Amino acid derived quinazolines as Rock/PKA inhibitors

Share Embed


Descrição do Produto

Bioorganic & Medicinal Chemistry Letters 23 (2013) 1592–1599

Contents lists available at SciVerse ScienceDirect

Bioorganic & Medicinal Chemistry Letters journal homepage: www.elsevier.com/locate/bmcl

Amino acid derived quinazolines as Rock/PKA inhibitors Sarwat Chowdhury a, Yen Ting Chen a, Xingang Fang a, Wayne Grant b, Jennifer Pocas b, Michael D. Cameron c,e, Claudia Ruiz c, Li Lin c, HaJeung Park d, Thomas Schröter b, Thomas D. Bannister a, Philip V. LoGrasso b,e, Yangbo Feng a,⇑ a

Medicinal Chemistry, Translational Research Institute, The Scripps Research Institute, 130 Scripps Way, 2A1, Jupiter, FL 33458, USA Discovery Biology, Translational Research Institute, The Scripps Research Institute, 130 Scripps Way, 2A1, Jupiter, FL 33458, USA Drug Metabolism and Pharmacokinetics, Translational Research Institute, The Scripps Research Institute, 130 Scripps Way, 2A1, Jupiter, FL 33458, USA d Crystallography/Modeling Facility, Translational Research Institute, The Scripps Research Institute, 130 Scripps Way, 2A1, Jupiter, FL 33458, USA e Department of Molecular Therapeutics, The Scripps Research Institute, 130 Scripps Way, 2A1, Jupiter, FL 33458, USA b c

a r t i c l e

i n f o

Article history: Received 30 November 2012 Revised 14 January 2013 Accepted 22 January 2013 Available online 31 January 2013 Keywords: Rock PKA Quinazoline Kinase inhibitor

a b s t r a c t SAR and lead optimization studies for Rock inhibitors based on amino acid-derived quinazolines are described. Studies demonstrated that these amino acid derived quinazolinones were mainly pan-Rock (I & II) inhibitors. While selectivity against other kinases could be achieved, selectivity for most of these compounds against PKA was not achieved. This is distinct from Rock inhibitors based on non-amino acid derived quinazolinones, where high selectivity against PKA could be obtained.22 The inhibitors presented here in some cases possessed sub-nanomolar inhibition of Rock, nanomolar potency in ppMLC cell based assays, low to fair cytochrome P-450 inhibition, and good human microsomal stability. Ó 2013 Elsevier Ltd. All rights reserved.

Rho-associated protein kinase (Rock), a serine-threonine kinase of the AGC family, is a major downstream effector of the Rho-mediated signaling cascade.1 Two isoforms of Rock (I & II) have been isolated that share 65% overall sequence homology and 92% in their ATP-binding domain.2 Upon activation following binding of Rho, Rock phosphorylates targets such as myosin light chain,3 myosin light chain phosphatase,4 LIM kinase5 and zipper-interacting kinase (ZIPK)6 which lead to cell signaling that primarily promotes changes in cell motility, adhesion,7 stress fiber formation,7 and force generation through smooth muscle contraction.8 Inhibition of Rock is thus considered to have potential therapeutic value for conditions such as hypertension,9,10 glaucoma,11 cancer metastasis12 and multiple sclerosis13 among others. Fasudil, the only clinically approved Rock inhibitor, has been used for the treatment of cerebral vasospasm in Japan since 1995.14 The proven safety and efficacy of Fasudil have engendered considerable interest in the development of Rock inhibitors for therapeutic purposes and consequently, several reports have disclosed the development of Rock inhibitors of various classes.15 Over the course of several years our group has reported Rock inhibitors of various scaffolds including benzadioxane, chroman, and tetrahydroisoquinoline amides,16 benzimidazoles,17 benzothiazoles,18 ureas,19 indazoles,20 and indole and azaindole-based21 com⇑ Corresponding author. Tel.: +1 561 228 2201. E-mail address: [email protected] (Y. Feng). 0960-894X/$ - see front matter Ó 2013 Elsevier Ltd. All rights reserved. http://dx.doi.org/10.1016/j.bmcl.2013.01.109

pounds. Recently we also reported a series of highly potent and selective quinazoline-based Rock inhibitors.22 Herein, we report our efforts towards development of quinazoline-based Rock inhibitors derived from amino acids. The reason for the introduction of an amino group (from amino acids) was mainly to enhance the aqueous solubility of Rock-II inhibitors, which is a highly desirable property for topical anti-glaucoma applications. The amino acid-quinazoline-based Rock inhibitors were designed to maximize three key interactions (Fig. 1): (i) hydrogen bond donor/acceptor interactions with the backbone (Glu-170 and Met-172) in the ATP hinge-binding site, (ii) electrostatic interaction with the conserved lysine side chain group (Lys-121), and (iii) hydrophobic interaction with the glycine-rich flexible P-loop as depicted in Figure 1a. We elected to probe three amino acid sub-structures (Fig. 1b): phenylglycine analogs 1, phenylalanine derivatives 2, and pyrrolidine-2-carboxylic acid based compounds (3). The later substructure, a b-amino acid, was derived from a Rock-II inhibitor discovered in our HTS campaign and showed promises in our earlier SAR studies.17b Synthesis of phenylalanine and phenylglycine-based Rock inhibitors began with HATU-mediated coupling of a desired amino acid 5 and 2-amino-5-bromobenzamide 4 (Scheme 1). The resulting amide 6 was then subjected to Suzuki heteroarylation under microwave irradiation to give the Boc-protected quinazolinone 8. The coupling reaction and the ring-closure reaction (to form the quinazolinone) were carried out in one-pot as described in

1593

S. Chowdhury et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1592–1599

(a)

Scheme 1. Deprotection of Boc by treatment with TFA in DCM at room temperature furnished the final compounds 1 and 2. The two-step synthesis consisting of an amide coupling followed by palladium-catalyzed Suzuki coupling (Scheme 1) was also adapted to access N-protected pyrrolidine-based compound 11 (Scheme 2). Boc-protected amino acids (9) were obtained from commercial sources while N–Bn or N–Me compounds were synthesized following established routes.23 Rock inhibitor 12 with the basic pyrrolidine ring was obtained from appropriate starting materials by Boc-deprotection using TFA in DCM at room temperature or debenzylation by hydrogenation using Pd(OH)2 in AcOH. Scheme 3 illustrates a representative procedure for the synthesis of N-substituted phenylalanine-based inhibitors. Briefly, EDC mediated coupling of compound 13 with a desired carboxylic acid followed by reverse phase preparative HPLC purification gave the desired product 14. In the cyclic b-aminoacid series, pyrrolidine N-functionalized Rock inhibitors were accessed first by treatment of 15 using TFA in DCM to remove the Boc protecting group as shown in Scheme 4. The free pyrrolidine NH group was then acylated (2i) by reaction with a desired acid chloride or treated with CDI followed by an

Lys-121 Glu-170 Met-172 O Het

R1

NH

p-loop

N N R2

(b) O Het

O

R1 Het

NH

NH

( )n

N

R1

N

NH2

NH

1, n= 0 2, n= 1

3

Figure 1. (a) Design rationale for amino acid derived quinazoline Rock inhibitors. (b) a-Amino acid derived inhibitors (1 & 2) and cyclic b-amino acid derived Rock inhibitors (3).

Br

R1 HATU, NMM,

+

NH2 NH2

BocHN

n = 0,1 COOH

4

DMF, rt 6 O Het

Het

7

O

NH NHBoc

N Pd(PPh3 )4 , K 2CO3 aq., EtOH:tol (3/2), 1 h, microwave, 140 o C,

n = 0,1

R1

O

O

NHBoc

N H

5

Het B

CONH2 O

Br

O

NH

TFA:DCM (1/1)

N

rt n = 0,1

R1

R1

NH2 n = 0,1

1 n= 0, 2 n= 1,

8 Scheme 1. Synthesis of Rock inhibitors 1 and 2.

O

R1

HOOC

Br

NH2

HATU, TEA, +

NH2

N 9 R2 2 R = Boc, Bn or Me

4

Het

7

O

O Het

2

R1 NH

N R2 R2 = Boc, Bn or Me 11

TFA:DCM (1/1), r.t (when R2 = Boc) H2, Pd(OH)2, AcOH, r.t (when R2 = Bn)

R1

NH N

R1

N R2 R = Boc, Bn or Me

10

N

Pd(PPh3)4, K2CO3 aq., EtOH:tol (3/2), 1 h, microwave, 140 °C,

N H

DMF, r.t

O

O Het B

CONH2 O

Br

12 NH

Scheme 2. Synthesis of cyclic b-amino acid derived Rock inhibitors.

1594

S. Chowdhury et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1592–1599

N HN

N

O NH N

N

O

NH2

O

HN NH

H N

N

HO

N O

DMF, NMM, HOBt, EDC, r.t, overnight

Cl

Cl

13

14

Scheme 3. Synthesis of N-substituted quinazolinone Rock inhibitors.

Br

O N H O

N Boc

NH2

N

1) TFA:DCM (1/1), r.t 2i) DMF, TEA, R3OCl, r.t or 2ii) CDI, THF, TEA, R3OH or R3NH2

NH N N

16

3) 7, Pd(PPh3)4, K2CO3 aq., EtOH:tol (3/2), 1 h, microwave, 140°C

15

O

HN

R

R = C(O)R3 (2i) R = C(O) OR3 or C(O)NR3 (2ii)

Scheme 4. Synthesis of N-substituted b-amino acid based inhibitors.

N

O

N HN

HN NH N

N

R4 N

NHBoc

F

BOP, R4NHCH3,TEA, DMF, r.t ,1 h

N

NHBoc

F

17 N

N

18

R4

HN N DCM/TFA (1:1), r.t, 1 h

N

NH2

F

19 Scheme 5. Synthesis of 4-amino quinazolines as Rock inhibitors.

alcohol or an amine to furnish target carbamate or urea compounds (2ii) (following procedures in Scheme 2). Finally, a onepot reaction of palladium mediated Suzuki heteroarylation and ring-closure was used to access the desired final product 16. Scheme 5 depicts a representative synthetic strategy employed to obtain 4-amino substituted quinazoline Rock inhibitors from their corresponding quinazolinones. Briefly, a solution of Boc-protected compound 17 in DMF was treated with BOP and a desired amine under basic conditions to smoothly furnish compound 18. Boc-deprotection by TFA as the final step followed by HPLC purification gave the target Rock inhibitor 19. Taking phenylalanine (2), phenylglycine (20), and pyrrolidine 3carboxylic acid (3) based compounds as model systems, our SAR began with the exploration of the hinge-binding groups in these three subclasses. As shown in Table 1, the unsubstituted pyrazole (in 22, 24, and 28) was a better hinge binding moiety in terms of Rock-II inhibition24 compared to the 3-methyl pyrazole (23, 27), the pyridine (21, 25), and the amino-pyrimidine (26) groups. Interestingly, compounds in sub-class 3 always gave higher Rock potency than those in sub-classes 2 and 20, probably due to a more rigid pyrrolidine structure in 3. It is important to point out that,

in all three classes, these inhibitors also exhibited high PKA inhibition which was comparable to the Rock-II potency (Table 1). PKA is always selected as the control to monitor the general kinase selectivity in our development of Rock inhibitors.16–22 Various SAR studies were thus carried out for all three sub-classes (2, 3, and 20) in order to increase selectivity against PKA. Since most amino acids contain at least one chiral center, determination of the preferred chirality (absolute configuration) is of high importance. Thus, pairs of stereoisomers in the phenylalanine series were prepared and evaluated. Data in Table 2 for three pairs of inhibitors (ca. 32 vs 33, 34 vs 35, and 36 vs 37) demonstrated that there were no significant differences between stereoisomers for both Rock and PKA activities. Chiral inhibitors for series 30 and 31 were not prepared but similar results are expected based on our docking studies. Therefore, the chirality of applied amino acids will not be stressed in future optimizations. Using pyrazole as the hinge-binder we also explored the P-loop binding region (see Fig. 1) by installing various substituents on the phenyl ring. As shown in Table 2, these substitutions increased the Rock potency for phenylglycine series 30 (compared to unsubstituted 24), slightly improved the potency for the phenylalanine ser-

1595

S. Chowdhury et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1592–1599 Table 1 Enzyme inhibition data for three sub-classes of Rock inhibitors

O Het

O Het

NH N

21

(S)-2

22

(S)-2

23

(±)-20

24

(±)-20

25

(±)-3

NH N

NH2

2

Structure

NH

Het

N NH2

ID

O OMe

NH

20

3

IC50a (nM)

Het

PKA/Rock-II

Rock-II

PKA

52

49

0.9

12

7

0.6

98

nd



NH N

43

20

0.5

N

23

29

1.3

6

10

1.6

7

7

1.0

1

4

4.0

N

NH N NH N

NH2

a

26

(±)-3

27

(±)-3

28

(±)-3

N N NH N NH N

Values are means of two or more experiments with errors within 35% of the mean. nd = not determined.

ies 29 (compared to compound 22), but had almost no effects on the Rock activity for pyrrolidine series 31 (compounds 49–51, compared to unsubstituted 28). Interestingly, substitutions in series 29 at 3- or 4- position by both electron-donating and electron-withdrawing groups all yielded potent Rock inhibitors and simultaneously reversed selectivity (Rock vs PKA) slightly favoring RockII as compared to unsubstituted compound 22. Similar selectivity pattern was also observed for the phenylglycine series 30 (45–48 vs 24). However, it should be pointed out that these modified compounds, including Rock inhibitors based on the pyrrolidines (49– 51), are still highly potent PKA inhibitors. In our recent report on quinazoline-based Rock inhibitors that were not derived from amino acids,22 we disclosed that substitution at the 8-position in the bicyclic quinazoline ring significantly increased selectivity against PKA. In our efforts to improve selectivity in this aminoacid-derived series, 8-substituted quinazoline compounds 36/37 and 47/48 were also prepared. As demonstrated by data in Table 2, this substitution was not able to enhance selectivity and still produced very potent PKA inhibitors in these amino acid derived quinazoline Rock inhibitors. In addition, benzyl substitution to the NH group of the pyrrolidine ring (52–54) produced inhibitors with lower potency against both Rock and PKA as compared to the free NH analogs (Table 2). Again, similar activities for Rock and PKA were observed, indicating that this substitution could not increase selectivity against PKA either. More N1-substitutions were evaluated for inhibitors from all three sub-classes (Table 3). Similar to compounds 52–54 (Table 2), a relatively large substitution in series 55 (14) and series 57 (65) reduced kinase activities. On the other hand, small un-branched substitutions in general were well tolerated by both Rock and PKA

(58, 63, and 64). Remarkably, the small and charged quaternary ammonium salt N,N-dimethyl pyrrolidine 63 also exhibited high Rock-II inhibition although with only a slight improvement in PKA selectivity. In contrast, potent Rock-II compounds (as compared to the parent unsubstituted structure 24) were obtained in the phenylglycine series 56. Among all candidates 59–62, compound 59 showed impressive picomolar Rock inhibitory activity in enzyme assays with fair selectivity against PKA. Therefore, phenylglycine 56 is more tolerant to N-substitutions than phenylalanine 55. Further exploration by modification of the quinazoline 4-position (group Y in Table 4) was also undertaken. In phenylalanine series 66, substitutions with amines in general reduced the RockII potency. Amine substitutions terminating in an N,N-dimethylamino group (68) was proved to be most detrimental for binding affinity while substitutions with smaller amino groups gave slightly improved selectivity at a cost of lower Rock-II inhibition (69). Interestingly, stereoisomers 69 and 71 showed significant difference in PKA selectivity while Rock-II inhibition remained unchanged. In the b-aminoacid series 67, as a general trend, both Rock and PKA potency fell off drastically as the size of amine group increased. A small dimethylamino substitution (73) showed high Rock-II potency as well as a remarkably high selectivity over PKA (ca. 110-fold). For Rock inhibitors based on amino acid derived quinazolines (the three sub-classes developed in our labs), compound 73 was the best overall in terms of both Rock-II affinity and PKA selectivity. Data in Table 5 seem to suggest that large substitutions on the 4-position of the quinazoline ring hurt the affinity to both Rock and PKA, but more significantly to Rock inhibitions. To help understand the observed SAR and to guide further optimizations, representative inhibitors from each sub-scaffold were

1596

S. Chowdhury et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1592–1599

Table 2 Enzyme inhibition data for optimized compounds

N

N R1 HN

O

HN

O

R1 NH

NH R2

NH2

32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 a

(S)-29 (R)-29 (S)-29 (R)-29 (S)-29 (R)-29 (S)-29 (S)-29 (S)-29 (R)-29 (R)-29 (R)-29 (R)-29 (±)-30 (±)-30 (±)-30 (±)-30 (±)-31 (±)-31 (±)-31 (±)-31 (±)-31 (±)-31

N N

R2

4-Cl 4-Cl 4-F 4-F 3-F 3-F 4-OMe 4-OEt 3,4-di-OMe 3,4-di-F 3-Cl 3-F 3-CF3 4-Cl 3-Cl 4-Cl 3-Cl 4-F 4-Cl 4-CF3 4-OMe 3-F 2-F

R3

31

30

R1

R1

NH

NH2

29

Structure

O

N

N R2

ID

N HN

R3

H H H H OMe OMe H H H H H H H H H OMe OMe H H H H H H

IC50a (nM)

— — — — — — — — — — — — — — — — — H H H Bn Bn Bn

PKA/Rock-II

Rock-II

PKA

6 4 9 8 8 7 4 19 39 7 13 10 16 10 13 9 8 2 2.0 3.0 1.3 1.7 4.8

Values are means of two or more experiments with errors within 35% of the mean. nd = not determined.

Table 3 Enzyme inhibition data for N-substituted Rock inhibitors

N

O

R1

HN

N HN

O

NH

R1

N X

X 57

56

55

R1

R1 NH

N X

Structure

O

NH

N

ID

N HN

IC50a (nM)

X

PKA/Rock-II

Rock-II

PKA

239

nd



14

17

1.2

7

6

21

3.5

18

96

5.3

O 14

(R)-55

4-Cl

N H

58

(R)-55

4-Cl

N H

59

(±)-56

4-Cl

N H

60

(±)-56

4-Cl

N H

61

(±)-56

4-Cl

N

O N

O

NH2

O N H

NH2

1597

S. Chowdhury et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1592–1599 Table 3 (continued) ID

R1

Structure

IC50a (nM)

X

PKA/Rock-II

Rock-II

PKA

20

28

1.4

O

a

62

(±)-56

4-Cl

N H

63

(±)-57

H

N

4

23

5.7

64 65

(±)-57 (±)-57

4-OMe 4-OMe

NMe NBn

5 88

26 117

5.2 1.3

NH2

Values are means of two or more experiments with errors within 35% of the mean. nd = not determined.

Table 4 Enzyme inhibition data for 4-substituted quinazoline Rock inhibitors

N

Y

N

R1

HN

Y

N

N

N 66

ID

a

Structure

R1

R1

HN

N NH2

NH 67

IC50a (nM)

Y

68

(R)-66

3-F

N

69

(R)-66

3-F

N

70

(R)-66

3-F

N

N

OH

OH OH

PKA/Rock-II

Rock-II

PKA

627

1160

1.8

27

44

1.6

159

nd

-

27

284

10.5

71

(S)-66

3-F

N

72

(S)-66

3-F

N

67

259

3.4

73

(±)-67

H

N

2.5

279

110

74

(±)-67

4-F

N

71

443

6.2

75

(±)-67

4-F

N H

371

884

2.4

N

Values are means of two or more experiments with errors within 35% of the mean. nd = not determined.

docked into a homology model of human Rock-II.16–22 Shown in Figure 2 are the docking modes of the phenylalanine based inhibitor 32 (Fig. 2A) and the pyrrolidine based compound 50 (Fig. 2B). Similar modes were also observed for the phenylglycine based Rock inhibitors. Several identical interactions were observed for both compounds 32 and 50, and these interactions were also obtained in the binding modes of non-amino acid based quinazoline Rock inhibitors.22 For example, the pyrazole moiety formed two hydrogen bonds with residues Glu170 and Met172 in the hinge region. The quinazolinone amide carbonyl group H-bound to the side chain amino group of Lys121, and the amide NH group formed an H-bond to residue Asp232. An important hydrophobic interaction was formed between the phenyl moiety and the hydrophobic binding pocket under the P-loop. Unlike the binding modes of non-amino acid based quinazoline Rock inhibitors, additional H-bonds were observed in these amino acid based quinazoline Rock inhibitors. The NH on the pyrrolidine ring of compound 50

formed an H-bond to residue Asn219, while the NH2 group of inhibitor 32 H-bound to both residue Asn219 and residue Asp232. The docking modes in Figure 2 explain very well the observed SAR. For example, a large Y substitution on series 66 and 67 reduced binding affinity because there was no space in the ATP pocket to accommodate such a large group. Functionalization to the NH2 group of series 56 and the pyrrolidine NH of series 57 did not affect much of the binding potency since these amino groups were orientated toward outside of the ATP pocket. On the other hand, functionalization to the NH2 of phenylalanine series 55 by a large group reduced the Rock potency significantly (14) because the NH2 group in this sub-scaffold bound toward the inside of the ATP pocket and there was not enough space to accommodate a large group. In addition to Rock-II inhibition, the Rock-I inhibitory activity was also assessed for amino acid derived quinazoline compounds. As shown in Table 5, all tested compounds were found to be panRock inhibitors with very similar potency for both Rock isoforms.

1598

S. Chowdhury et al. / Bioorg. Med. Chem. Lett. 23 (2013) 1592–1599

tude higher than their enzyme inhibition IC50 values, indicating a necessity to further improve the cell permeability of these Rock inhibitors. Interestingly, the two compounds having the highest cell potency (50, 64) are both from the cyclic b-amino acid pyrrolidine-quinazoline scaffold. Considering that compounds with the highest PKA selectivity (71, 73) are also from the same sub-class, this pyrrolidine-quinazoline scaffold will therefore be the focus for future optimizations. In vitro DMPK were assayed for most of these amino acid derived Rock inhibitors (Table 5). Basically, most inhibitors exhibited high stability in human microsomes. The low stability for 38, 59, and 64 is probably due to methoxy substitutions on the phenyl ring, or the demethylation on the pyrrolidine ring, or a combination. The rat microsomal stability was generally lower than that in human microsomes with the exception of compound 50, which exhibited an extremely high stability (t1/2 = 105 min) in rats. It is important to note that high stability is normally associated with free NH2 or NH groups present in the compounds. Remarkably, the stability of inhibitor 61 was extremely high in both human and rat microsomes (Table 5), which was probably due to its high CYP-450 inhibitions. As shown in Table 5, at 10 lM compound 61 strongly inhibited three of the four selected CYP-450 enzymes 1A2, 2C9, and 2D6, (P90% at 10 lM) and also moderately inhibited 3A4 (63% at 10 lM). Except for 48 and 61, the general CYP-450 inhibition profiles of inhibitors in Table 5 were fair to good. Indeed, most compounds had low inhibition of 3A4, the most important CYP enzyme. In summary, aminoacid-quinazoline chimeric scaffold-based kinase inhibitors with phenylglycine, phenylalanine, and pyrrolidine-2-carboxylic acid structures were synthesized and their Rock-II and PKA inhibitory properties were evaluated. These compounds generally exhibited pan-Rock inhibition as well as high PKA inhibitory properties although good selectivity against PKA could be achieved in a few cases. The overall best inhibitors (potency, stability, and CYP-450 inhibition profiles) developed for this scaffold so far were compounds 28, 32, 50, 59, and 64. These compounds might find applications that do not require high PKA selectivity, such as use as topical anti-glaucoma therapeutics.

Figure 2. Docking to a homology model of the Rock-II kinase for a phenylalanine based Rock inhibitor 32 (A), and a pyrrolidine based Rock inhibitor 50 (B).

While selectivity against PKA could be achieved in only a few cases, these Rock inhibitors were found to be indeed selective against many other kinases such as JNK1/2/3, p38a, MRCKa etc. (IC50 >1 lM, data not shown). The cell activities for those inhibitors which showed high enzyme potency were also evaluated in ppMLC target modulation cell assays.25 As shown in Table 5, compounds 50 and 64 exhibited high cell potency with IC50 values of
Lihat lebih banyak...

Comentários

Copyright © 2017 DADOSPDF Inc.