Synthesis and structure-Activity relationships of M2-Selective muscarinic receptor ligands in the 1-[4-(4-Arylsulfonyl)-phenylmethyl]-4-(4-piperidinyl)-piperazine family

Article abstract of DOI:10.1016/S0960-894X(02)00024-0

The synthesis and muscarinic binding properties of compounds based on the 1-[4-(4-arylsulfonyl)phenylmethyl]-4-(1-aroyl-4-piperidinyl)-piperazine skeleton are described. For compounds, substituted with appropriately configured methyl groups at the benzylic center and at the piperazine 2-position, high levels of selective, M₂ subtype affinity could be obtained, particularly when the terminal N-aroyl residue was ortho-substituted.

Full text of DOI:10.1016/S0960-894X(02)00024-0

Bioorganic & Medicinal Chemistry Letters 12 (2002) 795–798
Synthesis and Structure–Activity Relationships of M2-Selective
Muscarinic Receptor Ligands in the 1-[4-(4-Arylsulfonyl)-
phenylmethyl]-4-(4-piperidinyl)-piperazine Family
Stuart W. McCombie,a,* Sue-Ing Lin,a Jayaram R. Tagat,a Dennis Nazareno,a
Susan Vice,a Jennifer Ford,a Theodros Asberom,a Daria Leone,a
Joseph A. Kozlowski,a Guowei Zhou,a Vilma B. Ruperto,b
Ruth A. Duffyb and Jean E. Lachowiczb
aDepartment of Chemistry, Schering-Plough Research Institute, 2015 Galloping Hill Road,
Kenilworth, NJ 07033, USA
bDepartment of CNS Pharmacology, Schering-Plough Research Institute, 2015 Galloping Hill Road,
Kenilworth, NJ 07033, USA
Received 19 October 2001; accepted 21 December 2001
Abstract—The synthesis and muscarinic binding properties ofcompounds based on the 1-[4-(4-arylsulofnyl)phenylmethyl]-4-(1-
aroyl-4-piperidinyl)-piperazine skeleton are described. For compounds, substituted with appropriately configured methyl groups at
the benzylic center and at the piperazine 2-position, high levels ofselective, M 2 subtype affinity could be obtained, particularly when
the terminal N-aroyl residue was ortho-substituted. # 2002 Elsevier Science Ltd. All rights reserved.
Alzheimer’s Disease (AD), characterized by a pro-
gressive loss ofcognitive uf nction, is a serious medical
problem. Since AD is largely a disease ofthe elderly,
the percentage ofaffected persons is expected to grow,
as overall life expectancy continues to increase. Cur-
rently, therapeutic intervention in AD involves enhan-
cing levels ofthe key neurotransmitter acetylcholine
(ACh)1 by treatment with inhibitors ofacetyl-
cholinesterase, which retard the breakdown ofACh in
the synapse. An alternative way to elevate ACh levels is
via the cholinergic receptor system. Our approach has
enhanced selectivity over the postsynaptic M1 receptor
is obtained when R2 and R3 are (S)-Me and (R)-Me,
respectively. Initial modification ofthe aromatic ring
substitution pattern and the piperidine N-substituent led
to 2, which exhibited acceptable selectivity [Ki(M1)/
Ki(M2)=109] and also elevated brain acetylcholine
levels when administered orally to rats.
been to pursue antagonists ofthe presynaptic M
2
receptor.2,3 Selective binding to M2 relative to the other
muscarinic receptor subtypes is an important goal: the
postsynaptic M1 receptor is believed to mediate the effects
ofACh in learning and memory, and the peripheral M
receptor is involved in the regulation ofgastro-intestinal
functions.
3
In this paper, we describe chemical and biological
results which illustrate the optimization ofthe affinity
for the M2 receptor over the M1 and M3–M5 subtypes,
for the diaryl sulfone series. The effects of changes in the
In the accompanying paper,4 we described novel, M2-
selective ligands ofgeneral structure 1, and showed that
aryl ring ofthe ArSO moiety, the methylation pattern
2
on the piperazine ring, and the 1-substituent on the
piperidine ring are described. In the latter position, we
*Corresponding author. Tel.:+1-908-740-3455; fax:+1-908-740-7305;
e-mail: stuart.mccombie@spcorp.com
0960-894X/02/$ - see front matter # 2002 Elsevier Science Ltd. All rights reserved.
PII: S0960-894X(02)00024-0

796
S. McCombie et al. / Bioorg. Med. Chem. Lett. 12 (2002) 795–798
Scheme 2. (a) 1.1 equiv MeLi, then 1.2 equiv n-BuLi, THF-hexanes,
À70 ꢀC then 1.2 equiv 3,4-(methylenedioxy)-benzenesulfonyl fluoride,
À70 to À20 ꢀC; (b) steps c–f, Scheme 1; (c) di-tert-butyl dicarbonate,
Et3N, THF, rt, 5 h; (d) MeMgBr, THF, 0 ꢀC, 15 min; (e) CCl3CO2H,
ꢀ
.
NaBH3CN, CH2Cl2, rt, 2 h; (g) NaBH4, BF3 Et2O, DME, 80 C, 6 h.
Scheme 1. (a) 1,3-dibromo-5,5-dimethylhydantoin, MeSO3H, CH2Cl2,
rt, 48 h; or (b) PhI(OCOCF3)2, I2, CH2Cl2, rt, 48 h; (c) KOH, MeOH,
rt, 2 h; (d) (S)-MeCH(OSO2CF3)CO2Et, CH2Cl2, K2CO3–H2O, rt, 6 h;
(e) ClCH2COCl, 1,2-C2H4Cl2, 90 ꢀC, 4 h; (ꢀf) NH3, NaI, DMSO–H2O,
.
rt, 24 h; (g) NaBH4, BF3 Et2O, DME, 80 C, 6 h; (h) 1-(tert-butoxy-
carbonyl)-4-piperidone, NaBH(OAc)3, CH2Cl2, rt, 24 h; (i) n-BuLi,
THF-hexanes, À70 ꢀC, then ArSO2F, À70 to À20 ꢀC; (j) TFA, 1 h;
(k) RCOCl, CH2Cl2, K2CO3–H2O, rt, 2 h or RCO2H, EDCI, HOBt,
i-Pr2NEt, DMF, rt, 24 h.
show that certain ortho-substituted aromatic amides
confer exceptional selectivity for the M2 receptor.
Scheme 3. (a) KOH, MeOH, rt, 2 h; (b) (S)-MeCH(OSO2CF3)CO2Et,
CH2Cl2, K2CO3–H2O, rt, 6 h; (c) rac-MeCH(Cl)COCl, 1,2-C2H4Cl2,
90 ꢀC, 12 h; (d) NaN3, DMSO, 80 ꢀC, 6 h; (e) H2, Pd/C, EtOH, 50 psi,
Our initial objective was to develop a versatile synthesis
from available ‘chiral pool’ substances, such that inter-
mediates along the route could be diverted to generate a
variety ofsubstitution patterns. We chose diketopiper-
azines as the key intermediates, with late-stage intro-
duction ofthe arylsulofnyl substituent, as shown in
Scheme 1.
ꢀ
.
rt, 2 h; (f) NaBH4, BF3 Et2O, DME, 80 C, 6 h.
Alternatively, the same organolithium species could be
reacted with the appropriate aryl disulfide, and the
resulting product oxidized (HOAc, H2O2 or NaBO3) to
the sulfone. The same sulfides could also be obtained
from the iodocompound 7b and ArSH/K2CO3/CuI.
Direct reactions of 7b with ArSO2Na/CuI, a satisfac-
tory process in simple systems,8 gave low yields ofthe
sulfones.
Bromination or iodination of N-trifluoroacetyl-(S)-1-
phenylethylamine, crystallization to obtain the domi-
nant5 para isomers and subsequent alkaline hydrolysis
gave 4a and 4b, respectively. Sequential N-alkylation
with the triflate derived from ethyl (S)-lactate, N-chloro-
acetylation6 and final treatment with ammonia gave
the crystalline diketopiperazines 5a and 5b, which were
converted by reduction followed by reductive amination
to 7a and 7b, the first intermediates in the sequence
which needed chromatographic purification. The sul-
fones were prepared by Br(I)–Li exchange followed by
reaction with the appropriate arylsulfonyl fluoride.7
Yields were high except when the aryl residue carried an
oxy-substituent, meta to the fluorosulfonyl group. In
those cases, substantial amounts ofthe des-bromo
compound (X=H in structure 7a) were formed, and the
sulfonyl fluoride was decomposed. We ascribe this to
competing H-abstraction by the aryllithium species of
the proton between the two electron withdrawing
groups (–OR and –SO2F). In these cases, yields were
improved by adding 1.1 equivalents ofanhydrous mag-
nesium bromide to the intermediate aryl lithium, with
warming to 0 ꢀC before adding the sulfonyl fluoride. Fol-
lowing diaryl sulfone isolation, the sequence was com-
pleted by N-deprotection and amide formation under
standard conditions, to afford the target compounds 8.
Diketopiperazines were also the key to obtaining differ-
ent piperazine ring methylation patterns. The route to
the cis-2,3-dimethyl compound is shown in Scheme 2.
Activation ofthe lactam carbonyl permitted the highly
selective addition ofmethyl magnesium bromide to give
a
stable, N-tert-butoxycarbonyl hemiaminal. Mild
reduction, N-deprotection, and final reduction ofthe
remaining carbonyl group with borane afforded 10, f or
further elaboration. Only the cis isomer was produced
in the acyliminium ion reduction step, reflecting attack
ofthe hydride donor anti to the 2-methyl group.
To access 2,5-dimethylpiperazines (Scheme 3), the
chloroacetyl chloride in Scheme 2 was replaced with
racemic 2-chloropropionoyl chloride. Reaction ofthe
resulting amide 11 with ammonia was unsatisfactory,
but displacement with azide followed by catalytic
reduction gave the separable, diastereoisomeric 2,5-di-
substituted diketopiperazines. Final reduction with
borane afforded the piperazines 12 and 13, for further
elaboration.

S. McCombie et al. / Bioorg. Med. Chem. Lett. 12 (2002) 795–798
797
Using cloned human M1–M5 receptors, Ki values for
receptor binding were determined as previously descri-
bed.9 Briefly, membranes were incubated with [3H]-N-
methyl scopolamine and test compounds for 1 h at
room temperature, followed by rapid filtration and
scintillation counting to measure bound radioactivity.
Values shown are averages ooffur determinations;
inter-replicate variability was below 10%.
we then investigated diverse substitution patterns on the
aromatic ring ofthe arylsulfonyl group. Representative
compounds are shown in Table 2.
In general, we were not able to significantly improve
upon the affinity–selectivity profile exhibited by the best
compounds shown in Table 1, but some interesting
trends emerged. Comparison ofentries 19 and 30
shows that for o-toluamides, like sulfonamides, the
3,4-(methylenedioxy)-phenylsulfonyl compounds have
greater affinity than their p-methoxy counterparts.
2,4-Dimethoxy substitution (entry 31) also conferred
acceptable selectivity, but 2,3,4-trisubstitution was tol-
erated only when the substituents adjacent to the sulfo-
nyl group were tied back to form a bicyclic system
(entry 33 vs 32 and 35). Although earlier results3
had suggested that strongly electron-donating sub-
stituents were important when the piperidine carries
an N-sulfonyl substituent, this was not true for the
corresponding o-toluamides, where a m-methyl or
The results shown in Table 1, selected from a large
panel ofamides, in which the 3,4-(methylenedioxy)-
phenylsulfonyl moiety, the methylation pattern and the
absolute stereochemistry were kept constant, highlight
the structure–activity relationships.
Although the M2 affinity was not unduly sensitive to the
nature ofthe N-acyl group, the receptor subtype selec-
tivity was dramatically affected. Simple, aliphatic
amides (typified by 14) were not selective. In the aro-
matic amides, the substitution pattern proved to be the
key: a variety of ortho substituents conferred a high
degree ofselectivity ofr M
versus the M1 and M3
2
receptors, and acceptable selectivity versus M5 (entries
18–21). In general, relatively non polar substituents
were preferred, although the 2-hydroxy benzamide (23)
and particularly the 2-amino benzamide (21) were also
M2-selective. Compounds 24 and 25 show that large
substituents were tolerated with respect to binding affi-
nity, but did not affect selectivity in a consistent fashion.
For more highly substituted amides, 2,4- and 2,6-di-
substitution (17 and 27, respectively) were not helpful,
but the 2,3-(methylenedioxy)-benzamide 26 had excel-
lent selectivity. We also evaluated heterocyclic amides,
with mixed results. The thiophene analogue 28 ofthe
o-chlorobenzamide demonstrated high affinity and
selectivity, whereas the pyridine analogue 29 ofthe
o-toluamide was nonselective, and a relatively weak binder.
Table 2. Effects ofthe arylsulofnyl group on receptor binding and
selectivity
Entry
R2
R3
R4
R6
Ki(M2)(nM) M1/M2 M3/M2
19
30
31
32
33
34
35
36
37
H
H
OMe
OMe OMe OMe
–OCH2O– OMe
OMe
H
–OCH2O–
H
H
H
H
H
H
OMe
H
0.2
2.6
0.6
442
4.7
3.0
53
255
245
147
1
121
27
2
305
170
74
118
157
—
188
—
—
107
108
H
H
OMe
OMe
–OCH2O–
–OCH2O–
H
H
Me
CF3
H
H
0.4
1.8
Having established those structural requirements in the
terminal amide residue, which maximized M2 selectivity,
H
Table 1. Effects ofthe terminal amide on receptor binding and selectivity
Entry
R
Ki (M2) (nM)
M1/M2
M3/M2
M4/M2
M5/M2
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
cyclo-C3H5
C6H5
3,4-Cl2C6H3
2,4-Cl2C6H3
2-ClC6H4
1.3
0.6
0.8
0.3
0.1
0.2
0.1
0.5
4.4
1.0
0.1
0.2
0.1
0.8
0.3
77
12
32
8
66
298
255
333
239
13
84
22
266
330
37
23
29
11
17
69
74
176
70
5
15
33
73
38
57
83
0.3
3
4
2
7
4
6
12
7
3
7
6
9
17
8
7
23
15
7
5
25
23
72
24
5
18
6
47
80
2
2-MeC6H4
2-MeOC6H4
2-H2NC6H4
2-AcNHC6H4
2-HOC6H4
2-PhOC6H4
2-(PhCO2CH2)C6H4
2,3-(OCH2O)C6H3
2,6-Me2C6H3
3-Cl-2-thienyl
3-Me-2-pyridyl
413
8
45
3
4

798
S. McCombie et al. / Bioorg. Med. Chem. Lett. 12 (2002) 795–798
m-trifluoromethyl group was also compatible with high
levels ofselectivity (entries 36 and 37).
M1 receptor, or antagonizing the M2 receptor, see: Brinton,
R. D.; Yamazaki, R. S. Pharm. Res. 1998, 15, 386. (b) Bartus,
R. T.; Dean, R. L. III; Beer, B.; Lippa, A. S. Science 1982,
217, 408. (c) Doods, H. N. Drugs Fut. 1995, 20, 157. (d) Jaen,
J. C.; Davis, R. E. Annu. Rep. Med. Chem. 1994, 29, 23. (e)
Clader, J. W. Curr. Opin. Drug Discov. Develop. 1999, 2, 311.
(f) Baker, R.; Saunders, J. Annu. Rep. Med. Chem. 1989, 24, 31.
2. Kozlowski, J. A.; Lowe, D. B.; Guzik, H. S.; Zhou, G.;
Ruperto, V. B.; Duffy, R. A.; McQuade, R. A.; Crosby, G.,
Jr.; Billard, W.; Binch, H.; Lachowicz, J. E. Bioorg. Med.
Chem. Lett. 2000, 10, 2255.
Additional methyl groups on the piperazine ring had a
deleterious effect on selectivity. Reductive amination of
the 2,3-dimethyl piperazine derivative 10, and the
2,5-dimethylcompounds 12 and 13 with both cyclohex-
anone and 1-(o-toluoyl)-4-piperidinone gave the corre-
sponding piperidinyl-piperazines, which proved to have
high affinity, but poor subtype selectivity.
3. Wang, Y.; Chackalamannil, S.; Hu, Z.; Clader, J. W.;
Greenlee, W.; Billard, W.; Binch, H.; Crosby, G., Jr.; Ruperto,
V. B.; Duffy, R. A.; McQuade, R. A.; Lachowicz, J. E. Bioorg.
Med. Chem. Lett. 2000, 10, 2247.
4. Kozlowski, J. A.; Zhou, G.; Tagat, J. R.; Lin, S.; McCom-
bie, S. W.; Ruperto, V. B.; Duffy, R. A.; McQuade, R. A.;
Crosby, G. Jr.; Taylor, L. A.; Billard, W.; Binch, H. III;
Lachowicz, J. E. Bioorg. Med. Chem. Lett. 2002, 12, 791.
5. In the crude products, the p- to o- ratios were approxi-
mately 5:2 for bromination and 6:1 for iodination; small
amounts ofdi-halo compounds were also present.
6. N-Acylation ofthese hindered, moderately nucleophilic
aminoesters was slow, and proceeded poorly in the presence of
base, presumably because ofmore rapid of rmation and sub-
sequent polymerization ofchloroketene. Heating as indicated,
to dissociate intermediate hydrochloride salts, gave quantita-
tive conversion to the desired amides. After washing with
aqueous sodium bicarbonate solution and evaporation, the
crude amide was used directly in the next step.
7. The arylsulfonyl fluorides were obtained in excellent yields
by treating the corresponding chlorides with potassium fluo-
ride in aqueous acetone: Aitken, R.; Drysdale, A.; Ferguson,
M. G.; Lough, A. J. J. Chem. Soc., Perkin Trans. 1 1998, 875.
Convenient routes to the non-commercial arylsulfonyl chlor-
ides will be detailed elsewhere.
In conclusion, we have demonstrated that selective, high
affinity M2 receptor ligands may be obtained by manip-
ulation ofthe substituent pattern on the two, terminal
aromatic rings in a series of1-[ 1S-(4-[4-arylsulfonyl]-
phenyl)-ethyl]-2R-methyl-4-(1-aroyl-4-piperidinyl)-piper-
azines. Compounds with >100-fold selectivity for the
M2 receptor, relative to both the M1 and M3 subtypes,
were obtained, particularly when the piperidine nitrogen
bears an ortho substituted benzoyl group. Additional
methyl groups in the piperazine ring reduced the recep-
tor subtype selectivity.
Acknowledgements
The authors wish to thank the SPRI Physical-Analytical
Department for spectroscopic and analytical data, and
Drs. J. C. Clader, A. K. Ganguly, and C. Strader for
support and discussions.
References and Notes
8. Suzuki, H.; Abe, H. Tetrahedron Lett. 1995, 36, 6239.
9. Lachowicz, J. E.; Lowe, D.; Duffy, R. A.; Ruperto, V.;
Taylor, L. A.; Guzik, H.; Brown, J.; Berger, J. G.; Tice, M.;
McQuade, R.; Kozlowski, J.; Clader, J.; Strader, C. D.; Mur-
golo, N. Life Sci. 1999, 64, 535.
1. (a) For reviews describing potential approaches to the
treatment ofAlzheimer’s Disease by modiyfing acetylcholine
levels via the inhibition ofacetylcholinesterase, agonizing the