4-(3,4-Dihydro-1H-isoquinolin-2yl)-pyridines and 4-(3,4-Dihydro-1H-isoquinolin-2-yl)-quinolines as potent NR1/2B subtype selective NMDA receptor antagonists

Article abstract of DOI:10.1016/S0960-894X(03)00232-4

A series of 4-(3,4-dihydro-1H-isoquinolin-2yl)-pyridines and analogous quinolines was prepared and evaluated as NR1/2B subtype selective NMDA receptor antagonists. 2-Hydroxyalkylamino substitution combines high affinity with selectivity (vs α1 and M1 receptors) and activity in vivo.

Full text of DOI:10.1016/S0960-894X(03)00232-4

Bioorganic & Medicinal Chemistry Letters 13 (2003) 1759–1762
4-(3,4-Dihydro-1H-isoquinolin-2yl)-pyridines and 4-(3,4-Dihydro-
1H-isoquinolin-2-yl)-quinolines as Potent NR1/2B Subtype
Selective NMDA Receptor Antagonists
Bernd Buttelmann,a,* Alexander Alanine,a Anne Bourson,b Ramanjit Gill,b
Marie-Paule Heitz,a Vincent Mutel,b Emmanuel Pinard,a Gerhard Trubeb
and Rene Wylera
aPharma Division, Discovery Chemistry, F. Hoffmann-La Roche Ltd., CH-4070 Basel, Switzerland
bPharma Division, Preclinical CNS Research, F. Hoffmann-La Roche Ltd., CH-4070 Basel, Switzerland
Received 19 December 2002; revised 4 February 2003; accepted 26 February 2003
Abstract—A series of 4-(3,4-dihydro-1H-isoquinolin-2yl)-pyridines and analogous quinolines was prepared and evaluated as NR1/
2B subtype selective NMDA receptor antagonists. 2-Hydroxyalkylamino substitution combines high affinity with selectivity (vs a1
and M1 receptors) and activity in vivo.
# 2003 Elsevier Science Ltd. All rights reserved.
Introduction
Overactivation of N-methyl-d-aspartate (NMDA) recep-
tors by glutamate has been shown to play a key role in a
variety of acute and chronic neurodegenerative disorders.1
Whilst a number of subtype-unselective NMDA antago-
nists were found to be efficacious in animal models of
stroke,2 their clinical usefulness has been compromised by
dose-limiting side effects.3 Fortunately, antagonists of the
NR1/2B subtype4 of the NMDA receptor offer a mark-
edly improved safety window.5,6 In 1993 Ifenprodil (1)
was characterized as the first NR1/2B subtype selective
NMDA antagonist.7 Originally developed as an anti-
hypertensive, (1) also has a high affinity for a1 recep-
tors,8 an undesirable attribute in stroke indications.9
Meanwhile, a number of compounds structurally rela-
ted to (1) with reduced a1 affinity have been described.8
In the course of a medicinal chemistry program aimed
at identifying structurally novel NR1/2B subtype selec-
tive NMDA antagonists we could characterize 2-(3,4-
dihydro-1H-isoquinolin-2-yl)-quinolines (2) and 2-(3,4-
dihydro-1H-isoquinolin-2-yl)-pyridines (3) as repre-
sentatives of promising new classes.10,11 In this
communication we would like to disclose our efforts to
evaluate the regioisomeric analogues (e.g., 4) (Table 1).
Table 1. Binding affinities of reference compounds
Compd
Ki (nM)a NMDAb
Ki (nM)a a1c
1d
2
13
3.5
2
12
430
2300
3
aKi values are the medians of at least two dose–response curves.
bDisplacement of [3H]-25-6981.19
*Corresponding author. Tel.: +41-61-688-13-51; fax: +41-61-688-87-
14; e-mail: bernd.buttelmann@roche.com
cDisplacement of [3H]-prazosin.20
dCompound was synthesized at Hoffmann-La Roche.
0960-894X/03/$ - see front matter # 2003 Elsevier Science Ltd. All rights reserved.
doi:10.1016/S0960-894X(03)00232-4

1760
B. Buttelmann et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1759–1762
¨
Table 2. Binding affinities of compounds 14a–14h. Influence of sub-
stituents at 1,2,3,4-tetrahydroisoquinoline
Chemistry
The main synthetic scheme for the preparation of the
envisaged scaffolds is straightforward and uses estab-
lished synthetic methods as outlined in Scheme 1. Key
starting materials are 4-bromopyridines 5–8 and
4-chloroisoquinolines 9–11. These compounds are either
commercially available (5, 9 and 10) or known (6, 7 and
11).12À14 8 was obtained as the minor regioisomer by
Compd
R1
Ki [nM]a NMDAb Ki [nM] a a1c Ki [nM] a M1d
14a
14b
14c
14d
14e
14f
14g
14h
H
(rac)-1-Me
(rac)-4-Me
5-Cl
6-Cl
7-Cl
8-Cl
6,7-(OMe)2
8
5600
25
8
230
37
5800
ND
2,200
1,500
ND
1300
3800
ND
720
ND
900
660
ND
1000
650
ND
26
28,000
a,b,cSee Table 1.
dDisplacement of [3H]-pirenzipine.21
(Ki=8 nM) towards the NR1/2B subtype of the NMDA
receptor (Table 2). Interaction with a1 receptors is
marginal (Ki=5800 nM), however selectivity versus M1
receptors (Ki=720 nM) was determined to be insuffi-
cient. Thus our objecive was to elucidate the structural
requirements for an increased M1 selectivity (while
maintaining low a1 affinity). We first turned our atten-
tion to substituent effects at the 1,2,3,4-tetra-
hydroisoquinoline motif while keeping constant the
unsubstituted pyridine core. Introduction of a methyl
group in position 1 (14b) reduces NMDA-affinity
almost 1000-fold, suggesting that the dihedral angle
between pyridine and 1,2,3,4-tetrahydroisoquinoline
should not be widened. The positional isomer 14c
reveals that a 4-methyl group is much better tolerated.
However, a 3-fold drop in NMDA affinity (Ki=25 nM)
combined with an increased a1 affinity (Ki=2200 nM)
leads to an overall markedly reduced selectivity ratio.
This trend also holds true for 5-chloro substituted 14d.
The other 3 monochloro-1,2,3,4-terahydroisoquinolines
14e–14g exert only reduced NMDA affinity (Ki >25
nM). Notably the distal 6 position tolerates substitution
by chlorine the least (14e, Ki=230 nM). As introduction
of methoxy groups (14h) with an inverse electronic and
hydrophobic demand leads to a complete loss of
NMDA affinity (14h), we conclude that unsubstituted
1,2,3,4-tetrahydroisoquinolines are probably optimal,
both in terms of NMDA affinity and in selectivity.
Scheme 1. (a) NH4OH, 160 ꢀC (autoclave), 4%; (b) tetra-
hydroisoquinoline, 150 ꢀC, 12–88%; (c) EtCOCOCl, NEt3, rt, 63–
73%; (d) LAH, THF, 0 ꢀC, 47%; (e) LAH, THF, 75 ꢀC, 32–42%. (For
definitions of R1 see Table 2.)
partial aminolysis of dibromide 12.15 At elevated tem-
peratures the halogen in 4-position in 5–11 could be
displaced by variously substituted 1,2,3,4-tetra-
hydroisoquinolines 13a–h and compounds 14a–h to 20
were thus obtained. The respective 1,2,3,4-tetra-
hydroisoquinolines are either purchased from commer-
cial sources (13a and 13h) or known (13b to 13e), (13f,
13g).16,17 The free NH2-group in 17 and 20 could be
acylated to ethyloxamates (e.g. 21), which depending on
the reduction conditions could be reduced either to gly-
colamide 22 or to the ethanolamines 4 and 23.
Following an iterative approach (the unsubstituted
1,2,3,4-terahydroisoquinoline was kept constant), we
then turned our attention to varying the substitution
pattern at the pyridine as well as the isoquinoline core
(Table 3 and 4). Compared with 14a, the benzo-anne-
lated analogue 18 has less affinity towards the NR1/2B
subtype of the NMDA receptor (Ki=29 nM). As we
have recently shown, NMDA affinity in the structurally
related
2-(3,4-dihydro-1H-isoquinolin-2-yl)-quinoline
series (e.g., 2) critically depends on electron donating
substituents,10 we hoped to increase NMDA affinity by
introduction of a methyl (19) or amino-group (20). How-
ever, the results obtained with these compounds (Ki=75
nM and 19 nM, respectively) did not fully support this
hypothesis. Moreover, 20 has marked a1 affinity
Results and Discussion
The structurally most simple 4-(3,4-dihydro-1H-iso-
quinolin-2-yl)-pyridine 14a already displays high affinity

B. Buttelmann et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1759–1762
¨
1761
Table 5. In vivo potency of selected compounds
Table 3. Binding affinities of compounds 18–20, 23. Influence of
substituents at isoquinoline
Compd
ED50 [mg/kg]a
Sound induced seizures
1
2
3
16
<12
13
4
18
25
3
18
14a
14d
Compd
R4
Ki [nM]a
NMDAb
Ki [nM] a
a1c
Ki [nM] a
M1d
2
aCompounds were administered ip 30 min before testing in DBA/2
mice.
18
19
20
23
H
Me
29
75
19
22
3200
ND
120
2600
ND
ND
6700
NH2
NH(CH2)2OH
1700
however, that these series do not follow an identical
structure activity relationship. Glycolamide 22 and
notably ethanolamine 4 also bind strongly to the
NMDA receptor (Ki=15 nM and 2 nM, respectively).
As anticipated (vide supra) 4 exerts only a markedly
reduced affinity to a1 and M1 receptors (Ki >3000 nM);
this compound surpasses our selectivity ratio threshold
(1000-fold). Interestingly, also 21 and 22 meet this cri-
terion since both derivatives have virtually no affinity
towards a1 and M1 receptors (Ki >10,000 nM). From
these results it is apparent that polar functionalities in
this part of the molecule are not tolerated in the binding
pocket of M1 and a1 receptors.
a,b,cSee Table 1.
dSee Table 2.
Table 4. Binding affinities of compounds 4, 15–17, 21, 22, 24. Influ-
ence of substituents at pyridine
Compd R2
R3
Ki [nM]a
NMDAb
Ki [nM]a
a1c
Ki [nM]a
M1d
In vivo activity of key compounds was measured in
mice after ip administration using the standard sound-
induced seizures assay.18 As depicted in Table 5, the
2-ethanolamino substituted 4 - (3,4 - dihydro - 1H - iso-
quinolin-2-yl)-pyridine 4 and 4-(3,4-dihydro-1H-iso-
quinolin-2-yl)-quinoline 18 exhibit in vivo activity
comparable to the reference compounds (1–3). In this
respect the 2-unsubstituted congeners 14a and 14d were
found to be more potent.
4
Me
Me
H
Me
Me
NH(CH2)2OH
2
8
2
4
8
15
3500
5700
2100
1500
12,000
20,000
6800
ND
400
2600
11,000
60,000
15
16
17
21
22
H
NH2
NH2
NHCOCO2Et
Me NHCOCH2OH
a,b,cSee Table 1.
dSee Table 2.
(Ki=120 nM), rendering this compound less interesting.
In the case of the analogous 2-(3,4-dihydro-1H-iso-
quinolin-2-yl)-pyridines (e.g., 3) we have already shown
that replacement of an amino group by an ethanol-
amino side chain leads to an increase in selectivity
NMDA versus a1 and M1.11 This is also the case in the
isomeric series; the a1 affinity of 23 is reduced >10-fold
(Ki=1700 nM) when compared to 20. However, the
overall selectivity profile of 23 turned out not to be
superior to 18 (Table 3).
Having established that 4-(3,4-dihydro-1H-isoquinolin-
2-yl)-pyridines (e.g., 4) and 2-(3,4-dihydro-1H-iso-
quinolin-2-yl)-pyridines (e.g., 3) follow a similar SAR
pattern, we then turned our attention to the analogous
pyrimidines. We will report on these in due course.
Conclusion
4-(3,4-Dihydro-1H-isoquinolin-2-yl)-pyridines and 4-(3,4-
dihydro-1H-isoquinolin-2-yl)-quinolines were identified
as novel series of NR1/2B subtype selective NMDA
antagonists. The high affinity compounds 14a and 14d
were shown to be highly active in vivo. In close analogy to
structurally related isomeric derivatives additional
2-hydoxyalkylamino substitution at the pyridine core (as
in 4) combines high affinity at the NMDA receptor with
low muscarinic and adrenergic side effect liabilities.11
A similar optimization strategy was followed for the 4-
(3,4-dihydro-1H-isoquinolin-2-yl)-pyridines (Table 4).
Introduction of electron donating substituents in posi-
tion 2 of the pyridine core leads to compounds with
high affinity at the NR1/2B subtype of the NMDA
receptor (Me substituted 15: Ki=8 nM, NH2 substituted
16 Ki=2 nM). However, the envisaged 1000 fold selec-
tivity versus a1 and M1 receptors cannot be attained.
This is also the case for 2-NH2, 6-Me disubstituted
derivative 17. Surprisingly, replacement of the 2-NH2
group in 17 by ethyloxamate as exemplified by 21,
reduces NMDA affinity only 2 fold (Ki=8 nM). Based
on our understanding in the related 2-(3,4-dihydro-1H-
isoquinolin-2-yl)-quinolines (e.g., 2) we expected only
H-bond donating groups to be allowed as the terminal
substituent on the amino function.10 It is possible,
Acknowledgements
The skillful technical assistance of B. David and C.
Wunderlin is gratefully acknowledged. The authors
wish to thank W. Arnold, W. Meister for spectroscopic
characterization.

1762
B. Buttelmann et al. / Bioorg. Med. Chem. Lett. 13 (2003) 1759–1762
¨
References and Notes
M. P.; Mutel, V.; Pinard, E.; Trube, G.; Wyler, R. Bioorg.
Med. Chem. Lett. 2003, 13, 829.
12. Ochiai, S. Pharm. Bull 1954, 2, 147.
1. Choi, D. W. Neuron 1988, 1, 623.
2. McAuley, M. A. Cerebrovasc. Brain Metab. Rev. 1995, 7, 153.
3. Muir, K. W.; Lees, K. R. Stroke 1995, 26, 503.
4. Monyer, H.; Sprengel, R.; Schoepfer, R.; Herb, A.; Higu-
chi, M.; Lomeli, H.; Burnashev, N.; Sakmann, B.; Seeburg,
P. H. Science 1992, 256, 1217.
5. Gill, R.; Kemp, J. A.; Richards, J. G.; Kew, J. N. C. Curr.
Opin. Cardiovasc., Pulm. Renal Invest. Drugs 1999, 1, 576.
6. Bullock, M. R.; Merchant, R. E.; Carmack, C. A.; Dop-
penberg, E.; Shah, A. K.; Wilner, K. D.; Ko, G.; Williams,
S. A. Ann. N.Y. Acad. Sci. 1999, 890, 51.
13. Den Hertog, H. J. Recl. Trav. Chim. Pays-Bas 1945, 64, 85.
14. Hardman, R.; Partridge, M. W. J. Chem. Soc. 1958, 614.
15. Bernstein, J.; Stearns, B.; Dexter, M.; Lott, W. A. J. Am.
Chem. Soc. 1947, 69, 1147.
16. Bondinell, W. E.; Chapin, F. W.; Girard, G. R.; Kaiser,
C.; Krog, A.; Payloff, A. M.; Schwartz, M. S.; Silvestri, J. S.;
Vaidya, P.; Lam, B. L.; Wellman, G. R.; Pendleton, R. G. J.
Med. Chem. 1980, 23, 506.
17. Grunewald, G. L.; Sall, D. J.; Monn, D. A. J. Med. Chem.
1988, 31, 433.
7. Williams, K. Mol. Pharmacol. 1993, 44, 851.
8. Chenard, B. L.; Menniti, F. S. Curr. Pharm. Des. 1999, 5,
381.
18. Bourson, A.; Kapps, V.; Zwingelstein, C.; Rudler, A.;
Boess, F. G.; Sleight, A. J. Naunyn-Schmiedeberg’s Arch.
Pharmacol. 1997, 356, 820.
9. Bentue-Ferrer, D.; Decombe, R.; Reymann, J.-M.; Schatz,
C.; Allain, H. Clin. Neuropharmacol. 1999, 13 (Suppl. 3), S9.
10. Pinard, E.; Alanine, A.; Bourson, A.; Buttelmann, B.;
Gill, R.; Heitz, M. P.; Mutel, V.; Trube, G.; Wyler, R. Bioorg.
Med. Chem. Lett. 2002, 12, 2615.
19. Mutel, V.; Buchy, D.; Klingelschmidt, A.; Messer, J.;
Bleuel, Z.; Kemp, J. A. J. Neurochem. 1998, 70, 2147.
20. Greengrass, P.; Bremner, R. Eur. J. Pharmacol. 1979, 55,
323.
21. Hudkins, R. L.; Stubbins, J. F.; DeHaven-Hudkins, D. L.
Eur. J. Pharmacol. 1993, 231, 485.
11. Buttelmann, B.; Alanine, A.; Bourson, A.; Gill, R.; Heitz,