Synthesis and biological evaluation of novel 2,4-diaminoquinazoline derivatives as SMN2 promoter activators for the potential treatment of spinal muscular atrophy

J. Med. Chem. 2008, 51, 449–469

449

Synthesis and Biological Evaluation of Novel 2,4-Diaminoquinazoline Derivatives as SMN2

Promoter Activators for the Potential Treatment of Spinal Muscular Atrophy

John Thurmond,^†Matthew E. R. Butchbach,^‡Marty Palomo,^†Brian Pease,^†Munagala Rao,^†Louis Bedell,^†Monica Keyvan,^†

Grace Pai,^†Rama Mishra,^†Magnus Haraldsson,^§Thorkell Andresson,^§Gisli Bragason,^§Margret Thosteinsdottir,^§

Jon Mar Bjornsson,^§Daniel D. Coovert,^‡Arthur H. M. Burghes,^‡Mark E. Gurney, and Jasbir Singh*^,†

deCODE chemistry, Inc., 2501 DaVey Road, Woodridge, Illinois 60517, deCODE genetics, Inc., Sturlugata 8, IS-101, ReykjaVik, Iceland, Department

of Molecular and Cellular Biochemistry, College of Medicine and Public Health, The Ohio State UniVersity, Columbus, Ohio 43210

ReceiVed December 23, 2006

Proximal spinal muscular atrophy (SMA) is an autosomal recessive disorder characterized by death of motor

neurons in the spinal cord that is caused by deletion and/or mutation of the survival motor neuron gene

(SMN1). Adjacent to SMN1 are a variable number of copies of the SMN2 gene. The two genes essentially

differ by a single nucleotide, which causes the majority of the RNA transcripts from SMN2 to lack exon 7.

Although both SMN1 and SMN2 encode the same Smn protein amino acid sequence, the loss of SMN1 and

incorrect splicing of SMN2 have the consequence that Smn protein levels are insufﬁcient for the survival of

motor neurons. The therapeutic goal of our medicinal chemistry effort was to identify small-molecule activators

of the SMN2 promoter that, by up-regulating gene transcription, would produce greater quantities of full-

length Smn protein. Our initial medicinal chemistry effort explored a series of C5 substituted benzyl ether

based 2,4-diaminoquinazoline derivatives that were found to be potent activators of the SMN2 promoter;

however, inhibition of DHFR was shown to be an off-target activity that was linked to ATP depletion. We

used a structure-guided approach to overcome DHFR inhibition while retaining SMN2 promoter activation.

A lead compound 11a was identiﬁed as having high potency (EC₅₀) 4 nM) and 2.3-fold induction of the

SMN2 promoter. Compound 11a possessed desirable pharmaceutical properties, including excellent brain

exposure and long brain half-life following oral dosing to mice. The piperidine compound 11a up-regulated

expression of the mouse SMN gene in NSC-34 cells, a mouse motor neuron hybrid cell line. In type 1 SMA

patient ﬁbroblasts, compound 11a induced Smn in a dose-dependent manner when analyzed by immuno-

blotting and increased the number of intranuclear particles called gems. The compound restored gems numbers

in type I SMA patient ﬁbroblasts to levels near unaffected genetic carriers of SMA.

Introduction

SMA is caused by mutation or homozygous deletion of the

telomeric copy of the survival of motor neuron gene (SMN1)

on chromosome 5q13.⁴A second, duplicated copy of the gene,

designated as SMN2, is found adjacent on chromosome 5 in

the direction of the centromere.⁴The transcripts arising from

the SMN1 gene are spliced correctly to yield full-length Smn

mRNA. In contrast, a translationally silent single nucleotide

mutation (C f T) in exon 7 of the SMN2 gene leads to mis-

splicing of the SMN2 mRNA, deleting exon 7 from 80% to 90%

of SMN2 transcripts (termed the ∆7 form).^4–6Only ∼10–20%

of SMN2 gene transcripts correctly splice to produce full-length

mRNA.^4–8The absence of exon 7 from the SMN2 mRNA results

in the synthesis of truncated Smn protein with reduced ability

for self-oligomerization, causing instability and degradation of

the protein.^9,10Therefore, SMA patients, who lack the SMN1

gene, have low (full-length) Smn protein levels that are

insufﬁcient for motor neuron survival.^11,12

The severity of SMA disease correlates with the number of

SMN2 gene copies present in the affected child. The copy

number of the human SMN2 gene varies in the general

population and therefore the corresponding amount of full-length

Smn protein produced from these genes.^3,13–15The correlation

is not absolute because it is important to consider the number

of intact SMN2 genes and how much Smn protein the particular

SMN2 gene produces.^15,16In mice there is only one SMN

gene^17,18that is equivalent to SMN1. Homozygouos deletion

of the single mouse SMN gene results in embryonic lethality.¹⁹

A similar situation most likely exists in humans because no

individual entirely lacking SMN genes has been reported, but

Spinal muscular atrophy (SMA) is characterized by degenera-

tion of motor neurons in the anterior horn of the spinal cord,

atrophy of skeletal muscle, and consequent weakness. SMA

afﬂicts approximately 1 in 6000–8000 newborns and is the

leading hereditary cause of mortality in infants.^1,2The carrier

frequency for SMA is about 1 in 40 individuals.^2,3SMA is

classiﬁed into four types (types I-IV) according to age of onset

and severity of disease. Type I (acute, also called Werdnig-

Hoffmann disease) SMA typically presents itself within 6

months after birth; children never sit or stand and usually die

before the age of 2 without respiratory intervention. Type II

(intermediate) SMA patients have an onset after 6 months of

age and are usually diagnosed by 15 months. These children

may learn to sit but cannot stand or walk. Type III (mild) SMA,

often referred to as Kugelberg-Welander disease or juvenile

spinal muscular atrophy, is usually diagnosed between 2 and

17 years of age. These patients are able to sit and walk but

often become wheelchair-bound as the course of disease

progresses. Type IV SMA patients (adult form of SMA) are

characterized by age of onset at >30 years and show very mild

signs of muscle weakness.

* To whom correspondence should be addressed. Phone: 630-783-4915.

Fax: 630-783-4646. E-mail: jsingh@decode.com.

†

deCODE chemistry, Inc.

The Ohio State University.

deCODE genetics, Inc.

‡

§

10.1021/jm061475p CCC: $40.75

2008 American Chemical Society

Published on Web 01/19/2008

450 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 3

Thurmond et al.

good pharmacokinetic properties, and an increase in gems

numbers in cultured type I SMA patient ﬁbroblasts.

Chemistry

The 2,4-diaminoquinazoline derivatives functionalized at the

C5 position were prepared via cyclization of 6-substituted

2-aminobenzonitriles with chloroformamidine hydrochloride in

diglyme.³⁴The 6-substituted 2-nitrobenzonitriles were prepared

by displacement of one of the nitro groups in 2,6-dinitroben-

zonitrile. Reduction of the 6-substituted 2-nitrobenzonitriles to

6-substituted 2-aminobenzonitrile was accomplished using iron

powder in hydrochloric acid. Alternatively, a more efﬁcient

approach starting from 2,6-diﬂuorobenzonitrile (2), analogus to

the procedure reported by Hynes,³⁵was utilized to prepare

5-substituted 2,4-diaminoquinazolines 1, 5–9, and 11–13 as

outlined in Scheme 1. Reaction of 2,6-diﬂuorobenzonitrile (2)

with a wide variety of alcohols (aryl or heteroarylmethyl

alcohols; primary, secondary, or tertiary alkyl alcohols; R-alky-

larylmethyl alcohols, bicyclic secondary alcohols; and piperidine

methanol) (3) provided 2-ﬂuoro-6-(arylalkoxy/phenoxy)ben-

zonitriles (4). Cyclization of 2-ﬂuoro-6-(arylalkoxy/phenoxy)-

benzonitriles (4) to the corresponding 5-substituted 2,4-

diaminoquinazolines 1, 5–9, and 11–13 was achieved by heating

with guanidine carbonate in N,N-dimethylacetamide (Scheme

1). The alcohols used for reaction with 2 were either obtained

from commercial sources or prepared as outlined in Schemes

2–4.

Figure 1. HTS hits.³³

chromosomes lacking SMN genes do exist. In addition, deleting

the mouse SMN gene in a speciﬁc tissue causes destruction of

that speciﬁc cell type, again indicating SMN is an essential

gene.^20–22To rescue the embryonic lethality, SMN2 transgenes

have been introduced into SMN null mice, and indeed, two

copies of SMN2 result in severe SMA whereas eight copies

result in rescue of phenotype.^23–25Clearly, expression of full

length Smn at the right time can rescue the SMA phenotype in

mice. In addition, overexpression of the Smn form lacking exon

7 is not harmful but beneﬁcial to the SMA phenotype.²⁶

Expression of a mild missense mutation on the severe SMA

background results in mild SMA mice.²⁷Therefore, a therapeutic

agent that increases human SMN2 gene expression and conse-

quent production of (full-length) Smn protein has the potential

to reduce the clinical severity of SMA.

The therapeutic goal of increasing the level of Smn protein

might be achieved by (i) enhancing SMN2 gene transcription

or (ii) suppressing defective splicing of the SMN2 mRNA,

thereby increasing the number of full-length transcripts, or

alternatively a combination of steps i and ii. Small-molecule

activators of the SMN2 gene promoter, which enhance Smn

expression, thus represent a promising strategy for the treatment

of SMA. A series of compounds such as butyrate, 4-phenyl-

butrate, valproate, and aclarubicin have been reported to increase

Smn.^28–32However, these compounds need to be used at high

concentration, have poor metabolic properties, are relatively

nonspeciﬁc, or have toxicity problems.

Reduction of ketones 14 with a chiral reducing agent³⁶in

either catalytic or equimolar amounts of a chiral reagent

provided alcohols 15 and 17 of known chirality with very high

% ee (Scheme 2). Reaction of the chiral alcohols 15 and 17

with 2,6-diﬂuorobenzonitrile followed by cyclization with

guanidine carbonate yielded the chiral 5-substituted quinazolines

with known chirality. The enantiomeric purity of the ﬁnal

products (5-R-methylaryloxy 2,4-diaminoquinazolines (R)-7a

and (R)-7u, and (S)-7a and (S)-7u) was determined by chiral

HPLC (see Experimental section).

Cyclic tertiary alcohols (20b,c) were prepared by addition

of 3-chlorophenylmagnesium bromide to the corresponding

cyclic ketone (19b,c) (Scheme 3). These tertiary alcohols were

converted to the target products 8a-c using procedures outlined

in Scheme 1. The substituted 1-indanols (23) were prepared by

intramolecular cyclization of the phenyl substituted phenethyl

acid chloride (21) followed by reduction of the resulting

1-indanones (22, Scheme 4). 1-Indanols were reacted with 2

followed by cyclization of the corresponding intermediate 4 to

furnish the quinazoline analogues 6a,b.

Reverse benzyl analogues 10a-k were prepared from 2-bro-

momethyl-6-nitrobenzonitrile (Scheme 5). Substituted phenols

were reacted with 2-bromomethyl-6-nitrobenzonitrile 24 in the

presence of potassium carbonate to form the respective phe-

noxynitrobenzonitriles 25. The nitro group was reduced with

stannous chloride to provide aminobenzonitrile 26. Guanidine

hydrochloride was used for the cyclization reaction to afford

quinazoline derivatives 10a-k.

N-Substituted piperidine methyl ether derivatives were pre-

pared by functionalization of the piperidine nitrogen either

following the cyclization step to form quinazoline (route A) or

prior to quinazoline ring formation (route B). Route A was

amenable to parallel synthesis. It provided for rapid SAR studies.

For route A, the ﬂuorobenzonitrile intermediate 29 was cyclized

to form the quinazoline ring. Subsequent BOC deprotection

provided the free piperidine derivative 31. This key intermediate

was reacted with benzyl halides, acid chlorides, and sulfonyl

A cell-based assay for SMN2 promoter activation and the

details of an ultra-high-throughput screening (uHTS) campaign

of 558 000 compounds were described by Jarecki et al.³³This

effort led to the identiﬁcation of several small-molecule hits

representing nine scaffolds.³³The quinazoline-based compounds

1a and 1b (Figure 1) were among the most potent hits from the

uHTS campaign. When tested on ﬁbroblasts derived from SMA

type I patients, these compounds increased the relative abun-

dance of the full-length SMN transcript, increased Smn protein

levels, and increased the numbers of intranuclear structures

containing Smn, termed gems or Cajal bodies. Gems are

intranuclear concentrations of Smn found in most cells. Smn

functions in the assembly of small nuclear ribonucleoprotein

or snRNPs, which are important for splicing exons to create

mRNA. On the basis of our analysis of the chemical tractability

of the different uHTS hit series and given that the quinazoline-

based hits also possessed the attributes mentioned above, we

selected the quinazoline scaffold for SAR exploration. In this

paper, we describe the SAR and lead optimization of quinazoline

based activators of the SMN2 promoter. This has led to the

development of a series of 2,4-diaminoquinazoline analogues

with high potency in the cell-based SMN2 promoter assay, the

ability to activate the endogenous mouse SMN gene, oral

bioavailability, permeability across the blood-brain barrier,

Diaminoquinazoline DeriVatiVes as SMN2 Promoter ActiVator

Journal of Medicinal Chemistry, 2008, Vol. 51, No. 3 451

Scheme 1

Scheme 2^a

^aReagents: (a) borane-THF, (S)-MeCBS; (b) NaH, DMF, 2; (c) guanidine carbonate, DMA; (d) borane-THF, (R)-MeCBS.

Scheme 3^a

Scheme 4^a

^aReagents: (a) 2-Chlorophenylmagnesium bromide, THF.

^aReagents: (a) SOCl₂; (b) AlCl₃; CH₂Cl₂; (c) NaBH₄, MeOH.

chlorides to yield 11, 12, and 13, respectively. 4-Piperidinemeth-

anol was BOC-protected using di-tert-butyl dicarbonate in

dichloromethane. Subsequently, alcohol 28 was coupled to 2,6-

diﬂuorobenzonitrile to provide compound 29. For route B, the

BOC group was deprotected to give intermediate 32. It was

reacted with acid chlorides, sulfonyl chlorides, and benzyl

halides to give the N-functionalized piperidine intermediate 33.

6-Substituted 2-ﬂuorobenzonitrile 33x was further cyclized with

guanidine carbonate to yield 5-substituted 2,4-diaminoquinazo-

lines 11–13 (Scheme 6). This route was used for scale up of

selected analogues for generation of samples for PK studies.

Results and Discussions

uHTS Hits. Following the uHTS campaign described by

Jarecki et al.³³and before starting the focused medicinal

chemistry program, screening hits were resynthesized or ob-

tained from commercial sources. Their identity was conﬁrmed

by NMR, LC/MS, HPLC, and elemental analysis, whenever

appropriate. Retesting of hits 1a-d, (Table 1) in the NSC-34

SMN2 promoter assay veriﬁed that the EC₅₀obtained in our

assays compared well (r²) 0.98) with the corresponding data

reported by Jarecki et al.³³

Figure 2. Example of dose-response curve and key assay parame-

ters.

452 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 3

Thurmond et al.

Scheme 5^a

Table 1. Activity of C5-O-Aryl Analogues in the SMN2 Promoter

Assay

compd

Ar

EC₅₀(SD), µM

MaxFoldInd

1a

1c

1d

1e

1f

1g

1h

1i

4-Me phenyl

4-Cl phenyl

4-F phenyl

2-Cl phenyl

3-Cl phenyl

1-(5-isoquinoline)

2-(5-isoquinoline

1-(8-isoquinoline)

1.21

2.06(0.67)

1.32

0.51(0.05)

0.48(0.35)

1.03(0.71)

1.26(1.34)

2.44(0.58)

1.6

1.84

1.7

1.76

1.74

2.04

2.3

^aReagents: (a) K₂CO₃, ROH; (b) SnCl₂, HCl; (c) chloroformamidine

hydrochloride, diglyme.

Scheme 6^a

1.59

Table 2. Activity of C5-O-CH₂-Aryl Analogues in the SMN2 Promoter

Assay

compd

Ar

2-Cl phenyl

3-Cl phenyl

4-Cl phenyl

2-Me phenyl

3-Me phenyl

4-Me phenyl

2-F phenyl

EC₅₀(SD), µM

MFI

5a

5b

5c

5d

5e

5f

5g

5h

5i

5j

5k

5l

5m

5n

5o

5p

5q

5r

5s

0.14(0.04)

0.078(0.03)

0.18(0.1)

0.28

0.12

2.0

1.74

1.84

1.6

1.7

2.04

2.3

1.59

1.8

1.6

1.7

1.6

1.5

1.6

1.9

1.8

2.0

1.8

1.9

2.1

1.9

1.7

2.2

2.3

0.22

0.052(0.04)

0.055(0.01)

0.053(0.04)

0.065(0.04)

0.066(0.04)

0.30(0.23)

0.016(0.01)

0.023(0.02)

0.017(0.01)

0.014(0.01)

0.047(0.03)

0.12(0.06)

0.19(0.05)

0.25(0.01)

0.024(0.01)

0.073(0.02)

0.30(0.2)

3-F phenyl

4-F phenyl

2-OMe phenyl

3-OMe phenyl

4-OMe phenyl

2-CF₃phenyl

3-CF₃phenyl

4-CF₃phenyl

3,5-F2 phenyl

3,4-F2 phenyl

2,5-F2 phenyl

2,4-F2 phenyl

2,6-F2 phenyl

2-Py

3-Py

2-thiophenyl

2-furanyl

3-(5-methylisoxazolyl)

3-(1-isoquinoline)

^aReagents: (a) (BOC)₂O, TEA, CH₂Cl₂; (b) NaH, DMF; (c) HCl/p-

dioxane; (d) acid chloride, sulfonyl chloride, or benzyl halide, base; (e)

guanidine carbonate, DMA.

A sample dose–response curve for SMN2 promoter induction

in NSC-34 cells is shown in Figure 2. Two parameters used to

drive our SAR studies were (i) “EC₅₀” value, reﬂecting potency

of a compound, and (ii) “MaxFoldInd” (MFI), referring to the

maximum observed efﬁcacy for a promoter induction compared

to nonactivated cells. Both EC₅₀and maximum fold induction

(MFI) are highlighted for illustration in Figure 2. Lead

optimization was expected to provide maximum fold induction

at low (EC₅₀) concentration. Therefore, for generating a robust

SAR, we were optimizing both parameters.

5t

5u

5v

5w

5x

5y

5z

0.28(0.09)

0.22(0.06)

10-fold loss in activity (EC₅₀) 2.4 µM). The respective

sulfoxide was inactive.³⁷

SAR of C5-OAr and C5-OCH₂Ar. In addition to previously

reported³³tolyloxy analogue 1a, we initially prepared several

additional 5-aryloxyquinazoline analogues. As shown in Table

1, all of the substituted phenyl 1a–f and heteroaromatic 1g–I

analogues showed submicromolar to micromolar EC₅₀, chloro

analogues 1e and 1f being the best (EC₅₀) 0.52 and 0.48 µM,

respectively). Thioether analogue 1b and the corresponding

sulfoxide and sulfone were also prepared. Thioether 1b featured

EC₅₀) 0.22 µM, whereas the corresponding sulfone showed a

In order to explore expanded conformational space and

the electronic nature of the putative binding pocket around

C5 substituents, a series of 5-benzyloxy analogues were

prepared. As shown in Table 2, benzyloxy analogues provided

a signiﬁcant improvement of promoter activity compared to

the corresponding phenoxy series (see, for example, 1f vs

5b, 1c vs 5c, and 1d vs 5i). For the benzyloxy series,

incorporation of halogen and alkyl substituents enhanced

promoter activity. The observed activity pattern for halogen

Diaminoquinazoline DeriVatiVes as SMN2 Promoter ActiVator

Journal of Medicinal Chemistry, 2008, Vol. 51, No. 3 453

Table 3. Activity of C5-O-CH₂-Non-Aryl Analogues in the SMN2

Promoter Assay

Table 4. Activity of C5-O-CH(R)-Phenyl Analogues in the SMN2

Promoter Assay

SMN promoter

EC₅₀(SD), µM

compd

R1

R2

chirality^a

5b

7a

5c

7b

7c

7d

0.078(0.03)

0.021(0.01)

0.18(0.1)

0.080(0.06)

0.17

H

Me

H

Me

Et

tert-Bu

3-Cl

4-Cl

compd

non-aryl

EC₅₀(SD), µM

MaxFoldInd

RS

6a

6b

6c

6d

6e

6f

6g

6h

6j

Me

Et

n-Pr

n-butyl

n-octyl

i-Pr

sec-Bu

tert-Bu

0.17(0.04)

0.18(0.04)

0.14(0.04)

0.19(0.04)

0.55(0.4)

0.18(0.1)

0.097(0.04)

0.65

2.71

0.47

0.14(0.1)

0.48(0.06)

2.1

2.2

2.1

1.7

1.8

2.0

2.1

1.9

2.1

RS

7.76

^aCompounds are racemates.

Table 5. Activity of Racemic C5-O-CH(CH₃)-Phenyl Analogues in the

SMN2 Promoter Assay

cyclohexyl

cyclopentyl

2-(1-methylpiperidinyl)

2-THP

6k

6l

6m

(chloro), methyl, and methoxy substituents was meta > ortho

> para isomers (Table 2). In general, meta-substituted

benzyloxy analogues featured better activity. Both electron-

withdrawing and electron-donating meta-substituents groups

(e.g., Cl, OMe, and Me) were well tolerated. The corre-

sponding para isomers showed lower activity (e.g., 5b vs

5c, 2e vs 5f, and 5k vs 5l). Notably, triﬂuoromethyl

substituents enhanced activity compared with the correspond-

ing methyl isomer, suggesting the lipophilic nature of the

binding pocket. Considering that all monoﬂuoro isomers (2-,

3-, and 4-) had essentially identical activity, a number of

diﬂuoro analogues were synthesized and evaluated in the

promoter assay. The 3,5-diﬂuoro analogue 5p was 8 and 14

times more potent than the 2,6-diﬂuoro analogue 5t and 2,4-

diﬂuoro 5s, respectively. Compared to the substituted phenyl-

CH₂analogues, heterocyclic-CH₂ethers (thiophene 5w,

furane 5x, and isoxazole 5y) generally had poor activity,

3-pyridyl analogue 5v being an exception. The fused 3-py-

ridyl system, 5z, displayed a 10-fold loss in activity. Several

C5-OCH₂Ar/heterocyclic series provided potent analogues

(EC₅₀< 100 nM); however, the majority of these (with the

exception of 5g and 5v) did not afford more than 2-fold

induction of the SMN2 promoter above background.

With several potent analogues derived from the 5-benzyloxy

series in hand, we synthesized 24 analogues varying the

substitution pattern at the 6, 7, and 8 positions of the quinazoline

core. However, all of these analogues were inactive in the

promoter assay. In addition to the -O-Ar/heterocycle derived

analogues, -S(O)_n, n ) 0–2 (as mentioned above) and aza-

linked C5 analogues (data not shown) were also prepared. These

analogues were signiﬁcantly less active than the respective C5

benzyloxy analogues discussed above.

Nonaryl Ether Derivatives. 5-Ethoxyquinazoline 6b was as

active as reported by Jarecki et al.³In follow-up SAR studies,

a number of linear and branched alkoxy (C5-O-R) analogues

were prepared. Substitution with linear alkyl radicals (6a–d)

did not improve activity. Cycloalkyl and heterocyclic substi-

tutents (6j–m) also afforded disappointing results. On the other

hand, branched alkyl derived ethers improved biological activity;

e.g., (racemic) sec-Bu 6g had EC₅₀of ∼100 nM (Table 3).

r-Alkyls as Conformationally Constrained Analogues. On

the basis of the analysis of benzylic ethers and branched-alkyl

SMN promoter

EC₅₀(SD), µM

compd

R

chirality^a

7a

7b

7e

7f

7g

7j

7k

7m

7n

7p

7q

7r

7s

0.021 (0.01)

0.080 (0.06)

0.047 (0.03)

0.055 (0.04)

0.035 (0.04)

0.018 (0.01)

0.050 (0.07)

0.32 (0.5)

0.05 (0.04)

1.01 (0.3)

0.29 (0.2)

1.13

3-Cl

4-Cl

3-CF₃

3-F

3,5-F2

4-F

4-CF₃

3,4-Cl₂

3-F

RS

S

2-CF₃

2-Cl

3,4-(OCH₂O)

2-naphth

1-naphth

0.44 (0.4)

2.39

7t

S

^aWith the exception of 7s and 7t, which have (S)-chirality, all compounds

are racemates.

ethers, it was postulated that the conformational constraints

imposed by the R-substituent on the benzylic carbon might allow

preorganization of the C5 substituents. This may enhance

binding and show improved activity in the promoter assay.

Following this hypothesis, we prepared a series of phenyl

substituted R-methylbenzyl ether analogues. SAR data (Table

4) showed that an R-methyl substituent indeed improved potency

(for example, 5b vs 7a and 5c vs 7b). Larger alkyl substituents

were not tolerated. Both R-ethyl (7c) and R-tert-butyl (7d)

substitution resulted in a signiﬁcant drop in promoter activity.

For example, the tert-butyl analogue was ∼100-fold less active

than the methyl analogue 7b.

Encouraged by our initial results for chlorobenzyl analogues

(Table 4), we prepared a number of additional R-methyl

substituted molecules. As shown in Table 5, a number of 3-

and 4-substituted phenyl compounds also showed good activity.

Analogues 7p and 7q bearing ortho substituents displayed loss

in activity. Analogues modiﬁed with bicyclic aromatics 7r, 7s,

and 7t allowed for the exploration of conformational space.

Similar to the ortho substituents, 1-naphthyl analogue 7t led to

the greatest loss in activity compared to the 2-naphthyl analogue

7s. Overall, a number of 3- and 4-substituted (racemic) R-methyl

benzyl ethers provided SMN2 activators (as indicated by the

promoter assay) with EC₅₀in the 10–100 nM range (Table 5).

454 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 3

Thurmond et al.

Table 6. Activity of R,R-Disubstituted Benzyl Ether Analogues in the

SMN2 Promoter Assay

carbon of the phenyl group was recorded (Figure 3). Vectors

corresponding to conformations within 3 kcal/mol of the lowest

energy conformation for each compound were plotted. Figure

4 shows that regions of vectors shared by more active benzyl

ether analogues 5b [magenta] and 7a [blue] differ signiﬁcantly

from the phenyl ether analogue 1c [cyan]. For clarity, vectors

for the two indanol-derived products are shown in Figure 4b.

On the basis of our data, active 9a and inactive 9b compounds

access distinctly different conformational space. The region

highlighted with the yellow oval (Figure 4c) represents the

conformational space that is shared by all three active analogues

5b, 7a, and 9a. For the inactive indane analogue 9b, this

conformational space is not accessible. Thus, 7a represents the

potential “bioactive” orientation of C5 substituents. This

information was further used for the generation of a pharma-

cophore model of bioactivity.

SMN promoter

EC₅₀(SD), µM

O-C-C(aryl)

compd

-CH₂-X-CH₂-

(reference)

dimethyl [X ) absent]

Cy-hexyl [X ) (CH₂)₃)]

Cy-propyl [X ) bond]

angle, deg^a

5b

8a

8b

8c

0.078(0.03)

2.33

1.71

109.2

103.7

101.5

121.5

0.29(0.12)

^aValues derived from MMFF94 (SYBYL6.9) optimized geometries.

Overcoming DHFR Inhibition as an Off-Target

r,r-Disubstituted Analogues. Prior to embarking on enan-

tioselective synthesis of individual R-methyl enantiomers, it was

decided to evaluate the achiral R,R-disubstituted benzyl ether

analogues. The gem-dimethyl analogue 8a featuring an ad-

ditional C-methyl group to 6b resulted in about 100-fold loss

in activity. Several cyclcoalkyl analogues with varying

O-C-C(aryl) angle compared to the starting benzyl ether 5b

were synthesized (8b, 8c). The cyclopropyl moiety was expected

to have the least impact on the optimal topology of the appended

3-chlorophenyl group. The respective analogue 8c possessing

the largest angle (Table 7) was only 3-fold less active, while

the cyclohexyl analogue 8b was ∼20-fold less active compared

to benzyl ether 5b. Analogues 8a and 8b with similar values

for the O-C-C(aryl) angle had similar activity. Unfortunately,

none of the achiral analogues derived from R,R-substitution

provided improved SMN2 promoter activity (Table 6).

Activity. Selected quinazolines with SMN2 promoter EC₅₀

<

100 nM were proﬁled across a panel of cellular toxicity assays.

Speciﬁcally, we monitored proliferation, oxidative stress, mi-

tochondrial function (as assessed by MTT metabolism), and

energy metabolism (as assessed by measuring cellular ATP

levels) (see Table 8). Data for compounds 5b, 5c, and 7a

indicated that MTT metabolism and ATP depletion were the

most sensitive indicators of cellular toxicity.

This led to the hypothesis that 5b, 5c, and 7a either inhibit

synthesis of ATP or accelerate hydrolysis of ATP resulting in

cellular toxicity. Because the quinazoline core has been used

to derive potent dihydrofolate reductase inhibitors (DHFR), e.g.,

methotrexate (MTX), and the folate pathway is involved in the

synthesis of nucleoside triphosphates,⁴¹it seemed possible that

the cellular toxicity of the C5-quinazolines might be linked to

DHFR inhibition. This was indeed the case as shown in Table

9, as potent DHFR inhibitors also potently depleted cellular

ATP.

At this juncture, it was important to determine whether (a)

DHFR inhibition was a plausible mechanism for SMN2 gene

induction and, thus, if DHFR inhibition was linked to activity

in the promoter assay or (b) if DHFR inhibition was simply an

undesirable off-target activity. Methotrexate (MTX), a potent

inhibitor of human DHFR (IC₅₀) 1 nM), did not affect SMN2

promoter activity at up to 30 µM, indicating that DHFR

inhibition was an off-target activity.

Chiral r-Methyl Analogues. Because R-methyl was the

optimal alkyl group, both enantiomers of 7a were prepared and

assayed to determine chiral preference for activity. Considering

both the cell-based nature of the assay and about a 3- to 5-fold

variation from experiment to experiment, EC₅₀determinations

for the racemate and the two enantiomers were carried out side

by side in one experiment. The experiment (Table 7) showed

that the (S)-enantiomer (S)-7a was 16 times more active than

(R)-enantiomer (R)-7a. This (S)- vs (R)-enantiomeric preference

was conﬁrmed for the phenyl series (S)-7u vs (R)-7u as well.

Bioactive Orientation of the C5 Substituent. As mentioned

above, increase in bulk for the R-alkyl substituent reduced

activity. However, loss in activity going from a methyl to an

ethyl substituent could be rationalized in terms of (a) a

straightforward steric demand or (b) an unfavorable electronic

environment of the terminal CH₃group for the allowed

orientation [A], of the -CH₂-CH₃, as depicted in Scheme 7.

This analysis further suggested synthesis of the conformationally

constrained ethyl analogue C. Because meta substituents

provided greater activity, the corresponding isomeric cyclized

analogues D1 and D2 were considered (Scheme 8).

High-resolution crystallographic data for inhibitor bound

human DHFR are available in the public domain (Brookhaven

PDB), facilitating structure-guided optimization of the C5-

quinazolines in order to reduce or eliminate DHFR binding. As

shown in Figure 5, target-bound TMQ-DHFR [TMQ ) trime-

trexate] crystal data clearly showed H-bond interactions of the

2- and 4-amino groups of the quinazoline core with key active

site residues: Glu30, and Ile7 and Val115 of human DHFR.⁴²

Approximately 40 analogues representing functionalization of

the 2/4 amino groups had been prepared earlier in our program.

Unfortunately, any modiﬁcation or deletion of the C2- and/or

C4-amino (NH2) groups of the quinazoline analogues resulted

in loss of promoter activity (data not shown). Thus, isosteric

modiﬁcations of these amine functionalities did not represent a

viable option to overcome DHFR inhibition. Considering that

modiﬁcation at the C5 position of the 2,4-diaminoquinazoline

core allowed retention of promoter activity, our initial approach

was directed toward modiﬁcation of C5 substituents (juxtaposed

to the C4-NH₂group). We reasoned that these changes would

perturb and/or disfavor the observed H-bonding interaction of

the N⁴amine hydrogens with human DHFR residues, e.g., at

Between the two possible chloroindanol derived ether ana-

logues (Scheme 9), only the 6-chloroindanol derived analogue

6a was active. The 4-chloroindanol derivative 6b was inactive

in the promoter assay. Incidentally, both the constrained

analogue 6a and the benzyl ether 2b were equipotent.

Systematic conformational searches³⁸with the 3-chlorophenyl

1c, 3-chlorobenzyl 5b, 3-chloro-(S)-R-methylphenyl (S)-7a,

6-chloroindanyl 9a, and 4-chloroindanyl 9b analogues were

carried out with 10° increments using SYBYL. For each con-

formation, the vector tethering the oxygen atom to the para

Diaminoquinazoline DeriVatiVes as SMN2 Promoter ActiVator

Journal of Medicinal Chemistry, 2008, Vol. 51, No. 3 455

Table 7. SMN2 Promoter Assay Results for Chiral R-Methyl Benzyl Ether Analogues

compd

(see above left)

chirality

EC₅₀(SD), µM

(see above right)

chirality

EC₅₀(SD), µM

7a

(R)-7a

(S)-7a

RS

R

S

0.021 (0.01)

0.032

0.002

7u

(R)-7u

(S)-7u

RS

R

S

0.24

0.22 (0.08)

0.06 (0.05)

Scheme 7

Scheme 9

Scheme 8

isomeric benzyl ether series: 5 vs 10. It shows that not only did

the isomeric benzyl series 10 provide analogues with good

activity, these analogues also showed higher fold induction of

the SMN2 promoter. More importantly, these analogues dis-

played 10- to 20-fold improvement in selectivity over DHFR

inhibition⁴⁴compared with the benzyl ether analogues 5.

least Val115 or preferably both Ile7 and Val115 (proposals A,

B, Scheme 10). Proposal A involved incorporation of a suitably

placed hydrogen bond acceptor, whereas proposal B incorpo-

rated an H-bond donor/acceptor.⁴³

SAR for Reverse Benzyl Ethers. Initially, we planned to

prepare analogues with the oxygen (one or more) atom(s)

removed from the quinazoline core. As a result, several

analogues with the general structure referred to as “reverse

benzyl series” (Scheme 10) were prepared and tested in both

SMN2 promoter and DHFR assays. In these series, halogen

substituents (Cl and F) provided better activity than methyl and

methoxy, similar to the SAR data discussed above. Among

chloro isomers, the activity decreased in the following order:

meta > para > ortho. The respective para isomers were ranked

as follows: F > Cl > CH₃. OCH₃(Table 10). Among

disubstituted phenyl analogues, 2,4-diﬂuoro analogue 10b was

more potent. Data in Table 11 provide a comparison of two

Figure 3. Compounds used for systematic conformational search with

vectors plotted in Figure 4 highlighted.

456 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 3

Thurmond et al.

Figure 4. Vector maps corresponding to 3-chlorophenyl 1c [cyan], 3-chlorobenzyl 5b [magenta], 3-chloro-R-methylphenyl 7a [blue], 6-chloroindane

9a [green], and 4-chloroindane 9b [red]: (a) vectors corresponding to 1c, 5b, and 7a; (b) vectors corresponding to 9a and 9b; (c) vectors corresponding

to 5b, 7a, 9a, and 9b. Vector maps were generated using SYBYL 6.0/7.0.³⁹See Experimental Section for details.

,a

Table 8. Cellular Toxicity Screen Using Rat Hepatoma Cell Line⁴⁰

guided approach provided an efﬁcient means to overcome DHFR

off-target activity for this series of quinazolines.

cell no.

TC₅₀, µM

GST TC₅₀,

MTTTC₅₀,

ATP TC₅₀,

compd

µM

Piperidine Series. To improve further the separation between

SMN2 promoter induction and potential for DHFR inhibition,

a number of C5 substituents were investigated that would exceed

spatial limitations imposed by the DHFR active site (Figure 7).

Piperidine-derived analogues 11–13 are examples of these series.

Both synthetic ﬂexibility and the achiral nature of these agents

allowed for rapid SAR studies of promoter activity and DHFR

inhibition (Scheme 6). A diverse range of substituents were

explored to evaluate contribution of size and shape, H-bond

donor/acceptor features, and the ionic nature of the substituents.

In addition, the tether “X” was varied (e.g., X ) CH₂, CO, and

SO₂). Over 125 analogues were prepared to study SAR within

these series. Table 12 shows a selected set of N-substituted

piperidine analogues. In general, this pharmacophore provided

a number of potent analogues with good to excellent promoter

activity and sound maximum fold induction. The nature of the

“X” group also inﬂuenced the activity. Tertiary amine analogues

11 displayed the best activity compared to the corresponding

amide 12. These in turn were better than the corresponding

sulfonamides 13. A wide range of functional groups for the

tertiary amine analogues were tolerated, several analogues

showing EC₅₀< 100 nM. Although the preference for linker

“X” was CH₂> CO . SO₂, there were some notable exceptions

to this trend. This included the 3,4-diﬂuoro set, where the tertiary

amine analogue 11k had poor promoter activity compared with

the corresponding amide 12k and sulfonamide 13k. Steric bulk

explored with the 11e/12e/13e and 11f/12f/13f sets, where the

“R-Ph” substituent was replaced with 1- and 2-naththyl deriva-

tives, identiﬁed the 2-naphthyl pharmacophore as the worst,

irrespective of the nature of the linking group “X”. A number

of diﬂuoro derivatives, except 3,4-diﬂuoro, displayed similar

activity. The 2-ﬂuorophenyl substituent yielded the best EC₅₀.

Although amide derivatives in general had lower EC₅₀than the

respective tertiary amines, they afforded higher maximum fold

induction. It was gratifying to see that the majority of piperdine

analogues showed weak DHFR inhibition (IC₅₀. 125 µM)

except for the benzylpiperidine analogue 11b.

5c

5b

7a

32

5

1

71

29

55

36

8

4

0.1

0.3

1

^aGST, R-glutathione S-transferase; MTT, 3-[4,5-dimethylthiazol-2-yl]-

2,5-diphenyltetrazolium bromide; ATP, adenosine triphosphate; TC₅₀,

concentration that produced a half-maximal response, cellular toxicity.

Table 9. SMN2 Promoter Induction vs DHFR Inhibition

SMN2 promoter

DHFR IC₅₀, nM

(average n ) 2)

induction

compd

(EC₅₀, nM)

5b

7a

(S)-7a

98

34

25

1

78

21

2

methotrexate

inactive (no induction

up to 30 µM)

A number of molecules featuring desirable SMN2 promoter

activity and ∼100-fold range of DHFR inhibition were evaluated

in the cellular toxicity panel. A scatter plot of these data is shown

in Figure 6. The good correlation (r²) 0.87) for ATP depletion

TC₅₀vs DHFR IC₅₀further supported the hypothesis that cellular

toxicity was linked to DHFR inhibition. Thus, our structure-

A signiﬁcant portion of the derivatized piperidine analogues

showed good-to-excellent activity in the promoter assay, several

with maximum fold induction in the range of 2–2.5. The vast

majority of these compounds also showed good metabolic

stability. Selected compounds evaluated for human cytochrome

P450 (hCYP) inhibition also posed no concern. In order to

identify compounds with high brain exposure and to further

Figure 5. Human DHFR-bound ligand (1s3u.pdb). H-bonds of the

two amino groups of the quinazoline core and the DHFR amino acid

residues are highlighted.

Diaminoquinazoline DeriVatiVes as SMN2 Promoter ActiVator

Journal of Medicinal Chemistry, 2008, Vol. 51, No. 3 457

Scheme 10. Proposals To Disrupt Atypical Quinazoline-DHFR H-Bond Interactions

Table 10. Activity of C5-CH₂O-Ar (Reverse Benzyl) Analogues in the

SMN2 Promoter Assay

EC₅₀

(SD), µM

DHFR IC₅₀,

compd

Phenylsubstituents

MFI^a

µM

10a

10b

10c

10d

10e

10f

10g

10h

10i

3-Chloro

2-Chloro

2-Iodo

2-Fluoro

4-Methyl

4-Chloro

4-Fluoro

4-Methoxy

2,4-Diﬂuoro

0.042 (0.03)

1.16 (1.0)

0.39 (0.2)

0.019 (0.01)

0.41 (0.4)

0.19 (0.04)

0.35 (0.3)

1.40 (0.2)

0.042 (0.01)

2.0

2.2

1.9

2.1

2.2

2.0

2.2

2.1

2.2

8.4

Figure 6. Correlation of TC₅₀for ATP depletion using rat hepatoma

(H4IIE) cell line versus IC₅₀for DHFR inhibition for C5-functionalzed

2,4-diaminoquinazoline analogues. All of the analogues shown here

also have good activity in the SMN2 promoter assay.

39.0

5.1

^aMFI ) Maximum Fold Induction.

Table 11. Comparison of Two Isomeric Benzyl Ether Series

A-B ) -CH₂-O-

A-B ) -O-CH₂-

EC₅₀,

µM

DHFR

IC₅₀,

EC₅₀,

µM

DHFR

IC₅₀,

R

compd (MaxFoldInd) µM compd (MaxFoldInd) µM

2-F

2,4-F2 10i

3-Cl 10a

10d

0.019(2.1)

0.042(2.2)

0.042(2.0)

39.0

10.7

8.4

5g

5s

5b

0.053(1.8)

0.12(1.8)

0.078(1.8)

0.39

ND

0.28

prioritize these for in vivo experiments, active compounds were

dosed in mice at 2 mg/kg (iv). Concentrations of parent

compound were measured in plasma and brain samples using

LC/MS/MS at 15 min and 2 h postdose with n ) 4 mice per

time point. Data for these two time points helped sort com-

Figure 7. Size and shape of the pocket for a potential substituent projected

from C5-position of quinazoline is highlighted by an orange circle. This

orientation of quinazoline shown is identical to that of 2,4-diamino pteridine

(bicyclic core structure) of MTX and the A-ring containing 2,4-diamino

portion overlaps with the tetrahydrofolate-derived DHFR bound ligand in

1s3u.pdb. See Experimental Section for details.

pounds based on relative apparent t and relative trend for brain/

½

plasma exposure. A summary of these data for selected

compounds is shown in Table 13. Across three piperidine series,

good blood-brain barrier permeability was achieved if the polar

surface area (PSA) was below 90 Å².

Subsequently, full PK analysis for 11a was carried out

following iv and oral dosing. The resulting data reconﬁrmed

that 11a possessed a long brain half-life (t ≈ 19 h) (Table 14)

½

458 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 3

Thurmond et al.

Table 12. Activity of N-Substituted Piperidines Series in the SMN2 Promoter Assay

X ) CH₂

X ) CO

X ) SO₂

R

compd

EC₅₀(SD), nM

MFI^a

compd

EC₅₀(SD), nM

MFI^a

compd

EC₅₀(SD), nM

MFI^a

2-F

3-Cl

2-Cl

4-Cl

1-napth

2-napth

2-Me

3-Me

4-Me

2,4-F2

3,4-F2

2,3-F2

11a

11b

11c

11d

11e

11f

11g

11h

11i

4 (1)

2.4

2.2

2

1.8

2

1.7

1.9

2

12a

12b

12c

12d

12e

12f

12g

12h

12i

74

2.5

2.3

2.4

2.5

2.4

2.5

2.2

2.6

2.4

2.5

13a

13b

13c

13d

13e

13f

13g

13h

13i

29

172 (40)

897

603

350

464

545

221

551

2.1

1.9

2.1

2

1.9

2

54 (10)

38 (10)

47 (30)

43 (30)

148 (70)

18 (10)

22 (10)

24 (10)

18 (10)

322 (80)

38 (10)

60 (10)

49 (10)

81 (30)

35 (10)

151 (90)

160 (40)

110 (70)

133 (30)

89 (30)

104 (10)

67 (10)

2

1.8

1.9

2

11j

11k

11m

12j

12k

12m

13j

13k

82 (40)

133 (40)

2.1

2.2

^aMFI ) maximum fold induction.

Table 13. Plasma and Brain Levels at 15 min and 2 h and Key in Vitro Data for Selected Piperidine-Derived Analogues^a

plasma levels, ng/mL

brain levels, ng/g

SMN2 promoter

compd

X

R

EC₅₀, µM

DHFR IC₅₀, µM

15 min

120 min

15 min

120 min

11a

12c

12b

11b

13a

CH₂

CO

CH₂

SO₂

2-F

0.004

0.014

0.060

0.054

0.029

>125

55.2

38

240

225

462

279

21

20

17

610

25

1423

26

23

1150

305

1133

9

10

1518

106

2-Cl

3-Cl

2-F

>125

^aAll compounds were dosed at 2 mg/kg in 30% solutol.

Table 14. Pharmacokinetic Parameters for Piperidine Analog 11a Following iv and po Dosing to Mice^a

route (iv)

iv plasma

parameters

units

values

SD

route (po)

iv plasma

parameters

units

values

SD

t_1/2

h

NA

2406

83.8

38103

18.96

15724

110

2437

22.83

lC_max

t_1/2

AUC_last

AUC_INF

C_max

ng/mL

h

h·ng/mL

ng/g

110

NA

28.7

AUC_last

Cl_pred

V_ss,pred

t_1/2

AUC_last

Cl_pred

V_ss,pred

MRT_INF

h·ng/mL

(mL/h)/kg

mL/kg

h

hr·ng/g

(g/h)/kg

g/kg

9.95

74.0

2584

1.32

1118

8.6

2673

10893

765.00

18.80

15485

18910

27.30

45.6

3443

129

0.95

931

iv brain

t_1/2

h

AUC_last

AUC_INF

MRT_INF

h·ng/g

h

230

0.995

1041

1.22

h

^aCompound was dosed in 30% solutol at 2 mg/kg iv and in HPMC-Tween-80 at 10 mg/kg, orally.

with low plasma and brain clearance. This compound showed

very good brain exposure and overall coverage measured as

brain-to-plasma ratio based on C_maxor AUC, as both of these

ratios were about 5- to 6-fold higher in brain vs plasma.

Compound 11a also had very low potential for drug-drug

interaction because it did not inhibit any of the major human

cytochrome P450 enzymes (hCYP). Speciﬁcally, 11a inhibited

hCYP2D6 with an IC₅₀of 9.38 µM whereas inhibition of all

other enzymes tested (CYP1A2, CYP2C9, CYP2C19, and

CYP3A4) was <50% at 10 µM. Compound 11a displayed

attractive physicochemical properties, with 41.5 mg/mL solubil-

ity in water as a dihydrochloride salt.

Next, the piperidine analogue 11a was evaluated as a

modulator of (a) the endogenous SMN promoter in the mouse

NSC-34 cell line and (b) human Smn protein levels in SMA

type I patient ﬁbroblast cell lines. To test effects on the mouse

SMN promoter, 11a was incubated with NSC-34 cells for 18 h

using the conditions that were described for the promoter assay.

Cells from multiple wells were pooled, and RNA was isolated

for sequential cDNA generation and TaqMan analysis (see

Experimental Section for details). Compound 11a increased

mouse SMN mRNA about 2-fold in NSC-34 cells (Figure 8),

an effect similar in magnitude to the induction of the engineered

human SMN2 reporter gene. Although the mechanism of

promoter induction by 11a or the other quinazoline analogues

is unknown, this demonstrates that SMN promoter induction is

independent of chromosomal location or the local state of

chromatin activation. Trichostatin A, a histone deacetylase

inhibitor that potently induces SMN2 gene activity, was used

as a positive control in this experiment. Selected C5-subsituted

quinazolines did not inhibit histone deacetylase (data not shown)

as shown previously by Jarecki et al.³³

To examine the effect of these quinazoline derivatives on

human SMN expression, we treated a ﬁbroblast cell line from

Diaminoquinazoline DeriVatiVes as SMN2 Promoter ActiVator

Journal of Medicinal Chemistry, 2008, Vol. 51, No. 3 459

a type I SMA patient (line 3813) with different doses of 5b,

11a, or 11b. In the 3813 cell line, SMN1 has been deleted from

both chromosomes and SMN2 is present at low copy number.

For comparison, line 3814 is a ﬁbroblast line derived from the

carrier parent and represents 50% normal SMN levels due to

the deletion of the SMN1 gene on only one of the two parental

chromosomes. All three compounds increased amounts of SMN

protein in a dose-dependent manner (Figure 9C).

SMN localizes to discreet granules in the nucleus called gems.

The number of nuclei with gems and the number of gems per

cell is reduced in type 1 SMA cells.¹¹Compounds such as

butyrate, 4-phenylbutyrate, valproate, and alcarubicin increase

the number of gems in SMA ﬁbroblasts.^28–33In the experiment,

SMA patient ﬁbroblasts were treated with increasing doses of

5b, 11a, or 11b (10–1000 nM) for 5 days and the number of

nuclei with gems as well as the number of gems per 100 nuclei

were determined. All three compounds increased both the

number of cells with gems (Figure 9A) and the number of gems

(Figure 9A) in a dose-dependent manner. For instance, 100 nM

5b increased the number of gems by ∼2.9-fold (19.7 ( 2.6 for

Figure 8. Effect of compound 11a and its comparison with TSA on

mouse SMN mRNA expression in the murine NSC-34 cell line under

serum-free media conditions.

Figure 9. Effect of quinazoline derivatives on SMN expression in SMA patient ﬁbroblasts. (A) Change in gem number per cell and number of

gems per 100 nuclei after treatment with the drugs 5b, 11a, and 11b at the concentration indicated. / indicates p < 0.05. // indicates p < 0.005.

The cell 3814 is derived from the carrier parent and represents the levels trying to be obtained. (B) Examples of 3813 cell after treatment stained

with SMA antibody MANSMA2 (red). Arrow heads indicate gems, and arrow indicates cytoplasmic staining. (C) Western blot of SMN levels after

treatment with drug compounds as indicated. Notice dose-response to drugs.

460 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 3

Thurmond et al.

solvent A was aqueous 0.05% TFA and solvent C was methanol.

Chiral HPLC was performed with a Chiralcel OD-H column eluted

with a mixture of hexane and ethanol containing 0.1% triﬂuoroacetic

acid. Chiral HPLC was performed with a Chiralcel OD-H column

(250 mm × 4.6 mm, 5 µm, 254 nm) eluted with a mixture of hexane

(containing 0.1% triﬂuoroacetic acid) and ethanol (7% ethanol

isocratic elution over 45 min). Elemental analyses were carried out

by Galbraith Laboratories, Inc. (Knoxville, TN) or Midwest

Microlab, LLC (Indianapolis, IN). Mass spectra were obtained using

either APCI or electrospray ionization.

General Procedure A: Preparation of 5-Aryloxy or 5-Sub-

stituted Aryloxymethyl 2,4-Diaminoquinazoline. Step 1: Prepa-

ration of 2-Fluoro-6-(aryloxy/alkoxy)benzonitrile (4). A solution

of substituted phenol or substituted benzyl alcohol (10 mmol, 1

equiv) was added to a cooled (0 °C) slurry of sodium hydride

(washed w/hexanes) (0.407 g, 10 mmol, 1equiv) in DMF (7 mL)

under nitrogen atmosphere. The reaction mixture was slowly

warmed to room temperature and stirred for 1 h. The reaction

mixture was then added to a cooled (0 °C) solution of 2,6-

diﬂuorobenzonitrile 2 (7-10 mmol, 0.7–1.0 equiv) in DMF (5 mL)

and stirred for 3–5 h at room temperature. The reaction mixture

was poured on crushed ice–water, stirred, ﬁltered, and washed with

water. The resulting solid was ﬁltered and dried in vacuo to afford

2-ﬂuoro-6-(aryloxy/benzylox)benzonitrile intermediate.

Step 2. Cyclization of 2-Fluoro-6(aryloxy/substituted benzy-

loxy)benzonitrile (4) to 5-Ether Derivatives of 2,4-Diamino-

quinazoline. Guanidine carbonate (180 mg, 1.0 mmol, 2 equiv)

and 2-ﬂuoro-6(aryloxy/substituted benzyloxy)benzonitrile (0.5 mmol,

1 equiv) were heated at 140 °C in dimethylacetamide (2 mL) for

10 h and allowed to cool to room temperature overnight. The

reaction mixture was further cooled in the refrigerator. The resulting

solid was collected in vacuo and recrystallized from ethanol/water

to provide the quinazoline derivative.

5-(2-Chlorophenyloxy)quinazoline-2,4-diamine (1e). 2-Chlo-

rophenol (1.1 mL, 10 mmol) and sodium hydride (0.407 g, 10

mmol) was reacted with 2,6-diﬂuorobenzonitrile (1.39 g, 10 mmol)

in DMF (5 mL) using procedure A, step 1, to afford 0.746 g of

2-ﬂuoro-6-(2-chlorophenyloxy)benzonitrile (22% yield). 2-Fluoro-

6-(2-chlorophenyloxy)benzonitrile (229.3 mg, 0.9 mmol) and

guanidine carbonate (147.2 mg, 0.8 mmol) were reacted according

to procedure A, step 2. The solid was recrystallized from ethanol/

water, 1:1, and dried in vacuo to yield 119.7 mg of 5-(2-

chlorophenyloxy)quinazoline-2,4-diamine (46% yield). HPLC t_R)

8.6 min (system 1), 11.8 min (system 2). ¹H NMR (500 MHz,

DMSO-d₆) δ 7.66 (dd, J ) 1.5, 8.5 Hz, 1H), 7.44 (m, 1H), 7.32

(m, 3H), 7.19 (bs, 2H), 6.90 (dd, J ) 1.0, 8.5 Hz, 1H), 6.09 (s,

2H), 6.01 (dd, J ) 0.5, 8.0 Hz, 1H). MS m/z (ESI) 289 (M + H)⁺.

Compounds 1d and 1f were prepared analogous to the synthesis

of 1e.

5b versus 6.7 ( 0.7 for DMSO; n ) 3 per group; p ) 0.04)

while the same doses of 11a and 11b elevated the number of

gems by ∼5.5-fold (37.0 ( 0.6, n ) 3, p < 0.001) and ∼4.0-

fold (27.0 ( 2.1, n ) 3, p ) 0.026), respectively. Interestingly,

compounds 11a and 11b showed an increase in cytosolic

staining whereas 5b did not appreciably affect SMN staining

in the cytosol (Figure 9B).

These studies conﬁrmed that the piperidine 11a up-regulated

the mouse SMN promoter in NSC-34 cells, a mouse motor

neuron derived cell line. It also up-regulated the endogenous

human SMN2 promoter in its normal chromosomal context. The

compound increased full-length human Smn protein, as dem-

onstrated by an increase in gems number in type I SMA patient

ﬁbroblasts. Gems were restored to a level similar to carriers. In

vivo evaluation of compound 11a in a transgenic mouse SMA

disease model will be reported in due course.

Conclusion

Utilizing the uHTS hit reported by Jarecki et al.³³as the

starting point the uHTS hit reported by Jarecki et al.,³³our

ligand-based design strategy led to the identiﬁcation of a series

of C5 substituted 2,4-diaminoquinazolines (for example, 5b)

as inducers of the SMN2 promoter. Introduction of R-methyl

substituents provided conformationally constrained analogues

with improved activity (lower EC₅₀and higher maximum fold

induction) in the SMN2 promoter assay compared with 5b;

however, in vitro assessment of cellular toxicity led to identi-

ﬁcation of DHFR inhibition as an off-target activity. A focused,

structure-guided approach employing molecular modeling yielded

achiral piperidine series 11–13 decoupled from the DHFR

inhibition liability. Our best molecule 11a showed a promising

in vitro proﬁle (EC₅₀) 4 nM, EC₉₀) 70 nM, MaxFoldInd )

2.4 in the NSC-34 cell assay). Furthermore, 11a showed

desirable pharmaceutical properties and excellent brain exposure

(C_max) 3 µM, following 10 mg/kg po dose and t ) 19 h)

½

following oral dosing in mice. This agent also up-regulated the

mouse SMN promoter in a mouse motor neuron derived cell

line, NSC-34. In addition, compound 11a also up-regulated the

human SMN2 promoter in severe type I SMA patient ﬁbroblasts

and increased full-length human SMN protein as demonstrated

by an increase in intranuclear gems, or Cajal bodies. The

increase in the number of gems observed with the piperidine

analogue 11a was far superior to that observed for the initial

uHTS hit 1b. Gems were restored to the level of unaffected

genetic carriers of the heterozygous SMN1 gene deletion. We

anticipate that our lead candidate 11a will provide a tool for

exploring therapeutic beneﬁts of SMN2 promoter induction in

transgenic mouse models of SMA disease and ultimately in

human clinical trials.

5-(4-Fluorophenoxy)quinazoline-2,4-diamine (1d). Yield )

39%. HPLC t_R) 8.5 min (system 1), 11.6 min (system 2). ¹H

NMR (400 MHz, DMSO-d₆) δ 8.32 (s, 1H), 7.33–7.21 (br, m, 6H),

6.90 (d, J ) 8.4 Hz, 1H), 6.15 (d, J ) 8.0 Hz, 1H), 6.09 (br s, 2H).

MS m/z (ESI) 271 (M + H)⁺.

5-(3-Chlorophenyloxy)quinazoline-2,4-diamine (1f). Yield )

46%. HPLC t_R) 9.0 min (system 1), 12.4 min (system 2). ¹H

NMR (500 MHz, DMSO-d₆) δ 7.46 (t, J ) 8.0 Hz, 1H), 7.37 (t, J

) 8.0 Hz, 1H), 7.28 (m, 2H), 7.09 (m, 1H), 7.04 (bs, 1H), 6.97

(dd, J ) 1.0, 8.0 Hz, 1H), 6.31 (dd, J ) 1.0, 8.0 Hz, 1H), 6.10 (s,

2H), 6.01 (m, 1H). MS m/z (ESI) 289 (M + H)⁺.

Experimental Section

Chemistry. General Methods. All reagents and anhydrous

solvents were obtained from commercial sources and used without

further puriﬁcation unless otherwise noted. NMR spectra were

recorded at either 400 or 500 MHz in the solvent indicated, TMS

being an internal reference. Coupling constants (J) are given in

Hz. Column chromatography was carried out in the solvents

indicated with silica gel. HPLC purity of compounds was measured

with a reversed-phase HPLC (Zorbax SB-C₁₈column, 4.6 mm ×

150 mm, 5 µm, 254 nm) with two diverse solvent systems. In

system 1, compounds were eluted using a gradient elution 95/5 to

5/95 A/B over 20 min at a ﬂow rate of 1.0 mL/min, where solvent

A was aqueous 0.05% TFA and solvent B was acetonitrile (0.05%

TFA). In system 2, compounds were eluted using a gradient of

95/5 to 5/95 A/C over 20 min at a ﬂow rate of 1.0 mL/min, where

5-(2-Chlorobenzyloxy)quinazoline-2,4-diamine (5a). 2-Chlo-

robenzyl alcohol (1.54 g, 10.8 mmol), sodium hydride (60%, 432

mg, 10.8 mmol) in DMF (5 mL) were reacted with 2,6-diﬂuo-

robenzonitrile (1 g, 7.2 mmol) in DMF (20 mL) according to

procedure A. Puriﬁcation by recrystallization from cyclohexane

yielded 1.69 g of 2-ﬂuoro-6-(2-chlorobenzyloxy)benzonitrile (85%

yield). 2-Fluoro-6-(2-chlorobenzyloxy)benzonitrile (131.1 mg, 0.5

mmol) and guanidine carbonate (110 mg, 0.6 mmol) were reacted

according to procedure A. The solid was recrystallized from ethanol/

water, 1:1, and dried in vacuo to yield 50 mg of 5-(2-chloroben-

1

zyloxy)quinazoline-2,4-diamine (33% yield). HNMR (500 MHz,

Diaminoquinazoline DeriVatiVes as SMN2 Promoter ActiVator

Journal of Medicinal Chemistry, 2008, Vol. 51, No. 3 461

DMSO-d₆) δ 7.63 (dd, J ) 7.5, 2.5 Hz, 1H), 7.56 (dd, J ) 7.5, 1.5

Hz, 1H), 7.40 (m, 3H), 7.17 (br d, 2H), 6.81(d, J ) 8.5 Hz, 1H),

6.83 (d, J ) 8.0 Hz, 1H), 6.01 (br s, 2H), 5.33 (s, 2H). ¹³CNMR

(500 MHz, DMSO-d₆) δ 161.7, 160.7, 133.5, 133.0, 132.3, 130.9,

130.4, 129.7, 127.6, 117.5, 101.8, 101.3, 67.7. MS m/z (ESI) 299

(M - H)⁺. HPLC t_R) 9.1 min (system 1), 12.3 min (system 2).

5-(Furan-2-ylmethoxy)quinazoline-2,4-diamine (5x). Yield )

5%. ¹H NMR (500 MHz, DMSO-d₆) δ 7.75 (s, 1H), 7.38 (t, J ) 8.5

Hz, 1H), 7.17 (br d, 2H), 6.80 (d, J ) 8 Hz, 1H), 7.72 (d, J ) 8

Hz, 1H), 6.67 (d, J ) 3.5 Hz, 1H), 6.51 (m, 1H), 5.97 (s, 2H),

5.24 (s, 2H). MS m/z 257 (M + H)⁺HPLC t_R) 7.9 min (system

1), 10.8 min (system 2).

Compounds 5b-y were prepared similar to that described for

5-(5-Methylisoxazol-3-ylmethoxy)quinazoline-2,4-diamine

(5y). The target compound was prepared starting from (5-methyl-

isoxazol-3-yl)methanol according to procedure A. Yield ) 50%.

¹H NMR (500 MHz, DMSO-d₆) δ 7.40–7.20 (m, 3H), 6.81 (d, J

) 8.0 Hz, 1H), 6.61 (d, J ) 7.5 Hz, 1H), 6.35 (s, 3H), 5.96 (s,

2H), 5.33 (s, 2H), 2.43(s, 3H). MS m/z (ESI) 272 (M + H)⁺. HPLC

t_R) 7.6 min (system 1), 10.2 min (system 2).

5a.

5-(3-Chlorobenzyloxy)quinazoline-2,4-diamine (5b). Yield )

37%. ¹H NMR (500 MHz, DMSO-d₆) δ 7.60 (s, 1H), 7.45 (m,

3H), 7.33 (t, J ) 8.0 Hz, 1H), 7.23 (br d, 2H), 6.78 (d, J ) 8.5 Hz,

1H), 6.58 (d, J ) 8.0 Hz, 1H), 5.97 (br s, 2H), 5.29 (s, 2H).

¹³CNMR (500 MHz, DMSO-d₆) δ 161.7, 160.7, 155.9, 155.2,

139.0, 133.2, 132.3, 130.6, 128.1, 127.8, 126.6, 117.3, 102.1, 101.4,

69.0. MS m/z (ESI) 299 (M - H)⁺. FTIR 3515, 3397, 3345, 3120,

1652, 1615, 1596, 1575, 1552, 1500, 1479, 1435, 1407, 1371, 1356,

1254, 1178, 1130, 1081, 990, 863, 812, 784, 745, 686 cm^-1. Anal.

(C₁₅H₁₃ClN₄O) C, H, N.

5-(4-Chlorobenzyloxy)quinazoline-2,4-diamine (5c). Yield )

35%. ¹H NMR (500 MHz, DMSO-d₆) δ 7.48 (dd, J ) 9.0, 8.5 Hz,

3H), 7.32 (t, J ) 8.0, 8.5 Hz, 1H), 7.20 (bd, J ) 23.5 Hz, 2H),

6.77 (d, J ) 8.0 Hz, 1H), 6.58 (d, J ) 8.0 Hz, 1H), 5.95 (s, 2H),

5.27 (s, 2H). MS m/z 301 (M + H)⁺. HPLC t_R) 9.3 min (system

1), 12.5 min (system 2).

5-(Quinolin-3-ylmethoxy)quinazoline-2,4-diamine (5z). So-

dium borohydride (240 mg, 6.4 mmol) was added in portions to a

solution of 3-quinolinecarboxaldehyde (910 mg, 5.8 mmol) in

methanol at room temperature. After 3 h of stirring, the reaction

mixture was quenched with 10 mL (aqueous) of saturated NH₄Cl.

The mixture was extracted with ethyl acetate (3 × 30 mL).

Combined organics were washed with brine and dried over MgSO₄.

Crude quinolin-3-ylmethanol was obtained upon ﬁltration and

concentration in vacuo (795 mg, 86% yield). A solution of quinolin-

3-ylmethanol (785 mg, 4.9 mmol) in DMF was added to a cooled

(0 °C) slurry of sodium hydride (1 equiv) in DMF under nitrogen

atmosphere. The reaction mixture was slowly warmed to room

temperature and stirred for 45 min. In another vessel, a solution of

2,6-diﬂuorobenzonitrile (904 mg, 6.5 mmol) in DMF was chilled

to 0 °C, and sodium quinolin-3-ylmethoxide solution was added

over 20 min. The mixture was then stirred for 2 h at room

temperature. The reaction mixture was poured onto crushed

ice–water, stirred, ﬁltered, washed with water, and dried in vacuo

to afford 2-ﬂuoro-6-(quinolin-3-ylmethoxy)benzonitrile (1.1 g, 81%

yield). 2-Fluoro-6-(quinolin-3-ylmethoxy)benzonitrile (300 mg, 1.1

mmol) and guanidine carbonate (1.1 mmol) were heated at 150 °C

in dimethylacetamide for 5 h, then cooled back to room temperature

overnight. The reaction mixture was diluted with water and stirred

for 45 min, and the aqueous mixture was then extracted with ethyl

acetate (6 × 8 mL). Combined organic extracts were then washed

with water (3 × 5 mL) and brine and dried over MgSO₄. Following

ﬁltration and concentration in vacuo, the material was puriﬁed by

ﬂash silica gel and eluted with 5–10% MeOH/DCM gradient to

give 162 mg of 5-(quinolin-3-ylmethoxy)quinazoline-2,4-diamine

5-(3-Methylbenzyloxy)quinazoline-2,4-diamine (5e). Yield )

1

16%. H NMR (500 MHz, DMSO-d₆) δ 7.32 (m, 4H), 7.19 (m,

3H), 6.77 (d, J ) 8.5 Hz, 1H), 6.62 (d, J ) 8 Hz, 1H), 5.94 (s,

2H), 5.21 (s, 2H), 2.34 (s, 3H). MS m/z (ESI) 281 (M + H)⁺.

HPLC t_R) 9.2 min (system 1), 12.4 min (system 2).

5-(4-Methylbenzyloxy)quinazoline-2,4-diamine (5f). Yield )

39%. ¹H NMR (500 MHz, DMSO-d₆) δ 7.41 (d, J ) 7.5 Hz, 2H),

7.35 (t, J ) 8 Hz, 1H), 7.24 (d, J ) 8 Hz, 2H), 7.6 (br d, J ) 28

Hz, 2H), 6.77 (d, J ) 8.5 Hz, 1H), 6.63 (d, J ) 8 Hz, 1H), 5.92

(br s, 2 H), 5.20 (s, 2H), 2.32 (s, 3H). MS m/z 281 (M + H)⁺.

HPLC t_R) 9.2 min (system 1), 12.5 min (system 2).

5-(2-Fluorobenzyloxy)quinazoline-2,4-diamine (5g). Yield )

1

22%. H NMR (400 MHz, DMSO-d₆) δ 7.61 (m, 1H), 7.47 (m,

1H), 7.31 (m, 3H), 7.13 (br d, J ) 10.8 Hz, 2H), 6.79(d, J ) 9.2

Hz, 1H), 6.67 (d, J ) 8.0 Hz, 1H), 5.97 (br s, 2H), 5.32 (s, 2H).

MS m/z (ESI) 285 (M - H)⁺. HPLC t_R) 8.6 min (system 1),

11.7 min (system 2).

5-(3-Fluorobenzyloxy)quinazoline-2,4-diamine (5h). The crude

1

(46% yield). H NMR (400 MHz, DMSO-d₆) δ 9.09 (d, J ) 2.4

product was puriﬁed by ﬂash silica gel, eluted with 5–10% MeOH/

1

Hz, 1H), 8.51 (d, J ) 2.4 Hz, 1H), 8.03 (m, 2H), 7.80 (m, 1H),

7.65 (m, 1H), 7.36 (t, J ) 8.4, 8.0 Hz, 1H), 7.21 (bs, 1H), 7.13

(bs, 1H), 6.80 (dd, J ) 0.8 Hz, 1H), 6.72 (dd, J ) 0.8 Hz, 1H),

5.95 (bs, 2H), 5.51 (s, 2H). MS m/z (ESI) 319 (M + H)⁺.

DCM gradient, and dried in vacuo. Yield ) 15%. H NMR (400

MHz, DMSO-d₆) δ 7.51 (m, 2H), 7.36 (m, 2H), 7.31 (m, 2H), 7.22

(t, J ) 8.4, 0.8 Hz, 1H), 7.10 (m, 1H), 6.72 (dd, J ) 8.4, 0.8 Hz,

1H), 6.41 (d, J ) 7.6 Hz, 1H), 5.96 (bs, 2H), 5.72 (q, J ) 6.4 Hz,

1H), 1.69 (d, J ) 6.4 Hz, 3H). MS m/z (ESI) 290 (M - H)⁺. HPLC

t_R) 8.7 min (system 1), 11.7 min (system 2).

General Procedure B: Preparation of 5-Nonaryl Ethers

Derivatives of 2,4-Diaminoquinazolines 6a–m: 5-Methoxyquinazo-

line-2,4-diamine (6a). Sodium methoxide, 25 wt % in methanol

(0.31 mL, 1.5 mmol), was added dropwise to a cooled (0 °C)

solution of 2,6-dinitrobenzonitrile (260.0 mg, 11.3 mmol) in DMF

(1.5 mL) over 5 min. The reaction mixture was stirred at room

temperature for 4½ h, poured onto crushed ice-water, ﬁltered, and

washed with water. The beige solid was recrystallized from ethanol

to yield 173.0 mg of 2-methoxy-6-nitrobenzonitrile (67% yield).

2-Methoxy-6-benzonitrile (173.0 mg, 1 mmol), cyclohexene (0.33

mL, 3 mmol), and palladium, 10 wt %, on activated carbon (290

mg, 0.1 mmol) in ethanol (3.3 mL) were shaken at 80 °C for 7½

h. The reaction mixture was then diluted with ethanol (2.5 mL),

ﬁltered through a pad of Celite, washed with ethanol, and dried to

afford 87.1 mg of 2-amino-6-methoxybenzonitrile (57% yield).

2-Amino-6-methoxybenzonitrile (608.1 mg, 4 mmol) and guanidine

carbonate (825.6 mg, 4.6 mmol) were heated at 145 °C in

dimethylacetamide (10 mL) for 7 h. The reaction mixture was

diluted with water, stirred for 1 h, ﬁltered, washed with water, and

recrystallized with ethanol/water, 1:1, to yield 313.6 mg of

5-methoxyquinazoline-2,4-diamine (41% yield). ¹H NMR (400

MHz, DMSO-d₆) δ 7.35 (t, J ) 8.0 Hz, 2H), 7.15 (bs, 1H), 6.76

(d, J ) 7.6 Hz, 1H), 6.51 (d, J ) 8.0 Hz, 1H), 5.89 (bs, 2H), 3.89

5-(4-Fluorobenzyloxy)quinazoline-2,4-diamine (5i). Yield )

38%. ¹H NMR (400 MHz, DMSO-d₆) δ 7.59 (m, 2H), 7.32 (t, 1H),

7.26 (m, 2H), 7.17 (br s, 2H), 6.77(dd, J ) 8.4, 0.8 Hz, 1H), 6.61

(dd, J ) 7.6, 0.8 Hz, 1H), 5.95 (br s, 2H), 5.25 (s, 2H). MS m/z

(ESI) 285 (M + H)⁺. HPLC t_R) 8.8 min (system 1), 11.8 min

(system 2).

5-(Pyridin-3-ylmethoxy)quinazoline-2,4-diamine (5v). This

compound was prepared starting from 3-pyridinylcarbinol as

described for 5a above. Yield ) 24%. ¹H NMR (500 MHz, DMSO-

d₆) δ 8.75 (d, J ) 1.5 Hz, 1H), 8.57 (dd, J ) 1.0, 1.5 Hz, 1H),

7.94 (d, J ) 8.0 Hz, 1H), 7.45 (dd, J ) 5.0, 4.5 Hz, 1H), 7.34 (t,

J ) 8.0 Hz, 1H), 7.17 (bs, 2H), 6.78 (d, J ) 8.5 Hz, 1H), 6.64 (d,

J ) 8.0 Hz, 1H), 5.95 (s, 2H), 5.33 (s, 2H). MS m/z (ESI) 268 (M

+ H)⁺. HPLC t_R) 7.6 min (system 1), 10.2 min (system 2).

5-(Thiophen-2-ylmethoxy)quinazoline-2,4-diamine (5w). Yield

1

) 17%. H NMR (500 MHz, DMSO-d₆) δ 7.62 (dd, J ) 5.3, 1.5

Hz, 1H), 7.37 (t, J ) 6.5 Hz, 1H), 7.3 (dd, J ) 3.3, 1.0 Hz, 1H),

7.21 (s, 2H), 7.08 (dd, J ) 5.3, 4.0 Hz, 1H), 6.8 (d, J ) 8.5 Hz,

1H), 6.7 (d, J ) 8.0 Hz, 1H), 5.96 (s, 2H), 5.45 (s, 2H). MS m/z

(ESI) 273 (M + H)⁺. HPLC t_R) 8.3 min (system 1), 11.3 min

(system 2).

462 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 3

Thurmond et al.

(s, 3H). MS m/z (ESI) 191 (M + H)⁺. Anal. (C₉H₁₀N₄O·0.5H₂O)

C, H, N. HPLC t_R) 6.4 min (system 1), 8.6 min (system 2).

romethane) to yield the desired product 7c. Yield ) 28%. ¹H NMR

(500 MHz, DMSO-d₆) δ 7.44 (m, 4H), 7.35 (brd, 2H), 7.2 (t, J )

8.0 Hz, 1H), 6.69 (d, J ) 7.5 Hz, 1H), 6.35 (d, J ) 7.5 Hz, 1H),

5.99 (s, 2H), 5.48 (t, J ) 7.0 Hz, 1H), 2.06 (m, 1H), 1.92 (m, 1H),

0.96 (t, J ) 7.0 Hz, 3H). MS m/z (ESI) 330 (M + H)⁺. HPLC t_R

) 10.0 min (system 1), 13.1 min (system 2).

1

5-sec-Butoxyquinazoline-2,4-diamine (6g). Yield ) 15%. H

NMR (500 MHz, DMSO-d₆) δ 7.34 (t, J ) 8.0, 1H), 7.18 (s, 2H),

6.74 (d, J ) 8.0 Hz, 1H), 6.55 (d, J ) 8.0 Hz, 1H), 5.90 (s, 2H),

4.60 (m, 1H), 1.70 (m, 2H), 1.33 (d, J ) 6.0 Hz, 3H), 0.95 (t, J )

7.0 Hz, 3H). MS m/z (ESI) 233 (M + H)⁺.

5-[1-(4-Chlorophenyl)-2,2-dimethylpropoxy]quinazoline-2,4-

diamine (7d). Grignard reaction of trimethylacetaldehyde (240.0

mg, 2.8 mmol) with 4-chlorophenylmagnesium bromide in anhy-

drous ether afforded 0.58 g of 1-(4-chlorophenyl)-2,2-dimethyl-

propanol (>100% yield). A solution of 1-(4-chlorophenyl)-2,2-

dimethylpropanol (550.0 mg, 2.77 mmol) was subsequently reacted

as described above for the synthesis of 7a to provide the target

compound. Puriﬁcation by silica gel chromatography (5–10%

methanol in dichloromethane) yielded 270 mg of 5-[1-(4-chlo-

rophenyl)-2,2-dimethylpropoxy]quinazoline-2,4-diamine (67% yield).

¹H NMR (500 MHz, DMSO-d₆) δ 7.5 (brd, 2H), 7.4 (m, 4H), 7.17

(t, J ) 8.0 Hz, 1H), 6.69 (d, J ) 8.0 Hz, 1H), 6.31 (d, J ) 7.5 Hz,

1H), 6.08 (s, 2H), 5.36 (s, 1H), 1.01 (s, 9H). MS m/z (ESI) 357 (M

+ H)⁺. HPLC t_R) 10.8 min (system 1), 13.8 min (system 2).

5-[1-(3-Triﬂuoromethylphenyl)ethoxy]quinazoline-2,4-di-

amine (7e). The R-methyl-3-(triﬂuoromethyl)benzyl alcohol (0.6

mL, 3.9 mmol) was used for the synthesis using general procedure

C to provide the desired product 7e. Yield ) 13%. ¹H NMR (400

MHz, DMSO-d₆) δ 7.85 (s, 1H), 7.76 (d, J ) 7.2 Hz, 1H), 7.65

(m, 2H), 7.38 (bs, 1H), 7.3 (bs, 1H), 7.21 (t, J ) 8.4 Hz, 1H), 6.72

(dd, J ) 0.8, 8.4 Hz, 1H), 6.41 (d, J ) 7.6 Hz, 1H), 5.97 (bs, 2H),

5.83 (m, 1H), 1.72 (d, J ) 6.4 Hz, 3H). MS m/z (ESI) 350 (M +

H)⁺. HPLC t_R) 9.7 min (system 1), 12.7 min (system 2).

5-[1-(3-Fluorophenyl)ethoxy]quinazoline-2,4-diamine (7f).

5-(Tetrahydropyran-2-ylmethoxy)quinazoline-2,4-diamine (6m).

Yield ) 58%. ¹H NMR (500 MHz, DMSO-d₆) δ 7.46 (s, 1H), 7.34

(t, J ) 7.5 Hz, 1H), 7.21 (s, 1H), 6.77 (d, J ) 8 Hz, 1H), 6.52 (d,

J ) 8 Hz, 1H), 5.95 (s, 2H), 4.15 (d, J ) 9 Hz, 1H), 3.96 (m, 2H),

3.74 (s, 1H), 3.45 (t, J ) 10 Hz, 1H), 1.82 (m, 1H), 1.51 (m, 5H).

MS m/z (ESI) 275 (M + H)⁺.

General Procedure C: Synthesis of Racemic 5-(r-Substituted

benzyl ether)-2,4-Diaminoquinazolines 7a–t. 5-[1-(3-Chlorophe-

nyl)ethoxy]quinazoline-2,4-diamine (7a). Sodium hydride (60%,

7.92 mmol) was suspended in DMF and cooled to 0 °C. 1-(3-

Chlorophenyl)ethanol (1.24 g, 7.92 mmol) was dissolved in DMF

and added dropwise to the sodium hydride mixture. The solution

was allowed to warm to room temperature and stirred for 1 h. The

solution was then cooled to 0 °C. 2,6-Diﬂuorobenzonitrile (2) (0.9

equiv) in DMF was cooled to 0 °C, and the alcohol mixture was

added dropwise to the benzonitrile solution and stirred for 2 h. The

solution was poured over 100 mL of cooled water. The mixture

was extracted with ethyl acetate and solvent removed to yield 1.94 g

of 2-ﬂuoro-6-[1-(3-chlorophenyl)ethoxy]benzonitrile (89% yield).

2-Fluoro-6-[1-(3-chlorophenyl)ethoxy]benzonitrile (137 mg, 0.5

mmol) and guanidine carbonate (216 mg, 1.2 mmol) were heated

at 150 °C in dimethylacetamide for 6 h. The solvent was removed,

and water was added. The resulting solid was ﬁltered and dried in

vacuo to yield 146 mg of 5-[3-(4-chlorophenyl)propoxy]quinazo-

line-2,4-diamine (93% yield). ¹HNMR (500 MHz, DMSO-d₆) δ

7.54 (d, J ) 1.5 Hz, 1H), 7.36 (m, 3H), 7.29 (br s, 2H), 7.22 (t, J

) 8.5 Hz, 1H), 6.72 (d, J ) 8.0 Hz, 1H), 6.39 (d, J ) 7.5 Hz, 1H),

5.99 (br s, 2H), 5.70 (q, J ) 6.0 Hz, 1H), 1.68 (d, J ) 6.5 Hz, 3H).

¹³C NMR (500 MHz, DMSO-d₆) δ 161.8, 160.6, 155.2, 155.0,

144.7, 133.3, 132.1, 130.6, 127.7, 125.7, 124.4, 117.2, 103.3, 101.6,

75.3. MS m/z 316 (M + H)⁺. FTIR 3494, 3471, 3447, 3306, 3098,

2971, 2923, 1650, 1610, 1566, 1507, 1474, 1447, 1432, 1400, 1372,

1352, 1337, 1327, 1245, 1203, 1073, 813, 779, 695. The sample

was converted to the HCl salt using 4 N HCl in dioxane. Anal.

(C₁₆H₁₅ClN₄O·HCl) C, H, N, Cl.

1

Yield ) 43%. H NMR (400 MHz, DMSO-d₆) δ 7.42 (m, 3H),

7.31 (m, 2H), 7.22 (t, J ) 8.4, 8 Hz, 1H), 7.11 (m, 1H), 6.72 (dd,

J ) 0.8 Hz, 1H), 6.41 (d, J ) 7.6 Hz, 1H), 5.96 (bs, 2H), 5.72 (q,

J ) 6.4 Hz, 1H), 1.69 (d, J ) 6.4 Hz, 3H). MS (ESI) m/z 300 (M

+ H)⁺. HPLC t_R) 9.0 min (system 1), 12.0 min (system 2).

5-[1-(3,5-Diﬂuorophenyl)ethoxy]quinazoline-2,4-diamine (7g).

Sodium borohydride (267 mg, 7.0 mmol) was added in portions to

a solution of 3,5-diﬂuoroacetophenone (1 g, 6.4 mmol) in methanol

at room temperature. After 3 h of being stirred, the reaction mixture

was quenched with 10 mL (aqueous) saturated NH₄Cl. The mixture

was extracted with ethyl acetate (3 × 30 mL). Combined organics

were washed with brine and dried over MgSO₄. Crude 3,5-

diﬂuorophenylethanol was obtained upon ﬁltration and concentra-

tion in vacuo to a colorless oil (960 mg, 95% yield). A solution of

3,5-diﬂuorophenylethanol (960 mg, 6.1 mmol) was subsequently

reacted following general procedure C. The crude product was

puriﬁed by ﬂash silica gel and eluted with 5–10% MeOH/DCM

gradient to provide 145 mg of 5-[1-(3,5-diﬂuorophenyl)-1-meth-

ylethoxy]quinazoline-2,4-diamine (42% yield). ¹H NMR (400 MHz,

DMSO-d₆) δ 7.31 (bs, 2H), 7.24 (m, 2H), 7.15 (m, 1H), 6.74 (dd,

J ) 8.0, 7.6 Hz, 1H), 6.40 (d, J ) 7.6 Hz, 1H), 5.97 (bs, 2H), 5.72

(q, J ) 6.4 Hz, 1H), 1.69 (d, J ) 6.4 Hz, 3H). MS m/z (ESI) 315

(M - H)⁺. HPLC t_R) 9.2 min (system 1), 12.2 min (system 2).

5-[1-(4-Fluorophenyl)-1-methylethoxy]quinazoline-2,4-di-

amine (7j). The crude product was puriﬁed by ﬂash silica gel and

eluted with 5–10% MeOH/DCM gradient. 5-[1-(4-Fluorophenyl)-

1-methylethoxy]quinazoline-2,4-diamine was obtained in a 77%

5-[1-(4-Chlorophenyl)ethoxy]quinazoline-2,4-diamine (7b). So-

dium borohydride (26 mmol) was added portionwise to a cooled

(0 °C) solution of 4-chloroacetophenone (2.0 g, 12.94 mmol) in

ethanol under argon atmosphere. The reaction mixture was slowly

warmed to room temperature and stirred for 4 h. The reaction was

quenched with water, and the solvent was removed in vacuo. The

sample was extracted with dichloromethane, washed with water

and brine, and dried, and the solvent was removed to yield 2.0 g

of 1-(4-chlorophenyl)ethanol (98% yield). A solution of 1-(4-

chlorophenyl)ethanol (1.0 g, 6.39 mmol) was subsequently reacted

using general procedure C as described above for synthesis of 7a

1

to provide the target compound 7b. Yield ) 44%. H NMR (500

MHz, DMSO-d₆) δ 7.48 (d, J ) 8.0 Hz, 2H), 7.43 (d, J ) 8.5 Hz,

2H), 7.3 (brd, 2H), 7.21 (t, J ) 8.5 Hz, 1H), 6.7 (d, J ) 8.0 Hz,

1H), 6.38 (d, J ) 8.5 Hz, 1H), 5.94 (s, 2H), 5.71 (q, J ) 6.5 Hz,

1H), 1.67 (d, J ) 6.5 Hz, 3H). MS m/z (ESI) 315 (M + H)⁺. HPLC

t_R) 9.6 min (system 1), 12.7 min (system 2).

1

yield (275 mg). H NMR (400 MHz, DMSO-d₆) δ 7.51 (m, 2H),

7.4 (bs, 1H), 7.3 (bs, 1H), 7.19 (m, 3H), 7.2 (bs, 1H), 6.72 (dd, J

) 8.4, 0.8 Hz, 1H), 6.43 (d, J ) 7.6 Hz, 1H), 5.99 (bs, 2H), 5.72

(q, J ) 6.4 Hz, 1H), 1.68 (d, J ) 6.4 Hz, 3H). MS m/z (ESI) 297

(M - H)⁺. HPLC t_R) 9.012.0 min (system 1), 11.6 min

(system 2).

5-[1-(4-Chlorophenyl)propoxy]quinazoline-2,4-diamine (7c). So-

dium borohydride (2 equiv) was added portionwise to a cooled (0

°C) solution of 4-chloropropiophenone (3.0 g, 17.79 mmol) in

ethanol under argon atmosphere. The reaction mixture was slowly

warmed to room temperature and stirred for 4 h. The reaction was

quenched with water, and the solvent was removed. The sample

was extracted with dichloromethane, washed with water and brine,

and dried and the solvent was removed to yield 3.0 g of

1-(4-chlorophenyl)propanol (99% yield). The 1-(4-chlorophenyl)-

propanol (1.0 g, 5.86 mmol) was reacted following general

procedure C, and the ﬁnal product was isolated following puriﬁca-

tion by silica gel chromatography (5–10% methanol in dichlo-

5-[1-(4-Triﬂuoromethylphenyl)ethoxy]quinazoline-2,4-di-

amine (7k). The material was puriﬁed by ﬂash silica gel and eluted

with 5–10% MeOH/DCM gradient to give 177 mg (34% yield) of

5-[1-methyl-1-(4-triﬂuoromethylphenyl)ethoxy]quinazoline-2,4-di-

1

amine. H NMR (400 MHz, DMSO-d₆) δ 7.76 (d, J ) 8.4 Hz,

2H), 7.68 (d, J ) 8.4 Hz, 2H), 7.37 (bs, 1H), 7.3 (bs, 1H), 7.2 (t,

J ) 8.4, 8.0 Hz, 1H), 6.72 (dd, J ) 8.4, 0.8 Hz, 1H), 6.36 (d, J )

7.2 Hz, 1H), 5.97 (bs, 2H), 5.72 (q, J ) 6.0 Hz, 1H), 1.71 (d, J )

Diaminoquinazoline DeriVatiVes as SMN2 Promoter ActiVator

Journal of Medicinal Chemistry, 2008, Vol. 51, No. 3 463

6.4 Hz, 3H). MS m/z (ESI) 347 (M - H)⁺. HPLC t_R) 9.8 min

6.4 Hz, 3H). MS (ESI) m/z 315 (M + H)⁺. HPLC t_R) 9.5 min

(system 1), 12.6 min (system 2), chiral 99.8% ee, t_R) 26.22 min.

5-[(S)-1-(3-Chlorophenyl)ethoxy]quinazolin-2,4-diamine ((S)-

7a). To a solution of borane-tetrahydrofuran (77.6 mL, 77.62

mmol, Aldrich, 1 M solution in THF) and (R)-MeCBS (12.9 mL,

12.9 mmol, Aldrich, 1 M solution in toluene) was added a solution

of 3-chloroacetophenone (20.0 g, 129.37 mmol) in anhydrous

tetrahydrofuran slowly over 30 min at room temperature. After

complete addition, the reaction mixture was stirred for 10 min and

quenched with 2 N hydrochloric acid over 30 min. The reaction

mixture was extracted with ether, dried, ﬁltered, and concentrated

in vacuo to afford 20.8 g of (S)-1-(3-chlorophenyl)ethanol as a

viscous liquid (>100% yield). A solution of (S)-1-(3-chlorophe-

nyl)ethanol (20.4 g, 130.26 mmol) in DMF was reacted as described

above from the (R)-enantiomer ((R)-7a). The crude white solid

isolated following reaction with guanidine carbonate step was

recrystallized from methanol–water to yield 24.9 g of 5-(S)-1-(3-

chlorophenyl)ethoxyquinazoline-2,4-diamine (75% yield). ¹H NMR

(400 MHz, DMSO-d₆) δ 7.55 (d, J ) 1.6 Hz, 1H), 7.31–7.43 (m,

5H), 7.23 (t, J ) 7.2 Hz), 6.74 (dd, J ) 8.4, 0.8 Hz, 1H), 6.39 (d,

J ) 8.0 Hz, 1H), 6.05 (s, 2H), 5.7 (q, J ) 6.4 Hz, 1H), 1.69 (d, J

) 6.4 Hz, 3H). ¹³C NMR (100.5 MHz, DMSO-d₆) δ 161.87,

160.62, 155.17, 155.03, 144.7, 133.31, 132.18, 130.64, 127.68,

125.72, 124.42, 117.16, 103.33, 101.62, 75.37, 23.68. MS m/z (ESI)

315 (M + H)⁺. FT-IR 3509, 3392, 3350, 3308, 3164, 3134, 1642,

1590, 1569, 1550, 1499, 1443, 1252, 1075, 814 cm^-1. Mp 224–225

°C. Chiral HPLC (99.0% ee), t_R) 21.99 min. Anal. (C₁₆H₁₅ClN₄O)

C, H, N.

(system 1), 12.7 min (system 2).

5-((S)-1-Naphthalen-2-ylethoxy)quinazoline-2,4-diamine (7s).

General procedure C was used. Commercially available (S)-(-)-

R-methyl-2-naphthalene methanol was used as the starting material.

Following puriﬁcation by ﬂash silica gel and elution with 5–10%

MeOH/DCM gradient, 310 mg of 5-((S)-1-naphthalen-2-ylethoxy)-

quinazoline-2,4-diamine (37% yield) was obtained. ¹H NMR (500

MHz, DMSO-d₆) δ 7.97 (s, 1H), 7.94 (d, J ) 8.0 Hz, 1H), 7.89

(m, 2H), 7.61 (m, 1H), 7.51 (m, 3H), 7.29 (bs, 1H), 7.18 (t, J )

8.5, 8.0 Hz, 1H), 6.69 (d, J ) 8.5 Hz, 1H), 6.47 (d, J ) 8.0 Hz,

1H), 5.95 (bs, 2H), 5.85 (q, J ) 6.5 Hz, 1H), 1.78 (d, J ) 6.5 Hz,

3H). ¹³C NMR (500 MHz, DMSO-d₆) δ 161.9, 160.6, 155.3, 155.1,

139.5, 132.7, 132.5, 132.1, 128.5, 127.8, 127.6, 126.4, 126.1, 124.6,

123.6, 117.0, 103.5, 101.7, 76.4, 23.7. MS m/z (ESI) 332 (M +

H)⁺.

5-(1-Naphthalen-1-ylethoxy)quinazoline-2,4-diamine (7t). The

racemic R-methyl-1-naphthalene methanol (1.0 g, 5.8 mmol) was

used for the synthesis of this target analogue, and general procedure

C was utilized. Puriﬁcation by ﬂash silica gel and following elution

with 5–10% MeOH/DCM gradient provided 65 mg of 5-(1-

naphthalen-1-yl-ethoxy)quinazoline-2,4-diamine (17% yield). ¹H

NMR (400 MHz, DMSO-d₆) δ 8.32 (d, J ) 8.4 Hz, 1H), 7.98 (d,

J ) 8.0 Hz, 1H), 7.85 (d, J ) 8.0 Hz, 1H), 7.56 (m, 6H), 7.12 (t, J

) 8.0 Hz, 1H), 6.67 (d, J ) 9.2 Hz, 1H), 6.45 (m, 1H), 6.22 (d, J

) 7.6 Hz, 1H), 6.0 (br s, 2H), 1.81 (s, 3H). MS m/z (ESI) 331

(M - H)⁺.

5-[(R)-1-(3-Chlorophenyl)ethoxy]quinazolin-2,4-diamine, ((R)-

7a). To a solution of borane-tetrahydrofuran (0.647 mL, 0.647

mmol, Aldrich, 1 M solution in THF) and (S)-MeCBS (0.647 mL,

0.647 mmol, Aldrich, 1 M solution in toluene) was added

simultaneously a solution of 3-chloroacetophenone (1.00 g, 6.47

mmol) in anhydrous tetrahydrofuran (2.42 mL) and a solution of

borane-tetrahydrofuran (3.24 mL, 3.24 mmol, Aldrich, 1 M

solution in THF) over 30 min at ambient temperature. After

complete addition, the reaction mixture was stirred for 10 min,

quenched slowly at 0 °C with methanol (0.94 mL) and then

saturated HCl in ether (0.12 mL). After being stirred at 0 °C for 5

min and ambient temperature for 30 min, the solution was

concentrated in vacuo to an oil. The oil was twice diluted with

benzene and concentrated in vacuo, diluted with ether, and

concentrated in vacuo to an oil. Puriﬁcation by ﬂash chromatog-

raphy (silica gel, 20% ethyl acetate in hexane) provided (R)-1-(3-

chlorophenyl)ethanol as a clear colorless oil (0.784 g, 78% yield).

Optical rotation in CHCl₃at 20 °C at concentration 0.220 was

+40.9° (literature: +44°).⁴⁵A solution of (R)-1-(3-chlorophenyl)-

ethanol (0.730 g, 4.66 mmol) in DMF (5 mL) was added to a 0 °C

slurry of sodium hydride (0.207 g, 5.18 mmol) in DMF (5 mL)

under nitrogen atmosphere. The reaction mixture was warmed to

room temperature, stirred for 3 h, and then cooled with an ice bath.

The cold orange solution was added dropwise to a 0 °C solution of

2,6-diﬂuorobenzonitrile (0.720 g, 5.18 mmol) in DMF (5 mL). After

the mixture was stirred at ambient temperature for 18 h, the mixture

was diluted with water and extracted with ethyl acetate. The organic

layer was washed with water and brine, dried over MgSO₄, and

concentrated in vacuo to a yellow semisolid. Flash chromatography

puriﬁcation (silica gel, 50% hexane in DCM) afforded the 2-ﬂuoro-

5-[(R)-1-(3-chlorophenyl)ethoxy]benzonitrile as a pale-yellow oil

(0.645 g, 50% yield). A mixture of 2-ﬂuoro-5-[(R)-1-(3-chlorophe-

nyl)ethoxy]benzonitrile (0.630 g, 2.29 mmol) and guanidine carbon-

ate (0.988 g, 5.48 mmol) in dimethylacetamide (4.0 mL) was heated

at 145 °C for 7 h. The mixture was concentrated in vacuo to a

black solid. The solid was stirred in a mixture of water and ethyl

acetate, and then the tan solid was collected by ﬁltration in vacuo.

Flash chromatography puriﬁcation (silica gel, 10% methanol in

DCM) afforded 5-(R)-1-(3-chlorophenyl)ethoxyquinazoline-2,4-

diamine as an off-white solid (0.357 g, 49% yield). ¹H NMR (400

MHz, DMSO-d₆) δ 7.54 (d, J ) 1.6 Hz, 1H), 7.32–7.44 (m, 5H),

7.23 (t, J ) 8.4 Hz), 6.72 (dd, J ) 8.4, 0.8 Hz, 1H), 6.41 (d, J )

8.0 Hz, 1H), 6.03 (s, 2H), 5.71 (q, J ) 6.4 Hz, 1H), 1.68 (d, J )

5-(1-Phenylethoxy)quinazoline-2,4-diamine ((RS)-7u). Sodium

hydride (60%, 316 mg, 7.9 mmol) was suspended in DMF and

cooled to 0 °C. Starting with sec-phenethyl alcohol (966 mg, 7.9

mmol), synthesis of the target compound was carried out following

general procedure C described for 7a. The desired product was

isolated following recrystallization from ethanol/water to yield 49

mg of 5-(1-phenylethoxy)quinazoline-2,4-diamine (35% yield). ¹H

NMR (500 MHz, DMSO-d₆) δ 7.44 (d, J ) 7 Hz, 2H), 7.37 (m,

3H), 7.27 (m, 2H), 7.20 (t, J ) 8.5 Hz, 1H), 6.69 (d, J ) 8 Hz,

1H), 6.41 (d, J ) 7.5 Hz, 1H), 5.93 (s, 2H), 5.68 (q, J ) 6 Hz,

1H), 1.37 (d, J ) 6.5 Hz, 3H). MS m/z (ESI) 281 (M + H)⁺. HPLC

t_R) 8.9 min (system 1), 11.9 min (system 2).

5-((R)-1-Phenylethoxy)quinazoline-2,4-diamine ((R)-7u). Start-

ing with (R)-sec-phenethyl alcohol (0.33 mL, 4.1 mmol) and

following the procedure described above for the racemic product

(7u), 115 mg of 5-((R)-1-phenylethoxy)quinazoline-2,4-diamine

1

(21% yield) was isolated. H NMR (400 MHz, DMSO-d₆) δ 7.44

(m, 3H), 7.23 (t, J ) 8.0 Hz, 2H), 7.02 (dd, J ) 7.2, 4.2 Hz, 1H),

6.71(dd, J ) 7.6, 0.8 Hz, 1H), 6.42 (d, J ) 7.2 Hz, 1H), 6.19 (bs,

2H), 5.95 (bs, 2H), 5.69 (m, 1H), 1.68 (d, J ) 6.4 Hz, 3H). MS

m/z (ESI) 281 (M + H)⁺.

5-((S)-1-Phenylethoxy)quinazoline-2,4-diamine ((S)-7u). Fol-

lowing the procedure described above for ((R)-7u) enantiomer, (S)-

sec-phenethyl alcohol (611 mg, 5 mmol) provided 429 mg of 5-((S)-

1-phenylethoxy)quinazoline-2,4-diamine (77% yield). ¹H NMR

(400 MHz, DMSO-d₆) δ 7.92 (s, 1H), 7.44 (m, 7H), 6.70 (dd, J )

8.4, 0.8 Hz, 1 H), 6.43 (d, J ) 7.6 Hz, 1H), 6.03 (br s, 2H), 5.69

(q, J ) 6.4 Hz, 1H), 1.67 (d, J ) 6.4 Hz, 3H). MS m/z (ESI) 281

(M + H)⁺

. HPLC t_R) 8.9 min (system 1), 12.0 min (system 2).

5-[1-(3-Chlorophenyl)-1-methylethoxy]quinazoline-2,4-di-

amine (8a). To a cooled (-10 °C) and stirred solution of

3-chloroacetophenone (1.5 g, 9.7 mmol) in anhydrous ether (20

mL) was added dropwise methylmagnesium bromide (3.6 mL, 10.67

mmol, Aldrich, 3 M solution in ether). The reaction mixture was

slowly warmed to room temperature and stirred for 20 h. Ot was

cooled to 0 °C, and the reaction was quenched with saturated

ammonium chloride solution (6 mL). The sample was extracted

with ethyl acetate, washed with brine, dried, and concentrated in

vacuo to afford 1.55 g of 2-(3-chlorophenyl)propan-2-ol (94%

yield). 2-(3-Chlorophenyl)propan-2-ol (0.8 g, 4.69 mmol) was

reacted using general procedure A. The crude product from step 2

was puriﬁed by silica gel column chromatography using 5%

methanol in dichloromethane to afford 88 mg of 5-[1-(3-chlorophe-

464 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 3

Thurmond et al.

nyl)-1-methylethoxy]quinazoline-2,4-diamine (26% yield). ¹H NMR

(400 MHz, DMSO-d₆) δ 7.38–7.56 (m, 6H), 7.08 (t, J ) 8.8 Hz,

1H), 6.7 (d, J ) 9.6 Hz, 1H), 6.05 (s, 2H), 5.8 (d, J ) 8.4 Hz, 1H),

1.81 (s, 6H). MS m/z (ESI) 329 (M + H)⁺. HPLC t_R) 9.7 min

(system 1), 13.0 min (system 2).

°C, and the alcohol mixture was added dropwise to the benzonitrile

solution and stirred for 2 h. The solution was poured over 100 mL

of ice-cooled water. The solution was stored in the refrigerator for

3 h and solid was collected by ﬁltration and dried in vacuo to yield

2.97 g of 2-(6-chloroindan-1-yloxy)-6-ﬂuorobenzonitrile (76%

yield). 2-(6-Chloroindan-1-yloxy)-6-ﬂuorobenzonitrile (2.91 g, 10

mmol) was dissolved in dimethylacetamide (100 mL) with guani-

dine carbonate (3.82 g, 21 mmol). The vessel was purged with N₂,

sealed, and heated to 150 °C for 8 h and then cooled to room

temperature overnight. The reaction solution was slowly diluted

with 150 mL of water and stirred 1 h, then refrigerated for 1 h to

form a ﬁne brown precipitate that was removed by ﬁltration in

vacuo, washed with ethanol, and dried overnight under high vacuum

(10 Torr) at 30 °C. The dry solids were slurried in boiling methanol

and ﬁltered while hot to give 1.22 g of 5-(6-chloroindan-1-

5-[1-(3-Chlorophenyl)cyclohexyloxy]quinazoline-2,4-di-

amine (8b). To a cooled (0 °C) and stirred solution of cyclohex-

anone (0.3 g, 3.1 mmol) in anhydrous THF (10 mL) was added

dropwise 3-chlorophenylmagnesium bromide (3.4 mL, 10.67 mmol,

Aldrich, 1 M solution in THF). The reaction mixture was slowly

warmed to room temperature and stirred for 20 h. It was cooled to

0 °C, and the reaction was quenched with saturated ammonium

chloride solution (4 mL). The sample was extracted with ethyl

acetate, washed with brine, dried, and concentrated in vacuo to

afford 0.55 g of 1-(3-chlorophenyl)cyclohexanol (>100% yield).

1-(3-Chlorophenyl)cyclohexanol (0.5 g, 2.84 mmol) was reacted

using general procedure A. The crude product from step 2 was

puriﬁed by silica gel column chromatography using 10% methanol

in dichloromethane, giving 40 mg of 5-[1-(3-chlorophenyl)cyclo-

hexyloxy]quinazoline-2,4-diamine (24% yield). ¹H NMR (400

MHz, DMSO-d₆) δ 7.52–7.59 (m, 2H), 7.36–7.46 (m, 4H), 7.02

(t, J ) 8.0 Hz, 1H), 6.66 (dd, J ) 8.4, 0.8 Hz, 1H), 5.94 (s, 2H),

5.76 (dd, J ) 8.2, 1.2 Hz, 1H), 2.42 (m, 3H), 1.88–2.01 (m, 2H),

1.45–1.68 (m, 4H), 1.35–1.42 (m, 1H). MS m/z (APCI) 269 (M +

H)⁺.

1

yloxy)quinazoline-2,4-diamine (30% yield). H NMR (500 MHz,

DMSO-d₆) δ 7.55 (s, 1H), 7.41 (t, J ) 8.0 Hz, 1H), 7.41 (s, 1H),

7.11 (br s, 1H), 6.96 (br s, 1H), 6.82 (d, J ) 8.5 Hz, 1H), 6.77 (d,

J ) 8.0 Hz, 1H), 6.01 (s, 3H), 3.06 (m, 1H), 2.92 (m, 1H), 2.68

(m, 1H), 2.17 (m, 1H). ¹³C NMR (500 MHz, DMSO-d₆) δ 161.8,

160.6, 155.5, 155.2, 143.3, 143.0, 132.4, 131.2, 129.1, 126.8, 124.9,

117.3, 103.1, 101.7, 81.6, 31.9, 29.3. MS m/z (ESI) 410 (M + H)⁺.

HPLC t_R) 9.9 min (system 1), 16.1 min (system 2). FT-IR 3505,

3385, 3306, 3116, 2989, 2955, 2911, 1646, 1613, 1589, 1573, 1558,

1499, 1478, 1442, 1430, 1403, 1354, 1345, 1282, 1247, 1216, 1175,

1062, 813 cm^-1

.

5-[1-(3-Chlorophenyl)cyclopropoxy]quinazoline-2,4-di-

amine (8c). To a heated (50 °C) and vigorously stirred suspension

of samarium powder (1.0 g, 6.65 mmol) in anhydrous THF (3 mL)

were added dropwise a solution of methyl 3-chlorobenzoate (0.27

g, 1.58 mmol) and diiodomethane (1.27 g, 4.74 mmol) over 1.5 h.

The reaction mixture was stirred at room temperature for 1 h, cooled

to -10 °C, and then quenched with 1 N HCl solution (6 mL). The

sample was extracted with ether, washed with brine, dried, and

concentrated in vacuo. Puriﬁcation by silica gel column chroma-

tography using 1:1 DCM-Hex afforded 0.1 g of 1-(3-chlorophe-

nyl)cyclopropanol (37% yield). 1-(3-Chlorophenyl)cyclopropanol

(0.09 g, 0.53 mmol) was reacted using general procedure A. The

crude product from step 2 was puriﬁed by silica gel column

chromatography using 5–10% methanol in dichloromethane to

afford 30 mg of 5-[1-(3-chlorophenyl)cyclopropoxy]quinazoline-

5-(4-Chloroindan-1-yloxy)quinazoline-2,4-diamine (9b). 4-Chlo-

ro-1-indanone (689.8 mg, 4.1 mmol) was stirred in methanol (4

mL) and cooled to 0 °C. Sodium borohydride (156.6 mg, 4.1 mmol)

was added in small portions at 0 °C, and the mixture was stirred at

0 °C for ½ h. Water (5 mL) was added to the reaction mixture

when lots of white foam formed. The mixture was neutralized with

saturated ammonium chloride (10 mL). The solid was ﬁltered,

washed with water, and dried in vacuo to obtain 630.9 mg of white

solid as 4-chloro-1-indanol (90% yield). 4-Chloro-1-indanol (596

mg, 3.5 mmol) was reacted as described for 9a to provide the crude

product that was puriﬁed by column chromatography with DCM/

methanol, 9:1, to yield 90.9 mg of 5-(4-chloroindan-1-ylox-

y)quinazoline-2,4-diamine (16% yield). ¹H NMR (400 MHz,

DMSO-d₆) δ 7.46 (dd, J ) 8.4, 16.8 Hz, 3H), 7.35 (t, J ) 8.0 Hz,

1H), 7.25 (bs, 1H), 7.05 (bs, 1H), 6.83 (m, 2H), 6.18 (s, 2H), 6.11

(m, 1H), 3.10 (m, 1H), 2.98 (m, 1H), 2.74 (m, 1H), 2.19 (m, 1H).

MS m/z (ESI) 327 (M + H)⁺. HPLC t_R) 9.9 min (system 1),

13.2 min (system 2).

1

2,4-diamine (54% yield). H NMR (400 MHz, DMSO-d₆) δ 7.67

(s, 2H), 7.53–7.58 (m, 1H), 7.26–7.46 (m, 3H), 6.84 (dd, J ) 8.2,

0.8 Hz, 1H), 6.41 (s, 2H), 6.32–6.4 (m, 2H), 1.72 (d, J ) 7.2 Hz,

4H). MS m/z (ESI) 327 (M + H)⁺. HPLC t_R) 9.8 min (system

1), 13.0 min (system 2).

General Procedure D: Synthesis of 5-(Aryloxymethyl)quinazo-

line-2,4-diamine (10a-i). 5-(3-Chlorophenoxymethyl)quinazo-

line-2,4-diamine (10a). 3-Chlorophenol (0.27 g, 2.07 mmol) and

potassium carbonate (1 equiv) were added to a cooled (0 °C) and

stirred solution of 2-bromomethyl-6-nitrobenzonitrile⁴⁴(0.5 g, 2.07

mmol) in DMF under nitrogen atmosphere. The reaction mixture

was stirred at 0 °C for 1.5 h, then diluted with pyridine (1.5 mL)

and water, stirred for 1 h, ﬁltered, and dried in vacuo. Puriﬁcation

by silica gel chromatography (1:1 dichloromethane in hexanes)

yielded 185 mg of 2-(3-chlorophenoxymethyl)-6-nitrobenzonitrile

(31% yield). A solution of tin(II) chloride (0.4 g, 1.73 mmol) and

concentrated hydrochloric acid (1.2 mL) was cooled (15 °C), and

a solution of 2-(3-chlorophenoxymethyl)-6-nitrobenzonitrile (100.0

mg, 0.35 mmol) was added while stirring. The reaction mixture

was slowly warmed to room temperature and stirred for 3 h. The

reaction mixture was poured over crushed ice and potassium

hydroxide solution. The mixture was stirred, extracted with dichlo-

romethane, ﬁltered, and concentrated in vacuo. Puriﬁcation by silica

gel chromatography (dichloromethane) yielded 38 mg of 2-amino-

6-(3-chlorophenoxymethyl)benzonitrile (43% yield). 2-Amino-6-

(3-chlorophenoxymethyl)benzonitrile (35.0 mg, 0.14 mmol) and

chloroformamidine hydrochloride (23.0 mg, 0.2 mmol) were heat-

ed at 140 °C in diglyme for 3 h. The reaction mixture was diluted

with water, stirred for 2 h, ﬁltered, washed with water, and

concentrated in vacuo. Puriﬁcation by silica gel chromatography

(5–10% methanol in dichloromethane) yielded 20 mg of 5-(3-

chlorophenoxymethyl)quinazoline-2,4-diamine (48% yield). HPLC

5-(6-Chloroindan-1-yloxy)quinazoline-2,4-diamine (9a). 3-(4-

Chlorophenyl)propionic acid (3.83 g, 20.7 mmol) was dissolved

in 50 mL of thionyl chloride and stirred at room temperature

overnight. The excess thionyl chloride is removed in vacuo to give

3-(4-chlorophenyl)propionyl chloride as a pale-yellow oil, which

was used without puriﬁcation (4.21 g, ∼100% yield). 3-(4-

Chlorophenyl)propionyl chloride (4.21 g, 20.7 mmol) dissolved in

25 mL of dichloromethane was slowly added to a cold (0 °C), stirred

suspension of aluminum chloride (2.77 g, 20.7 mmol) in 75 mL of

dichloromethane. After being stirred for 10 min at 0 °C, the reaction

mixture was heated to reﬂux for 4 h and then cooled to room

temperature overnight. The mixture was diluted with 100 mL of

water, and the layers were separated. The organic layer was washed

with 0.1 M sodium hydroxide and brine and then dried over

magnesium sulfate. The dried solution is concentrated in vacuo to

give 2.51 g of 6-chloroindan-1-one (72% yield). Sodium borohy-

dride (1.1 equiv) was added portionwise to a solution of 6-chlor-

oindan-1-one (2.51 g, 15 mmol) in methanol. The reaction mixture

was stirred overnight at room temperature and then quenched with

water and extracted with ethyl acetate. Solvent was removed in

vacuo to yield 2.03 g of 6-chloroindan-1-ol (80% yield). 6-Chlor-

oindan-1-ol (2.11 g, 12.5 mmol) was dissolved in DMF and added

dropwise to a stirred suspension of sodium hydride (60%, 1 equiv)

in DMF at 0 °C. The solution was allowed to warm to room

temperature and stirred for 1 h. The solution was then cooled to 0

°C. 2,6-Diﬂuorobenzonitrile (0.9 equiv) in DMF was cooled to 0

Diaminoquinazoline DeriVatiVes as SMN2 Promoter ActiVator

Journal of Medicinal Chemistry, 2008, Vol. 51, No. 3 465

t_R) 9.3 min (system 1), 12.5 min (system 2). ¹H NMR (400 MHz,

DMSO-d₆) δ 7.5 (dd, J ) 8.6, 7.2 Hz, 1H), 7.37 (t, J ) 8.4 Hz,

1H), 7.27 (m, 2H), 7.21 (d, J ) 6.0 Hz, 1H), 7.07–7.12 (m, 2H),

6.91 (s, 2H), 6.27 (s, 2H), 5.44 (s, 2H). MS m/z (ESI) 301 (M +

H)⁺.

5-(2-Chlorophenoxymethyl)quinazoline-2,4-diamine (10b).

The target compound was prepared from 2-chlorophenol (0.11 g,

0.83 mmol) using general procedure D. Puriﬁcation by silica gel

chromatography (5% methanol in dichloromethane) yielded 50 mg

of 5-(2-chlorophenoxymethyl)quinazoline-2,4-diamine (54% yield).

¹H NMR (400 MHz, DMSO-d₆) δ 7.44–7.5 (m, 3H), 7.36 (ddd, J

) 8.0, 8.0, 1.6 Hz, 1H), 7.26 (dd, J ) 8.4, 1.2 Hz, 1H), 7.21 (dd,

J ) 8.4, 1.2 Hz, 1H), 7.04 (ddd, J ) 7.6, 7.6, 1.6 Hz, 1H), 6.93 (s,

2H), 6.14 (s, 2H), 5.51 (s, 2H). MS m/z (ESI) 301 (M + H)⁺.

HPLC t_R) 8.9 min (system 1), 12.2 min (system 2).

5-(2-Iodophenoxymethyl)quinazoline-2,4-diamine (10c). The

target compound was prepared from 2-iodophenol (0.18 g, 0.87

mmol) following general procedure D. Puriﬁcation by silica gel

chromatography (5% methanol in dichloromethane) yielded 65 mg

(49%) of 10c. ¹H NMR (400 MHz, DMSO-d₆) δ 7.82 (d, J ) 7.6

Hz, 1H), 7.49 (t, J ) 7.6 Hz, 1H), 7.42 (t, J ) 7.2 Hz, 1H), 7.27

(m, 3H), 6.86 (m, 3H), 6.16 (s, 2H), 5.47 (s, 2H). MS m/z (ESI)

393 (M + H)⁺. HPLC t_R) 9.3 min (system 1), 12.6 min

(system 2).

4-(2,4-Diaminoquinazolin-5-yloxymethyl)piperidine-1-carboxylic acid

tert-butyl ester (16.4 g, 44.0 mmol) was suspended in 80 mL of

dioxane, and 4 M HCl in dioxane (176 mL, 176 mmol) was added,

changing the mixture from homogeneous to heterogeneous. After

the mixture was stirred for 5 h at room temperature, solids were

ﬁltered off and rinsed once with ether. Solids were triturated with

1 N NaOH for 30 min. Solids were collected by ﬁltration and dried

in vacuo to yield 5-(piperidin-4-ylmethoxy)quinazoline-2,4-diamine

31 (9.05 g, 75% yield).

General Procedure E: Functionalization of 5-(Piperidin-4-

ylmethoxy)quinazoline-2,4-diamine. To a mixture of 5-(piperidin-

4-ylmethoxy)quinazoline-2,4-diamine (31) (4.4 mmol, 1 equiv) and

triethylamine (1.2 mL, 8.8 mmol, 2 equiv) in 10 mL of DMF was

added a benzyl halide (4.8 mmole, 1.1 equiv), and the mixture was

stirred at 60 °C for 16 h. The mixture was was cooled to room

temperature and concentrated in vacuo. Residue was taken up in

30 mL of dichloromethane and 2 mL of methanol, creating a

homogeneous mixture. Product 11 was puriﬁed via ﬂash chroma-

tography using a 5–10% MeOH/DCM with 0.1% NH₄OH gradient.

Material was then repuriﬁed by trituration with 1 N NaOH and

ethanol for 1.5 h.

[5-(1-(2-Fluorobenzyl)piperidin-4-ylmethoxy]quinazoline-2,4-

diamine (11a). The title compound was prepared using general

procedure E from 31 and 2-ﬂuorobenzyl bromide (0.6 mL, 4.8

mmol), and solids were collected by ﬁltration and dried in vacuo

to obtain 5-[(1-(2-ﬂuorobenzyl)piperidin-4-ylmethoxy]quinazoline-

5-(2-Fluorophenoxymethyl)quinazoline-2,4-diamine (10d).

The target compound was prepared from 2-ﬂuorophenol (0.07 g,

0.62 mmol) following general procedure D. Puriﬁcation by silica

gel chromatography (5% methanol in dichloromethane) yielded 65

1

2,4-diamine, 740 mg (44% yield). H NMR (500 MHz, DMSO-

d₆) δ 7.40 (t, J ) 7.5 Hz, 1H), 7.31 (m, 1H), 7.18 (m, 3H), 6.77

(d, J ) 8.0 Hz, 1H), 6.52 (d, J ) 8.0 Hz, 1H), 5.98 (br s, 2H), 3.97

(d, J ) 6.0 Hz, 2H), 3.53 (s, 2H), 2.85 (br d, J ) 11.5 Hz, 2H),

2.01 (t, J ) 10 Hz, 2H), 1.84 (m, 1H), 1.73 (d, J ) 12 Hz, 2H),

1.34 (m, 2H). MS m/z (ESI) 382 (M + H)⁺. Anal. (C₂₁H₂₄FN₅O)

C, H, N.

1

mg (46%) of 10d. H NMR (400 MHz, DMSO-d₆) d 7.42–7.48

(m, 2H), 7.22–7.29 (m, 2H), 7.13–7.2 (m, 2H), 6.99–7.06 (m, 1H),

6.94 (s, 2H), 6.14 (s, 2H), 5.48 (s, 2H). MS m/z (ESI) 286 (M +

H)⁺. HPLC t_R) 8.6 min (system 1), 11.6 min (system 2).

5-(2,4-Diﬂuorophenoxymethyl)quinazoline-2,4-diamine (10i).

The target compound was prepared from 2,4-diﬂuorophenol (0.11

g, 0.83 mmol) following general procedure D. Puriﬁcation by silica

gel chromatography (5% methanol in dichloromethane) yielded 72

mg (52%) of 10i. ¹H NMR (400 MHz, DMSO-d₆) δ 7.43–7.51

(m, 2H), 7.3–7.38 (m, 1H), 7.27 (dd, J ) 8.4, 1.2 Hz, 1H), 7.14

(d,J ) 6.8 Hz, 1H), 7.05–7.11 (m, 1H), 7.04 (s, 2H), 6.25 (s, 2H),

5.47 (s, 2H). MS m/z (ESI) 303 (M + H)⁺.

5-[1-(3-Chlorobenzyl)piperidin-4-ylmethoxy]quinazoline-2,4-

diamine (11b). The title compound employing general procedure

E was prepared using 3-chlorobenzyl bromide in DMF (4 mL).

The crude solid was puriﬁed by silica gel column chromatography

using 3–10% methanol in dichloromethane to afford 1.82 g of 5-[1-

(3-chlorobenzyl)piperidin-4-ylmethoxy]quinazoline-2,4-diamine (50%

1

yield). HPLC t_R) 7.5 min (system 1), 10.2 min (system 2). H

NMR (400 MHz, DMSO-d₆) δ 7.25–7.38 (m, 5H), 7.2 (s, 2H),

6.77 (dd, J ) 8.0, 1.2 Hz, 1H), 6.53 (dd, J ) 8.0, 0.8 Hz, 1H),

5.95 (s, 2H), 3.99 (d, J ) 6.4 Hz, 2H), 3.48 (s, 2H), 2.84 (d, J )

11.2 Hz, 2H), 1.7–2.07 (m, 5H), 1.3–1.42 9m, 2H). MS m/z (ESI)

399 (M + H)⁺.

Synthesis of 5-(Piperidin-4-ylmethoxy)quinazoline-2,4-di-

amine (31). 4-Piperidinemethanol 27 (1.5 g, 13.0 mmol) was

dissolved in a mixture of dichloromethane and triethylamine (2.7

mL, 19.5 mmol). Di-tert-butyl dicarbonate (3.1 g, 14.3 mmol) was

added and stirred for 2.5 h at room temperature. The mixture was

poured over dilute acetic acid, and the organic layer was separated.

Organics were washed with water, saturated sodium bicarbonate

and brine, dried over MgSO₄, and concentrated in vacuo. Crude

material was puriﬁed by ﬂash chromatography using 1–5%

methanol/DCM gradient to afford 4-hydroxymethylpiperidine-1-

carboxylic acid tert-butyl ester 28 (2.5 g, 89% yield). A solution

of 4-hydroxymethylpiperidine-1-carboxylic acid tert-butyl ester (1

g, 4.6 mmol) in DMF was added dropwise to a suspension of

sodium hydride (60%, 1 equiv) in DMF that was cooled to 0 °C.

The solution was allowed to warm to room temperature and stirred

for 1 h. 2,6-Diﬂuorobenzonitrile 2 (0.9 equiv) in DMF was cooled

to 0 °C, and the alcohol mixture was added dropwise to the

benzonitrile solution and stirred for 2 h. The solution was poured

over 100 mL of cooled water. The solution was stored in the

refrigerator for 3 h, and the solid was collected by ﬁltration. 4-(2-

Cyano-3-ﬂuorophenoxymethyl)piperidine-1-carboxylic acid tert-

butyl ester 29 (1.22 g, 81% yield) was obtained after puriﬁcation

by column chromatography (15–20% ethyl acetate/hexanes gradi-

ent). The cyclization of 4-(2-cyano-3-ﬂuorophenoxymethyl)piperi-

dine-1-carboxylic acid tert-butyl ester 29 (17.9 g, 53.5 mmol) was

carried out using guanidine carbonate (53.5 mmol) and heating at

140 °C in dimethylacetamide for 4 h. The mixture was cooled and

triturated with water for 45 min. Solids were ﬁltered off and

triturated with ethanol. Solids were collected by ﬁltration and dried

in vacuo to afford 16.4 g of 4-(2,4-diaminoquinazolin-5-yloxy-

methyl)piperidine-1-carboxylic acid tert-butyl ester 30 (82% yield).

[5-(1-(2-Chlorobenzyl)piperidin-4-ylmethoxy]quinazoline-2,4-

diamine (11c). The title compound employing general procedure

E was prepared using 2-chlorobenzyl bromide (0.5 mL, 3.7 mmol)

in DMF (4 mL). Solids were ﬁltered off, rinsed once with methanol

and twice with diethyl ether, and dried in vacuo to yield the title

1

compound (684 mg, 47% yield). H NMR (400 MHz, DMSO-d₆)

δ 7.49 (d, J ) 7.6 Hz, 1H), 7.41 (d, J ) 7.6 Hz, 1H), 7.32 (m,

3H), 7.2 (br s, 2H), 6.76 (d, J ) 7.6 Hz, 1H), 6.53 (d, J ) 7.6 Hz,

1H), 5.94 (br s, 2H), 3.99 (d, J ) 6.0 Hz, 2H), 3.57 (s, 2H), 2.89

(m, 2H), 2.08 (t, J ) 12 Hz, 2H), 1.89 (m, 1H), 1.75 (d, J ) 12

Hz, 2H), 1.36 (m, 2H). MS m/z (ESI) 398 (M + H)⁺.

General Procedure F: Functionalization of 5-(Piperidin-4-

ylmethoxy)quinazoline-2,4-diamine Using Polymer-Bound Re-

agents: 5-[1-(4-Chlorobenzyl)piperidin-4-ylmethoxy]quinazoline-

2,4-diamine Hydrochloride (11d). 5-(Piperidin-4-ylmethoxy)quinazo-

line-2,4-diamine (75 mg, 0.27 mmol) (31) was shaken for 60 h at

60 °C in the presence of 3-(morphonlino)propylpolystyrene sul-

fonamide (PS-NMM) (240 mg, 0.6 mmol) and 3,4-diﬂuorobenzyl

bromide (66 mg, 0.32 mmol) in 4 mL N,N-DMF. Tris-(2-

aminoethyl)aminomethyl polystyrene (PS-trisamine) (168 mg, 0.6

mmol) was then added to the mixture, and shaking was continued

for an additional 2 h. Resins were ﬁltered off and rinsed with

methanol. The ﬁltrate was concentrated in vacuo, and the residue

was triturated with 2 mL of 1 N NaOH and ethanol (1:1). Solids

were collected by ﬁltration to yield 5-[1-(3,4-diﬂuorobenzyl)pip-

eridin-4-ylmethoxy]quinazoline-2,4-diamine, which was stirred in

466 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 3

Thurmond et al.

the presence of 4 M HCl in dioxane (4 equiv). Solids were ﬁltered,

rinsed once with ether, and dried in vacuo to give 5-[1-(4-

chlorobenzyl)piperidin-4-ylmethoxy]quinazoline-2,4-diamine hy-

[5-(1-(2-Fluorobenzenesulfonyl)piperidin-4-ylmethoxy]quinazo-

line-2,4-diamine (13a). The title compound employing general

procedure E was prepared using 2-ﬂuorobenzenesulfonyl chloride

(196 mg, 1.0 mmol) in 2.5 mL of DMF to yield crude product

(331 mg, 85% yield). Following puriﬁcation by silica gel column

(5% MeOH/CH₂Cl₂+ 1% NH₄OH), the fractions containing the

desired product were reduced in vacuo to ∼5 mL and diluted with

2 M HCl in Et₂O (5 mL). The solvents were completely removed

under N₂purge. The resulting semisolid was triturated with absolute

EtOH and stored at 2 °C for 72 h. The resulting solids were ﬁltered

and dried under vacuum at 30 °C overnight to yield 182 mg of

5-[1-(2-ﬂuorobenzenesulfonyl)piperidin-4-ylmethoxy]quinazoline-

1

drochloride (66 mg, 56% yield). H NMR (400 MHz, DMSO-d₆)

δ 12.9 (s, 1H), 11.3 (s, 1H), 8.82 (s, 1H), 8.23 (s, 1H), 7.7 (m,

4H), 7.52 (d, J ) 6.8 Hz, 2H), 7.02 (m, 2H), 4.66 (br s, 3H), 4.26

(d, J ) 4.8 Hz, 2H), 4.17 (d, J ) 6.4 Hz, 2H), 3.34 (br d, J ) 11.6

Hz, 2H), 2.87 (m, 2H), 2.2 (m, 1H), 1.81 (m, 4H). MS m/z (ESI)

398 (M + H)⁺.

5-[1-(1-Naphthalen-1-ylmethylpiperidin-4-ylmethoxy]quinazo-

line-2,4-diamine (11e). The title compound employing general

procedure F was prepared using 1-(chloromethyl)naphthalene (57

1

2,4-diamine hydrochloride (yield ) 43%.). H NMR (500 MHz,

1

mg, 0.32 mmol) in 4 mL of DMF to yield 94 mg (84% yield). H

DMSO-d₆) δ 12.93 (s, 1H), 8.95 (s, 1H), 8.16 (s, 1H), 7.80 (t, J )

7.5 Hz, 1H), 7.77 (dt, J ) 7.5, 2.5 Hz, 1H), 7.68 (t, J ) 8.5 Hz,

1H), 7.50 (t, J ) 8.5 Hz, 1H), 7.45 (t, J ) 7.5 Hz, 1H), 6.99 (d, J

) 8.5 Hz, 1H), 4.14 (d, J ) 6.5 Hz, 2H), 3.74 (br d, J ) 12.0 Hz,

2H), 2.54 (m, 2H), 2.07 (m, 1H), 1.84 (br d, J ) 12 Hz, 2H), 1.33

(m, 2H). ¹³C NMR (500 MHz, DMSO-d₆) δ 162.3, 159.2, 157.2,

156.8, 154.2, 141.0, 136.2, 135.8, 130.8, 125.1, 124.2, 117.6, 117.5,

108.6, 106.8, 99.6, 73.1, 45.1, 33.4, 27.8. MS m/z (ESI) 432 (M +

H)⁺.

NMR (400 MHz, DMSO-d₆) δ 8.26 (d, J ) 8.0 Hz, 1H), 7.85 (m,

2H), 7.53 (m, 3H), 7.44 (d, J ) 4.8 Hz, 1H), 7.33 (t, J ) 8.4 Hz,

1H), 7.19 (br s, 2H), 6.75 (d, J ) 8.4 Hz, 2H), 6.51 (d, J ) 8.0 Hz,

1H), 5.94 (br s, 2H), 3.79 (d, J ) 6.4 Hz, 2H), 3.87 (s, 2H), 2.92

(d, J ) 11.2 Hz, 2H), 2.07 (t, J ) 11.6 Hz, 2H), 1.91 (m, 1H),

1.74 (d, J ) 12.4 Hz, 2H), 1.3 (m, 2H). MS m/z (ESI) 414 (M +

H)⁺.

5-[1-(2-Naphthalen-1-ylmethylpiperidin-4-ylmethoxy]quinazo-

line-2,4-diamine (11f). The title compound employing general

procedure F was prepared using 2-(bromomethyl)naphthalene (77

5-[4-(3-Chlorobenzenesulfonyl)piperidin-4-ylmethoxy]quinazo-

line-2,4-diamine (13b). Compound 31 (50 mg, 0.18 mmol) was

shaken for 60 h at room temperature in the presence of 3-(mor-

pholino)propylpolystyrene sulfonamide (PS-NMM) (160 mg, 0.4

mmol) and 3-chlorobenzenesulfonyl chloride (78 mg, 0.37 mmol)

in 4 mL of DMF. Tris-(2-aminoethyl)aminomethyl polystyrene (PS-

trisamine) (112 mg, 0.4 mmol) was then added to the mixture, and

shaking was continued for an additional 2 h. Resins were ﬁltered

off and rinsed with methanol. Filtrate was concentrated at re-

duced pressure, and residue was triturated with 2 mL of 1 N NaOH

and ethanol (1:1). Solids were collected by ﬁltration and dried in

vacuo to yield 80 mg of 5-[4-(3-chlorobenzenesulfonyl)piperidin-

4-ylmethoxy]quinazoline-2,4-diamine (99% yield). ¹H NMR (400

MHz, DMSO-d₆) δ 7.82 (m, 1H), 7.72 (m, 3H), 7.32 (t, J ) 8.4

Hz, 1H), 7.11 (br s, 2H), 6.75 (d, J ) 8.0 Hz, 1H), 6.50 (d, J ) 7.6

Hz, 1H), 5.93 (br s, 2H), 3.96 (d, J ) 6.4 Hz, 2H), 3.72 (br d, J )

11.6 Hz, 2H), 2.32 (t, J ) 12 Hz, 2H), 1.84 (m, 3H), 1.36 (m, 2H).

MS m/z (ESI) 448 (M + H)⁺.

Biological Assays. SMN2 Promoter Assay. (NSC-34/

ꢀ-Lactamase Assay). The 3.4 kb SMN2 promoter NSC-34/ꢀ-

lactamase assay developed by Jarecki et al.³³was used with the

following modiﬁcations. In brief, 55 000 cells/well were plated into

a 96-well cell culture plate and incubated along with test compounds

for 19 h at 37 °C, 90% humidity, and 5% CO₂. The total assay

volume was 100 µL. Test compound concentrations were varied,

usually in the concentration range 2 pM to 90 µM (dose curves).

At the end of the incubation, 14 µL of CCF2/AM (dye, purchased

from Invitrogen)⁴⁶was added to each of the assay wells and the

assay plates were incubated for another 3 h before the plates are

read on a ﬂuorometer (LJL Analyst). The CCF2/AM dye was

prepared according to an assay kit from Invitrogen (K1085). The

resulting raw data were processed using IDBS activity base to

determine EC₅₀values and other curve parameters. For each sample,

there are two readouts: the substrate at 520 nm (green) and the

product at 450 nm (blue). The ratio of the blue to the green signal

for each sample was calculated, and this value was then divided

by a similar value calculated for the negative control (which contains

no compound). The resulting value is called fold induction (FoldInd)

and was plotted against compound concentration to generate a dose

curve. TSA (trichostatin A, a histone deacetylase inhibitor) was

used as a positive control during the initial stages of the project

and was subsequently replaced with compound 2b as the SAR

elaboration continued. The EC₅₀for each curve is calculated from

the curve along with other curve parameters, including maximum

fold induction (MaxFoldInd, MFI), which is the maximum induction

apparent on the curve.

1

mg, 0.35 mmol) in 4 mL of DMF to yield 75 mg (67% yield). H

NMR (400 MHz, DMSO-d₆) δ 7.88 (m, 3H), 7.79 (s, 1H), 7.48

(m, 3H), 7.34 (t, J ) 8.0 Hz, 1H), 7.22 (br s, 2H), 6.76 (d, J ) 8.0

Hz, 1H), 6.53 (d, J ) 7.6 Hz, 1H), 5.96 (br s, 2H), 3.99 (d, J ) 6.0

Hz, 2H), 3.64 (s, 2H), 2.92 (m, 2H), 2.03 (t, J ) 10.8 Hz, 2H),

1.91 (m, 1H), 1.74 (d, J ) 12.0 Hz, 2H), 1.37 (m, 2H). MS m/z

(ESI) 414 (M + H)⁺.

[4-(2,4-Diaminoquinazolin-5-yloxymethyl)piperidin-1-yl](2-

ﬂuorophenyl)methanone (12a). The title compound employing

general procedure F was prepared using 2-ﬂuorobenzoyl chloride

1

(57 mg, 0.36 mmol) in 4 mL of DMF to yield 43 mg (60%). H

NMR (400 MHz, DMSO-d₆) δ 7.49 (m, 1H), 7.33 (m, 4H), 7.17

(br s, 2H), 6.77 (d, J ) 8.0 Hz, 1H), 6.54 (d, J ) 8.0 Hz, 1H), 5.94

(br s, 2H), 4.56 (br d, J ) 12.8 Hz, 1H), 4.03 (d, J ) 6.0 Hz, 2H),

3.42 (br d, J ) 12.8 Hz, 1H), 3.12 (m, 1H), 2.86 (t, J ) 11.2 Hz,

1H), 2.22 (br s, 1H), 1.89 (br d, J ) 13.6 Hz, 1H), 1.73 (br d, J )

11.2 Hz, 1H), 1.26 (m, 2H). MS m/z (ESI) 396 (M + H)⁺.

(3-Chlorophenyl)-[4-(2,4-diaminoquinazolin-5-yloxymeth-

yl)piperidin-1-yl]methanone (12b). 2-Fluoro-6-(piperidin-1-yl-

methoxy)benzonitrile hydrochloride (0.7 mmol) was suspended in

2 mL of tetrahydrofuran, and in one portion triethylamine was added

(0.3 mL, 2.2 mmol) to the mixture at room temperature. 3-Chlo-

robenzoyl chloride (129 mg, 0.7 mmol) was then added, and the

mixture was stirred at room temperature for 16 h. Reaction was

quenched with 2 mL of 1 N HCl. The mixture was extracted twice

with 10 mL of ethyl acetate, and organic layers were combined.

The mixture was then washed with saturated NaHCO₃and brine

and dried over MgSO₄. The mixture was ﬁltered and concentrated

in vacuo to an oil. Material was puriﬁed via ﬂash silica gel

chromatography using 0.5–2% methanol/dichloromethane gradient

to afford 200 mg of 2-[1-(3-chlorobenzoyl)piperidin-4-ylmethoxy-

6-ﬂuorobenzonitrile (77% yield). The cyclization of 2-[1-(3-

chlorobenzoyl)piperidin-4-ylmethoxy-6-ﬂuorobenzonitrile (194 mg,

0.5 mmol) was carried out using guanidine carbonate (1 equiv

mmol) and heating at 140 °C in dimethylacetamide for 4 h. The

mixture was cooled and triturated with water for 45 min. Solids

were ﬁltered off and triturated with ethanol. Solids were collected

by ﬁltration and dried in vacuo to afford 169 mg of (3-chlorophe-

nyl)-[4-(2,4-diaminoquinazolin-5-yloxymethyl)piperidin-1-yl]meth-

anone (82% yield). HPLC t_R) 8.9 min (system 1), 12.3 min

1

(system 2). H NMR (400 MHz, DMSO-d₆) δ 7.51 (m, 3H), 7.35

(m, 2H), 7.19 (br d, J ) 16.4 Hz, 2H), 6.77 (dd, J ) 8.4, 0.8 Hz,

1H), 6.57 (d, J ) 8.0 Hz, 1H), 5.98 (br s, 2H), 4.51 (br s, 1H),

4.03 (d, J ) 6.4 Hz, 2H), 3.55 (br s, 1H), 3.11 (m, 1H), 2.84 (br

s, 1H), 2.22 (m, 1H), 1.88 (br s, 1H), 1.75 (br s, 1H), 1.33 (br s,

2H). MS m/z (ESI) 412 (M - H)⁺.

DHFR Assay. The assay is an adaptation for 96-well plate format

from Appleman et al.⁴⁷Human, recombinant dihydrofolate reduc-

tase (Sigma, D6566) was incubated along with test compounds (119

Diaminoquinazoline DeriVatiVes as SMN2 Promoter ActiVator

Journal of Medicinal Chemistry, 2008, Vol. 51, No. 3 467

pM to 125 µM) and NADPH (Sigma, N-7505) in 96-well plates.

After a 10 min preincubation at room temperature, dihydrofolic

acid (Sigma, D7006) was added and the kinetic traces at 340 nm

were measured on a Spectra Max 190 plate reader at room

temperature. Reaction rates were determined from the slopes of

the linear portions of the progress curves. The ﬁnal assay conditions

were 0.014 units of DHFR, 300 µM NADPH, 30 µM dihydrofolic

acid, 50 mM Tris, pH 7.6, 100 mM NaCl, and 1.25% DMSO. Each

compound was measured in duplicate on every assay plate. Each

assay plate contained 12 positive controls (DMSO only) and 12

negative controls (no dihydrofolic acid). All data were percent-

normalized with respect to controls and presented as percent

inhibition. Inhibition data were analyzed using Prism and ﬁt to a

standard four-parameter logistic equation to calculate IC₅₀values.

In Vitro Studies Using NSC-34 Murine Cell Line. Test

compound and TSA (trichostatin A, Sigma, T-8552) were formu-

lated in DMSO and added to NSC-34 cell culture for ﬁnal

concentrations of 500 nM, 50 nM (test compound), and 100 nM

(TSA), respectively. Final concentration of DMSO was 0.5%. The

cells were incubated for 18 h and then harvested for RNA isolation

using RNeasy Mini Kit (Qiagen) according to the manufacturer’s

protocol. RNA quality and quantity were measured on an Agilent

2100 bioanalyzer (Agilent Technologies). cDNA synthesis was

performed using High Capacity cDNA Archive kit (Applied

Biosystems) according to the manufacturer’s protocol. Gene expres-

sion analysis was performed using TaqMan Assays-on-Demand

Gene Expression Products (Applied Biosystems) according to the

manufacturer’s protocol (SMN assay no. Mm00488315_m1). Each

assay was run in triplicate, and the data were normalized against

an internal housekeeping gene, B-actin (ACTB assay no. 4352933E).

The relative difference in expression was calculated using the

expression 2e(-∆C_T). Statistical comparisons between groups in

the gene expression studies were performed using Student’s t test,

and an associated probability of <5% was considered signiﬁcant.

Biochemical Analysis in SMA Patient Fibroblasts. Cell lines

derived from skin ﬁbroblasts of a type I SMA patient (line 3813)

and from a SMA carrier (line 3814) have been described previ-

ously.⁴⁸For immunoblot analyses, cells were plated into 12-well

plates at a density of 4000 cells/cm². For immunoﬂuorescence

experiments, cells were seeded onto glass coverslips coated with

1% gelatin at a density of 4000 cells/cm². Drug compounds or

DMSO was added to medium at a 1:1000 dilution. Medium was

changed daily, and fresh compound was added. Immunoblot of

ﬁbroblast extracts was performed as described previously.¹¹Sixty

micrograms of protein extract were loaded per lane of a 12%

polyacrylamide gel. The primary antibodies used in these experi-

ments were a mouse anti-SMN mAb (MANSMA2 (8F7),⁴⁹1:100)

and a mouse anti-ꢀ-tubulin mAb (Tub2.1, Sigma-Aldrich, 1:1000).

Immunostaining of ﬁbroblast cells was accomplished as described

previously⁴⁸with modiﬁcation. Brieﬂy, cells grown on gelatinized

coverslips were ﬁxed with ﬁxative buffer⁵⁰(2% paraformaldehyde,

400 µM CaCl₂, 50 mM sucrose in 100 mM sodium phosphate

buffer, pH 7.4) for 30 min at room temperature, rinsed with

phosphate buffered saline (PBS), and permeabilized with ice-cold

acetone for 10 min. After drying for at least 30 min at room

temperature, the cells were rehydrated with PBS for 10 min and

then blocked with 5× BLOCK⁴⁸(5% horse serum, 5% fetal bovine

serum, and 0.1% bovine serum albumin in PBS) for 60 min at room

temperature. The cells were incubated overnight with primary

antibody solution (mouse anti-SMN MANSMA 2, diluted 1:200

in 1× BLOCK) at 4 °C. The cells were then washed (3 × 10 min)

with PBS and incubated with secondary antibody solution (bioti-

nylated goat antimouse IgG (Jackson ImmunoResearch) diluted

1:400 with 1× BLOCK) for 60 min at room temperature. The cells

were then washed with PBS (3 × 10 min) and then incubated with

AlexaFluor 594-conjugated strepavidin (Invitrogen) diluted 1:200

with PBS for 60 min at room temperature. Cells were then

counterstained with Hoescht 33342 (1 µg/mL) in PBS for 5 min.

After thorough washing with PBS, coverslips were mounted onto

glass slides with ImmuMount (Shandon Lipshaw) and stored at 4

°C until analysis.

For gem counting, immunolabeling was visualized using a Nikon

ﬂuorescent microscope, and the number of gems in 100 randomly

selected nuclei and the number of cells with gems were collected.

Confocal images were obtained using a Zeiss 510 META laser

scanning confocal microscope (Campus Microscopy and Imaging

Facility, The Ohio State University) as a series of optic sections

(thickness of 500 nm/section) taken through the z-axis. These stacks

of sections were then compressed using LSM 5 image browser (Carl

Zeiss, Inc.).

Statistical Analysis. All parametric data were expressed as mean

standard error. Comparisons made between parametric data were

made using one-way ANOVA with a Bonferonni post hoc test. All

statistical analyses were performed using SPSS, version 14.

Pharmacokinetics in Mice. NMRI mice were dosed using the

formulation and dose as indicated, and plasma and brain samples

at the time indicated were processed.

Plasma Sample Processing. An aliquot of 50 µL of plasma

sample was pipetted into a 96-well plate. A solution (150 µL) of

ice-cold acetonitrile containing internal standard was added to each

sample. The plate was vortex-mixed for 3 min and then allowed to

stand for 30 min in a refrigerator (2-8 °C). Samples were

centrifuged for 10 min at 4100 rpm. The supernatant (130 µL) from

each sample was transferred to new 96-well plates, and 130 µL of

mobile phase was added to the supernatant and vortex-mixed for 3

min. Subsequently, the solutions were centrifuged for 10 min at

4100 rpm and an amount of 20 µL of each supernatant was injected

into the LC-MS-MS system for quantiﬁcation. Concentration of

a given compound in plasma is reported as ng/mL.

Brain Sample Processing. Two volumes (mL) of ringer acetate

were added for each gram of brain tissue. The brain tissue sample

was homogenized using a Tissue-Tearor hand-held homogenizer

(BioSpec Products, OK). The sample was homogenized for 30 s in

an ice–water bath. The homogenized brain tissue (2.5 mL sample

from each 1 g of brain tissues) was transferred to an Eppendorf

tube and stored in a -25 °C freezer until analyzed. After thawing

at room temperature, an aliquot of 50 µL of homogenized brain

was pipetted into a 96-well plate. A solution (150 µL) of ice-cold

acetonitrile containing internal standard was added to each sample.

The plate was vortex-mixed for 3 min and then allowed to stand

for 30 min in a refrigerator (2-8 °C). These samples were then

processed as described above for plasma samples. Concentration

of the compound in brain is reported as ng/g. All pharmacokinetic

parameters reported were calculated using WinNoLin (version 5.0).

Computational Chemistry and Molecular Modeling Me-

thods. Conformational Analysis. The 2D molecular models of

molecules 1c, 2b, 4a, 6a, and 6b were constructed using ISIS draw,

and a small structure data ﬁle (sd) was generated. This sd ﬁle was

then subjected to CONCORD 6.0 in the SYBYL7.0³⁹environment,

and 3D structures were generated. The energy minimization of all

the molecules were carried out using MMFF94 with conjugate

gradient method having a gradient norm of 0.001 kcal/mol.

Gasteiger-Huckel charges were assigned to all the molecules. All

the minimized structures were aligned by atom-to-atom map of the

common ether oxygen and quinazoline core atoms. The confor-

mational search analysis of all the above molecules was performed

in SYBYL7.0 using the Systematic Search module. All rotatable

bonds (RB) were analyzed, and they were given a torsional

increment angle of 10°. The “compute energy” and the “electro-

static” options incorporated in the systematic search were used to

generate the conformers with an energy cutoff of 7.0 kcal/mol. At

the same time, the coordinate maps were generated assigning the

origin and the target atomic centers (as shown in Figure 3). Finally,

for all conformations within 3.0 kcal/mol from the lowest energy

conformer, a sweep graph was created by deﬁning the trajectory

type with the sequence of atoms and the vectors maps were

generated.

Analysis of the human DHFR Cocrystal Structure. The human

DHFR-bound ligand (1s3u.pdb) from Protein Data Bank was

analyzed in SYBYL7.0. Atom type and the bond type for the bound

ligand were modiﬁed as needed. By use of the BIOPOLYMER

module of SYBYL7.0, all hydrogen atoms were added to the protein

468 Journal of Medicinal Chemistry, 2008, Vol. 51, No. 3

Thurmond et al.

(13) Vitali, T.; Sossi, V.; Tiziano, F.; Zappata, S.; Giuli, A.; Paravatou-

Petsotas, M.; Neri, G.; Brahe, C. Detection of the survival motor

neuron (SMN) genes by FISH: further evidence for a role for SMN2

in the modulation of disease severity in SMA patients. Hum. Mol.

Genet. 1999, 8, 2525–2532.

and the ligand. All nonhydrogen atoms were deﬁned as an

aggregate, and hydrogens were minimized using the TRIPOS force

ﬁeld. The hydrogen bonding network was created between the

bound ligand and the interacting residues in the active site.

(14) Campbell, L.; Potter, A.; Ignatius, J.; Dubowitz, V.; Davies, K.

Genomic variation and gene conversion in spinal muscular atrophy:

implications for disease process and clinical phenotype. Am. J. Hum.

Genet. 199761, 40–50.

(15) Feldkotter, M.; Schwarzer, V.; Wirth, R.; Wienker, T. F.; Wirth, B.

Quantitative analyses of SMN1 and SMN2 based on real-time light

Cycler PCR: fast and highly reliable carrier testing and prediction of

severity of spinal muscular atrophy. Am. J. Hum. Genet. 2002, 70,

358–368.

(16) Arkblad, E. L.; Darin, N.; Berg, K.; Kimber, E.; Brandberg, G.;

Lindberg; Holmberg, E.; Tulinius, M.; Nordling, M. Multiplex ligation-

dependent probe ampliﬁcation improves diagnostics in spinal muscular

atrophy. Neuromuscular Disord. 2006, 16, 830–838.

(17) DiDonato, C. J.; Chen, X.; Noya, D.; Kroenberg, J. R.; Nadeau, J.;

Simard, L. R. Cloning, characterization, and copy number of the

murine survival motor neuron gene: homolog of the spinal muscular

atrophy-determining gene. Genome Res. 1997, 7, 339–351.

Acknowledgment. The authors gratefully acknowledge the

scientiﬁc contributions of Denise Anderson for running mass

spectral analyses, Dr. Jun Zhang and Emmanuel Onua for

ADME support, Dr. Nelson Zhao and Amanda Heimann for

acquiring NMR spectra, Kevin Schilling and Brian Bock for

performing analytical HPLC analyses, and Jim Nie for carrying

out chiral HPLC analyses. We also thank Dr. Glenn E. Morris

for generously providing us with the MANSMA2 (8F7) mouse

anti-SMN antibody, and we thank the Campus Microscopy and

Imaging Facility (The Ohio State University) for providing

access to the confocal microscope. The authors also thank Drs.

Jill Jarecki, Christopher Spancake, David Zembower, Brent

Stockwell, and Brian Pollok for helpful discussions and Audrey

Lewis for her enthusiastic encouragement of this work. The

project was supported in its entirety by funding from the

Families of SMA.

(18) Viollet, L.; Bertrandy, S.; Brunialti, A. L.; Lefebvre, S.; Burlet, P.;

Clermont, O.; Cruaud, C.; Guenet, J. L.; Munnuch, A.; Melki, J. cDNA

isolation, expression, and chromosomal localization of the mouse

survival motor neuron gene (Smn). Genomics 1997, 40, 185–188.

(19) Schrank, B.; Gotz, R.; Gunnerson, J. M.; Ure, J. M.; Toyka, K.; Smith,

A.; Sendtner, M. Inactivation of the survival motor neuron gene,a

candidate gene for human spinal muscular atrophy, leads to massive

cell death in early mouse embryos. Proc. Natl. Acad. Sci. U.S.A. 1997,

94, 9920–9925.

(20) Vitte, J. M.; Davoult, B.; Natacha Roblot, N.; Mayer, M.; Joshi, V.;

Courageot, S.; Tronche, F.; Vadrot, J.; Moreau, M. H.; Kemeny, F.;

Melki, J. Deletion of murine Smn exon 7 directed to liver leads to

severe defect of liver development associated with iron overload. Am. J.

Pathol. 2004, 165, 1731–1741.

Supporting Information Available: Synthetic procedures for

compounds 1g-i, 5d,j-u, 6c-f,h,j-l, 7m-r, 10e-h,j-k, 11g-m,

12c-m, 13c-k; table listing HPLC retention times and purity data;

assay details for SMN2 promoter assay; assay description for

metabolic stability. This material is available free of charge via

the Internet at http://pubs.acs.org.

References

(21) Cifuentes-Diaz, C.; Frugier, T.; Tiziano, F. D.; Lacene, E.; Roblot,

N.; Joshi, V.; Moreau, M. H.; Melki, J. Deletion of murine SMN exon

7 directed to skeletal muscle leads to severe muscular dystrophy. J. Cell

Biol. 2001, 152, 1107–1114.

(22) Frugier, T.; Tiziano, F. D.; Cifuentes-Diaz, C.; Miniou, P.; Roblot,

N.; Dierich, A.; Le Meur, M.; Melki, J. Nuclear targeting defect of

SMN lacking the C-terminus in a mouse model of spinal muscular

atrophy. Hum. Mol. Genet. 2000, 9, 849–858.

(23) Hsieh-Li, H. M.; Chang, J. G.; Jong, Y. J.; Wu, M. H.; Wang, N. M.;

Tsai, C. H.; Li, H. A mouse model for spinal muscular atrophy. Nat.

Genet. 2000, 24, 66–70.

(24) Monani, U. R.; Sendtner, M.; Coovert, D. D.; Parsons, D. W.;

Andreassi, C.; Le, T. T.; Jablonka, S.; Schrank, B.; Rossol, W.; Prior,

T. W.; Morris, G. E.; Burghes, A. H. M. The human centromeric

survival motor neuron gene (SMN2) rescues embryonic lethality in

Smn^-/-mice and results in a mouse with spinal muscular atrophy.

Hum. Mol. Genet. 2000, 9, 333–339.

(25) Butchbach, M. E. R.; Burghes, A. H. M. Perspectives on models of

spinal muscular atrophy for drug discovery. Drug DiscoVery Today:

Dis. Models 2004, 1, 151–156.

(26) Le, T. T.; Pham, L. T.; Butchbach, M. E. R.; Zhang, H. L.; Monani,

U. R.; Coovert, D. D.; Gavrilina, T. O.; Xing, L.; Bassell, G. J.;

Burghes, A. H. M. SMN∆7, the major product of the centromeric

survival motor neuron (SMN2) gene, extends survival in mice with

spinal muscular atrophy and associates with full-length SMN. Hum.

Mol. Genet. 2005, 14, 845–857.

(1) Pears, J. Incidence, prevalence, and gene frequency studies of chronic

childhood spinal muscular atrophy. J. Med. Genet. 1978, 15, 409–

413.

(2) Roberts, D. F.; Chavez, J.; Court, S. D. The genetic component in

child mortality. Arch. Dis. Child. 1970, 45, 33–38.

(3) McAndrew, P. E.; Parsons, D. W.; Simard, L. R.; Rochette, C.; Ray,

P. N.; Mendell, J. R.; Prior, T. W.; Burghes, A. H. Identiﬁcation of

proximal spinal muscular atrophy carriers and patients by analysis of

SMNT and SMNC gene copy number. Am. J. Hum. Genet. 1997, 60,

1411–1422.

(4) Lefebvre, S.; Burglen, L.; Reboullet, S.; Clermont, O.; Burlet, P.;

Viollet, L.; Benichou, B.; Cruaud, C.; Millasseau, P.; Zeviani, M.; et

al. Identiﬁcation and characterization of a spinal muscular atrophy-

determining gene. Cell 1995, 80, 155–165.

(5) Lorson, C. L.; Hahnen, E.; Androphy, E. J.; Wirth, B. A single

nucleotide in the SMN gene regulates splicing and is responsible for

spinal muscular atrophy. Proc. Natl. Acad. Sci. U.S.A. 1999, 96, 6307–

6311.

(6) Monani, U. R.; Lorson, C. L.; Parsons, D. W.; Prior, T. W.; Androphy,

E. J.; Burghes, A. H.; McPherson, J. D. A single nucleotide difference

that alters splicing patterns distinguishes the SMA gene SMN1 from

the copy gene SMN2. Hum. Mol. Genet. 1999, 8, 1177–1183.

(7) Cartegni, L.; Krainer, A. R. Disruption of an SF2/ASF-dependent

exonic splicing enhancer in SMN2 causes spinal muscular atrophy in

the absence of SMN1. Nat. Genet. 2002, 30, 377–384.

(8) Gennarelli, M.; Lucarelli, M.; Capon, F.; Pizzuti, A.; Merlini, L.;

Angelini, C.; Novelli, G.; Dallapiccola, B. Survival motor neuron gene

transcript analysis in muscle from spinal muscular atrophy patients.

Biochem. Biophys. Res. Commun. 1995, 213, 342–348.

(9) Lorson, C. L.; Strasswimmer, J.; Yao, J. M.; Baleja, J. D.; Hahnen,

E.; Wirth, B.; Le, T.; Burghes, A. H.; Androphy, E. J. SMN

oligomerization defect correlates with spinal muscular atrophy severity.

Nat. Genet. 1998, 19, 63–66.

(10) Lorson, C. L.; Androphy, E. J. An exonic enhancer is required for

inclusion of an essential exon in the SMA-determining gene SMN.

Hum. Mol. Genet. 2000, 9, 259–265.

(11) Coovert, D. D.; Le, T. T.; McAndrew, P. E.; Strasswimmer, J.;

Crawford, T. O.; Mendell, J. R.; Coulson, S. E.; Androphy, E. J.; Prior,

T. W.; Burghes, A. H. The survival motor neuron protein in spinal

muscular atrophy. Hum. Mol. Genet. 1997, 6, 1205–1214.

(27) Monani, U. R.; Pastore, M. T.; Gavrilina, T. O.; Jablonka, S.; Le,

T. T.; Andreassi, C.; DiCocco, J. M.; Lorson, C.; Androphy, E. J.;

Sendtner, M. A transgene carrying an A2G missense mutation in the

SMN gene modulates phenotypic severity in mice with severe (type

I) spinal muscular atrophy. J. Cell Biol. 2003, 160, 41–52.

(28) Andreassi, C.; Angelozzi, C.; Tiziano, F. D.; Vitali, T.; De Vincenzi,

E.; Boninsegna, A.; Villanova, M.; Bertini, E.; Pini, A.; Neri, G.;

Brahe, C. Phenylbutyrate increases SMN expression in vitro: relevance

for treatment of spinal muscular atrophy. Eur. J. Hum. Genet,. 2004,

12, 59–65.

(29) Andreassi, C.; Jarecki, J.; Zhou, J.; Coovert, D. D.; Monani, U. R.;

Chen, X.; Whitney, M.; Pollok, B.; Zhang, M.; Androphy, E.; Burghes,

A. H. M. Aclarubicin treatment restores SMN levels to cells derived

from type I spinal muscular atrophy patients. Hum. Mol. Genet. 2001,

10, 2841–2849.

(12) Lefebvre, S.; Burlet, P.; Liu, Q.; Bertrandy, S.; Clermont, O.; Munnich,

A.; Dreyfuss, G.; Melki, J. Correlation between severity and SMN

protein level in spinal muscular atrophy. Nat. Genet. 1997, 16, 265–

269.

(30) Brichta, L.; Hofmann, Y.; Hahnen, E.; Siebzehnrubl, F. A.; Raschke,

H.; Blumcke, I.; Eyupoglu, I. Y.; Wirth, B. Valproic acid increases

the SMN2 protein level: a well-known drug as a potential therapy for

spinal muscular atrophy. Hum. Mol. Genet. 2003, 12, 2481–2489.

Diaminoquinazoline DeriVatiVes as SMN2 Promoter ActiVator

Journal of Medicinal Chemistry, 2008, Vol. 51, No. 3 469

(31) Chang, J. G.; Hseih-Li, H. M.; Jong, Y. J.; Wang, N. M.; Tsai, C. H.;

Li, H. Treatment of spinal muscular atrophy by sodium butyrate. Proc.

Natl. Acad. Sci. U.S.A. 2001, 98, 9808–9813.

(32) Sumner, C. J.; Huynh, T. N.; Markowitz, J. A.; Perhac, J. S.; Hill, B.;

Coovert, D. D.; Schussler, K.; Chen, X.; Jarecki, J.; Burghes, A. H. M.;

Taylor, J. P.; Fischbeck, K. H. Valproic acid increases SMN levels in

spinal muscular atrophy patient cells. Ann. Neurol. 2003, 54, 647–

654.

(33) Jarecki, J.; Chen, X.; Bernardino, A.; Coovert, D. D.; Whitney, M.;

Burghes, A.; Stack, J.; Pollok, B. A. Diverse small-molecule modula-

tors of SMN expression found by high-throughput compound screen-

ing: early leads towards a therapeutic for spinal muscular atrophy.

Hum. Mol. Genet. 2005, 14, 2003–2018.

(34) Harris, N. V.; Smith, C.; Bowden, K. Antifolate and antibacterial

activities of 5-substituted 2,4-diaminoquinazolines. J. Med. Chem.

1990, 33, 434–444.

(35) Hynes, J. H.; Tomazic, A.; Parrish, C. A. Further studies on the

synthesis of quinazoline from 2-ﬂuorobenzonitriles. J. Heterocycl.

Chem. 1991, 28, 1357–1363.

(36) Nettles, S. M.; Matos, K.; Burkhardt, E. R.; Rouda, D. R.; Corella,

J. A. Role of NaBH₄stabilizer in the oxazaborolidine-catalyzed

asymmetric reduction of ketones with BH₃-THF. J. Org. Chem. 2002,

67, 2970–2976.

(37) A similar trend was also observed with couple of other sulphides and

sulphones. In addition, in vitro metabolic stability studies using rat

and human liver mircosomal preparation revealed that the thioethers

disappeared rapidly, most likely because of oxidation at sulfur to

provide sulfoxide and/or sulphone.

(38) For conformational analysis the (S)-enantiomer of analogues 7a, 9a

and 9b were used using Systematic Search sub-routine of SYBYL.

(39) SYBYL Molecular Modeling Systems, versions 6.9 and 7.0; Tripos

Associates: St. Louis, MO.

pyrimidine nucleotide synthesis and metabolism. Arabidopsis Book

2002, 39 (1), 1–20.

(42) Cody, V.; Luft, J. R.; Walt, P.; Gangjee, A.; Queener, S. F. Structure

determination of tetrahydroquinazoline antifolates in complex with

humane and Pneumocystis carinii dihydrofolate reductase: correlations

between enzyme selectivity and stereochemistry. Acta Crystallogr.

2004, D60, 646–655.

(43) Analogues representing this approach will be reported in a future

communication elsewhere.

(44) It is worth pointing out that during the docking studies for some

selected analogues corresponding to the reverse benzyl ether series

(10) using the DHFR active site (using coordinates of 1s3u.pdb), an

intramolecular hydrogen bond was observed between the benzyl ether

oxygen and HN of the 4-amino group instead of the typical H-bond

of N4-H with Val115 carbonyl. However, this in silico observation

has not been conﬁrmed with experimental data.

(45) Ashton, W. T.; Hynes, J. B. Synthesis of 5-substituted quinazolines

as potential antimalarial agents. J. Med Chem. 1973, 16, 1233.

(46) (a) Zlokarnik, G. Fusions to beta-lactamase as a reporter for gene

expression in live mammalian cells. Methods Enzymol. 2000, 326, 221–

244. (b) Zlokarnik, G.; Negulescu, P. A.; Knapp, T. E.; Mere, L.;

Burres, N.; Feng, L.; Whitney, M.; Roemer, K.; Tsien, R. Y.

Quantitation of transcription and clonal selection of single living cells

with ꢀ-lactamase as reporter. Science 1998, 279, 84–88.

(47) Appleman, J. R.; Howell, E. E.; Kraut, J.; Kuhl, M.; Blakley, R. L.

J. Biol. Chem. 1988, 263, 9187–9198.

(48) Coovert, D. D.; Le, T. T.; McAndrew, P. E.; Strasswimmer, J.;

Crawford, T. O.; Mendell, J. R.; Coulson, S. E.; Androphy, E. J.; Prior,

T. W.; Burghes, A. H. M. The survival motor neuron protein in spinal

muscular atrophy. Hum. Mol. Genet. 1997, 6, 1205–1214.

(49) Young, P. J.; Le, T. T.; Thi Man, N.; Burghes, A. H. M.; Morris,

G. E. The relationship between SMN, the spinal muscular atrophy

protein, and nuclear coiled bodies in differentiated tissues and cultured

cells. Exp. Cell Res. 2000, 25, 365–374.

(40) (a) Cockerell, G. L; McKim, J. M.; Vonderfecht, S. L. Strategic

importance of research support through pathology. Toxicol. Pathol.

2002, 30 (1), 4–7. (b) McKim, J. M.; Wilga, P. C.; Pregenzer, J. F.;

Petrella, D. K. A biochmeical approach to in-vitro toxicity testing.

Pharm. DiscoVery 2005, 1, 30–36.

(41) (a) McGuir, J. J. Anticancer antifolates: current status and future

directions current pharmaceutical design. Curr. Pharm. Des. 2003, 9

(31), 2593–2613. (b) Moffatt, B. A.; Ashihara, H. Purine and

(50) Butchbach, M. E. R.; Guo, H.; Lin, C. G. Methyl-ꢀ-cyclodextrin but

not retinoic acid reduces EAAT3-mediated glutamate uptake and

increases GTRAP3-18 expression. J. Neurochem. 2003, 84, 891–894.

JM061475P

Article Doi

Synthesis and biological evaluation of novel 2,4-diaminoquinazoline derivatives as SMN2 promoter activators for the potential treatment of spinal muscular atrophy

DOI: 10.1021/jm061475p

Source and publish data:

Authors:

Article abstract of DOI:10.1021/jm061475p

Full text of DOI:10.1021/jm061475p

Products guided by the article

R&D Labs maybe for 872180-68-2

Relevant to this article

Hot Product