d. The regio-
isomeric 30-substituted phenanthridinone 16f was essen-
tially inactive relative to its 40-substituted isomers.
Several of the Y5 antagonists from the carbazole, phe-
nanthridinone and imidazolylcarbazole series were
examined in vivo in a feeding model using 24-h fasted
rats. Compounds were dosed orally in a methylcellulose
vehicle, food was re-introduced, and total food intake
was measured 3 h later. As seen in Table 4, compounds
considered active in this assay were those that produced
statistically significant reductions in food intake
of525% relative to control animals. With the exception
Table 3. Y5 binding and functional data for 17b–17e
Compd
Avg. Y5 Ki,
nM (n)
Avg. Ca2+ IC50
nM (n)
,
17a
17b
17c
17d
17e
17f
10 (2)
15 (4)
101
242 (1)
Table 2. Y5 binding and functional data for 16a–f
Compd
R1
R2
Avg. Y5 Ki,
nM (n)
Avg. Ca2+ IC50
nM (n)
,
16a
Et
43 (2)
12 (4)
166 (1)
40-
967 (1)
16b
16c
16d
16e
16f
Et
37 (3)
40-
40-
15 (2)
1.5 (2)
8 (2)
Et
41 (4)
362 (3)
351 (3)
46 (3)
i-Pr
i-Pr
Et
38 (4)
40-
16 (4)
40-
30-
2165 (2)

1992
M. Hammond et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1989–1992
Table 4. Food intake effects of analogues in ratsa
structure. Neither of the new series contains the toxic
3-amino-9-ethylcarbazole core, yet each matches the
lead series in affinity for the Y5 receptor. Although
compounds were found to be potent Y5 antagonists,
their in vitro affinity did not translate into consistent in
vivo activity.
Compd
24 h Food-Deprived FI (% )
1j
ꢁ7/ꢁ59*/ꢁ16
ꢁ20/ꢁ20/ꢁ51*
ꢁ14/ꢁ23/ꢁ40*
ꢁ34*/ꢁ40*
16b
16e
17a
17b
17e
+18/+25
ꢁ20/+4/ꢁ7
References and Notes
*Considered active in this assay.
aAll compounds dosed orally at 40 mg/kg in male Sprague–Dawley
rats. Data presented as percent change (ꢁ=reduction; +=increase)
in food intake relative to control animals.
1. (a) Michel, M. C.; Beck-Sickinger, A.; Cox, H.; Doods,
H. N.; Herzog, H.; Larhammar, D.; Quirion, R.; Schwartz, T.;
Westfall, T. Pharm. Rev. 1998, 50, 143. (b) Gehlert, D. R.
Proc. Soc. Exp. Biol. Med. 1998, 218, 7.
2. Stanley, B. G.; Leibowitz, S. F. Life Sci. 1984, 35, 2635.
3. (a) Wyss, P.; Stricker-Krongrad, A.; Brunner, L.; Miller, J.;
Crossthwaite, A.; Whitebread, S.; Criscione, L. Regul. Pept.
1998, 75-76, 363. (b) Gerald, C.; Walker, M. W.; Criscione, L.;
Gustafson, E. L.; Batzl-Hartmann, C.; Smith, K. E.; Waysse,
P.; Durkin, M. M.; Laz, T. M.; Linemeyer, D. L.; Schaffhau-
ser, A. O.; Whitebread, S.; Hofbauer, K. G.; Taber, R. I.;
Branchek, T. A.; Weinshank, R. L. Nature (London) 1996,
382, 168. (c) Cabrele, C.; Langer, M.; Bader, R.; Wieland,
H. A.; Doods, H. N.; Zerbe, O.; Beck-Sickinger, A. G. J. Biol.
Chem. 2000, 275, 36043. (d) Parker, E. M.; Balasubramaniam,
A.; Guzzi, M.; Mullins, D. E.; Salisbury, B. G.; Sheriff, S.;
Witten, M. B.; Hwa, J. J. Peptides 2000, 21, 393. (e) McCrea,
K.; Wisialowski, T.; Cabrele, C.; Church, B.; Beck-Sickinger,
A.; Kraegen, E.; Herzog, H. Regul. Pept. 2000, 87, 47.
4. Hammond, M. IDrugs 2001, 4, 920.
5. (a) Block, M. H.; Boyer, S.; Brailsford, W.; Brittain, D. R.;
Carroll, D.; Chapman, S.; Clarke, D. S.; Donald, C. S.; Foote,
K. M.; Godfrey, L.; Ladner, A.; Marsham, P. R.; Masters,
D. J.; Mee, C. D.; O’Donovan, M. R.; Pease, J. E.; Pickup,
A. G.; Rayner, J. W.; Roberts, A.; Schofield, P.; Suleman, A.;
Turnbull, A. V. J. Med. Chem. 2002, 45, 3509. (b) Rudolf, K.;
Hurnaus, R.; Eberlein, W.; Engel, W.; Wielans, H.-A.; Krist,
B. WO 0296902, Chem. Abstr. 2003, 138, 14006.
6. Elliott, R. L.; Griffith, D. A.; Hammond, M. U. S. Patent 6
399 631, 2002 Chem. Abstr 2002, 137, 6086.
7. Huff, J.; Haseman, J. Environ. Health Perspect. 1991, 96, 23.
8. Harayama, T.; Akiyama, T.; Kawano, K. Chem. Pharm.
Bull. 1996, 44, 1634.
9. Elliott, R. L.; Oliver, R. M.; Hammond, M.; Patterson,
T. A.; She, L.; Hargrove, D. M.; Martin, K. A.; Maurer, T. S.;
Kalvass, J. C.; Morgan, B. P.; DaSilva-Jardine, P. A.; Ste-
venson, R. W.; Mack, C. M.; Cassella, J. V. J. Med. Chem.
2003, 46, 670.
10. (a) Youngman, M. A.; McNally, J. J.; Lovenberg, T. W.;
Reitz, A. B.; Willard, N. M.; Nepomuceno, D. H.; Wilson,
S. J.; Crooke, J. J.; Rosenthal, D.; Vaidya, A. H.; Dax, S. L.
J. Med. Chem. 2000, 43, 213. (b) Kordik, C. P.; Luo, C.;
Zanoni, B. C.; Lovenberg, T. W.; Wilson, S. J.; Vaidya, A. H.;
Crooke, J. J.; Rosenthal, D. I.; Reitz, A. B. Bioorg. Med.
Chem. Lett. 2001, 11, 2287.
of the imidazolylcarbazole 17a, which produced sig-
nificant reductions in food intake both times it was
tested, the other five antagonists, 1j, 16b, 16e, 17b
and 17e, did not have consistent anorectic effects (n
of at least 2). Compound 17b was also tested in an
overnight feeding model wherein the compound was
dosed 2 h before the start of the dark cycle and food
intake was measured 13 h later, and was found to
have no effect on food intake.
To determine whether the variable activity of these
compounds could be due to poor bioavailability, we
examined the pharmacokinetic properties of the carba-
zole 1j and the imidazolylcarbazole 17b in separate
studies. Carbazole 1j was found to have excellent expo-
sure, having a Cmax of 3133 ng/mL and a bioavailability
of >50% when dosed orally at 10 mg/kg. When dosed at
40 mg/kg, 17b was found to have a moderate plasma
exposure of 350 ng/mL and a bioavailability of 13%,
suggesting that poor exposure could account for the lack
of activity in some cases, but not all. Although CSF and/
or brain exposure data is not available for these com-
pounds, we know that in other chemically distinct Y5
antagonist series known to be well-exposed in both the
CNS and periphery we have observed a similar absence
of in vivo activity.9 By contrast, we have been able to
demonstrate that these compounds will block the orexi-
genic effect of a Y5 selective agonist.9
The ability of small molecule Y5 antagonists to block
food intake under ‘normal’ (non agonist-induced) con-
ditions has yet to be demonstrated unambiguously.
Although selective Y5 antagonists with potent anorectic
effects have been reported,10 recently there have been
studies disclosed in which potent, selective Y5 antago-
nists with good central exposure are capable of block-
ing agonist-induced feeding but have no effect on
normal food intake.11 Other Y5 antagonists with good in
vivo anorectic potency have more recently been found to
lack selectivity for the Y5 receptor.12 While it is clear that
NPY and its peptide agonists produce robust orexigenic
effects in rodents, blockade of the Y5 receptor does not
appear to unequivocally produce anorectic effects.
11. (a) Kanatani, A.; Ishihara, A.; Iwaasa, H.; Nakamura, K.;
Okamoto, O.; Hidaka, M.; Ito, J.; Fukuroda, T.; MacNeil,
D. J.; Van der Ploeg, L. H. T.; Ishii, Y.; Okabe, T.; Fukami,
T.; Ihara, M. Biochem. Biophys. Res. Commun. 2000, 272, 169.
(b) Turnbull, A. V.; Ellershaw, L.; Masters, D. J.; Birtles, S.;
Boyer, S.; Carroll, D.; Clarkson, P.; Loxham, S. J. G.;
McAulay, P. Diabetes 2002, 51, 2441.
12. Della Zuana, O.; Sadlo, M.; Germain, M.; Feletou, M.;
Chamorro, S.; Tisserand, F.; de Montrion, C.; Boivin, J. F.;
Duhault, J.; Boutin, J. A.; Levens, N. Int. J. Obes. 2001, 25, 84.
In summary, two series of potent Y5 antagonists
have been developed from an aminocarbazole lead