Optimization of Blood-Brain Barrier Permeability with Potent and Selective Human Neuronal Nitric Oxide Synthase Inhibitors Having a 2-Aminopyridine Scaffold

Article

pubs.acs.org/jmc

Cite This: J. Med. Chem. XXXX, XXX, XXX −XXX

Optimization of Blood−Brain Barrier Permeability with Potent and

Selective Human Neuronal Nitric Oxide Synthase Inhibitors Having a

2‑Aminopyridine Scaﬀold

Ha T. Do,^†Huiying Li,^‡Georges Chreiﬁ,^‡Thomas L. Poulos,*^,‡and Richard B. Silverman*^,†

^†Department of Chemistry, Department of Molecular Biosciences, Chemistry of Life Processes Institute, Center for Molecular

Innovation and Drug Discovery, Center for Developmental Therapeutics, Northwestern University, 2145 Sheridan Road, Evanston,

Illinois 60208-3113, United States

^‡Departments of Molecular Biology and Biochemistry, Pharmaceutical Sciences, and Chemistry, University of California, Irvine,

California 92697-3900, United States

S

* Supporting Information

ABSTRACT: Eﬀective delivery of therapeutic drugs into the

human brain is one of the most challenging tasks in central

nervous system drug development because of the blood−brain

barrier (BBB). To overcome the BBB, both passive

permeability and eﬄux transporter liability of a compound

must be addressed. Herein, we report our optimization related

to BBB penetration of neuronal nitric oxide synthase (nNOS)

inhibitors toward the development of new drugs for

neurodegenerative diseases. Various approaches, including

enhancing lipophilicity and rigidity of new inhibitors and

modulating the pK_aof amino groups, have been employed. In addition to determining inhibitor potency and selectivity, crystal

structures of most newly designed compounds complexed to various nitric oxide synthase isoforms have been determined. We

have discovered a new analogue (21), which exhibits not only excellent potency (K_i< 30 nM) in nNOS inhibition but also a

signiﬁcantly low P-glycoprotein and breast-cancer-resistant protein substrate liability as indicated by an eﬄux ratio of 0.8 in the

Caco-2 bidirectional assay.

Neuronal nitric oxide synthase (nNOS) has been validated

as a promising therapeutic target in the development of new

treatments for neurodegenerative diseases.⁸⁻¹⁰In brain, nitric

oxide (NO) produced by nNOS participates in neuronal

transmission.¹¹The overproduction of NO in cells, however, is

harmful. In particular, excess NO formed by upregulated

nNOS in the CNS can cause excessive nitration and

nitrosylation of proteins, leading to their misfolding and

aggregation.¹²Additionally, the reaction of NO with super-

oxide anion creates a strongly oxidizing reagent, peroxynitrite,

which, together with reactive oxygen species,^13,14damages

DNA and causes lipid peroxidation. These processes lead to

nerve cell death and impairment in neuronal transmission,

which are commonly observed in symptoms of neuro-

degenerative diseases.^15,16Limiting NO production through

inhibition of nNOS, therefore, could be an important approach

to protect neurons and slow the progression of neuro-

degenerative diseases.^17,18

INTRODUCTION

■

Neurodegenerative diseases, such as Alzheimer’s, Parkinson’s,

and Huntington’s diseases, are characterized by a gradual

degeneration and death of neurons in the central nervous

system (CNS), causing problems in muscular movements and

mental functioning of patients. Despite unmet medical needs,

comprehensive treatments for these diseases are still very

limited.^1,2One of the most diﬃcult challenges in CNS drug

development is eﬀective delivery of therapeutic drugs into the

human brain, mainly because of the blood−brain barrier

(BBB) located at the interface of blood vessels and brain

tissues.³The BBB is composed of a layer of endothelial cells

with tight junctions that prevents the access of external toxins

and therefore protects the brain and preserves its optimal

physiological environment. This cell layer, however, also limits

the access of therapeutic drugs into the brain.⁴The major

pathway for CNS drugs to cross the BBB is by passive diﬀusion

through its lipid membrane. In addition to the tight junctions

of endothelial cells, high expression levels of eﬄux transporters,

especially P-glycoprotein (P-gp) and breast-cancer-resistant

protein (BCRP), contribute greatly to the limited brain

exposure of CNS drugs.⁵Consequently, it becomes necessary

in CNS drug development to increase passive permeability and

lower P-gp- and BCRP-mediated eﬄux.^6,7

nNOS is a homodimeric enzyme with each monomer

containing one C-terminal reductase domain and one N-

terminal oxygenase domain. The C-terminal reductase domain

Received: December 29, 2018

A

DOI: 10.1021/acs.jmedchem.8b02032

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Figure 1. Structural modiﬁcations of lead compound 1: (i) enhancing rigidity with a C−C triple bond; (ii) enhancing lipophilicity and rigidity by

incorporating a pyrrolidine ring; (iii) enhancing lipophilicity by incorporating more ﬂuorine atoms into the middle linker; (iv) diﬂuorobenzene

linker incorporated with a C−C triple bond; (v) diﬂuorobenzene linker incorporated with a pyrrolidine ring; and (vi) modulating pK_aof the amino

tail group.

consists of nicotinamide adenine dinucleotide phosphate

(NADPH), ﬂavin adenine dinucleotide, and ﬂavin mono-

nucleotide binding sites, whereas the N-terminal oxygenase

domain contains a structural zinc at the dimer interface, a

tetrahydrobiopterin (H₄B) cofactor, and a heme as the

catalytic center, which is also the substrate, ^L-arginine (L-

Arg), binding site. These two domains are connected to each

other by a calmodulin (CaM) binding motif. When CaM is

bound in the presence of the calcium inﬂux, electron ﬂow is

facilitated from the reductase domain to the heme active site in

the oxygenase domain, where ^L-Arg gets oxidized in a two-step

reaction to generate ^L-citrulline and release NO.^17,19

Competing against L-Arg binding with a compound at the

active site is one of the fundamental approaches to inhibit

nNOS.¹⁸The challenges of this task involve not only inhibitor

potency but also selectivity for nNOS over both endothelial

NOS (eNOS) and inducible NOS (iNOS), the two isoforms

that share very similar structural features in the active site to

that of nNOS.^20,21It is necessary to avoid over-inhibition of

these two nitric oxide synthase (NOS) isoforms since eNOS

inhibition can result in cardiovascular failure, whereas iNOS

inhibition can cause a disruption in the immune system.²²

In recent years, our eﬀorts in achieving nNOS inhibitors

with excellent potency and high isoform selectivity have led to

a promising class of molecules bearing a 2-aminopyridine

scaﬀold. Using this molecular scaﬀold, we have obtained nNOS

inhibitors that exhibit excellent activity at concentrations in the

sub-30 nM range.^17,23,24Our ﬁrst generation of nNOS

inhibitors bearing a 2-aminopyridine scaﬀold, however, showed

poor predicted permeation through the BBB as revealed by

very little Caco-2 permeability.²⁵Recently, we have been able

to improve cell membrane permeability of our 2-aminopyridine

nNOS inhibitors, while retaining their high inhibitory activity.

In our previous report, we obtained a new lead compound (1,

Figure 1), which shows excellent potency and selectivity to

human nNOS (K_{i (hnNOS)}= 30 nM; hnNOS/heNOS = 2799)

and displays an eﬄux ratio (ER) of 5.9 in the Caco-2 assay.²⁶

To move forward in CNS drug development, cell membrane

permeability of these 2-aminopyridine nNOS inhibitors must

be further improved with a required ER of <2.5 for being a

likely CNS (+) drug.^7,27

Here, we report our optimization toward improving cell

membrane permeability and reducing P-gp and BCRP

substrate liability of 2-aminopyridine nNOS inhibitors, using

1 as the lead compound for numerous chemical modiﬁcations.

Insights into understanding the structural eﬀects on the activity

and permeability of analogues have been obtained through

various medicinal chemistry approaches, including enhancing

the lipophilicity and rigidity of new analogues, along with

modulating the pK_aof basic amino groups (Figure 1). These

structural modiﬁcations have been centralized on enhancing

the disposition of nNOS inhibitors into the brain while

preserving their potency and selectivity comparable to those of

1. Moreover, in this work, we aim, for the ﬁrst time, to

investigate the inhibition studies of potential compounds using

all human NOS isoforms, which helps to provide not only a

direct comparison of isoform selectivity but also more robust

data for clinical studies, if any, of the studied nNOS inhibitors

should be advanced to a later stage of drug development.

B

DOI: 10.1021/acs.jmedchem.8b02032

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

a

Scheme 1. Synthesis of 2 and 3

a

Reagents and conditions: (a) (i) n-BuLi 1.6 M/tetrahydrofuran (THF), THF, −78 → −20 °C, 15 min, (ii) 23, THF, −78 → −20 °C, 20 min; (b)

25a or 25b, Pd(PPh₃)₄, CuI, triethylamine (TEA)/dimethylformamide (DMF) (9:1), microwave, 120 °C, 30 min; (c) 20% triﬂuoroacetyl (TFA),

CH₂Cl₂, room temperature (RT), 1 h; (d) NH₂OH·HCl, EtOH/H₂O (2:1), 100 °C, 20 h.

a

Scheme 2. Synthesis of Pyrrolidinoalkynes

a

Reagents and conditions: (a) BH₃1 M/THF, THF, 0 °C → RT; (b) Dess−Martin periodinane, CH₂Cl₂, RT, 3 h; (c) 31, MeOH, RT, 15 h.

a

Scheme 3. Synthesis of Pyrrolidine Analogues 4−9

a

Reagents and conditions: (a) 32a−c, Pd(PPh₃)₄, CuI, TEA/DMF (9:1), microwave, 120 °C, 30 min; (b) 20% TFA, CH₂Cl₂, RT, 1 h; (c) Pd/C,

H₂, MeOH, RT, 20 h; (d) (i) HCHO 37% in H₂O, (ii) NaBH₄, MeOH; (e) NH₂OH·HCl, EtOH/H₂O (2:1), 100 °C, 20 h.

designed that increased lipophilicity and rigidity of lead

compound 1 (Figure 1). Rigidity was increased by adding a

C−C triple bond to the tail end to give analogues 2 and 3.

Analogues 4−9 with a pyrrolidine ring in the tail introduce

RESULTS AND DISCUSSION

■

Overall Strategy. While lead compound 1 binds tightly to

nNOS and exhibits excellent n/e selectivity, 1 has a high ER

(Table 3). To increase brain penetration, molecules were

C

DOI: 10.1021/acs.jmedchem.8b02032

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

a

Scheme 4. Synthesis of Di- and Triﬂuorobenzene Analogues 10−14

a

Reagents and conditions: (a) (i) n-BuLi 1.6 M/THF, THF, −78 → −20 °C, 15 min, (ii) 36a−e, THF, −78 → −20 °C, 20 min; (b) 25b,

Pd(PPh₃)₄, CuI, TEA/DMF (9:1), microwave, 120 °C, 30 min; (c) Pd/C, H₂, MeOH, RT, 20 h; (d) NH₂OH·HCl, EtOH/H₂O (2:1), 100 °C, 20

h.

a

Scheme 5. Synthesis of Analogues 15−18 with Enhanced Lipophilicity and Rigidity

a

Reagents and conditions: (a) 25a,b or 32a,b, Pd(PPh₃)₄, CuI, TEA/DMF (9:1), microwave, 120 °C, 30 min; (b) 20% TFA, CH₂Cl₂, RT, 1 h; (c)

Pd/C, H₂, MeOH, RT, 20 h; (d) (i) HCHO 37% in H₂O, (ii) NaBH₄, MeOH; (e) NH₂OH·HCl, EtOH/H₂O (2:1), 100 °C, 20 h.

greater lipophilicity and reduce the number of rotatable bonds.

Diﬀerent enantiomers of the pyrrolidine ring were also studied

to investigate the eﬀect of chirality. To increase permeability

and decrease metabolism, the pK_aof the tail was decreased by

adding a ﬂuorine atom to the C4 position on the pyrrolidine

ring (8 and 9). The lipophilicity of the new analogues was

enhanced by incorporation of additional ﬂuorine atoms onto

the middle ﬂuorobenzene linker of 1 (compounds 10−14).

Compounds with an X-,Y-diﬂuorobenzene linker (Figure 1)

were identiﬁed as the best in both permeability and structural

binding properties. Rigidity in the tail was built into 10 to give

15 and 16. The optimal X-,Y-diﬂuorobenzene linker was then

merged with the pyrrolidine tail in the 4−9 series to give 17

and 18. Finally, the pK_aof the cyclic amino group, i.e.,

pyrrolidine ring, in the tail chain was modulated by diﬀerent

heterocycles, including morpholine (19) and azetidine (21), as

well as introducing an electron-withdrawing group into the

pyrrolidine ring (20). The predicted physicochemical proper-

ties of these newly designed compounds can be found in Table

S3 in the Supporting Information. All new analogues were

tested for nNOS inhibition and selectivity over eNOS and

iNOS, whereas cell membrane permeability was investigated

using a parallel artiﬁcial membrane permeability for blood−

brain barrier (PAMPA-BBB) assay. Compounds with high

potency, selectivity, and permeability were further examined in

a Caco-2 bidirectional assay to evaluate their P-gp and BCRP

substrate liability.

Chemistry. The synthesis of compounds 2 and 3 with

enhanced rigidity in the tail chain is shown in Scheme 1.

Deprotonation of pyrrole-protected 2,4-dimethylpyridine 22

by n-BuLi followed by a reaction of the generated anion with

electrophile 23 provided intermediate 24 with the 2-amino-

pyridine head and the middle linker coupled. Sonogashira

coupling of 24 with either N-Boc-N-methyl-propargylamine

(25a) or 3-dimethylamino-1-propyne (25b) aﬀorded alkynes

carrying a Boc-protected secondary (26a) or tertiary (26b)

amine. Boc-deprotection of 26a followed by pyrrole depro-

D

DOI: 10.1021/acs.jmedchem.8b02032

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

a

Scheme 6. Synthesis of 19−21

a

Reagents and conditions: (a) 44a or 44b, Pd(PPh₃)₄, CuI, TEA/DMF (9:1), microwave, 120 °C, 30 min; (b) 44c, Pd(PPh₃)₂Cl₂, CuI, PPh₃,

DEA/DMF (1:1), microwave, 120 °C, 20 min; (c) 45a → 46a: (i) 20% TFA, CH₂Cl₂, RT, 1 h, (ii) Pd/C, H₂, MeOH, RT, 20 h, (iii) HCHO 37%

in H₂O, NaBH₄, MeOH; (d) 45b,c → 46b,c: (i) Pd(OH)₂/C, H₂(1 atm), MeOH, RT, 20 h, (ii) HCHO 37% in H₂O, NaBH₄, MeOH; (e)

NH₂OH·HCl, EtOH/H₂O (2:1), 100 °C, 20 h.

tection yielded target compound 2, whereas pyrrole depro-

tection of 26b generated compound 3.

previously synthesized 37a with diﬀerent alkynes (25a,b and

32a,b) yielded intermediates 39 and 41−43. Boc-deprotection

of 39 generated intermediate 40, which underwent pyrrole

deprotection along with 41 to yield the two desired

compounds (15 and 16) bearing the alkyne amino group in

the tail chain. On the other hand, 42 and 43 underwent a

sequence of reactions, including Boc-deprotection, hydro-

genation with Pd/C, methylation with formaldehyde, and

pyrrole deprotection, to give the target products (17 and 18)

containing a pyrrolidine ring in their tail chain.

The synthesis of pyrrolidine analogues 4−9 was carried

through the preparation of pyrrolidinoalkynes 32a−c (Scheme

2). The pure enantiomers of these pyrrolidinoalkynes were

directly synthesized from their corresponding aldehydes (30a−

c) using Seyferth−Gilbert homologation. Aldehydes 30a−c

were prepared from the oxidation of their corresponding

alcohols, obtained from either commercial sources (29a and

29b) or by being synthesized (29c) from a carboxylic acid

precursor (28). Scheme 3 shows the synthetic routes for 4−9

from 32a−c. Sonagashira coupling of intermediate 24 with

diﬀerent pyrrolidinoalkynes (32a−c), followed by a sequence

of Boc-deprotection and hydrogenation, gave intermediates

34a−c. Pyrrole deprotection of these intermediates provided

secondary amine analogues 4, 6, and 8. In a separate pathway,

methylation of the secondary amino group using form-

aldehyde/NaBH₄, followed by a removal of the pyrrole-

protecting group, generated tertiary amine analogues 5, 7, and

9.

The synthesis of analogues containing di- and triﬂuor-

obenzene linkers (10−14) was started with the preparation of

linker components (36a−e) from commercially available

sources (see the Supporting Information). Following a general

synthetic route for nNOS inhibitors as shown in Scheme 1, the

synthesized linkers (36a−e) were coupled with pyrrole-

protected 2-aminopyridine head 22 through carbon−carbon

bond formation using n-BuLi. Generated intermediates (37a−

e) then underwent Sonogashira coupling with alkyne 25b to

keep the tail chain the same as that of lead compound 1.

Alkyne reduction of intermediates 38a−e followed by a pyrrole

deprotection yielded desired products 10−14 (Scheme 4).

To obtain the highest possible improvement in cell

membrane permeability, a new set of compounds with both

enhanced lipophilicity and increased rigidity were synthesized.

New analogues were designed to have a diﬂuorobenzene

middle linker and a tail chain containing either a pyrrolidine

ring or an alkyne amino group in the tail chain to increase their

lipophilicity or reduce the number of rotatable bonds,

respectively. The synthesis of these analogues (15−18) is

shown in Scheme 5. Sonogashira coupling reactions of

The synthesis of compounds 19−21 to modulate the pK_aof

the amino group in the tail chain ﬁrst involved the preparation

of the corresponding alkyne-functionalized tail chains (44a−

c), whose syntheses can be found in the Supporting

Information. Sonogashira coupling again was used to attach

these tail chains to intermediate 37a. The target compounds

(19−21) were obtained by subsequent reactions including (i)

Boc/Cbz deprotection, (ii) hydrogenation of the alkyne, (iii)

methylation of the secondary amine to the tertiary amine by

HCHO/NaBH₄, and (iv) pyrrole deprotection. It is worth

mentioning that these reactions worked eﬃciently for the

synthesis of 19, whereas the removal of the Cbz group in the

synthesis of 20 and 21 required the use of Pd(OH)₂/C under

H₂gas. The use of Pd/C, H₂to remove the Cbz group did not

yield the desired products, even at a pressure up to 110 psi

hydrogen gas. Additionally, Sonogashira coupling of azetidine

alkyne 44c required the use of a diﬀerent Pd catalyst and base

(i.e., Pd(PPh₃)₂Cl₂and diethylamine, respectively) to obtain a

reasonable yield and less inseparable byproducts (Scheme 6).

Biological Activity. The inhibitory activity and selectivity

of new analogues 2−21 were determined using the NO

hemoglobin (Hb) capture assay, and the results are

summarized in Table 1 along with those of lead compound

1. These compounds were ﬁrst tested against the more easily

puriﬁed rat and human nNOS and murine iNOS to evaluate

potency and selectivity. Relative to 1 with a K_{i (hnNOS)}= 30 nM,

eight of the new compounds did not exhibit a signiﬁcant

decrease in potency, eight had a 2- to 3-fold decrease, and four

exhibited a 4- to 5-fold decrease in potency. The parallel

artiﬁcial membrane permeability for blood−brain barrier

(PAMPA-BBB) assay was used to evaluate the eﬀects of our

E

DOI: 10.1021/acs.jmedchem.8b02032

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

the tail amine (9), introducing more ﬂuorines (10−14), and

combining modulation of pK_a, lipophilicity, and additional

ﬂuorines (15−21), help improve permeability. It is evident

that subtle changes in structure can result in signiﬁcant

permeability changes. For example, the addition of a single

methyl group to 4 and 6 to give 5 and 7, respectively, results in

an approximately 2-fold increase in permeability.

The most promising compounds were selected for more

detailed analysis by comparing selectivity against the other

NOS isoforms, and the results are shown in Table 2.

Eﬄux Transporter Liability. Because compounds 18 and

21 stand out as the most selective nNOS inhibitors (Table 2)

in addition to having excellent cell permeability properties

(Table 1), these two compounds were selected for further

studies on the potential for BBB penetration. P-glycoprotein

(P-gp) and breast-cancer-resistant protein (BCRP) are eﬄux

transporters that are highly expressed at the BBB to remove

harmful molecules, including potential drugs, out of the

brain.^5,30Evaluation of the potential of a compound as an

eﬄux transporter substrate, therefore, is one of the crucial steps

in CNS drug development. The P-gp and BCRP substrate

liability of a compound can be evaluated through an eﬄux ratio

(ER) obtained from a Caco-2 bidirectional assay,³¹which

measures the ability of compounds to cross a monolayer of

colon cells with expressed P-gp and BCRP from two directions,

either from apical to basal (A → B) or from basal to apical (B

→ A) wells. An ER is then determined by a ratio of the

apparent permeability (P_app) of B → A over A → B.

Compounds with an ER larger than 3 are often considered

as substrates of P-gp and BCRP with limited retention in the

brain.⁷

Table 1. Rat and Human nNOS Potency and Eﬀective

Permeability (P_e) in the PAMPA-BBB Assay of 1−21

a

K_i(nM)

rat

human

nNOS

cmpd. nNOS

hn/rn P_e(10⁻⁶cm s⁻¹) (PAMPA-BBB)

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

26

32

58

31

26

28

33

65

128

19

60

79

47

45

27

81

27

13

62

55

26

30

58

124

68

52

59

1.2

1.8

2.1

2.2

2.0

2.1

2.5

2.0

1.6

1.5

0.6

2.0

1.2

1.0

1.6

1.0

1.4

1.9

1.4

1.6

0.9

14.80 0.69

15.52 0.02

18.76 0.03

5.18 0.05

13.30 1.44

5.52 0.22

13.61 0.24

12.34 0.06

18.90 0.08

15.90 0.48

15.38 0.50

16.35 0.44

18.62 0.20

18.70 0.20

18.62 0.48

18.64 0.31

17.91 0.60

17.00 1.00

21.10 1.12

17.67 1.80

16.32 0.35

84

129

204

29

36

157

58

46

43

84

37

21

89

87

23

a

K_ivalues were calculated from the IC₅₀values of a dose−response

curve using the Cheng−Prusoﬀ equation. Six- to nine concentrations

were tested, and the IC₅₀value was calculated from an average of at

least two duplicates. The standard errors were less than 10%. The

ratio hn/rn was determined to evaluate the potential translation of

these inhibitors from preclinical data to a clinical study. This ratio was

aimed to be as close to 1.0 as possible so that there would be little to

no signiﬁcant diﬀerence in the amount of inhibitors used in rat and

human dosage. The eﬀective permeability (P_e) was obtained after 17

h incubation of a compound through a layer of porcine brain lipid and

was reported as an average of triplicate with a standard deviation.

As shown in Table 3, lead compound 1 has a high eﬄux

ratio (ER = 5.9) and is, therefore, not a good candidate as a

Table 3. Caco-2 Apparent Permeability and Eﬄux Ratio

(ER) of Selected nNOS Inhibitors with Control

Compounds

apparent permeability

a

(P_app, 10⁻⁶cm s⁻¹

)

structural modiﬁcations on cell membrane permeability of 2−

21, which helps to understand the structure−permeability

relationship of these compounds. From the PAMPA-BBB

assay, a compound is predicted to have a good BBB

penetration and is classiﬁed as a CNS (+) molecule if its

eﬀective permeability (P_e) is larger than 4.0 × 10⁻⁶cm

s⁻¹.^26,28,29The results in Table 1 reveal that compounds 2−21

exhibit a predicted CNS (+) characteristic, in which 4−8

exhibit lower P_ethan that of lead compound 1, whereas 2, 3,

and 9−21 display comparable or higher permeability than that

of 1. In most cases, the strategies we employed, such as

increasing the rigidity of the tail (2 and 3), altering the pK_aof

compd.

mean A → B

mean B → A

eﬄux ratio

1

18

21

9.2 0.3

1.1 0.1

17.05 0.08

37.18

54.2 17.6

2.3 0.2

13.71 0.07

20.39

5.9

2.1

0.8

0.55

1.47

b

metoprolol

c

atenolol

0.39

<0.17

0.58

d

erythromycin

13.39

>78.76

a

b

c

Apparent permeability value. High permeability control. Low

d

permeability control. High eﬄux control.

Table 2. Potency and Selectivity of Selected Compounds with Other NOS Isoforms

K_i(nM)

selectivity

cmpd.

human nNOS

human eNOS

human iNOS

murine iNOS

hnNOS/heNOS

hnNOS/hiNOS

rnNOS/miNOS

1

10

14

15

17

18

21

30

29

46

43

37

21

23

83 976

35 028

50 280

16 960

35 758

25 548

21 980

3501

1312

2509

3059

1635

2222

1780

3857

823

2756

1208

1093

394

966

1216

956

117

45

54

71

44

149

43

47

112

84

99

2119

3028

2187

1290

2060

106

77

79

F

DOI: 10.1021/acs.jmedchem.8b02032

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Figure 2. Inhibitor binding environment in the structure of (A) hnNOS-2 (Protein Data Bank (PDB) code: 6NG1) and (B) hnNOS-3 (PDB code:

6NG2). The omit Fo − Fc electron density for the bound ligand is displayed at the contour level of either 2.5σ (regular) or 3.5σ (polder). Major

hydrogen bonds are depicted with dashed lines. The heme propionates are labeled in red. This and all other structural ﬁgures in this study have

similar representations and were prepared with PyMol (www.pymol.org).

CNS drug. Compound 18 gives a substantially improved ER,

although membrane penetration is not ideal, whereas

compound 21 has excellent membrane penetration and gives

a very low ER of 0.8, which compares favorably with the CNS

drug, metoprolol, with an ER of 0.55. For comparison, two

additional non-CNS drugs are listed in Table 3, both of which

exhibit much higher ERs than those of compound 21.

Structural Analysis: Increasing Lipophilicity (Com-

pounds 4−7). To enhance the lipophilicity of 1 and,

therefore, potentially increase its permeability, compounds

4−9, with a pyrrolidine ring at the tail, were designed and

synthesized. Although blocking the H-bonding capability of the

pyrrolidine N atom with a methyl group, changing 4 and 6 to 5

and 7, makes drastic changes in permeability, there is little

eﬀect on interactions with the enzyme. Compounds 4 and 5

bind very similarly to that of 1 with the tail pyrrolidine

similarly positioned to the tertiary amine of 1 for interactions

with heme propionate A (Figure 3A,B). Changing the chirality

of the 4 pyrrolidine, however, to give 6 does result in a

substantially diﬀerent binding mode to hnNOS (Figure 3C).

Now the pyrrolidine with a diﬀerent chirality can no longer

properly interact with the heme propionates but approaches

Asp602 instead. This also requires a fairly substantial

repositioning of the central ﬂuorobenzene, which is now

roughly perpendicular to the orientation in 4. These

diﬀerences have little eﬀect on K_i(Table 1). That these

compounds can have such diﬀerent binding modes yet retain

similar potency is likely the result of multiple ways the tail

amino group can be stabilized by either the heme propionates

or the Asp602 carboxylate. The binding of compounds 4−7 to

rnNOS shares more or less the same bent down mode (Figure

S3) with the pyrrolidine interacting with heme propionates,

which is in agreement with their similar K_ivalues (Table 1).

Binding of this series of compounds to heNOS provided

some surprises. For compounds 4, 6, and 8, two molecules of

inhibitor bind (Figure S4). Using 6 as an example (Figure

S4B), one 6 molecule binds as expected in the active site but

the second 6 molecule binds in the H₄B site such that the

aminopyridine of the inhibitor takes the place of the H₄B. This

requires Arg365, which normally interacts with the H₄B, to

swing out of the way to make room for the second inhibitor

molecule. The movement of Arg365 also provides an opening

for Zn²⁺to bind and coordinate with Asp369 and His461 of

molecule B of the dimer. We have observed this type of two

inhibitor binding together with formation of the Zn²⁺site in

previous studies.³²The new observation from the present

study is that the binding of Zn²⁺is clearly associated with the

change in Arg365 and not the displacement of H₄B by a

second inhibitor molecule since in the heNOS-9 complex

(Figure S4D), a second inhibitor molecule does not bind but

Structural Analysis: Lead Compound 1. Our early

crystallographic work focused on rat nNOS (rnNOS) and

bovine eNOS (beNOS), but we have more recently focused on

the human isoforms (hnNOS and heNOS). In the current

study, a total of 45 crystal structures using the various nNOS

and eNOS isoforms have been solved. Here, we focus on a few

selected human isoform structures with greater details

provided in the Supporting Information. The crystal structure

of lead compound 1 complexed with hnNOS and heNOS was

previously published, and here, we provide a brief summary

relevant to the current study. As with all our inhibitors, the

aminopyridine end of 1 stacks over the heme and H-bonds

with the buried conserved active site Glu. However, the mode

of binding of the rest of 1 is quite diﬀerent in hnNOS and

heNOS (Figure S1). The central ﬂuorobenzene in hnNOS

points “up” toward Tyr567, whereas the tail tertiary amine

interacts with the H₄B and heme propionate A (Figure S1).

The location of the ﬂuorobenzene requires that Gln483 adopts

a new rotamer (the “out” rotamer), which places the side chain

amide parallel and in contact with the ﬂuorobenzene group. In

heNOS, the electron density of 1 is poorly deﬁned, but it

nevertheless is clear that the central ﬂuorobenzene ring does

not point up but adopts a “bent down” binding mode and lies

parallel to the heme. The tertiary amino tail density is poorly

deﬁned, suggesting weaker interactions with the heme

propionates than in hnNOS. This diﬀerence in the orientation

of the central ﬂuorobenzene groups is observed in many of the

structures solved as part of this study.

Structural Analysis: Rigidifying Lead Compound 1

(Compounds 2 and 3). Rigidifying the tail end of 1 as in

compounds 2 and 3 has little eﬀect on the binding mode to

hnNOS (Figure 2). The approximately 4-fold lower aﬃnity of

3 relative to that of 1 is most likely associated with the inability

of the more rigid tail to maximize electrostatic interactions

with heme propionate A. The binding mode changes for 2 and

3 in rnNOS are described in Figure S2.

G

DOI: 10.1021/acs.jmedchem.8b02032

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

8 and 9, exhibit much better permeability than that of their

parent compounds, 4 or 5, but show a 2-fold and a 4-fold drop,

respectively, in both rnNOS and hnNOS inhibitory activities

compared with 4 and 5 (Table 1). This reduction is very likely

the result of a decrease in the pK_aof the tail amino group,

thereby diminishing electrostatic interactions with the heme

propionates (Figure 4), which we also found in an earlier

study.²⁶The addition of the ﬂuorine atom to 5 to give 9 also

causes a substantial change in binding mode to hnNOS.

Although the pyrrolidine of 5 is stabilized by the heme

Figure 3. Inhibitor binding environment in the structure of (A)

hnNOS-4 (PDB code: 6NG4), (B) hnNOS-5 (PDB code: 6NG5),

and (C) hnNOS-6 (PDB code: 6NG6). Note that the binding mode

is changed for 4 and 6 because of the chirality diﬀerence.

Arg365 still moves, enabling Zn²⁺to bind. It is quite possible

that this Zn²⁺site is the one associated with the well-known

ability of transition metals to inhibit NOS activity.³³The

binding of Zn²⁺requires disruption of the Arg365−H₄B

interaction, and this interaction is known via mutagenesis

studies to be essential for NOS activity.³⁴

Addition of Fluorine to Decrease pK_a(Compounds 8

and 9). Structural modiﬁcations of compounds that improve

permeability may not correlate with inhibitor potency.

Analogues that carry a ﬂuorine atom on the pyrrolidine ring,

Figure 4. Inhibitor binding environment in the structure of (A)

rnNOS-8 (PDB code: 6NGQ), (B) rnNOS-9 (PDB code: 6NGR),

and (C) hnNOS-9 (PDB code: 6NG7).

H

DOI: 10.1021/acs.jmedchem.8b02032

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Figure 5. Inhibitor binding environment in the structure of (A) hnNOS-10 (PDB code: 6NG8) and (B) hnNOS-14 (PDB code: 6NGD). The

ﬂuorine atom positions on the middle benzene ring are marked in red in this and all following ﬁgures.

propionate (Figure 3B), the extra ﬂuorine atom is likely the

reason that the pyrrolidine of 9 approaches Asp602 instead of

the heme propionate (Figure 4C). The ﬂuorobenzene ring

orientation is almost perpendicular to each other in these two

modes, the Gln483 being in the “out” or “in” rotamer

accordingly. A similar binding mode transition in hnNOS was

also observed above from compound 4 to 6 because of the

chirality change on the pyrrolidine (Figure 3).

reduce the number of rotatable bonds and further increase

lipophilicity. We did not attempt to get X-ray crystal data for

16, but the structures of 15, 17, and 18 bound to hnNOS all

exhibit very similar “up” binding modes (Figure 7), and the

three bound to rnNOS are in the bent down modes (Figure

S6). In both cases, the tail amino group is stabilized by its

interaction with heme propionate A via a bridging water.

Therefore, it is not surprising that the inhibitory potencies of

all three are very similar in the low nanomolar range. In the

PAMPA-BBB assay, with a rigid alkyne tail, both 15 and 16

show enhanced permeability with P_evalues ca. 18 × 10⁻⁶cm

s⁻¹.

Fluorination of the Central Benzene Linker (Com-

pounds 10−14). To further enhance bioavailability, addi-

tional ﬂuorine atoms were added to the middle benzene linker

(compounds 10−14), and this caused signiﬁcant changes in

inhibitor binding modes. With compounds where the multiple

ﬂuorine atoms are on the same side (W, X, and Y position) of

the central benzene ring (10 and 14, see Figure 1), the “up”

binding mode is very similar to lead compound 1 in hnNOS

(Figure 5). In addition, the inhibitory potencies are similar, but

the permeability substantially improves, with additional

ﬂuorines (compare 1 and 14, Table 1). When the additional

ﬂuorines are on the opposite sides of the benzene ring, the

binding mode changes. In 11, the Y-,Z-diﬂuorobenzene ring

binds “bent down” and is oriented such that the Y-ﬂuorine is

about 3.8 Å from the heme iron for possible electrostatic

interactions (Figure 6A). For compound 12, its central X-,Z-

diﬂuorobenzene is much farther from the heme iron because

the 2-carbon linker from the aminopyridine to diﬂuorobenzene

bends in a diﬀerent way from that in 11 (Figure 6B).

Compound 13 shares the same ﬂuorobenzene position with

12, but the ring ﬂips 180° relative to 13 (Figure 6C). For

compounds 11, 12, and 13, electrostatic stabilization of the tail

tertiary amino group determines potency. The tail amino

group of 11 is near heme propionate D, whereas that of

compound 13 is ∼3.4 Å from Asp602. The tail tertiary amino

group of 12 is not close to any negatively charged group.

Therefore, of the 10−14 compounds, 12 has the weakest

electrostatic interactions with neighbors, which possibly

accounts for the decrease in potency. The changes in binding

modes for compounds 10−14 in rnNOS are less dramatic

(Figure S5), which are reﬂected in similar potencies (Table 1).

Building on Compound 10 (Compounds 15−18). On

the basis of the information accumulated thus far, 10 was

chosen as the scaﬀold for further modiﬁcation because of its

excellent inhibitory potency and permeability properties. The

new analogues were designed with an alkyne amino group (15,

16) or a pyrrolidine ring (17, 18) incorporated in the tail to

Modiﬁcation of Compound 18 (Compounds 19−21).

Because of its excellent potency, selectivity, and cell

permeability properties, 18 was selected for further mod-

iﬁcation. Various heterocycles, including morpholine, (4-

ethoxy)pyrrolidine, and azetidine, as found in compounds

19, 20, and 21, respectively, were utilized to modulate the

basicity of the tertiary amino group of the pyrrolidine ring,

which could potentially enhance its permeability and protect it

from metabolism.³⁵Crystal structures show that compounds

18 (Figure 7C) and 19 (Figure 8A) adopt the lead compound

1 binding mode in hnNOS with the tail end of the inhibitor

near the heme propionates. The decrease in potency of 19 is

very likely the result of the lower pK_aof the morpholine amino

group in 19. On the other hand, compound 19 exhibits the

best cell permeability properties with a P_evalue of 21.1 × 10⁻⁶

cm s⁻¹. This indicates that achieving both good potency and

high permeability can be a delicate balance. Although

compounds 20 and 21 bind to hnNOS in the “up” mode,

the orientation of the central diﬂuorobenzene is almost

perpendicular to that found in 18 and 19, so Gln483 is not

required to adopt the “out” rotamer to make room for the

inhibitor. The altered diﬂuorobenzene orientation was

observed for compounds 6 (Figure 3C) and 9 (Figure 4C)

as well. The positions of the tail amino group have two

diﬀerent preferences. For compounds 6 and 9, the tail amine is

near Asp602, which results in rather poor potency, whereas

both the (4-ethoxy)pyrrolidine of 20 and the azetidine of 21

are stabilized by heme propionate A. The better binding

potency of 21 is possibly the result of higher basicity of its

azetidine nitrogen, thereby providing better electrostatic

stabilization. Most importantly, compound 21 not only has

excellent membrane penetration but also gives a very low ER of

0.8 (Table 3), which means it is potentially a good CNS drug.

I

DOI: 10.1021/acs.jmedchem.8b02032

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Figure 7. Inhibitor binding environment in the structure of (A)

hnNOS-15 (PDB code: 6NGE), (B) hnNOS-17 (PDB code: 6NGF),

and (C) hnNOS-18 (PDB code: 6NGH).

Figure 6. Inhibitor binding environment in the structure of (A)

hnNOS-11 (PDB code: 6NGA), (B) hnNOS-12 (PDB code: 6NGB),

and (C) hnNOS-13 (PDB code: 6NGC).

heNOS selectivity in the 900−1200 range (Table 2). For

human iNOS, the use of the diﬂuorobenzene linker tends to

decrease selectivity compared to that of lead compound 1,

which is consistent with the observation for rnNOS selectivity

over murine iNOS (Table 1). Unlike other analogues, 18 and

21 retain comparable hnNOS/hiNOS selectivity to that of 1.

To understand the structural basis for isoform selectivity,

crystal structures of the most selective inhibitors bound to

heNOS are required. However, those selective inhibitors that

bind poorly to heNOS tend not to give clear electron density.

In the series listed in Table 2, the best heNOS crystal

structures were obtained for 21. In contrast to the up mode

binding of the central diﬂuorobenzene in hnNOS, this ring

binds heNOS in the “bent down” mode in contact with the

In rnNOS, all three compounds bind in the “bent down” mode

(Figures S7 and 9A) with their tail ring nitrogen next to

propionate A. The variation in potency reﬂects the strength of

this electrostatic interaction.

Isoform Selectivity. In addition to enhancing potency and

cell permeability, it is important to retain selectivity for nNOS

over the other NOS isoforms, especially eNOS. Because of

both potency and bioavailability, six compounds, 10, 14, 15,

17, 18, and 21, were selected for estimating selectivity. With

the exception of 15, all of these inhibitors exhibit hnNOS/

J

DOI: 10.1021/acs.jmedchem.8b02032

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

propionate D with close contacts with heme and Met570, the

benzene in heNOS sits above two propionates. The reason

behind this preference lies in the amino acid diﬀerences

nearby. Where heNOS has Val104 and Phe105, rnNOS has

Met336 and Leu337 (Met341 and His342 in hnNOS). The

larger Met336 in rnNOS allows good van der Waals contacts

with the methyl group of the azetidine, placing the latter near

heme propionate A, whereas the contact with the bulky

Phe105 in heNOS pulls the azetidine the opposite way. On the

other hand, the tail azetidine of 21 in hnNOS comes within 3.8

Å of heme propionate A. Therefore, in both rnNOS and

hnNOS, the 21 tertiary amine tail interacts more favorably

with heme propionate A than in heNOS, resulting in better

potency and n/e selectivity.

CONCLUSIONS

■

In summary, we report our optimization of nNOS inhibitors

bearing a 2-aminopyridine scaﬀold with emphasis on

improving cell membrane permeability of these inhibitors to

be able to cross the blood−brain barrier, while retaining

excellent inhibition activity and high isoform selectivity. A new

series of potent and selective human nNOS inhibitors have

been designed and synthesized by employing various medicinal

chemistry approaches involving enhancing the lipophilicity,

increasing molecular rigidity, and modulating the pK_aof the

basic amino tail group of lead molecule 1. Crystal structures

show that the central ﬂuorobenzene can adopt quite diﬀerent

orientations in the active site, but this variability does not alter

the binding aﬃnity between hnNOS and rnNOS. As in

previous studies, aﬃnity diﬀerences are controlled mainly by

electrostatic and van der Waals interactions with the tail end of

the inhibitor. The most promising inhibitor, 21, with one

additional ﬂuorine atom on the ﬂuorobenzene middle linker

compared with lead compound 1 and an azetidine ring in the

tail displays excellent inhibition for human nNOS (K_i= 23

nM) and high selectivity over human eNOS (hn/he = 956)

and human iNOS (hn/hi = 77) but also exhibits a great

potential of brain penetration. Caco-2 bidirectional assays

reveal that 21 has an eﬄux ratio of only 0.8, which is

signiﬁcantly lower than that of lead compound 1 (ER = 5.9)

and the suggested ER of <2.5 for CNS (+) drugs. The Caco-2

bidirectional assay also revealed that 21 has high cell

membrane permeability with the P_appvalue of 17.0 × 10⁻⁶

cm s⁻¹, which is in good agreement with the eﬀective

permeability (P_e= 16.3 × 10⁻⁶cm s⁻¹) determined by the

PAMPA-BBB assay. Our results herein provide the basis for

further exploration of the 2-aminopyridine nNOS inhibitors in

CNS drug development and additional insights into the

strategies to overcome the BBB using medicinal chemistry

approaches.

Figure 8. Inhibitor binding environment in the structure of (A)

hnNOS-19 (PDB code: 6NGI), (B) hnNOS-20 (PDB code: 6NHB),

and (C) hnNOS-21 (PDB code: 6NHC).

EXPERIMENTAL SECTION

■

Chemistry. General Procedures. All reagents, unless speciﬁed,

were obtained from Sigma-Aldrich, Combi-blocks, and Oakwood

Chemical Companies. Anhydrous solvents (THF, CH₂Cl₂, MeCN,

and DMF) were puriﬁed before use by passing through a column

composed of activated alumina and a supported copper redox catalyst.

Sonogashira coupling was carried out in the Biotage Initiator

microwave using Biotage microwave vials (0.5−2, 2−5, and 10−20

mL). Thin-layer chromatography was performed on silica gel 60 F254

precoated plates (0.25 mm) from SiliCycle, and components were

visualized by ultraviolet light (254 nm) and/or KMnO₄or ninhydrin

stain. Flash column chromatography was performed on an Agilent

971-FP automated ﬂash puriﬁcation system with a Varian column

heme propionates. As noted earlier, all of the compounds in

the current study also bind to rnNOS in the bent down mode.

This “up” and “bent down” diﬀerence has little to do with

potency between hnNOS and rnNOS (Table 1). What

requires explanation is why 21 exhibits a hnNOS/heNOS

selectivity of 956 as well as high rnNOS/heNOS selectivity.

Although 21 binds to both rnNOS and heNOS in a bent down

mode, the exact position of the central diﬂuorobenzene is quite

diﬀerent (Figure 9A,B, respectively). Although the benzene

ring in rnNOS is sequestered within the arm of heme

K

DOI: 10.1021/acs.jmedchem.8b02032

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

Figure 9. Inhibitor binding environment in the structure of (A) rnNOS-21 (PDB code: 6NHD) and (B) heNOS-21 (PDB code: 6NHF).

station and various SiliCycle cartridges (4−80 g, 40−63 μm, 60 Å).

concentrated, and a crude product was subjected to an alkyne

reduction without any puriﬁcation.

¹H and ¹³C NMR spectra were recorded on a Bruker Avance-III NMR

The crude product (1 equiv) was dissolved in MeOH (0.1 M), and

the solution was degassed for 5 min, followed by an addition of 10 wt

% Pd/C. The reaction was run at RT for 20 h under a hydrogen

balloon (1 atm). The crude mixture was then ﬁltered through a pad of

celite, and the ﬁltrate was concentrated under reduced pressure.

The crude reduction product was diluted with MeOH (0.24 M),

followed by an addition of 37% HCHO in H₂O (3 equiv). The

reaction was run at RT for 5 min. The reaction was brought to 0 °C,

and NaBH₄(3 equiv) was added slowly. The reaction was further run

for 2 h at RT. Upon completion, the reaction was quenched with

water, and the methanol was removed under reduced pressure. The

aqueous mixture was extracted with ethyl acetate three times, and the

organic layers were combined, dried over Na₂SO₄, and concentrated

to give the crude product, which was subjected to pyrrole

deprotection without any puriﬁcation using the protocol in general

procedure A to give 5, 7, 9, 17, 18, and 19. For compounds 4, 6, and

8, which carry only the secondary amine in the tail, the syntheses

followed the same protocol except the reductive amination with

HCHO was omitted.

General Procedure E: Cbz Deprotection, Reductive Amination,

and Pyrrole Deprotection. The starting material (45b,c, 1 equiv) was

dissolved in MeOH (0.1 M). The solution was degassed for 5 min,

and 10% Pd(OH)₂/C was added. The reaction was run at RT for 24 h

under a hydrogen balloon (1 atm). After completion, the reaction

mixture was ﬁltered through a pad of celite, and the ﬁltrate was

concentrated under reduced pressure to give the crude product, which

was subjected to reductive amination with formaldehyde and pyrrole

deprotection following the same protocols described in general

procedure D to give 20−21.

spectrometer at 500 and 126 MHz, respectively, in CDCl₃or CD₃OD.

Chemical shifts were reported in parts per million, multiplicities are

indicated by s = singlet, d = doublet, t = triplet, q = quartet, sep =

septet, dd = doublet of doublet, dt = doublet of triplet, m = multiplet,

and br = broad resonance. Coupling constants “J” were reported in

hertz. High-resolution mass spectral (HRMS) data were obtained on

an Agilent 6210 LC-TOF spectrometer in the positive ion mode using

electrospray ionization (ESI) with an Agilent G1312A high-perform-

ance liquid chromatography (HPLC) pump and an Agilent G1367B

autoinjector at the Integrated Molecular Structure Education and

Research Center (IMSERC), Northwestern University. The purity of

compounds was tested using a reserved-phase analytical Agilent

Inﬁnity 1260 HPLC with an Agilent Poroshell 120 EC-C18 column,

detecting with UV absorbance at 254 nm. All compounds undergoing

biological testing were >95% pure.

General Procedure A: Pyrrole Deprotection. In a microwave vial,

starting materials 26b or 41 (1 equiv) and NH₂OH·HCl (3−4 equiv)

were added. They are diluted with EtOH/water (2:1) to form a 0.16

M solution. The microwave vial was capped, and the reaction mixture

was run at 100 °C for 20 h. The cap was removed, and the reaction

mixture was concentrated under reduced pressure. The crude product

mixture was puriﬁed by reversed ﬂash chromatography to give ﬁnal

products 3 or 16.

General Procedure B: Boc and Pyrrole Deprotection. The starting

material 26a or 39 (1 equiv) was dissolved in CH₂Cl₂(0.1 M),

followed by the addition of TFA (1.1 equiv) at 0 °C, and the reaction

was run at RT. After stirring at RT for 1 h, the crude product was

concentrated under reduced pressure, diluted back with CH₂Cl₂, and

washed with sat. NaHCO₃. The organic layer was dried over Na₂SO₄

and concentrated to give crude products, which were submitted to

pyrrole deprotection following the protocol in general procedure A to

give 2 or 15.

General Procedure C: Alkyne Reduction and Pyrrole Depro-

tection. The starting material 38a−e (1 equiv) was dissolved in

MeOH (0.1 M). The solution was degassed for 5 min, and 10 wt %

Pd/C was added. The reaction was run at RT for 20 h under a

hydrogen balloon (1 atm). The crude mixture was then ﬁltered

through a pad of celite, and the ﬁltrate was concentrated under

reduced pressure. Without any puriﬁcation, the crude product was

subjected to the pyrrole deprotection following the protocol in

general procedure A to give 10−14.

General Procedure D: Boc-Deprotection, Alkyne Reduction,

Reductive Amination, and Pyrrole Deprotection. The starting

material (33a−c, 42, 43, or 45a) was dissolved in CH₂Cl₂(0.1 M),

and TFA (1.1 equiv) was added at 0 °C. The reaction was run at RT

for 1 h; then, the solvent and TFA were removed under reduced

pressure. The crude mixture was diluted back with CH₂Cl₂and

washed with saturated Na₂CO₃. The organic layer was then

6-(3-Fluoro-5-(3-(methylamino)prop-1-yn-1-yl)phenethyl)-4-

methylpyridin-2-amine (2). Compound 2 (74 mg, 50% for two

steps) was prepared from 26b (237 mg, 0.5 mmol) according to

1

general procedure B. H NMR (500 MHz, CD₃OD) δ 7.08 (s, 1H),

6.98−6.91 (m, 2H), 6.29 (s, 1H), 6.27 (s, 1H), 3.58 (s, 2H), 2.92

(dd, J = 6.5, 9.4 Hz, 2H), 2.79 (dd, J = 6.3, 9.2 Hz, 2H), 2.48 (s, 3H),

2.17 (s, 3H). ¹³C NMR (126 MHz, CD₃OD) δ 162.4 (d, J_C−F= 246.2

Hz), 157.6, 154.4, 147.9, 143.2 (d, J_C−F= 8.1 Hz), 127.8 (d, J_C−F

=

2.9 Hz), 123.1 (d, J_C−F= 10.1 Hz), 116.5 (d, J_C−F= 21.5 Hz), 116.3

(d, J_C−F= 23.5 Hz), 113.6, 109.6, 86.5 (d, J_C−F= 3.6 Hz), 79.3, 38.0,

33.7, 33.6, 31.5, 20.6. HRMS-ESI: calculated for C₁₈H₂₀FN₃[M +

H]⁺298.1714, found 298.1716.

L

DOI: 10.1021/acs.jmedchem.8b02032

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

procedure D. ¹H NMR (500 MHz, CD₃OD) δ 7.05 (s, 1H), 6.89 (t, J

= 10.9 Hz, 2H), 6.68 (s, 1H), 6.63 (s, 1H), 3.59−3.45 (m, 2H),

3.40−3.34 (m, 1H), 3.10−2.98 (m, 4H), 2.76 (t, J = 8.0 Hz, 2H),

2.36 (s, 3H), 2.32−2.23 (m, 1H), 2.18−1.91 (m, 4H), 1.79−1.62 (m,

1H). ¹³C NMR (126 MHz, CD₃OD) δ 163.1 (d, J_C−F= 244.9 Hz),

157.7, 154.4, 148.4, 143.4 (d, J_C−F= 7.8 Hz), 142.6 (d, J_C−F= 7.8

Hz), 124.2, 113.6, 112.9 (d, J_C−F= 21.8 Hz), 112.8 (d, J_C−F= 21.6

Hz), 109.4, 60.1, 44.9, 34.0, 33.9, 33.4, 32.0, 29.7, 23.1, 20.5. HRMS-

ESI: calculated for C₂₀H₂₆FN₃[M + H]⁺328.2184, found 328.2180.

6-(3-(3-(Dimethylamino)prop-1-yn-1-yl)-5-ﬂuorophenethyl)-4-

methylpyridin-2-amine (3). Compound 3 (132 mg, 90%) was

prepared from 26b (184 mg, 0.47 mmol) according to general

procedure A. H NMR (500 MHz, CD₃OD) δ 7.34 (s, 1H), 7.23−

1

7.14 (m, 2H), 6.69 (s, 1H), 6.62 (s, 1H), 4.36 (s, 2H), 3.38−3.28 (m,

2H), 3.13−2.97 (m, 8H), 2.36 (s, 3H). ¹³C NMR (126 MHz,

CD₃OD) δ 166.4 (d, J_C−F= 246.4 Hz), 161.6, 158.4, 151.9, 147.2 (d,

J_C−F= 8.1 Hz), 131.8 (d, J_C−F= 2.8 Hz), 126.7 (d, J_C−F= 10.2 Hz),

120.8 (d, J_C−F= 21.8 Hz), 120.4 (d, J_C−F= 23.7 Hz), 117.5, 113.5,

92.1 (d, J_C−F= 3.5 Hz), 81.8, 69.4, 45.5 (2C), 37.6, 37.5, 24.5.

HRMS-ESI: calculated for C₁₉H₂₂FN₃[M + H]⁺312.1871, found

312.1874.

(S)-6-(3-Fluoro-5-(2-(1-methylpyrrolidin-2-yl)ethyl)phenethyl)-4-

methylpyridin-2-amine (7). Compound 7 (42 mg, 24% for four

steps) was prepared from 33b (257 mg, 0.5 mmol) according to

1

general procedure D. H NMR (500 MHz, CD₃OD) δ 7.07 (s, 1H),

6.95−6.88 (m, 2H), 6.70 (s, 1H), 6.63 (s, 1H), 3.71 (ddd, J = 5.0, 8.0,

11.5 Hz, 1H), 3.34−3.30 (m, 1H), 3.17 (dt, J = 8.4, 11.4 Hz, 1H),

3.05 (s, 4H), 2.94 (s, 3H), 2.84−2.67 (m, 2H), 2.45−2.38 (m, 1H),

2.36 (s, 3H), 2.33−2.26 (m, 1H), 2.21−2.04 (m, 2H), 1.98−1.80 (m,

2H). ¹³C NMR (126 MHz, CD₃OD) δ 163.0 (d, J_C−F= 244.7 Hz),

157.6, 154.4, 148.3, 143.4 (d, J_C−F= 7.8 Hz), 142.6 (d, J_C−F= 7.8

Hz), 124.2, 113.6, 113.0 (d, J_C−F= 21.4 Hz), 112.8 (d, J_C−F= 21.6

Hz), 109.5, 68.7, 55.9, 38.5, 34.1, 33.9, 31.9, 31.7, 29.3, 21.1, 20.6.

HRMS-ESI: calculated for C₂₁H₂₈FN₃[M + H]⁺342.2340, found

342.2341.

(R)-6-(3-Fluoro-5-(2-(pyrrolidin-2-yl)ethyl)phenethyl)-4-methyl-

pyridin-2-amine (4). Compound 4 (78 mg, 33% for three steps) was

prepared from 33a (362 mg, 0.7 mmol) according to general

procedure D. H NMR (500 MHz, CD₃OD) δ 7.06 (s, 1H), 6.94−

1

6.86 (m, 2H), 6.68 (s, 1H), 6.63 (s, 1H), 3.56−3.50 (m, 1H), 3.36−

3.34 (m, 1H), 3.23 (q, J = 7.3 Hz, 1H), 3.11−2.98 (m, 4H), 2.76 (t, J

= 8.0 Hz, 2H), 2.36 (s, 3H), 2.31−2.22 (m, 1H), 2.17−1.94 (m, 4H),

1.79−1.66 (m, 1H). ¹³C NMR (126 MHz, CD₃OD) δ 161.5 (d, J_C−F

= 244.8 Hz), 156.1, 152.9, 146.8, 141.8 (d, J_C−F= 7.8 Hz), 141.0 (d,

J

_C−F= 7.9 Hz), 122.7, 112.1, 111.4 (d, J_C−F= 21.6 Hz), 111.3 (d, J_C−F

= 21.4 Hz), 107.9, 58.5, 43.3, 32.5, 32.4, 31.9, 30.5, 28.2, 21.6, 19.0.

HRMS-ESI: calculated for C₂₀H₂₆FN₃[M + H]⁺328.2184, found

328.2185.

6-(3-Fluoro-5-(2-((2R,4S)-4-ﬂuoropyrrolidin-2-yl)ethyl)-

phenethyl)-4-methylpyridin-2-amine (8). Compound 8 (62 mg, 30%

for three steps) was prepared from 33c (310 mg, 0.6 mmol) according

1

to general procedure D. H NMR (500 MHz, CD₃OD) δ 7.08 (s,

1H), 6.88 (dd, J = 17.5, 9.6 Hz, 2H), 6.69 (s, 1H), 6.61 (s, 1H), 5.44

(dt, J = 3.4, 52.3 Hz, 1H), 3.89−3.80 (m, 1H), 3.71 (ddd, J = 3.9,

13.9, 34.7 Hz, 1H), 3.64−3.49 (m, 1H), 3.09−2.99 (m, 4H), 2.85−

2.71 (m, 2H), 2.60−2.48 (m, 1H), 2.34 (s, 3H), 2.20 (ddt, J = 7.0,

9.3, 14.0 Hz, 1H), 2.14−1.91 (m, 2H). ¹³C NMR (126 MHz,

CD₃OD) δ 162.9 (d, J_C−F= 244.5 Hz), 157.5, 154.2, 148.2, 143.0 (d,

(R)-6-(3-Fluoro-5-(2-(1-methylpyrrolidin-2-yl)ethyl)phenethyl)-4-

methylpyridin-2-amine (5). Compound 5 (41 mg, 15% for four

steps) was prepared from 33a (401 mg, 0.8 mmol) according to

J

_C−F= 7.4 Hz), 142.4 (d, J_C−F= 7.6 Hz), 124.1, 113.5, 112.9 (d, J_C−F

= 21.6 Hz), 112.8 (d, J_C−F= 21.8 Hz), 109.4, 91.9 (d, J_C−F= 175.9

Hz), 58.5, 51.0 (d, J_C−F= 24.7 Hz), 37.5 (d, J_C−F= 20.9 Hz), 33.9,

33.8, 32.9, 31.9, 20.5. HRMS-ESI: calculated for C₂₀H₂₅F₂N₃[M +

H]⁺346.2089, found 346.2088.

1

general procedure D. H NMR (500 MHz, CD₃OD) δ 7.06 (s, 1H),

6.91 (dd, J = 9.6, 20.1 Hz, 2H), 6.68 (s, 1H), 6.63 (s, 1H), 3.72−3.68

(m, 1H), 3.37−3.36 (m, 1H), 3.23−3.12 (m, 1H), 3.09−2.99 (m,

4H), 2.94 (s, 3H), 2.84−2.67 (m, 2H), 2.47−2.38 (m, 1H), 2.36 (s,

3H), 2.34−2.27 (m, 1H), 2.22−2.03 (m, 2H), 1.97−1.78 (m, 2H).

¹³C NMR (126 MHz, CD₃OD) δ 163.1 (d, J_C−F= 244.7 Hz), 157.7,

154.4, 148.4, 143.3 (d, J_C−F= 7.5 Hz), 142.6 (d, J_C−F= 7.8 Hz),

124.2, 113.6, 113.0 (d, J_C−F= 21.4 Hz), 112.8 (d, J_C−F= 21.6 Hz),

109.5, 68.7, 55.9, 38.5, 34.0, 33.9, 31.9, 31.7, 29.2, 21.1, 20.5. HRMS-

ESI: calculated for C₂₁H₂₈FN₃[M + H]⁺342.2340, found 342.2342.

6-(3-Fluoro-5-(2-((2R,4S)-4-ﬂuoro-1-methylpyrrolidin-2-yl)ethyl)-

phenethyl)-4-methylpyridin-2-amine (9). Compound 9 (30 mg, 24%

for four steps) was prepared from 33c (180 mg, 0.35 mmol)

1

according to general procedure D. H NMR (500 MHz, CD₃OD) δ

7.08 (s, 1H), 6.94 (d, J = 9.9 Hz, 1H), 6.90 (d, J = 9.7 Hz, 1H), 6.68

(s, 1H), 6.64 (s, 1H), 5.45 (dd, J = 5.1, 52.5 Hz, 1H), 3.94 (t, J = 15.0

Hz, 1H), 3.58−3.46 (m, 2H), 3.07−3.03 (m, 4H), 3.02 (s, 3H),

2.97−2.69 (m, 3H), 2.36 (s, 3H), 2.24−2.09 (m, 2H), 2.01−1.94 (m,

1H). ¹³C NMR (126 MHz, CD₃OD) δ 163.1 (d, J_C−F= 244.7 Hz),

(S)-6-(3-Fluoro-5-(2-(pyrrolidin-2-yl)ethyl)phenethyl)-4-methyl-

pyridin-2-amine (6). Compound 6 (47 mg, 28% for three steps) was

prepared from 33b (257 mg, 0.5 mmol) according to general

M

DOI: 10.1021/acs.jmedchem.8b02032

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

157.6, 154.4, 148.3, 143.0 (d, J_C−F= 7.7 Hz), 142.7 (d, J_C−F= 7.7

Hz), 124.2, 113.6, 113.0 (d, J_C−F= 21.4 Hz), 112.9 (d, J_C−F= 21.6

Hz), 109.5, 90.6 (d, J_C−F= 176.0 Hz), 67.5, 61.6 (d, J_C−F= 23.5 Hz),

38.4, 36.8 (d, J_C−F= 22.7 Hz), 34.1, 33.9, 32.6, 31.5, 20.5. HRMS-

ESI: calculated for C₂₁H₂₇F₂N₃[M + H]⁺360.2246, found 360.2246.

6-(3-(3-(Dimethylamino)propyl)-2,5,6-triﬂuorophenethyl)-4-

methylpyridin-2-amine (13). Compound 13 (10 mg, 24% for two

steps) was prepared from 38d (50 mg, 0.12 mmol) according to

general procedure C ¹H NMR (500 MHz, CD₃OD) δ 7.24 (ddd, J =

6.9, 8.8, 10.6 Hz, 1H), 6.71 (s, 1H), 6.54 (s, 1H), 3.23−3.13 (m, 4H),

3.03 (t, J = 7.4 Hz, 2H), 2.91 (s, 6H), 2.73 (t, J = 7.8 Hz, 2H), 2.34 (s,

3H), 2.08−1.96 (m, 2H). ¹³C NMR (126 MHz, CD₃OD) δ 156.1,

153.0, 152.9 (ddd, J_C−F= 1.3, 5.0, 242.0 Hz), 146.0, 145.8 (ddd, J_C−F

= 8.8, 13.9, 245.7 Hz), 145.1 (ddd, J_C−F= 2.5, 12.6, 244.4 Hz), 121.9

(dt, J_C−F= 5.1, 19.2 Hz), 115.2 (dd, J_C−F= 16.6, 23.1 Hz), 114.5 (dd,

J_C−F= 5.9, 19.5 Hz), 112.2, 108.3, 55.3, 40.6, 30.2, 23.3, 23.0, 20.2,

19.0. HRMS-ESI: calculated for C₁₉H₂₄F₃N₃[M + H]⁺352.1995,

found 352.1996.

6-(5-(3-(Dimethylamino)propyl)-2,3-diﬂuorophenethyl)-4-meth-

ylpyridin-2-amine (10). Compound 10 (25 mg, 39% for two steps)

was prepared from 38a (80 mg, 0.2 mmol) according to general

1

procedure C. H NMR (500 MHz, CD₃OD) δ 7.13−7.06 (m, 2H),

6.71 (s, 1H), 6.60 (s, 1H), 3.21−3.15 (m, 2H), 3.15−3.03 (m, 4H),

2.92 (s, 6H), 2.70 (t, J = 7.7 Hz, 2H), 2.36 (s, 3H), 2.11−2.01 (m,

2H). ¹³C NMR (126 MHz, CD₃OD) δ 157.6, 154.4, 150.1 (dd, J_C−F

= 246.6, 13.3 Hz), 147.9, 147.3 (dd, J_C−F= 244.4, 12.6 Hz), 137.4−

137.2 (m), 128.7 (d, J_C−F= 12.5 Hz), 125.3 (t, J_C−F= 3.0 Hz), 115.3

(d, J_C−F= 17.5 Hz), 113.6, 109.6, 56.9, 42.1, 32.7, 31.1, 27.5, 25.7,

20.5. HRMS-ESI: calculated for C₁₉H₂₅F₂N₃[M + H]⁺334.2089,

found 334.2089.

6-(5-(3-(Dimethylamino)propyl)-2,3,4-triﬂuorophenethyl)-4-

methylpyridin-2-amine (14). Compound 14 (48 mg, 36% for two

steps) was prepared from 38e (160 mg, 0.37 mmol) according to

1

general procedure C H NMR (500 MHz, CD₃OD) δ 7.15 (td, J =

7.6, 2.4 Hz, 1H), 6.64 (s, 1H), 6.58 (s, 1H), 3.24−3.16 (m, 2H),

3.12−2.98 (m, 4H), 2.91 (s, 6H), 2.77 (t, J = 7.7 Hz, 2H), 2.34 (s,

3H), 2.11−2.01 (m, 2H). ¹³C NMR (126 MHz, CD₃OD) δ 156.5,

6-(3-(3-(Dimethylamino)propyl)-2,6-diﬂuorophenethyl)-4-meth-

ylpyridin-2-amine (11). Compound 11 (35 mg, 35% for two steps)

was prepared from 38b (122 mg, 0.3 mmol) according to general

155.2, 149.3, 148.1 (dd, J_C−F= 245.7, 10.1 Hz, 2C), 139.6 (td, J_C−F

250.7, 16.4 Hz), 124.8 (dd, J_C−F= 7.6, 3.8 Hz), 124.3 (dd, J_C−F

=

13.9, 5.0 Hz), 124.0 (dd, J_C−F= 13.9, 3.8 Hz), 113.6, 109.2, 56.8, 42.1,

33.4, 27.4, 24.7, 24.6, 20.4. HRMS-ESI: calculated for C₁₉H₂₄F₃N₃[M

+ H]⁺352.1995, found 352.1997.

1

procedure C H NMR (500 MHz, CD₃OD) δ 7.27 (q, J = 7.9 Hz,

1H), 6.91 (t, J = 8.8 Hz, 1H), 6.73 (s, 1H), 6.49 (s, 1H), 3.24−3.19

(m, 2H), 3.14 (t, J = 7.3 Hz, 2H), 3.02 (t, J = 7.3 Hz, 2H), 2.92 (s,

6H), 2.73 (t, J = 7.7 Hz, 2H), 2.32 (s, 3H), 2.04 (t, J = 8.2 Hz, 2H).

¹³C NMR (126 MHz, CD₃OD) δ 160.0 (dd, J_C−F= 244.8, 8.2 Hz),

159.2 (dd, J_C−F= 245.6, 8.2 Hz), 157.5, 154.4, 147.8, 129.3 (dd, J =

6.6, 10.0 Hz), 123.0 (dd, J = 3.7, 17.0 Hz), 114.4 (t, J = 20.6 Hz),

113.7, 110.7 (dd, J = 3.6, 22.3 Hz), 109.7, 56.9, 42.1, 31.9, 25.1 (d, J =

2.4 Hz), 24.8, 21.4, 20.5. HRMS-ESI: calculated for C₁₉H₂₅F₂N₃[M +

H]⁺334.2089, found 334.2090.

6-(2,3-Diﬂuoro-5-(3-(methylamino)prop-1-yn-1-yl)phenethyl)-4-

methylpyridin-2-amine (15). Compound 15 (80 mg, 63% for two

steps) was prepared from 39 (198 mg, 0.4 mmol) according to

general procedure B. ¹H NMR (500 MHz, CD₃OD) δ 7.42−7.33 (m,

2H), 6.72 (s, 1H), 6.59 (s, 1H), 4.18 (s, 2H), 3.16 (dd, J = 6.3, 9.0

Hz, 2H), 3.06 (dd, J = 6.2, 9.0 Hz, 2H), 2.85 (s, 3H), 2.35 (s, 3H).

¹³C NMR (126 MHz, CD₃OD) δ 157.6, 154.5, 149.9 (dd, J_C−F

=

248.4, 13.6 Hz), 149.6 (dd, J_C−F= 250.6, 13.0 Hz), 147.5, 129.8 (d,

J

_C−F= 13.5 Hz), 129.6−129.4 (m), 118.9 (d, J_C−F= 19.3 Hz), 117.9−

117.8 (m), 113.6, 109.8, 85.7, 79.1, 38.0, 32.4, 31.5, 27.2, 20.5.

HRMS-ESI: calculated for C₁₈H₁₉F₂N₃[M + H]⁺316.1620, found

316.1618.

6-(3-(3-(Dimethylamino)propyl)-2,5-diﬂuorophenethyl)-4-meth-

ylpyridin-2-amine (12). Compound 12 (32 mg, 42% for two steps)

was prepared from 38c (93 mg, 0.23 mmol) according to general

1

procedure C H NMR (500 MHz, CD₃OD) δ 7.11−6.92 (m, 2H),

6.72 (s, 1H), 6.59 (s, 1H), 3.25−3.18 (m, 2H), 3.14−3.01 (m, 4H),

2.92 (s, 6H), 2.75 (t, J = 7.9 Hz, 2H), 2.36 (s, 3H), 2.13−2.01 (m,

2H). ¹³C NMR (126 MHz, CD₃OD) δ 158.4 (d, J_C−F= 242 Hz),

157.6, 155.2 (d, J_C−F= 241 Hz), 154.4, 147.9, 128.9 (dd, J = 8.1, 19.2

Hz), 128.2 (dd, J = 8.2, 19.3 Hz), 115.0 (ddd, J = 4.7, 20.3, 24.5 Hz,

2C), 113.6, 109.6, 56.9, 42.1, 32.5, 27.7, 25.4, 24.5, 20.6. HRMS-ESI:

calculated for C₁₉H₂₅F₂N₃[M + H]⁺334.2089, found 334.2092.

6-(5-(3-(Dimethylamino)prop-1-yn-1-yl)-2,3-diﬂuorophenethyl)-

4-methylpyridin-2-amine (16). Compound 16 (65 mg, 66%) was

prepared from 41 (122 mg, 0.3 mmol) according to general procedure

N

DOI: 10.1021/acs.jmedchem.8b02032

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

A. ¹H NMR (500 MHz, CD₃OD) δ 7.45−7.38 (m, 2H), 6.70 (s, 1H),

6.60 (s, 1H), 4.35 (s, 2H), 3.15 (t, J = 7.7 Hz, 2H), 3.10−2.98 (m,

8H), 2.36 (s, 3H). ¹³C NMR (126 MHz, CD₃OD) δ 157.7, 154.5,

150.0 (dd, J_C−F= 248.8, 13.7 Hz), 149.8 (dd, J_C−F= 250.9, 12.9 Hz),

147.6, 129.8 (d, J_C−F= 13.5 Hz), 129.7 (t, J_C−F= 3.5 Hz), 119.1 (d,

J_C−F= 19.5 Hz), 117.5 (dd, J_C−F= 8.8, 4.7 Hz), 113.6, 109.8, 87.4,

77.7, 46.9, 41.5 (2C), 32.4, 27.2, 20.5. HRMS-ESI: calculated for

C₁₉H₂₁F₂N₃[M + H]⁺330.1776, found 330.1777.

113.6, 109.6, 67.6, 63.7, 63.3, 54.0, 39.9, 32.7, 30.1, 28.0, 27.5, 20.5.

HRMS-ESI: calculated for C₂₁H₂₇F₂N₃O [M + H]⁺376.2195, found

376.2197.

6-(5-(2-((2S,4R)-4-Ethoxy-1-methylpyrrolidin-2-yl)ethyl)-2,3-di-

ﬂuorophenethyl)-4-methylpyridin-2-amine (20). Compound 20 (22

mg, 41% for three steps) was prepared from 45b (75 mg, 0.13 mmol)

1

according to general procedure E. H NMR (500 MHz, CD₃OD) δ

6.96 (ddd, J = 11.3, 7.3, 2.2 Hz, 1H), 6.82 (t, J = 5.1, 2.2 Hz, 1H),

6.28 (s, 1H), 6.27 (s, 1H), 3.98 (td, J = 6.6, 6.1, 3.5 Hz, 1H), 3.47 (q,

J = 6.9 Hz, 2H), 3.18 (d, J = 11.0 Hz, 1H), 3.00 (t, J = 7.8 Hz, 2H),

2.82 (d, J = 7.8 Hz, 2H), 2.68−2.59 (m, 1H), 2.54−2.47 (m, 1H),

2.42 (dt, J = 13.5, 7.4 Hz, 1H), 2.35 (dd, J = 11.0, 5.8 Hz, 1H), 2.31

(s, 3H), 2.17 (s, 3H), 2.06−1.97 (m, 1H), 1.60−1.51 (m, 2H), 1.19

(t, J = 7.0 Hz, 3H). ¹³C NMR (126 MHz, CD₃OD) δ 159.3, 157.5,

150.1 (dd, J_C−F= 245.6, 13.3 Hz), 149.7, 147.0 (dd, J_C−F= 242.9, 12.8

Hz), 138.3 (d, J_C−F= 5.2 Hz), 130.3 (d, J_C−F= 12.7 Hz), 125.1 (t,

J_C−F= 3.3 Hz), 114.2 (d, J_C−F= 17.3 Hz), 113.2, 106.7, 76.3, 65.4,

63.9, 62.2, 39.0, 38.4, 37.3, 34.4, 31.6, 28.6, 19.6, 14.3. HRMS-ESI:

calculated for C₂₃H₃₁F₂N₃O [M + H]⁺404.2508, found 404.2510.

(R)-6-(2,3-Diﬂuoro-5-(2-(1-methylpyrrolidin-2-yl)ethyl)-

phenethyl)-4-methylpyridin-2-amine (17). Compound 17 (55 mg,

46% for four steps) was prepared from 42 (173 mg, 0.33 mmol)

1

according to general procedure D. H NMR (500 MHz, MeOD) δ

7.18−7.06 (m, 2H), 6.71 (s, 1H), 6.60 (s, 1H), 3.77−3.67 (m, 1H),

3.35−3.30 (m, 1H), 3.21−3.15 (m, 1H), 3.15−3.02 (m, 4H), 2.95 (s,

3H), 2.79−2.65 (m, 2H), 2.47−2.38 (m, 1H), 2.36 (s, 3H), 2.33−

2.24 (m, 1H), 2.22−2.03 (m, 2H), 1.95−1.80 (m, 2H). ¹³C NMR

(126 MHz, MeOD) δ 157.6, 154.4, 150.1 (dd, J_C−F= 246.9, 13.0 Hz),

147.9, 147.3 (dd, J_C−F= 244.4, 13.9 Hz), 137.8−137.1 (m), 128.7 (d,

J_C−F= 12.4 Hz), 125.3 (d, J_C−F= 3.2 Hz), 115.2 (d, J_C−F= 17.3 Hz),

113.6, 109.6, 68.6, 55.9, 38.5, 32.7, 31.9, 31.2, 29.2, 27.5, 21.1, 20.5.

HRMS-ESI: calculated for C₂₁H₂₇F₂N₃[M + H]⁺360.2246, found

360.2247.

(S)-6-(2,3-Diﬂuoro-5-(2-(1-methylazetidin-2-yl)ethyl)phenethyl)-

4-methylpyridin-2-amine (21). Compound 21 (15 mg, 11% for three

steps) was prepared from 45c (213 mg, 0.4 mmol) according to

general procedure E. ¹H NMR (500 MHz, CD₃OD) δ 6.93 (ddd, J =

11.3, 7.3, 2.2 Hz, 1H), 6.79 (d, J = 6.1 Hz, 1H), 6.27 (s, 2H), 3.38 (td,

J = 7.7, 2.3 Hz, 1H), 3.05 (qd, J = 7.9, 5.9 Hz, 1H), 2.99 (t, J = 7.8 Hz,

2H), 2.89−2.84 (m, 1H), 2.82 (d, J = 7.8 Hz, 2H), 2.58−2.45 (m,

2H), 2.32 (s, 3H), 2.17 (s, 3H), 2.08−2.01 (m, 1H), 1.93−1.84 (m,

1H), 1.83−1.67 (m, 2H). ¹³C NMR (126 MHz, CD₃OD) δ 159.4,

(S)-6-(2,3-Diﬂuoro-5-(2-(1-methylpyrrolidin-2-yl)ethyl)-

phenethyl)-4-methylpyridin-2-amine (18). Compound 18 (64 mg,

38% for four steps) was prepared from 43 (244 mg, 0.47 mmol)

1

according to general procedure D. H NMR (500 MHz, CD₃OD) δ

7.15−7.07 (m, 2H), 6.70 (s, 1H), 6.60 (s, 1H), 3.71 (ddd, J = 5.0, 8.0,

11.6 Hz, 1H), 3.21−3.09 (m, 4H), 3.05 (dd, J = 6.6, 9.3 Hz, 2H), 2.95

(s, 3H), 2.80−2.63 (m, 2H), 2.46−2.38 (m, 1H), 2.34−2.24 (m, 1H),

2.21−2.04 (m, 2H), 1.94−1.80 (m, 2H). ¹³C NMR (126 MHz,

CD₃OD) δ 157.7, 154.4, 150.1 (dd, J_C−F= 246.8, 13.4 Hz), 147.9,

147.3 (dd, J_C−F= 244.4, 13.9 Hz), 137.5, 128.7 (d, J_C−F= 12.5 Hz),

125.2, 115.2 (d, J_C−F= 17.3 Hz), 113.6, 109.6, 68.7, 55.9, 38.5, 32.7,

31.9, 31.2, 29.2, 27.5, 21.1, 20.5. HRMS-ESI: calculated for

C₂₁H₂₇F₂N₃[M + H]⁺360.2246, found 360.2245.

157.6, 150.1 (dd, J_C−F= 245.6, 13.2 Hz), 149.6, 147.0 (dd, J_C−F

=

242.9, 12.8 Hz), 138.2−138.1 (m), 130.3 (d, J_C−F= 12.7 Hz), 125.1

(t, J_C−F= 3.2 Hz), 114.2 (d, J_C−F= 17.1 Hz), 113.2, 106.7, 68.0, 52.4,

43.4, 37.3, 37.2, 30.5, 28.6, 23.6, 19.6. HRMS-ESI: calculated for

C₂₀H₂₅F₂N₃[M + H]⁺346.2089, found 346.2091.

NOS Enzyme Inhibition Assay. The NOS inhibitory activity of 2−

21 was measured by the hemoglobin (Hb) NO capture assay

following a protocol described previously.^26,36Brieﬂy, the assay was

done in 100 mM 4-(2-hydroxyethyl)-1-piperazine-ethanesulfonic acid

(HEPES) buﬀer with 10% glycerol (pH 7.4) at 37 °C in the presence

of 10 μM ^L-Arg, 10 μM H₄B, 100 μM NADPH, 0.83 mM CaCl₂, 320

units/mL calmodulin, and 3 μM human oxyhemoglobin. The

concentration of ^L-Arg, 10 μM, was used as it is suﬃcient not to

cause NOS uncoupling and is close to the K_mvalues of all three NOS

isoforms where competitive inhibitors can be detected eﬀectively. The

assay was performed in 96-well plates using a Biotek Gen5 microplate

reader. NO production was kinetically monitored at 401 nm for 6

min. Rat nNOS,³⁷human nNOS,³⁸murine macrophage iNOS,³⁹

human iNOS,⁴⁰and human eNOS⁴¹are expressed in Escherichia coli

and puriﬁed as previously reported. The inhibition constants (K_i) for

all NOSs were calculated from the IC₅₀values of the dose−response

curves using the Cheng−Prusoﬀ equation: K_i= IC₅₀/(1 + [S]/K_m)⁴²

and K_m(human nNOS: 1.6 μM; rat nNOS: 1.3 μM; murine iNOS:

8.2 μM; bovine eNOS: 1.7 μM; human eNOS: 3.9 μM; human iNOS:

8.0 μM).^43,44Dose−response curves were constructed from seven to

nine test concentrations (200 μM to 50 nM), and IC₅₀values were

calculated by nonlinear regression using GraphPad Prism software.

(S)-6-(2,3-Diﬂuoro-5-(2-(4-methylmorpholin-3-yl)ethyl)-

phenethyl)-4-methylpyridin-2-amine (19). Compound 19 (68 mg,

28% for four steps) was prepared from 45a (346 mg, 0.65 mmol)

1

according to general procedure D. H NMR (500 MHz, CD₃OD) δ

7.19−7.03 (m, 2H), 6.70 (s, 1H), 6.59 (s, 1H), 4.15 (dd, J = 13.2, 3.5

Hz, 1H), 4.09−4.01 (m, 1H), 3.93−3.77 (m, 1H), 3.63 (dd, J = 13.2,

10.4 Hz, 1H), 3.50 (dd, J = 12.9, 2.2 Hz, 1H), 3.30−3.25 (m, 1H),

3.17−3.02 (m, 4H), 2.99 (s, 3H), 2.80−2.59 (m, 2H), 2.36 (s, 3H),

2.32−2.23 (m, 1H), 1.93−1.78 (m, 1H). ¹³C NMR (126 MHz,

CD₃OD) δ 157. 7, 154.4, 150 (dd, J_C−F= 248.2, 13.9 Hz), 147.9,

147.4 (dd, J_C−F= 244.4, 12.6 Hz), 137.2 (d, J_C−F= 5.0 Hz), 128.8 (d,

J_C−F= 12.5 Hz), 125.3 (d, J_C−F= 3.3 Hz), 115.3 (d, J_C−F= 17.8 Hz),

O

DOI: 10.1021/acs.jmedchem.8b02032

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

The calculated standard deviations from dose−response curves of the

assays were less than 10% with all NOSs.

are crucial for the proper folding of the full-length hiNOS that uses

CaM as a tightly bound subunit.

PAMPA-BBB Assay. The PAMPA-BBB assay was performed

following a protocol described previously.²⁶Brieﬂy, the assay was

done in 10 mM phosphate-buﬀered saline (PBS) buﬀer (pH 7.5), and

compounds were tested at a concentration of 200 μM. The donor

plate was ﬁrst coated with 4 μL of porcine brain lipid (20 mg/mL in

dodecane), followed by an addition of 250 μL of a test compound.

The acceptor plate was ﬁlled with 250 μL of PBS, and the donor plate

was carefully placed on top of the acceptor plate to make a

“sandwich”. The plate was incubated at 25 °C for 17 h in a saturated

humidity atmosphere with an orbital agitation at 100 rpm. Verapamil

and theophylline were used as a positive and a negative control,

respectively. After incubation, 150 μL of test solution was taken from

each well from both sides (donor and acceptor) and transferred to the

UV plate for measurement. The eﬀective permeability (P_e) was

c a l c u l a t e d u s i n g t h e f o l l o w i n g e q u a t i o n : ^{4 5}

Inhibitor Complex Crystal Preparation. The sitting drop vapor

diﬀusion method was used to grow crystals at 4 °C for the heme

domains of rnNOS (8 mg/mL containing 20 mM histidine), the

hnNOS K301R/R354A/G357D mutant (10 mg/mL), and heNOS (7

mg/mL). The crystal growth conditions were as described

previously.³⁸The only exception is that the pH for the heNOS

crystal growth is 7.5 rather than 6.5 as mistakenly reported there.

Fresh crystals were ﬁrst passed stepwise through cryoprotectant

solutions. The pH of the ﬁnal soaking solution for rnNOS was

adjusted from 5.8 to 6.5, 7.0 (in 2-(N-morpholino)ethanesulfonic

acid), and 7.5 (in HEPES) and that for hnNOS from 7.2 to 7.5 (in

HEPES), and for heNOS, the Bis−Tris buﬀer at pH 7.5 was

unchanged. At pH 7.5, crystals were soaked with 5−10 mM inhibitor

for 2−4 h at 4 °C before being ﬂash cooled with liquid nitrogen and

stored until data collection. The presence of an acetate ion near the

heme active site in bovine eNOS caused interference in the binding

mode of some inhibitors. The high concentration of magnesium

acetate in the heNOS growth conditions may also introduce an

acetate near the active site that may inﬂuence the binding mode of

inhibitors.⁴⁶To avoid having this acetate in the structure, the 250 mM

magnesium acetate in the crystallization well solution was replaced

with 100 mM MgCl₂in the cryosoaking solution.

X-ray Diﬀraction Data Collection, Data Processing, and

Structural Reﬁnement. The cryogenic (100 K) X-ray diﬀraction

data were collected remotely at the Stanford Synchrotron Radiation

Lightsource (SSRL) or Advanced Light Source (ALS) through data

collection control software Blu-Ice⁴⁷and a crystal-mounting robot.

When a charge-coupled device (CCD) detector was used, 100−125°

of data were typically collected with 0.5° per frame. If a Pilatus pixel

array detector was used, 140−160° of ﬁne-sliced data were collected

with 0.2° per frame. Raw CCD data frames were indexed, integrated,

and scaled using iMOSFLM,⁴⁸but the pixel array data were preferably

processed with XDS⁴⁹and scaled with Aimless.⁵⁰The binding of

inhibitors was detected by initial diﬀerence Fourier maps calculated

with REFMAC.⁵¹The inhibitor molecules were then modeled in

Coot⁵²and reﬁned using REFMAC or PHENIX.⁵³The crystal

packing of the MgCl₂-soaked heNOS crystals was changed slightly,

resulting in a symmetry change from the orthorhombic P2₁2₁2₁

reported previously³⁸to monoclinic P2₁, with a β angle only 0.6−

0.7° oﬀ compared to the original 90°. Therefore, a molecular

replacement calculation with PHASER-MR⁵⁴was needed to solve the

structure. In the P2₁space group, there are two heNOS dimers in the

asymmetric unit. Disordering in portions of inhibitors bound in the

NOS active sites was often observed, sometimes resulting in poor

density quality. However, partial structural features were usually still

visible if the contour level of the sigmaA weighted 2m|Fo| − D|Fc|

map was dropped to 0.5σ, which aﬀorded the building of reasonable

models into the disordered regions. Water molecules were added in

PHENIX and checked by Coot. The TLS⁵⁵protocol was

implemented in the PHENIX reﬁnements with each subunit as one

TLS group. The omit Fo − Fc density maps were calculated by the

Polder map facility in PHENIX for the bound inhibitors.⁵⁶The

reﬁned structures were validated in Coot before deposition in the

Protein Data Bank. The crystallographic data collection and

reﬁnement statistics are reported in Table S2.

Ä

É

Å

Ñ

V_A·V_D

(V_A+ V_D)

V_A+ V_D

(1 − R)·V_D

C_A(t)

C_D(0)

2.303

A·(t − τ_ss

Å

Ñ

Å

P =

·

lg 1 −

·

, where P is the

e

Å

Ñ

e

(

) (

)

Ñ

)

Å

Ñ

Å

Ç

Ñ

Ö

eﬀective permeability (cm s⁻¹); V_Aand V_Dare the volume of the

acceptor and donor wells (0.25 cm³), respectively; C_A(t) is the

concentration of the acceptor well at time t; C_D(0) and C_D(t) are the

concentrations of the donor well at t₀and t, respectively; A is the ﬁlter

well area (0.21 cm²); t is the incubation time (s); τ_ssis the time to

reach a steady state (usually very short compared with the incubation

time); and R is the retention membrane factor and was calculated

Ä

É

Å

Ñ

Å

C_D(t)

C_D(0)

V_A^C_A^(t)

Ñ

Å

Ñ

Å

using the following equation: R = 1 −

−

·

. P_ewas

Å

Ñ

V_DC (0)

D

Å

Ç

Ñ

Ö

reported as an average of triplicate with a standard deviation.

Caco-2 Assay. The bidirectional Caco-2 assay was performed by

Sai Life Sciences, Pune, India, or Chempartner, Shanghai, China.

Brieﬂy, the assay was done in Hank’s balanced salt solution buﬀer (pH

7.4) in 90 min at 37 °C. Compounds were tested at a concentration of

5 μM (0.1% dimethyl sulfoxide). Studied compounds were applied to

either the apical (A → B direction) or the basal (B → A direction)

side. The apparent permeability (P_app) was calculated using the

following equation: P_app= (dQ/dt)/C₀A, where dQ/dt is the change

of test compound concentration in the receiver chamber over time, C₀

is the initial concentration of the compounds in the donor well, and A

is the ﬁlter well area (0.7 cm²). The eﬄux ratio is deﬁned by the ratio

of the apparent permeability of B →A over that of A → B. An ER

value above 2.5 indicates that a compound is possibly a substrate of P-

gp or other active eﬄux transporters.

Expression and Puriﬁcation of Full-Length hiNOS. The plasmid

of human iNOS with an N-terminal 6-His tag built in a pCWori

vector was a generous gift from Dr. Paul R. Ortiz de Montellano’s

laboratory (University of California, San Francisco). The E. coli

competent cells BL21(DE3) were ﬁrst transformed with the plasmid

of human CaM to prepare the CaM-plasmid-containing competent

cells, which were then transformed a second time with the hiNOS

plasmid and grown on an agar plate containing both ampicillin and

chloramphenicol. The resulting colonies were used to inoculate the

overnight Luria-Bertani culture containing both antibiotics. On the

next day, each 1 L of terriﬁc broth culture (containing 0.5 mM CaCl₂)

was inoculated with 2 mL of starter culture in the presence of 100 μg/

mL ampicillin and 35 μg/mL chloramphenicol. The incubation was

continued at 37 °C with 220 rpm shaking until the optical density at

600 nm reached 2.0 or higher. At that time, the protein expression

was induced with 0.5 mM isopropyl β-^D-1-thiogalactopyranoside, 0.4

mM δ-aminolevulinic acid, and 3 μM riboﬂavin. The incubation was

continued for another 40 h at 25 °C and 100 rpm before cell

harvesting. The hiNOS protein puriﬁcation protocols were similar to

those³⁸used for both hnNOS and heNOS through two aﬃnity

columns, Ni nitrilotriacetic acid and 2′,5′-adenosine diphosphate

(2′,5′-ADP) Sepharose, in sequence. To remove the oxidized

nicotinamide adenine dinucleotide phosphate used in the ADP

column elution, the ﬁnal step of puriﬁcation was either a Superdex

200 column (GE Healthcare) or simply a small desalting column

(10DG, BioRad) depending on the purity of the protein from the

earlier two columns. The coexpression and copuriﬁcation with CaM

ASSOCIATED CONTENT

■

S

* Supporting Information

The Supporting Information is available free of charge on the

ACS Publications website at DOI: 10.1021/acs.jmed-

chem.8b02032.

Crystal structures of rat nNOS with compounds 2−7,

10−15, and 17−20 and heNOS with compounds 2, 4, 6,

8, 9, 11, 13, and 14; crystallographic data collection and

reﬁnement statistics for rat nNOS, human nNOS, and

human eNOS; predicted physicochemical properties of

P

DOI: 10.1021/acs.jmedchem.8b02032

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

glycoprotein; BCRP, breast-cancer-resistant protein; ER, eﬄux

ratio; P_e, eﬀective permeability; P_app, apparent permeability

1−21, detailed synthetic procedures and analytical data

for all precursors and intermediates; HPLC traces of

compounds 2−21 (PDF)

REFERENCES

■

Molecular formula strings (CSV)

(1) Gribkoff, V. K.; Kaczmarek, L. K. The need for new approaches

in CNS drug discovery: Why drugs have failed, and what can be done

to improve outcomes. Neuropharmacology 2017, 120, 11−19.

(2) Nutt, D. J.; Attridge, J. CNS drug development in Europe 

past progress and future challenges. Neurobiol. Dis. 2014, 61, 6−20.

(3) Di, L.; Rong, H.; Feng, B. Demystifying brain penetration in

central nervous system drug discovery. J. Med. Chem. 2013, 56, 2−12.

(4) Banks, W. A. From blood−brain barrier to blood−brain

interface: new opportunities for CNS drug delivery. Nat. Rev. Drug

Discovery 2016, 15, 275.

Accession Codes

PDB codes for X-ray structures described in this study are as

follow: hnNOS-2, 6NG1; hnNOS-3, 6NG2; hnNOS-4, 6NG4;

hnNOS-5, 6NG5; hnNOS-6, 6NG6; hnNOS-9, 6NG7;

hnNOS-10, 6NG8; hnNOS-11, 6NGA; hnNOS-12, 6NGB;

hnNOS-13, 6NGC; hnNOS-14, 6NGD; hnNOS-15, 6NGE;

hnNOS-17, 6NGF; hnNOS-18, 6NGH; hnNOS-19, 6NGI;

hnNOS-20, 6NHB; hnNOS-21, 6NHC; rnNOS-2, 6NGJ;

rnNOS-3, 6NGK; rnNOS-4, 6NGL; rnNOS-5, 6NGM;

rnNOS-6, 6NGN; rnNOS-7, 6NGP; rnNOS-8, 6NGQ;

rnNOS-9, 6NGR; rnNOS-10, 6NGS; rnNOS-11, 6NGT;

rnNOS-12, 6NGU; rnNOS-13, 6NGV; rnNOS-14, 6NGW;

rnNOS-15, 6NGX; rnNOS-17, 6NGY; rnNOS-18, 6NGZ;

rnNOS-19, 6NH0; rnNOS-20, 6NHE; rnNOS-21, 6NHD;

heNOS-2, 6NH1; heNOS-4, 6NH2; heNOS-6, 6NH3;

heNOS-8, 6NH4; heNOS-9, 6NH5; heNOS-11, 6NH6;

heNOS-13, 6NH7; heNOS-14, 6NH8; heNOS-21, 6NHF.

̈

(5) Loscher, W.; Potschka, H. Drug resistance in brain diseases and

the role of drug efflux transporters. Nat. Rev. Neurosci. 2005, 6, 591.

(6) Di, L.; Kerns, E. H. Drug-like Properties: Concepts, Structure

Design, and Methods from ADME to Toxicity Optimization, 2nd ed.;

Elsevier: Amsterdam, 2016; p 329.

(7) Rankovic, Z. CNS drug design: balancing physicochemical

properties for optimal brain exposure. J. Med. Chem. 2015, 58, 2584−

2608.

(8) Maccallini, C.; Amoroso, R. Targeting neuronal nitric oxide

synthase as a valuable strategy for the therapy of neurological

disorders. Neural Regener. Res. 2016, 11, 1731−1734.

́

(9) Drechsel, D. A.; Estevez, A. G.; Barbeito, L.; Beckman, J. S.

AUTHOR INFORMATION

■

Nitric oxide-mediated oxidative damage and the progressive demise of

motor neurons in ALS. Neurotoxic. Res. 2012, 22, 251−264.

(10) de la Torre, J. C.; Stefano, G. B. Evidence that Alzheimer’s

disease is a microvascular disorder: the role of constitutive nitric

oxide. Brain Res. Rev. 2000, 34, 119−136.

Corresponding Authors

*E-mail: poulos@uci.edu. Tel: +1 949 824 7020 (T.L.P.).

*E-mail: Agman@chem.northwestern.edu. Tel: +1 847 491

5653. Fax: +1 847 491 7713 (R.B.S.).

(11) Garthwaite, J.; Boulton, C. L. Nitric oxide signaling in the

central nervous system. Annu. Rev. Physiol. 1995, 57, 683−706.

(12) Uehara, T.; Nakamura, T.; Yao, D.; Shi, Z.-Q.; Gu, Z.; Ma, Y.;

Masliah, E.; Nomura, Y.; Lipton, S. A. S-Nitrosylated protein-

disulphide isomerase links protein misfolding to neurodegeneration.

Nature 2006, 441, 513−517.

ORCID

Thomas L. Poulos: 0000-0002-5648-3510

Richard B. Silverman: 0000-0001-9034-1084

Notes

The authors declare no competing ﬁnancial interest.

(13) Li, J.; Wuliji, O.; Li, W.; Jiang, Z.-G.; Ghanbari, H. A. Oxidative

stress and neurodegenerative disorders. Int. J. Mol. Sci. 2013, 14,

24438−24475.

ACKNOWLEDGMENTS

■

(14) Chen, X.; Guo, C.; Kong, J. Oxidative stress in neuro-

degenerative diseases. Neural Regener. Res. 2012, 7, 376−385.

We are grateful for the generous support from the National

Institutes of Health (R01 GM049725 to R.B.S., GM057353 to

T.L.P.). This work made use of the IMSERC at Northwestern

University, which has received support from the Soft and

Hybrid Nanotechnology Experimental (SHyNE) Resource

(NSF NNCI-1542205); the State of Illinois and International

Institute for Nanotechnology (IIN). We also thank the

beamline staﬀ at SSRL and ALS for their assistance during

the remote X-ray diﬀraction data collections. H.L. would like

to thank Carla Plaza for her assistance in protein preparations,

which provided samples for both the structure determinations

and the inhibitory assays.

́

(15) Torreilles, F.; Salman-Tabcheh, S.; Guerin, M.-C.; Torreilles, J.

Neurodegenerative disorders: the role of peroxynitrite. Brain Res. Rev.

1999, 30, 153−163.

(16) Bolanos, J. P.; Almeida, A.; Stewart, V.; Peuchen, S.; Land, J.

̃

M.; Clark, J. B.; Heales, S. J. R. Nitric oxide-mediated mitochondrial

damage in the brain: mechanisms and implications for neuro-

degenerative diseases. J. Neurochem. 1997, 68, 2227−2240.

(17) Mukherjee, P.; Cinelli, M. A.; Kang, S.; Silverman, R. B.

Development of nitric oxide synthase inhibitors for neurodegenera-

tion and neuropathic pain. Chem. Soc. Rev. 2014, 43, 6814−6838.

(18) Silverman, R. B. Design of selective neuronal nitric oxide

synthase inhibitors for the prevention and treatment of neuro-

degenerative diseases. Acc. Chem. Res. 2009, 42, 439−451.

(19) Alderton, W. K.; Cooper, C. E.; Knowles, R. G. Nitric oxide

synthases: structure, function and inhibition. Biochem. J. 2001, 357,

593−615.

ABBREVIATIONS

■

NO, nitric oxide; nNOS, neuronal nitric oxide synthase; iNOS,

inducible nitric oxide synthase; eNOS, endothelial nitric oxide

synthase; rnNOS, rat neuronal nitric oxide synthase; hnNOS,

human neuronal nitric oxide synthase; hiNOS, human

inducible nitric oxide synthase; heNOS, human endothelial

nitric oxide synthase; ^L-Arg, L-arginine; 2′,5′-ADP, 2′,5′-

adenosine diphosphate; CaM, calmodulin; DNA, deoxyribo-

nucleic acid; H₄B, (6R)-5,6,7,8-tetrahydrobiopterin; HEPES,

4-(2-hydroxyethyl)-1-piperazine-ethanesulfonic acid; NADPH,

reduced nicotinamide adenine dinucleotide phosphate; BBB,

blood−brain barrier; CNS, central nervous system; PAMPA,

parallel artiﬁcial membrane permeability assay; P-gp, P-

(20) Moore, P. K.; Handy, R. L. C. Selective inhibitors of neuronal

nitric oxide synthase  is no NOS really good NOS for the nervous

system? Trends Pharmacol. Sci. 1997, 18, 204−211.

(21) Li, H.; Wang, H.-Y.; Kang, S.; Silverman, R. B.; Poulos, T. L.

Electrostatic control of isoform selective inhibitor binding in nitric

oxide synthase. Biochemistry 2016, 55, 3702−3707.

(22) Barbanti, P.; Egeo, G.; Aurilia, C.; Fofi, L.; Della-Morte, D.

Drugs targeting nitric oxide synthase for migraine treatment. Expert

Opin. Invest. Drugs 2014, 23, 1141−1148.

́

(23) Ji, H.; Stanton, B. Z.; Igarashi, J.; Li, H.; Martasek, P.; Roman,

L. J.; Poulos, T. L.; Silverman, R. B. Minimal pharmacophoric

Q

DOI: 10.1021/acs.jmedchem.8b02032

J. Med. Chem. XXXX, XXX, XXX−XXX

Journal of Medicinal Chemistry

Article

elements and fragment hopping, an approach directed at molecular

diversity and isozyme selectivity. design of selective neuronal nitric

oxide synthase inhibitors. J. Am. Chem. Soc. 2008, 130, 3900−3914.

S. K.; Weber, P. C. Structural characterization of nitric oxide synthase

isoforms reveals striking active-site conservation. Nat. Struct. Biol.

1999, 6, 233.

́

(42) Cheng, Y.-C.; Prusoff, W. H. Relationship between the

inhibition constant (Ki) and the concentration of inhibitor which

causes 50 percent inhibition (IC50) of an enzymatic reaction.

Biochem. Pharmacol. 1973, 22, 3099−3108.

(24) Kang, S.; Li, H.; Tang, W.; Martasek, P.; Roman, L. J.; Poulos,

T. L.; Silverman, R. B. 2-Aminopyridines with a truncated side chain

to improve human neuronal nitric oxide synthase inhibitory potency

and selectivity. J. Med. Chem. 2015, 58, 5548−5560.

̈

(43) Leber, A.; Hemmens, B.; Klosch, B.; Goessler, W.; Raber, G.;

́

(25) Wang, H.-Y.; Qin, Y.; Li, H.; Roman, L. J.; Martasek, P.;

Mayer, B.; Schmidt, K. Characterization of recombinant human

endothelial nitric-oxide synthase purified from the yeast pichia

pastoris. J. Biol. Chem. 1999, 274, 37658−37664.

Poulos, T. L.; Silverman, R. B. Potent and selective human neuronal

nitric oxide synthase inhibition by optimization of the 2-amino-

pyridine-based scaffold with a pyridine linker. J. Med. Chem. 2016, 59,

4913−4925.

(44) Fossetta, J. D.; Niu, X. D.; Lunn, C. A.; Zavodny, P. J.; Narula,

S. K.; Lundell, D. Expression of human inducible nitric oxide synthase

in Escherichia coli. FEBS Lett. 1996, 379, 135−138.

(26) Do, H. T.; Wang, H.-Y.; Li, H.; Chreifi, G.; Poulos, T. L.;

Silverman, R. B. Improvement of cell permeability of human neuronal

nitric oxide synthase inhibitors using potent and selective 2-

aminopyridine-based scaffolds with a fluorobenzene linker. J. Med.

Chem. 2017, 60, 9360−9375.

(27) Wager, T. T.; Hou, X.; Verhoest, P. R.; Villalobos, A. Moving

beyond rules: the development of a central nervous system

multiparameter optimization (CNS MPO) approach to enable

alignment of druglike properties. ACS Chem. Neurosci. 2010, 1,

435−449.

(28) Di, L.; Kerns, E. H.; Fan, K.; McConnell, O. J.; Carter, G. T.

High throughput artificial membrane permeability assay for blood−

brain barrier. Eur. J. Med. Chem. 2003, 38, 223−232.

(29) Di, L.; Kerns, E. H.; Bezar, I. F.; Petusky, S. L.; Huang, Y.

Comparison of blood−brain barrier permeability assays: in situ brain

perfusion, MDR1-MDCKII and PAMPA-BBB. J. Pharm. Sci. 2009, 98,

1980−1991.

(30) Davis, T. P.; Sanchez-Covarubias, L.; Tome, M. E. P-

glycoprotein trafficking as a therapeutic target to optimize CNS

drug delivery. Adv. Pharmacol. 2014, 71, 25−44.

(31) Patil, A. G.; D’Souza, R.; Dixit, N.; Damre, A. Validation of

quinidine as a probe substrate for the in vitro P-gp inhibition assay in

Caco-2 cell monolayer. Eur. J. Drug Metab. Pharmacokinet. 2011, 36,

115.

(32) Delker, S. L.; Xue, F.; Li, H.; Jamal, J.; Silverman, R. B.; Poulos,

T. L. Role of zinc in isoform-selective inhibitor binding to neuronal

nitric oxide synthase. Biochemistry 2010, 49, 10803−10810.

(33) Perry, J. M.; Zhao, Y.; Marletta, M. A. Cu²⁺and Zn²⁺inhibit

nitric-oxide synthase through an interaction with the reductase

domain. J. Biol. Chem. 2000, 275, 14070−14076.

(34) Wang, Z.-Q.; Tejero, J.; Wei, C.-C.; Haque, M. M.; Santolini, J.;

Fadlalla, M.; Biswas, A.; Stuehr, D. J. Arg375 tunes tetrahydrobiopter-

in functions and modulates catalysis by inducible nitric oxide

synthase. J. Inorg. Biochem. 2012, 108, 203−215.

(35) St. Jean, D. J.; Fotsch, C. Mitigating heterocycle metabolism in

drug discovery. J. Med. Chem. 2012, 55, 6002−6020.

(36) Hevel, J. M.; Marletta, M. A. Nitric-oxide synthase assays.

Methods Enzymol. 1994, 233, 250−258.

(37) Roman, L. J.; Sheta, E. A.; Martasek, P.; Gross, S. S.; Liu, Q.;

Masters, B. S. High-level expression of functional rat neuronal nitric

oxide synthase in Escherichia coli. Proc. Natl. Acad. Sci. 1995, 92,

8428−8432.

(38) Li, H.; Jamal, J.; Plaza, C.; Pineda, S. H.; Chreifi, G.; Jing, Q.;

Cinelli, M. A.; Silverman, R. B.; Poulos, T. L. Structures of human

constitutive nitric oxide synthases. Acta Crystallogr., Sect. D: Biol.

Crystallogr. 2014, 70, 2667−2674.

(39) Hevel, J. M.; White, K. A.; Marletta, M. A. Purification of the

inducible murine macrophage nitric oxide synthase. Identification as a

flavoprotein. J. Biol. Chem. 1991, 266, 22789−22791.

́

̋ ́

(45) Muller, J.; Esso, K.; Dargo, G.; Konczol, A.; Balogh, G. T.

̈

̈ ̈

Tuning the predictive capacity of the PAMPA-BBB model. Eur. J.

Pharm. Sci. 2015, 79, 53−60.

(46) Cinelli, M. A.; Li, H.; Pensa, A. V.; Kang, S.; Roman, L. J.;

́

Martasek, P.; Poulos, T. L.; Silverman, R. B. Phenyl ether- and aniline-

containing 2-aminoquinolines as potent and selective inhibitors of

neuronal nitric oxide synthase. J. Med. Chem. 2015, 58, 8694−8712.

(47) McPhillips, T. M.; McPhillips, S. E.; Chiu, H.-J.; Cohen, A. E.;

Deacon, A. M.; Ellis, P. J.; Garman, E.; Gonzalez, A.; Sauter, N. K.;

Phizackerley, R. P.; Soltis, S. M.; Kuhn, P. Blu-ice and the distributed

control system: software for data acquisition and instrument control

at macromolecular crystallography beamlines. J. Synchrotron Radiat.

2002, 9, 401−406.

(48) Battye, T. G. G.; Kontogiannis, L.; Johnson, O.; Powell, H. R.;

Leslie, A. G. W. iMOSFLM: a new graphical interface for diffraction-

image processing with MOSFLM. Acta Crystallogr., Sect. D: Biol.

Crystallogr. 2011, 67, 271−281.

(49) Kabsch, W. XDS. Acta Crystallogr., Sect. D: Biol. Crystallogr.

2010, 66, 125−132.

(50) Evans, P. Scaling and assessment of data quality. Acta

Crystallogr., Sect. D: Struct. Biol. 2006, 62, 72−82.

(51) Murshudov, G. N.; Vagin, A. A.; Dodson, E. J. Refinement of

macromolecular structures by the maximum-likelihood method. Acta

Crystallogr., Sect. D: Struct. Biol. 1997, 53, 240−255.

(52) Emsley, P.; Cowtan, K. Coot: model-building tools for

molecular graphics. Acta Crystallogr., Sect. D: Biol. Crystallogr. 2004,

60, 2126−2132.

(53) Adams, P. D.; Afonine, P. V.; Bunkoczi, G.; Chen, V. B.; Davis,

I. W.; Echols, N.; Headd, J. J.; Hung, L.-W.; Kapral, G. J.; Grosse-

Kunstleve, R. W.; McCoy, A. J.; Moriarty, N. W.; Oeffner, R.; Read, R.

J.; Richardson, D. C.; Richardson, J. S.; Terwilliger, T. C.; Zwart, P.

H. PHENIX: a comprehensive Python-based system for macro-

molecular structure solution. Acta Crystallogr., Sect. D: Struct. Biol.

2010, 66, 213−221.

(54) McCoy, A. J.; Grosse-Kunstleve, R. W.; Adams, P. D.; Winn, M.

D.; Storoni, L. C.; Read, R. J. Phaser crystallographic software. J. Appl.

Crystallogr. 2007, 40, 658−674.

(55) Winn, M. D.; Isupov, M. N.; Murshudov, G. N. Use of TLS

parameters to model anisotropic displacements in macromolecular

refinement. Acta Crystallogr., Sect. D: Struct. Biol. 2001, 57, 122−133.

(56) Liebschner, D.; Afonine, P. V.; Moriarty, N. W.; Poon, B. K.;

Sobolev, O. V.; Terwilliger, T. C.; Adams, P. D. Polder maps:

improving OMIT maps by excluding bulk solvent. Acta Crystallogr.,

Sect. D: Struct. Biol. 2017, 73, 148−157.

(40) Li, H.; Raman, C. S.; Glaser, C. B.; Blasko, E.; Young, T. A.;

Parkinson, J. F.; Whitlow, M.; Poulos, T. L. Crystal structures of zinc-

free and -bound heme domain of human inducible nitric-oxide

synthase: implications for dimer stability and comparison with

endothelial nitric-oxide synthase. J. Biol. Chem. 1999, 274, 21276−

21284.

(41) Fischmann, T. O.; Hruza, A.; Niu, X. D.; Fossetta, J. D.; Lunn,

C. A.; Dolphin, E.; Prongay, A. J.; Reichert, P.; Lundell, D. J.; Narula,

R

DOI: 10.1021/acs.jmedchem.8b02032

J. Med. Chem. XXXX, XXX, XXX−XXX

Article Doi

Optimization of Blood-Brain Barrier Permeability with Potent and Selective Human Neuronal Nitric Oxide Synthase Inhibitors Having a 2-Aminopyridine Scaffold

DOI: 10.1021/acs.jmedchem.8b02032

Source and publish data:

Authors:

Article abstract of DOI:10.1021/acs.jmedchem.8b02032

Full text of DOI:10.1021/acs.jmedchem.8b02032

Products guided by the article

R&D Labs maybe for 438050-04-5

Relevant to this article

Hot Product