HT2C Receptor Antagonists 321 Indoline Containing Structures

The search for selective 5-HT2C receptor antagonists has been extensively pursued by researchers at GlaxoSmithKline. The evolution of those studies is graphically summarized by the Structures I to VI depicted in Fig. 3.1. Early studies were aimed to obtain 5-HT2C agents with selectivity over 5-HT2A receptors, because both 5-HT2

Dipartimento Farmaco-Chimico, Universitá degli Studi di Barí, via Orabona, 4, 70125, Bari, Italy e-mail: [email protected]

G. Di Giovanni et al. (eds.), 5-HT2C Receptors in the Pathophysiology of CNS Disease, 29

The Receptors 22, DOI 10.1007/978-1-60761-941-3_3, © Springer Science + Business Media, LLC 2011

Fig. 3.1 The evolution of the first template of 5-HT2C receptor antagonists developed by GlaxoSmithKline

receptor subtypes are localized within the central nervous system, in contrast to the 5-HT2B subtype, which is localized peripherally. Later, the selectivity over 5-HT2B receptor was considered important because this receptor is notably expressed in heart valves and pulmonary arteries and could be implicated in cardiopulmonary toxicity. Particular attention has been devoted to the pharmacokinetic properties of the new chemical entities, and therefore, structural features that were not strictly necessary for the interaction with the receptor have been incorporated. The starting point was the 5-HT2C receptor antagonist 1 (SB-206553) (pK = 7.9) with 160-fold selectivity over 5-HT2A receptor. This compound suffered from metabolic demethy-lation to a nonselective compound and therefore, to circumvent its metabolic liability, the replacement of the pyrroloindole ring was investigated. Modeling studies on a range of isomeric ^-substituted pyrrole analogues of 1, in which the pyrrole ring was fused across each of the 4-5, 5-6, and 6-7 bonds of the indoline, provided a definition of an "allowed" volume for 5-HT2C receptor affinity, which included a volume that was "disallowed" at the 5-HT2A receptor. These findings were rationalized by considering key differences in the sequences of the 5-HT2C and 5-HT2A receptors in a region adjacent to the indole N-methyl group in the proposed binding mode of 1. These sequence differences would be expected to give rise in the 5-HT2A receptor to a smaller binding pocket which can less easily accommodate the indole N-methyl group of 1, thus leading to the observed selectivity. On such basis, a series of mono- and disubstituted indolines were studied (Fig. 3.1, Structure I). 5,6-Disubstitution pattern gave the best results in term of affinity at 5-HT2C receptor (pKi values >8) and selectivity over 5-HT2A receptor (>30-fold). In particular, an electron-withdrawing group at the 6 position was preferred and a correlation between 5-HT2C affinity and increasing lipophilicity was apparent at the 5 position. The properties of the substituent in 5 position have been evaluated in depth (compounds 2-17, Table 3.1). In particular, the size and shape of the 5-substituent was varied in order to probe the crucial 5-HT2C-allowed/5-HT2A-disallowed region in combination with the small, lipophilic, electron-withdrawing Cl or CF3 substituent in 6 position. The most interesting results were obtained in the case of the 5-alkyl-thio- and 5-alkyloxy-disubstituted compounds. Several of these compounds combined high 5-HT2C affinity (pKi > 8) with >100-fold selectivities over 5-HT2A. Moreover, it was evidenced, among the 6-CF3-substituted analogues, that the size and shape of the 5 substituent was crucial to the selectivity of the compounds. The best substituents were the thioethyl (11), thiopropyl (12), and i-propoxy (17). A further increase of the size of the substituent led to a drop in affinity and selectivity (compounds 13 and 14). Molecular modeling on the 5-CH3S-6-CF3 derivative 10 (SB-221284) revealed that these substituents optimally interact/occupy the crucial 5-HT2C-allowed/5-HT2A-disallowed volume. The introduction of a 6-substituent also has the added beneficial effect of restricting the rotation of the 5-substituent to favor those conformations in which the alkyl group occupies the crucial region. Compound 10 was manually docked into a model of the 5-HT2C receptor.

The proposed binding mode was very similar to that proposed for 1 with the urea carbonyl oxygen double-hydrogen bonding to the hydroxyl side chains of Ser-312 and Ser-315. The 3-pyridyl ring occupies a lipophilic pocket defined by the side

Table 3.1 Binding affinities of 5-HT receptor antagonistsa

Compound

R5

R6

5-HT2C

5-HT2A

5-HT2B

2

ch3

Cl

8.2

6.8

-

3

ch2ch3

Cl

8.3

6.4

-

4

CH2CH2CH3

Cl

7.7

5.9

-

5

CH(CH3)2

Cl

7.7

5.7

-

6

C(CH3)3

Cl

7.7

6.2

6.8

7

och2ch3

Cl

7.6

5.4

-

8

OCH(CH3)2

Cl

7.8

<5.2

-

9

SCH3

Cl

8.2

5.6

-

10 (SB-221284)

sch3

CF3

8.6

6.4

7.9

11

sch2ch3

CF3

8.5

5.5

8.0

12

sch2ch2ch3

CF3

8.2

<5.2

7.8

13

SCH2CH3)2

CF3

7.5

5.3

-

14

CH(CH3)2

CF3

7.6

<5.2

-

15

och3

CF3

8.0

6.0

-

16

och2ch3

CF3

8.2

5.8

-

17

OCH(CH3)2

CF3

8.5

5.8

8.4

aReceptor binding assays were performed by using HEK-293 cell lines expressing recombinant HT, and 5-HT receptors. Determination of affinities was made using [3H]

human 5-HT2C, 5 mesulergine for 5 receptors

HT2C receptors, [3H]ketanserin for 5-HT2A receptors, and [3H]-5-HT for 5-HT2B

aReceptor binding assays were performed by using HEK-293 cell lines expressing recombinant HT, and 5-HT receptors. Determination of affinities was made using [3H]

human 5-HT2C, 5 mesulergine for 5 receptors

HT2C receptors, [3H]ketanserin for 5-HT2A receptors, and [3H]-5-HT for 5-HT2B

chains of the aromatic residues Phe-508, Trp-613, Phe-616, and Phe-617. Within this pocket the 3-pyridyl ring is able to form both p-p stacking and edge-to-face aromatic interactions with several of the aromatic residues lining the pocket. These interactions probably contribute significantly to the overall binding of these ligands to the receptor. The substituted indoline is placed in another pocket, the boundary of which is defined by residues Val-212, Phe-311, Val-608, Phe-609, Met-612, and Tyr-715. This pocket is also very lipophilic in nature, although less aromatic than the 3-pyridyl binding pocket. In the 5-HT2A receptor sequence both the 212 and 608 residues are Leu. These differences would be expected to lead to binding pockets of reduced size, and it was proposed that these steric differences in the receptors might account for the observed 5-HT2C specificity (Bromidge et al. 1998).

Due to the synthetic complexity of indolines with Structure I (Fig. 3.1), the more accessible equivalently substituted phenyl ureas were evaluated (Fig. 3.1, Structure II). However, this structural simplification was detrimental for affinity (pK. < 7.8) (Bromidge et al. 1999).

Although indolines 2, 10, and 15 displayed high affinity and specificity for the 5-HT2C receptor, they potently inhibited a number of human cytochrome P450 enzymes, in particular, the CYP1A2 isoform, and this precluded further development. Introduction of steric hindrance around the pyridine nitrogen of 10 revealed that the unhindered nitrogen was responsible for the P450 inhibitory activity. Modeling studies revealed that the lipophilic pocket that was occupied by the pyridine ring was quite deep. Therefore, the substitution of the pyridyl ring with a variety of aryl groups was investigated with the twofold aim to reduce the cytochrome P450 inhibitory activity and to increase the 5-HT2C affinity (Fig. 3.1, Structure III). Although this modification was generally tolerated for the 5-HT2C affinity, none of the compounds displayed acceptable CYP1A2 activity. Alternatively, the replacement of the pyridyl with phenyl ring was accomplished (Fig. 3.1, Structure IV). In particular, the phenyl ring was decorated with various pyridyl substituents in order to retain good solubility and efficacious interaction with the 5-HT2C receptor. Compounds 18-26 (Table 3.2) generally retained good 5-HT2C affinity and selectivity over 5-HT2A receptors combined with reduced P450 liability. In particular, increasing the torsion angle between the two aromatic rings greatly increased the selectivity over 5-HT2A receptors. The most notable example was compound 23, which displayed sub-nanomolar 5-HT2C affinity along with >20,000-fold selectivity over 5-HT2A receptor. Modeling studies of 23 showed an almost orthogonal relationship between the two aryl rings. However, this group of compounds yet displayed poor oral activity or inhibitory activity of other P450 enzymes.

In order to improve duration of action in vivo, a series of analogues of compound 19 (Table 3.2, R3 = 3-Py, R4 = R5 = H) was studied. A variety of substituents capable of blocking the metabolism of the electron-rich phenyl ring (R4 or R5 = Cl, F, Br, CH3, OCH3) were well tolerated, affording good to excellent 5-HT2C affinity and increased selectivity over 5-HT2A. Incorporating 4,5-disubstitution produced an additive effect on activity leading to very high 5-HT2C affinities and selectivities over 5-HT2A (compounds 27 and 28, Table 3.2). These compounds demonstrated low inhibition of CYP1A2 and potent oral activity in the rat.

Many of the compounds described in Tables 3.1 and 3.2 possessed excellent 5-HT2C affinity and selectivity over 5-HT2A but lacked selectivity over the 5-HT2B receptor. Receptor-ligand modeling work suggested that the lipophilic pocket within the 5-HT2C receptor was still not fully exploited. This model suggested that introducing a linker group between the aromatic rings (Fig. 3.1, Structure V) would more optimally occupy this hydrophobic region (compounds 29-30, Table 3.3). The bispyridyl ether 29 showed excellent 5-HT2C affinity and almost 100-fold selectivity over 5-HT2A receptor. Increasing the torsion angle between the two aryl rings by introduction of a 2-methyl substituent into the terminal pyridyl ring was beneficial for 5-HT2C affinity and increased tenfold the selectivity over 5-HT2A receptor (compound 30). In addition, 80-fold selectivity over the 5-HT2B receptor was achieved. Modification of the substitution pattern of the indoline ring afforded compounds 31 (SB-243213) and 32 (SB-242084), which displayed optimal affinity for 5-HT2C receptor, >100-fold selectivity over 5-HT2A and 5-HT2B receptors, and suitable properties for in vivo evaluation (Bromidge et al. 2000a).

The structure-activity relationships of compounds 31 (SB-243213) and 32 (SB-242084) have been studied further (Fig. 3.1, Structure V). The results from the indoline substitution optimization overlapped those obtained for the compounds

Table 3.2 Binding affinities of 5-HT receptor antagonistsa

Compound

R3

R4

R5

5-HT2C

5-HT2A

5-HT2B

CYP1A2 IC50 (mmol/L)

18

2-Py

H

H

7.7

6.3

-

3%

19

3-Py

H

H

9.0

6.7

8.1

4

20

4-CH3-3-Py

H

H

9.2

6.6

-

8%

21

2-CH3-3-Py

H

H

8.3

6.3

7.8

2%

22

2,4-di-CH3-3-Py

H

H

8.6

6.7

7.9

-

23

4-CH3-3-Py

CH3

H

9.5

<5.2

8.4

>100

24

4-CH3-3-Py

Cl

H

9.2

<6.0

-

11%

25

H3

3-Py

H

7.8

5.4

8.0

0%

26

H

4-Py

H

8.3

<5.0

7.9

-

27

3-Py

CH3

F

9.2

5.8

8.1

-

28

3-Py

4,5-OCH2CH2

9.3

6.6

8.1

2%

aReceptor binding assays were performed by using HEK-293 cell lines expressing recombinant human 5-HT2C, 5-HT2A, and 5-HT2B receptors. Determination of affinities was made using [3H] mesulergine for 5-HT2C receptors, [3H]ketanserin for 5-HT2A receptors, and [3H]-5-HT for 5-HT2B receptors. The cytochrome P450 inhibitory potential was determined using isoform-selective assays and heterologously expressed human CYP1A2

Table 3.3 Binding affinities of 5-HT2C receptor antagonistsa

Compound

R!

R2

R3

5-HT2C

5-HT2A

5-HT2B

CYP1A2 IC50 (mmol/L)

29

H

CF3

och3

8.9

7.0

7.7

4%

30

CH3

CF3

och3

9.2

6.1

7.3

>100

31(SB-243213)

ch3

CF3

CH3 3

9.0

6.8

7.0

>100

32(SB-242084)

ch3

Cl

ch3

9.0

6.8

7.0

>100

aReceptor binding assays were performed by using HEK-293 cell lines expressing recombinant human 5-HT2C, 5-HT2A, and 5-HT2B receptors. Determination of affinities was made using [3H] mesulergine for 5-HT2C receptors, [3H]ketanserin for 5-HT2A receptors, and [3H]-5-HT for 5-HT2B receptors. The cytochrome P450 inhibitory potential was determined using isoform-selective assays and heterologously expressed human CYP1A2

with Structure I (Fig. 3.1): an electron-withdrawing group in R6 and an alkyl group in R5 were preferred. Optimization of the terminal aryl ring was focused on 2-and 4-positions because of its role on selectivity. Considering the R2 substituent, the replacement of the methyl with ethyl or «-propyl or chloro was well tolerated with respect to the affinity, whereas the introduction of an ¿-propyl group was detrimental for affinity. Shifting of the methyl in 4 position (R2= H, R4= CH3) or its replacement with a Cl maintained excellent 5-HT2C affinity (pK = 9.3). However, these modifications had a negative effect on selectivity. As for the central pyridine ring, introduction of a methyl or its replacement by various diazines was well tolerated for 5-HT2C receptor affinity (pK = 8.5-8.7), but produced a loss in 5-HT2B selectivity (Bromidge et al. 2000b).

As already discussed, the to-pyridyl ether moiety in Structure V (Fig. 3.1) occupies a lipophilic pocket defined by side-chain aromatic residues on transmembrane helices 5 and 6. The size and shape of this binding pocket was further explored by incorporating longer linker groups between the aromatic rings and by evaluation of various terminal heteroaryl groups (Fig. 3.1, Structure VI). Affinity data indicated that the unsubstituted 2-pyridylmethoxy substituent (Ar-X) optimally fills the lipophilic binding pocket (R=CF3: pK = 9.3; R=Cl: pK = 8.6). Finally, modifications of both length and nature of the linker were detrimental of affinity and selectivity (Bromidge et al. 2000c).

Defeat Drugs and Live Free

Defeat Drugs and Live Free

Being addicted to drugs is a complicated matter condition that's been specified as a disorder that evidences in the obsessional thinking about and utilization of drugs. It's a matter that might continue to get worse and become disastrous and deadly if left untreated.

Get My Free Ebook


Post a comment