Discovery of substituted biphenyl imidazoles as potent, bioavailable bombesin receptor subtype-3 agonists

Article abstract of DOI:10.1016/j.bmcl.2010.01.154

We report SAR studies on a novel non-peptidic bombesin receptor subtype-3 (BRS-3) agonist lead series derived from high-throughput screening hit RY-337. This effort led to the discovery of compound 22e with significantly improved potency at both rodent and human BRS-3.

Full text of DOI:10.1016/j.bmcl.2010.01.154

Bioorganic & Medicinal Chemistry Letters 20 (2010) 1913–1917
Contents lists available at ScienceDirect
Bioorganic & Medicinal Chemistry Letters
journal homepage: www.elsevier.com/locate/bmcl
Discovery of substituted biphenyl imidazoles as potent, bioavailable
bombesin receptor subtype-3 agonists
a
a
a
a
a
Shuwen He a, , Peter H. Dobbelaar , Jian Liu , Tianying Jian , Iyassu K. Sebhat , Linus S. Lin ,
Allan Goodman b, Cheng Guo b, Peter R. Guzzo b, Mark Hadden b, Alan J. Henderson b, Megan Ruenz b,
Bruce J. Sargent b, Larry Yet b, Theresa M. Kelly c, Oksana Palyha c, Yanqing Kan c, Jie Pan c, Howard Chen c,
Donald J. Marsh c, Lauren P. Shearman d, Alison M. Strack d, Joseph M. Metzger d, Scott D. Feighner c,
Carina Tan c, Andrew D. Howard c, Constantin Tamvakopoulos a, Qianping Peng a, Xiao-Ming Guan c,
Marc L. Reitman c, Arthur A. Patchett a, Matthew J. Wyvratt Jr. a, Ravi P. Nargund a
*
a Department of Medicinal Chemistry, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
b Albany Molecular Research Institute, 26 Corporate Circle, Albany, NY 12203, USA
c Department of Obesity Research, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
d Department of Pharmacology, Merck Research Laboratories, PO Box 2000, Rahway, NJ 07065, USA
a r t i c l e i n f o
a b s t r a c t
Article history:
We report SAR studies on a novel non-peptidic bombesin receptor subtype-3 (BRS-3) agonist lead series
derived from high-throughput screening hit RY-337. This effort led to the discovery of compound 22e
with significantly improved potency at both rodent and human BRS-3.
Received 5 January 2010
Revised 21 January 2010
Accepted 27 January 2010
Available online 4 February 2010
Ó 2010 Elsevier Ltd. All rights reserved.
Keywords:
Bombesin receptor subtype-3
BRS3
Agonist
Biphenyl imidazole
Obesity has become a major global health issue causing the
World Health Organization (WHO) to officially declare obesity a
disease.1 Obesity causes or exacerbates many health problems,
such as hypertension, type 2 diabetes mellitus, and cardiovascular
disease. Current drugs approved for the chronic treatment of obes-
ity have suboptimal tolerability and limited efficacy.2 In order to
address this unmet medical need, we are interested in developing
therapies based on novel mechanisms.
Bombesin receptor subtype-3 (BRS-3 or BB3), a G-protein cou-
pled receptor (GPCR), belongs to the bombesin receptor family.3
Bombesin is a peptide (pGlu-Gln-Arg-Leu-Gly-Asn-Gln-Trp-Ala-
Val-Gly-His-Leu-Met-NH2), originally isolated from the skin of
the European frog Bombina bombina. Two mammalian bombesin-
related peptides have been identified, gastrin-releasing peptide
(GRP) and neuromedin B (NMB). The biological effects of these
peptides are mediated by the bombesin family receptors (GRPR
or BB1 and NMBR or BB2). BRS-3 is primarily expressed in the cen-
tral nervous system, particularly the hypothalamus.4,5 The natural
ligand for the BRS-3 is unknown and, despite its name, BRS-3 does
not bind bombesin with high affinity.
Validation of BRS-3 as a potential target for the treatment of
obesity comes from rodent genetics and pharmacology. Mice lack-
ing BRS-3 develop metabolic defects and obesity.6,7 They are
hyperphagic with reduced metabolic rate and reduced core tem-
perature. Complementing the BRS-3 knock-out mouse data, Merck
scientists demonstrated BRS-3 antagonist Bantag-1, when admin-
istered intracerebroventricularly, increases food intake and body
weight in rats.8 Peptides with functional activity at human BRS-3
have been reported.9 More recently, small molecule agonists based
on an omeprazole lead also appeared.10 In this paper, we will de-
scribe Merck’s early efforts at identifying small molecule BRS-3
agonists.
A high-throughput screening campaign of the Merck sample
collection identified racemic RY-337. The compound was particu-
larly appealing due to its low molecular weight and non-peptidic
structure. In vitro binding and functional assay data for RY-337
are summarized in Table 1.11 RY-337 is completely inactive on hu-
man BB1 and BB2 receptors and is 10-fold more potent on rodent
BRS-3 compared to the human receptor.
* Corresponding author. Tel.: +1 732 594 0881; fax: +1 732 594 3007.
E-mail address: shuwen_he@merck.com (S. He).
0960-894X/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved.
doi:10.1016/j.bmcl.2010.01.154

1914
S. He et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1913–1917
Table 1
Guided by functional activity assays on human and mouse
receptors, we embarked on an SAR study to improve the human
potency of this structural series.
Binding affinity and functional activitya of RY-337 at human, rat and mouse bombesin
family receptors
The SAR study began with the branch side chain. The com-
pounds and data are listed in Table 2. The synthesis of the analogs
was similar to Scheme 1, except 12f, which required a slightly dif-
ferent synthetic route for its key aldehyde intermediate 17
(Scheme 2). From Table 2, it was clear that the length of the side
chain had minimal effect on activity at human BRS-3. The bulky
iso-propyl group drastically reduced the potency at the mouse
receptor. The discovery that removal of the side chain (12f) main-
tained good potency and that the compound had a reasonable PK
profile in rats (Table 3) led to the decision to continue the SAR
study without the branch side chain.
HN
N
MW=406.53
COOH
OH
RY-337
Receptor
Bindinga IC50
(nM)
Functional EC50
(nM)
Activationc at
10 M (%)
l
We next examined effects of the linkage between the imidazole
and biphenyl moiety (Table 4). Shortening the linkage to methy-
lene (18a) completely abolished activity at both human and mouse
receptors. Extension to a three carbon linkage (18b) had a smaller
effect on potency. Finally, rigidifying with a trans C–C double bond
(18c) significantly reduced the potency. Therefore two carbon link-
age appeared to be optimal.
We next turned our attention to replacement of the n-butyl
chain on the imidazole ring (Table 5). While there was little
improvement with a phenyl substituent (19a), we were excited
to find that a benzyl group (19b) tremendously improved potency
at the mouse receptor. More importantly, the change also im-
proved potency at the human receptor. Extending the length of
the linker (19c) reduced the potency.
Encouraged by these results, we synthesized a series of com-
pounds with different substitutions on the phenyl ring (Table 6,
20a–p). Introduction of an electron-donating methoxy group re-
duced the potency. The same is largely true for the electron-with-
drawing nitrile, although meta-CN substitution maintained most of
the activity. Among analogs bearing a Me group at different posi-
tions, ortho-Me substitution (20g) maintains similar potency to
the un-substituted analog (19b), while other positional isomers
diminished activity at the human receptor. Chloro substitution re-
hBB1
hBB2
hBRS-3
rBRS-3
mBRS-3
>10,000
>10,000
3676
ND
NDb
ND
1897
174.3
237.5
ND
ND
74
101
98
ND
a
b
c
Data are averages of at least three repeated measurements.
Not determined.
The percent activation is the maxim activation of tested compound relative to
that of dY-peptide.
Synthesis of RY-337 is outlined in Scheme 1. The key step is the
formation of imidazole 11 from the reaction of aldehyde 10 and an
a
-hydroxylketone 6 in the presence of copper (II) acetate and
ammonium acetate in acetic acid12 with concomitant loss of the
TBS protecting group.
a-Hydroxylketone 6 was conveniently pre-
pared from commercially available diol 5.13a Alkylation of the
imine derived from aldehyde 9 with benzylic bromide 4 followed
by mild acidic hydrolysis produced aldehyde 10.14
a-Hydroxylke-
tones can conveniently be prepared in one step from a variety of
commercially available starting materials, such as alkenes, 1,2-
diols, carboxylic acids,
gent strategy enabled rapid synthesis of new analogs.
a
-halo-ketones, and alkynes.13 This conver-
CH2OH
CH2Br
I
I
a, b
c
d
CO2H
CO2
t-Bu
CO2
t-Bu
CO2t-Bu
1
2
3
4
e
OH
OH
OH
O
5
6
f
g, h
O
HO
OH
HO
OTBS
OTBS
7
8
9
HN
CHO
l
m
N
i, j, k
RY-337
CO2t-Bu
OTBS
OH
CO2t-Bu
10
11
Scheme 1. Reagents and conditions: (a) (COCl)2, DMF (cat.), CH2Cl2,, 0 °C; (b) t-BuOH, pyridine, 0 °C to rt; (c) 4-(hydroxylmethyl) phenyl-boronic acid, Pd(PPh3)4, Na2CO3,
DME–H2O, reflux; (d) NBS, PPh3, CH2Cl2, À78 °C to rt; (e) NaBrO3, NaHSO3, CH3CN–H2O; (f) NaH, TBSCl, THF; (g) DMSO, (COCl)2 À78 °C; (h) Et3N, À78 °C to rt; (i)
cyclohexylamine, C6H6; (j) LDA, 4, THF; (k) 0.5 N HCl (aq); (l) Cu(OAc)2, 6, NH4OAc, HOAc, reflux 0.5 h; (m) TFA.

S. He et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1913–1917
1915
Table 2
Activity of 12 at human and mouse BRS-3a
HN
N
R
12
COOH
R
Human functional EC50 (nM) (%activation)
Mouse functional EC50 (nM) (%activation)
12a
12b
12c
12d
12e
12f
n-Butyl
n-Propyl
Ethyl
Methyl
iso-Propyl
H
2656 (66%)
1183 (96%)
2722 (76%)
2760 (79%)
4618 (60%)
2707 (78%)
527 (96%)
165 (102%)
340 (104%)
184 (110%)
2240 (87%)
93 (116%)
a
Data are averages of at least three repeated measurements.
CO2H
a
b
d
OH
OH
O
Br
Br
B
O
13
14
15
OH
O
I
c
CO2t-Bu
CO2t-Bu
CO2t-Bu
2
16
17
Scheme 2. Reagents and conditions: (a) LAH, ether, 0 °C to rt; (b) bis(pinacolate)diboron, Pd(dppf), KOAc, DMSO; (c) 15, Pd(dppf), Na2CO3 (2 N),DMF, 85 °C; (d) TEMPO (cat.),
trichloroisocyanuric acid, CH2Cl2, 0 °C to rt.
Table 5
Activity of 19 at human and mouse BRS-3a
Table 3
Pharmacokinetic data for 12fa
R
PK parameter
Rata
HN
F (%)
16
1.3
0.1
Cl (mL minÀ1 kgÀ1
)
N
Vdss (L kgÀ1
)
t1/2 (h)
AUCn (lM h/mpk)
1.56
6.48
19
COOH
a
Compound dosed in Sprague–Dawley rats as a solution in EtOH/PEG400/water
R
Human functional EC50
(nM) (%activation)
Mouse functional EC50
(nM) (%activation)
(10:40:50) at 1 mg/kg, iv and 4 mg/kg, po.
19a
19b
19c
Ph
4949 (30%)
205 (109%)
1323 (86%)
872 (101%)
11 (111%)
174 (89%)
–CH2Ph
–CH2CH2Ph
Table 4
Activity of 18 at human and mouse BRS-3a
a
Data are averages of at least three repeated measurements.
H
N
L
N
sults in a loss of potency. Finally, compounds with mono-F substi-
tution maintain potency at both ortho (20m) and meta (20n) posi-
tions. The ortho, meta di-F substituted benzyl analog (20p)
appeared optimal in this limited series.
18
COOH
L
Human functional EC50
(nM) (%activation)
Mouse functional EC50
(nM) (%activation)
In order to further improve the potency at the human receptor,
we continued the SAR study with aliphatic substituents (Table 7).
Since the benzyl group gave better potency than the n-butyl substi-
tuent, we decided to synthesize analogs with –CH2-cycloalkyl sub-
stitution. A small cyclopropyl ring (21a) gave poor activity while
larger sized rings (21b–e) offered single digit nanomolar potency
at the mouse receptor and significantly improved human potency.
18a CH2
18b (CH2)3
18c CH@CH
(trans)
>10,000 (0%)
5598 (54%)
4376 (8%)
>10,000 (0%)
382 (101%)
10,000 (39%)
a
Data are averages of at least three repeated measurements.

1916
S. He et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1913–1917
Table 6
Comparison of activity of 20 with 19b on human and mouse BRS-3a
X
HN
N
20
COOH
X
Human functional EC50 (nM) (%activation)
Mouse functional EC50 (nM) (%activation)
19b
20a
20b
20c
20d
20e
20f
20g
20h
20i
20j
20k
20l
20m
20n
20o
20p
H
205 (109%)
1724 (88%)
4124 (73%)
5360 (44%)
5479 (51%)
812 (95%)
5887 (25%)
127 (93%)
4307 (87%)
2663 (88%)
292 (100%)
825 (91%)
2288 (89%)
177 (108%)
159 (107%)
1601 (84%)
87 (104%)
11 (111%)
163 (101%)
79 (80%)
513 (88%)
642 (139%)
20 (102%)
723 (85%)
9.3 (106%)
75 (89%)
162 (120%)
22 (109%)
31 (98%)
102 (105%)
9.5 (121%)
7.9 (108%)
50 (104%)
8 (110%)
o-OMe
m-OMe
p-OMe
o-CN
m-CN
p-CN
o-Me
m-Me
p-Me
o-Cl
m-Cl
p-Cl
o-F
m-F
p-F
o, m-di-F
a
Data are averages of at least three repeated measurements.
Table 7
Activity of 21 at human and mouse BRS-3a
R
HN
N
O
21h : R=
21
COOH
R
Human functional EC50 (nM) (%activation)
Mouse functional EC50 (nM) (%activation)
21a
21b
21c
21d
21e
21f
–CH2-cyclopropyl
–CH2-cyclobutyl
–CH2-cyclopentyl
–CH2-cyclohexyl
-CH2-cycloheptyl
–CH2–CH2-cyclohexyl
–Cyclohexyl
1004 (96%)
588 (105%)
133 (102%)
41 (104%)
88 (106%)
642 (94%)
375 (102%)
338 (96%)
72 (144%)
7.1 (99%)
7.9 (94%)
7.3 (102%)
5.7 (102%)
33 (101%)
14.7 (116%)
23 (106%)
21g
21h
See Scheme above
a
Data are averages of at least three repeated measurements.
In particular, compound 21d with –CH2-cyclohexyl substitution
maintained 41 nM EC50 at hBRS-3. We also tried changing the
length of the linkage between the imidazole ring and the cyclo-
hexyl ring. Both extending and shortening (21f–g) the linkage re-
duced potency. Introduction of an oxygen in the cyclohexyl ring
(21h) did not offer much advantage in terms of potency.
potency and compound 22e stood out as the most potent
compound at human BRS-3 in this series. Further alterations
(e.g., 22f) reduced potency. As outlined in Table 9, 22c demon-
strated good rat PK properties.
We tested several potent compounds in animal models to
demonstrate efficacy on food intake. When dosed intracerebro-
Finally we studied uncyclized alkyl substituents (Table 8).
Simple addition of a Me group on the n-butyl chain (22a) signif-
icantly improved potency on human and mouse receptors. Add-
ing a second Me group further increased the potency (22b).
Truncating one methylene (22c) had no effect on activity at
mouse BRS-3 but resulted in reduced activity at the human
receptor. Further shortening of the linkage (22d) resulted in a
large loss of potency. Further homologation of 22c recovered
ventricularly in rat at 25 lg, compound 22c reduced food intake
by 29% compared to animals treated with vehicle (20% PG-
water). However, none of the compounds exhibited efficacy in
diet induced obese (DIO) mice when dosed orally. Although
potent with good plasma drug levels after oral dosing, further
investigation suggested that the compounds do not reach suffi-
cient brain levels to interact with the target BRS-3 receptors in
the hypothalamus.

S. He et al. / Bioorg. Med. Chem. Lett. 20 (2010) 1913–1917
1917
Table 8
Comparison of activity of 22 with 12f on human and mouse BRS-3a
R
R
HN
N
12f :
22
22d :
22a :
COOH
Et
Et
22b :
22c :
22e:
23f :
Et
Human functional EC50 (nM) (%activation)
Mouse functional EC50 (nM) (%activation)
12f
22a
22b
22c
22d
22e
22f
2707 (78%)
119 (112%)
39 (102%)
106 (103%)
4334 (45%)
25 (101%)
76 (97%)
93 (117%)
10 (111%)
6.2 (99%)
6.3 (105%)
2745 (108%)
9.6 (94%)
30 (116%)
a
Data are averages of at least three repeated measurements.
8, 461; (d) Boyle, R. G.; Humphries, J.; Mitchell, T.; Showell, G. A.; Apaya, R.;
Iijima, H.; Shimada, H.; Arai, T.; Ueno, H.; Usui, Y.; Sakaki, T.; Wada, E.; Wada, K.
J. Pept. Sci. 2005, 11, 136.
Table 9
Pharmacokinetic data for 22c
PK parameter
Rata
10. Carlton, D. L.; Collin-Smith, L. J.; Daniels, A. J.; Deaton, D. N.; Goetz, A. S.;
Laudeman, C. P.; Littleton, T. R.; Musso, D. L.; Morgan, R. J. O.; Szewczyk, J. R.;
Zhang, C. Bioorg. Med. Chem. Lett. 2008, 18, 5451.
F (%)
19
Cl (mL minÀ1 kgÀ1
)
1.02
0.128
1.54
9.41
11. (a) For human BRS-3 binding assays, 1–4
from NFAT-CHO cells expressing the receptor were incubated with 0.3 pM
125I]-[D-Tyr6,b-Ala11,Phe13,Nle14]-Bombesin (6–14) 125I-dY-peptide) and
various concentrations of test compounds in 200 of binding buffer
lg of membrane protein obtained
Vdss (L kgÀ1
)
t1/2 (h)
AUCn (lM h/mpk)
[
(
lL
(50 mM Tris, pH 7.2, 5 mM MgCl2, 0.1% BSA). After a 2 h incubation at room
temperature, the binding reaction was terminated by filtering through a GF/c
filter and washing the filter with PBS using a Packard 96-well Harvester. The
amount of radioligand bound to the receptor was determined by measurement
of the radioactivities on the filter through liquid scintillation counting. The
nonspecific binding was defined as the binding in the presence of 100 nM
unlabeled dY-bombesin. The data in % inhibition of binding was plotted versus
the log molar concentration of receptor ligand (compound). The IC50 was
reported as the inflection point of the resulting sigmoidal curve.
a
Compound dosed in Sprague–Dawley rats as a solution in EtOH/PEG400/water
(10:40:50) at 1 mg/kg, iv and 4 mg/kg, po.
In summary, we report the SAR of a non-peptidic BRS3 agonist
lead series. This work culminated in compounds with much im-
proved potency on both rodent and human receptors (e.g., 22e)
and with good rodent PK profiles. Efforts at addressing the low
brain penetration will be reported in the subsequent paper.
(b) The functional assay is an aequorin bioluminescence assay. It was
performed in 96-well format using
a Wallac Microbeta luminometer
equipped with microinjector module. Compounds in DMSO (0.5% final
concentration) were titrated in the plates at 2Â concentration in a volume of
0.1 mL ECB buffer (20 mM HEPES, pH 7.4, 140 mM NaCl, 20 mM KCl, 1 mM
MgCl2, 1 mM CaCl, 5 mM glucose, 0.1 mg/ml BSA). The HEK293AEQ cells from
lines expressing either human, rat or mouse BRS-3 (20,000 per well) were
charged with coelenterazine (Molecular Probes) and then injected in 0.1 mL
ECB buffer into the compound containing wells. The bioluminescence was
monitored for 30 s, or alternatively, total bioluminescence was determined
over 10 min. The bioluminescent readings were plotted versus the log molar
concentration of receptor ligands (compounds). The EC50 for activation was
reported as the inflection point of the resulting sigmoidal curve. The percent
activation is the maxim activation of tested compound relative to that of dY-
peptide.
References and notes
1. Obesity: Preventing and Managing the Global Epidemic, Report of a WHO
Consultation, World Health Organization, Geneva, 2004.
2. Bays, H. E. Obes. Res. 2004, 12, 1197.
3. For a general review: Jensen, R. T.; Battey, J. F.; Spindel, E. R.; Benya, R. V.
Pharmacol. Rev. 2008, 60, 1.
4. Ohki-Hamazaki, H.; Wada, E.; Matsui, K.; Wada, K. Brain Res. 1997, 762, 165.
5. Liu, J.; Lao, Z. J.; Zhang, J.; Schaeffer, M. T.; Jiang, M. M.; Guan, X. M.; Van der
Ploeg, L. H.; Fong, T. M. Biochemistry 2002, 41, 8954.
6. Ohki-Hamazaki, H.; Watase, K.; Yamamoto, K.; Ogura, H.; Yamano, M.; Yamada,
K.; Maeno, H.; Imaki, J.; Kikuyama, S.; Wada, E.; Wada, K. Nature 1997, 390, 165.
7. Ladenheim, E. E.; Hamilton, N. L.; Behles, R. R.; Bi, S.; Hampton, L. L.; Battey, J.
F.; Moran, T. H. Endocrinology 2008, 149, 971.
8. Guan, X.-M.; Chen, H.; Dobbelaar, P. H.; Dong, Y.; Fong, T. M.; Gagen, K.; Gorski,
J.; He, S.; Howard, A. D.; Jian, T.; Jiang, M.; Kan, Y.; Kelly, T. M.; Kosinski, J.; Lin,
L. S.; Liu, J.; Marsh, D. J.; Metzger, J. M.; Miller, R.; Nargund, R. P.; Palyha, O.;
Shearman, L.; Shen, Z.; Stearns, R.; Strack, A. M.; Stribling, S.; Tang, Y. S.; Wang,
S.-P.; White, A.; Yu, H.; Reitman, M. L. Cell Metab. 2010, 11, 101.
9. (a) Mantey, S. A.; Weber, H. C.; Sainz, E.; Akeson, M.; Ryan, R. R.; Pradhan, T. K.;
Searles, R. P.; Spindel, E. R.; Battey, J. F.; Coy, D. H.; Jensen, R. T. J. Biol. Chem.
1997, 272, 26062; (b) Mantey, S. A.; Coy, D. H.; Pradhan, T. K.; Igarashi, H.; Rizo,
I. M.; Shen, L.; Hou, W.; Hocart, S. J.; Jensen, R. T. J. Biol. Chem. 2001, 276, 9219;
(c) Weber, D.; Berger, C.; Heinrich, T.; Eickelmann, P.; Antel, J. J. Pept. Sci. 2002,
(c) The binding protocols for human BB1 & BB2 are the same as for BRS-3
except that less protein (membrane) is needed for these two receptors. Both
used 0.5 lg per well instead of the 2 lg typically used for BRS-3.
12. Liverton, N. J.; Butcher, J. W.; Claiborne, C. F.; Claremon, D. A.; Libby, B. E.;
Nguyen, K. T.; Pitzenberger, S. M.; Selnick, H. G.; Smith, G. R.; Tebben, A.; Vacca,
J. P.; Varga, S. L.; Agarwal, A.; Dancheck, K.; Forsyth, A. J.; Fletcher, D. S.; Frantz,
B.; Hanlon, W. A.; Harper, C. F.; Hofsess, S. J.; Kostura, M.; Jiunn Lin, J.; Luell, S.;
O’Neill, E. A.; Orevillo, C. J.; Pang, M.; Parsons, J.; Rolando, A.; Sahly, Y.; Visco, D.
M.; O’Keefe, S. J. J. Med. Chem. 1999, 42, 2180.
13. (a) Sakaguchi, S.; Kikuchi, D.; Ishii, Y. Bull. Chem. Soc. Jpn. 1997, 70, 2561; (b)
Srinivasan, N. S.; Lee, D. G. Synthesis 1979, 520; (c) Wissner, A. J. Org. Chem. 1979,
44, 4617; (d) Zhu, G.; Casalnuovo, A. L.; Zhang, X. J. Org. Chem. 1998, 63, 8100.
14. Stevens, C.; De Kimpe, N. Synlett 1991, 351.