Discovery and Optimization of Small Molecule Splicing Modifiers of Survival Motor Neuron 2 as a Treatment for Spinal Muscular Atrophy

Article abstract of DOI:10.1021/acs.jmedchem.6b00460

The underlying cause of spinal muscular atrophy (SMA) is a deficiency of the survival motor neuron (SMN) protein. Starting from hits identified in a high-throughput screening campaign and through structure-activity relationship investigations, we have developed small molecules that potently shift the alternative splicing of the SMN2 exon 7, resulting in increased production of the full-length SMN mRNA and protein. Three novel chemical series, represented by compounds 9, 14, and 20, have been optimized to increase the level of SMN protein by >50% in SMA patient-derived fibroblasts at concentrations of <160 nM. Daily administration of these compounds to severe SMA Δ7 mice results in an increased production of SMN protein in disease-relevant tissues and a significant increase in median survival time in a dose-dependent manner. Our work supports the development of an orally administered small molecule for the treatment of patients with SMA.

Full text of DOI:10.1021/acs.jmedchem.6b00460

Subscriber access provided by - Access paid by the | UCSB Libraries
Article
Discovery and optimization of small molecule splicing modifiers of
survival motor neuron 2 (SMN2) as a treatment for spinal muscular atrophy
Matthew G Woll, Hongyan Qi, Anthony Turpoff, Nanjing Zhang, Xiaoyan Zhang, Guangming Chen,
Chunshi Li, Song Huang, Tianle Yang, Young-Choon Moon, Chang-Sun Lee, Soongyu Choi,
Neil G. Almstead, Nikolai N. Naryshkin, Amal Dakka, Jana Narasimhan, Vijayalakshmi Gabbeta,
Ellen Welch, Xin Zhao, Nicole Risher, Josephine Sheedy, Marla Weetall, and Gary M. Karp
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b00460 • Publication Date (Web): 14 Jun 2016
Downloaded from http://pubs.acs.org on June 15, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Discovery and optimization of small molecule
splicing modifiers of survival motor neuron 2
(SMN2) as a treatment for spinal muscular atrophy
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Matthew G. Woll, Hongyan Qi, Anthony Turpoff, Nanjing Zhang, Xiaoyan Zhang, Guangming
Chen, Chunshi Li, Song Huang, Tianle Yang, YoungꢀChoon Moon, ChangꢀSun Lee, Soongyu
Choi, Neil G. Almstead, Nikolai A. Naryshkin, Amal Dakka, Jana Narasimhan, Vijayalakshmi
Gabbeta, Ellen Welch, Xin Zhao, Nicole Risher, Josephine Sheedy, Marla Weetall, Gary M.
Karp*
PTC Therapeutics, Inc., 100 Corporate Court, South Plainfield, NJ 07080, United States
ABSTRACT
The underlying cause of spinal muscular atrophy (SMA) is a deficiency of the survival motor
neuron (SMN) protein. Starting from hits identified in a high throughput screening campaign
and through structure–activity relationship (SAR) investigations, we have developed small
molecules that potently shift the alternative splicing of the SMN2 exon 7, resulting in increased
production of the fullꢀlength SMN mRNA and protein. Three novel chemical series, represented
by compounds 9, 14, and 20, have been optimized to increase the level of SMN protein by >50%
in SMA patient−derived fibroblasts at concentrations < 160 nM. Daily administration of these
compounds to severe SMA ∆7 mice results in an increased production of SMN protein in disease
relevant tissues and a significant increase in median survival time in a dose dependent manner.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Our work supports the development of an orally administered small molecule for the treatment of
patients with SMA.
INTRODUCTION
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Spinal muscular atrophy (SMA) is caused by a genetic defect in the SMN1 gene, rendering it
1
,2
incapable of producing the survival motor neuron protein (SMN). Humans usually have two
or more copies of the paralogous SMN2 gene, but this gene produces much less functional SMN
protein due to an alternative splicing event of exon 7. This alternative pathway is the result of a
translationally synonymous C to T mutation in exon 7. As a result, the majority of transcripts
produced from SMN2 lack exon 7 (SMN2ꢀꢁ7), and encode a truncated SMNꢀꢁ7 protein that is
3
,4
rapidly degraded.
Antisense oligonucleotides (ASOs) that target the ISSꢀN1 regulatory element near exon 7 of
SMN2 preꢀmRNA can shift the outcome of the splicing reaction toward the generation of full
5
ꢀ7
length SMN2 mRNA. Initial reports with 2’ꢀOꢀmethoxyꢀethyl ribose (MOE) phosphorothioate
ASOs that were delivered by intracerebroventricular (ICV) injection to neonatal mice with a
severe SMA phenotype at postnatal day 1 (PND1) produced an increase in SMN protein levels in
8
neuronal tissue, but only modestly increased the median survival time of the mice. In order to
achieve maximum survival benefit, coꢀadministration of the ASO via subcutaneous (SC)
9
injection was required. Phosphorodiamidate morpholino oligonucleotides (PMOs) targeting ISSꢀ
1
0ꢀ
N1 demonstrated a much greater survival benefit than MOE ASOs after a single ICV injection.
12
Additionally, PMOs that target the E1 intronic repressor near exon 7 have shown survival
1
3
benefit after a single ICV dose.
ACS Paragon Plus Environment

Page 3 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Despite the promising results of the ASOs, there remained a need to discover and develop
compounds with improved activity and pharmaceutical properties. Small molecules possess the
inherent advantage of oral administration and broad peripheral and CNS distribution. Several
classes of small molecules (e.g., hydroxyurea, salbutamol, tetracycline derivatives) have been
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
4ꢀ20
reported to modify the alternative splicing of SMN2 exon 7.
However, the biological activity
of all these compounds is weak compared to that of ASOs. We recently disclosed the structures
of three small molecules that shift SMN2 splicing toward the production of fullꢀlength SMN2
2
1
mRNA with high selectivity. Administration of these compounds to SMA mice led to an
increase in SMN protein levels, improvement of motor function, protection of the neuromuscular
circuit, and an extended life span. Recently, another series of small molecules demonstrated a
22
long term survival benefit in SMA mice. The compounds were shown to enhance the
recruitment of the U1 snRNP to the (nonꢀcanonical) 5’ splice site in SMN2 intron 7. Our
compounds similarly enhance the U1 – preꢀmRNA interaction at the 5’ splice site of SMN2
intron 7.
Herein we describe the discovery and chemical optimization of the three chemical series
2
1
disclosed in our previous publication. Through chemical modification of hits identified in a
high throughput screen of our chemical library, we identified molecules that demonstrate both an
increase in the full length SMN2 transcript (SMNꢀFL) and a concomitant decrease in SMN2ꢀ∆7.
A detailed structure‒activity relationship (SAR) investigation improved the potency of the lead
molecules nearly three orders of magnitude (Figure 1). Increases in the level of fullꢀlength SMN2
mRNA resulted in significantly increased SMN protein levels in SMA patient−derived cells and
in mice with a severe SMA phenotype (∆7 mice). Moreover, as a consequence of the increased
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
level of SMN protein, administration of these compounds significantly prolongs survival in ∆7
mice.
Figure 1. Three novel chemical series discovered in lead optimization
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
RESULTS AND DISCUSSION
Chemistry. In order to synthesize various 3,7ꢀdisubstituted coumarins, an appropriately
substituted ethyl 2ꢀaryl/heteroarylꢀacetate (1a for 3ꢀbenzoxazolyl coumarins) can undergo aldol
condensation with an appropriately substituted salicylaldehyde (2 for 7ꢀpiperazinyl coumarins).
The transient condensation product undergoes ringꢀclosing transesterification to give coumarin 3,
as shown in Scheme 1. Removal of the Boc protecting group with hydrogen chloride in 1,4ꢀ
dioxane provides 4.
a
Scheme 1. The synthesis of 3,7ꢀdisubstituted coumarin 4
ACS Paragon Plus Environment

Page 5 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
a
Conditions: (a) 2 (1.05 equiv), piperidine (0.25 equiv), AcOH (0.1 equiv), EtOH, reflux, 24 h,
9
3%; (b) 4 N HCl in 1,4ꢀdioxane, rt, 3 h, 85%.
An alternative strategy to generate 3,7ꢀdisubstituted coumarins, when it is desired to fix the 3ꢀ
position as a specific imidazoꢀheterocycle, can be found in Scheme 2. Aldol condensation of
ethyl acetoacetate with 4ꢀfluorosalicaldehyde, followed by ringꢀclosing transesterification yields
coumarin 5. Treatment of 5 with bromine provides the αꢀbromoketone 6. When 6 and 2ꢀaminoꢀ
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
,5ꢀdimethylpyrazine are combined at elevated temperatures, an alkylationꢀcondensation reaction
occurs to generate 7. The nucleophilic displacement of fluorine at the 7ꢀposition of 7 by (S)ꢀ2ꢀ
methylpiperazine provides 8. A diverse set of amines can be introduced at this stage of the
synthesis. Additional diversification of 8 was achieved through reductive amination or alkylation
of the secondary amine. The Nꢀethyl variant 9 was obtained after treatment of 8 with
acetaldehyde and sodium triacetoxyborohydride.
a
Scheme 2. The synthesis of 3,7ꢀdisubstituted coumarin 9
a
Conditions: (a) ethyl acetoacetate (1 equiv), piperidine (0.1 equiv), rt, 10 min, 95%; (b) Br2
(1.05 equiv), CHCl , rt, 1 h, 72%; (c) 2ꢀaminoꢀ3,5ꢀdimethylpyrazine (0.85 equiv), CH CN, 120
3
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
°
C, 20 min, 90%; (d) (S)ꢀ2ꢀmethylpiperazine (2.1 equiv), K CO (4 equiv), DMSO, 120 °C, 2 h,
2
3
6
4%; (e) CH CHO (2 equiv), NaBH(OAc) (3 equiv), CH Cl :MeOH (9:1), rt, 16 h, 72%.
3 3 2 2
Using a similar synthetic strategy as found in Scheme 2, isocoumarins could be generated from
intermediate 12 (Scheme 3). The generation of the isocoumarin core, however, required a distinct
synthetic pathway from the coumarins. The lactone architecture was generated via a Sonogashira
coupling of 3ꢀbutynꢀ2ꢀol to 2ꢀiodoꢀ5ꢀfluoroꢀbenzoic acid. Under the reaction conditions, the
alkyne intermediate undergoes nucleophilic attack by the orthoꢀcarboxylate, giving isocoumarin
10. Oxidation of 10 with manganese (IV) oxide gives methyl ketone 11, which can be
brominated alpha to the ketone to give 12. Intermediate 12 was converted to 14 in a similar
fashion as found in Scheme 2. Interestingly, the treatment of 13 with 1ꢀmethylꢀpiperazine leads
initially to lactone ring opening. Next, the fluorine on the ringꢀopened intermediate is displaced
by an additional 1ꢀmethylꢀpiperazine molecule. Finally, the isocoumarin lactone ring reforms
and subsequently remains intact, due to the increased stability provided by the electron donating
amine.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Scheme 3. The synthesis of 3,7ꢀdisubstituted isocoumarin 14
a
Conditions: (a) 3ꢀbutynꢀ2ꢀol (2.3 equiv), ZnCl (1 equiv), Pd(PPh ) (0.05 equiv), Et N (3
2
3 4
3
equiv), DMF, 100 °C, 2 h, 84% ; (b) MnO (10 equiv), CH Cl , rt, 48 h, 66%; (c) Br (1.1 equiv),
2
2
2
2
ACS Paragon Plus Environment

Page 7 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
CHCl , rt, 1 h, 96%; (d) 2ꢀaminoꢀ3,5ꢀdimethylpyrazine (1.1 equiv), CH CN, 100 °C, 16 h, 93%;
3
3
(e) 1ꢀmethylꢀpiperazine (3 equiv), NMP, 180 °C, 24 h, 57%.
Access to 2,7ꢀdisubstituted pyridopyrimidinones required a different synthetic approach than the
previously described routes for coumarins and isocoumarins. Diversity is achieved by
condensing ethyl βꢀketoaryl esters (e.g. 18) with 2ꢀaminoꢀ5ꢀfluoropyridine to generate 19 as
depicted in Scheme 4. The fluorine at the 7ꢀposition in 19 can undergo nucleophilic displacement
with amines as described in previous Schemes to give 20. Access to intermediate 18 begins with
the tandem alkylationꢀcondensation reaction of diethyl 1Hꢀpyrazoleꢀ3,5ꢀdicarboxylate with
chloroacetone in the presence of ammonium acetate to provide 15. The 4ꢀhydroxyl group of 15
can be converted to chlorine in high yield with phosphorus oxychloride. A Suzuki coupling using
methaneboronic acid converts the chlorine to methyl (17). Claisen condensation with tertꢀbutyl
acetate provides a tertꢀbutylꢀβꢀketoꢀintermediate that is transformed to 18 via transesterification
in ethanol.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
Scheme 4. The synthesis of 2,7ꢀdisubstituted pyridopyrimidinone 20
a
Conditions: (a) chloroacetone (1 equiv), K CO (1.1 equiv), acetone, 30 °C, 6.5 h; (b) NH OAc
2
3
4
(
20 equiv), AcOH, reflux, 60 h, 64% over 2 steps; (c) POCl , reflux, 16 h, 85%; (d) MeB(OH)2
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
(
3 equiv), Pd(PPh ) (0.03 equiv), K CO (5 equiv), DMF, 100 °C, 15 h, 82%; (e) i) tertꢀ
3
4
2
3
butylacetate (2 equiv), LDA (2.4 equiv), −78 °C to −30 °C, 97%, ii) EtOH, 120 °C, 1 h, 98%; (f)
2
ꢀaminoꢀ5ꢀfluoropyridine (1.2 equiv), pyridinium pꢀtoluenesulfonate (0.05 equiv), 130 °C, 8 h,
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
1%; (g) 1ꢀmethylꢀpiperazine (10 equiv), DMAc, 120 °C, 15 h, 80%.
HTS and lead discovery. In order to identify compounds that increase the inclusion of exon 7
during SMN2 preꢀmRNA splicing, a HEK293H cell line was generated that contained an SMN2
gene fragment (from exon 6 to the 5’ region of exon 8), followed by the firefly luciferase coding
sequence. Luciferase is expressed only when splicing includes SMN2 exon 7 in the mRNA,
otherwise the luciferase coding sequence is out of the translational reading frame with respect to
the initiation codon upstream of the SMN2 sequence. The ratio of the SMN2ꢀꢁ7 transcript to the
SMN2ꢀFL transcript in this minigene construct is ~97:3. Full correction of splicing would result
in a ~30ꢀfold increase in the SMN2ꢀFL transcript. In the type 1 SMA patient−derived fibroblasts
(GM03813) the ratio of the SMN2ꢀꢁ7 transcript to the SMN2ꢀFL transcript is ~1:1. We reasoned
that screening compounds using the minigene construct could be advantageous as the larger
dynamic range in the minigene assay would allow for the identification of compounds with weak
activity. Using the minigene assay to screen our library of ~200,000 compounds, we identified
molecules that increased the luciferase signal more than 3ꢀfold above the DMSO control level.
All compounds identified in the primary screen were confirmed in the minigene assay for
concentrationꢀresponse activity by measuring the relative amount of SMN2ꢀFL mRNA using the
reverse transcriptionꢀquantitative polymerase chain reaction (RTꢀqPCR). Compounds were
compared by measuring the concentration required to increase the amount of SMN2ꢀFL to a level
1
.5 fold that of the baseline (EC1.5X RNA). Several hits were identified for further evaluation.
ACS Paragon Plus Environment

Page 9 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
For each hit we designed dozens of structurally similar analogs in an effort to establish a
structure−activity relationship (SAR) and identify a lead series to optimize for development. A
coumarin hit (21) showed promising activity (EC1.5X RNA = 0.22 µM). Several analogs of 21 were
synthesized by modifying the 7ꢀposition of the coumarin, exchanging the diethylamino
substituent with structurally similar moieties. Upon initial inspection it appeared problematic to
make even minor structural modifications at the 7ꢀposition without negatively impacting activity
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(Table 1, compounds 4, 22ꢀ24). We observed that analog 4, containing a piperazine at the 7ꢀ
position, induced a maximum amount of SMN2ꢀFL (max fold RNA) that was much greater than
that induced by the other analogs, even though the compound had an EC1.5X RNA considerably
lower than that of 21.
Further insight was achieved when we evaluated the EC1.5X RNA without normalization to
GAPDH mRNA (a standard reference mRNA). Although GAPDH normalization is used to
remove some of the artifacts of PCR amplification, it can produce an artificially high increase in
the mRNA of interest in cases when GAPDH mRNA decreases at a relative rate greater than that
of the mRNA of interest. (This phenomenon may occur at concentrations at which the compound
exhibits cytotoxicity). Analysis of the SMN2ꢀFL data without GAPDH correction (Table 1, Not
GAPDH normalized) showed that compound 21 was weakly active, whereas 4 retained
substantial activity. Based on these data, we reasoned that the basic amine found in 4 was a
critical component of the pharmacophore and focused our efforts on synthesizing additional 7ꢀ
piperazino coumarin analogs.
Table 1. In vitro activity of 7ꢀsubstituted coumarins in the SMN2 minigene assay
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
GAPDH normalized
Not GAPDH normalized
Compd
R
EC1.5x RNA
EC1.5x RNA
(ꢀM)
Max fold
4.5
Max fold
1.2
(
ꢀM)
21
22
0.22
22
>32
>32
>32
1.5
1.0
1.2
1.0
2
3
4
>32
2
>32
8.7
1.0
25
>32
10
1.1
7
4
Lead optimization. Although compound 4 increased SMN2ꢀFL levels in the minigene assay, its
potency was modest (EC1.5X RNA ~ 10 µM). Additionally, compound 4 did not induce a
measurable increase in the level of SMN protein in SMA patient−derived fibroblasts, likely due
23
to its weak activity. We pursued potency improvements by modifying the three heterocyclic
constituents of the molecule. For the purpose of clarity we adopted the nomenclature shown in
Figure 2. The coumarin moiety is referred to as the “core”, which is flanked on the “leftꢀside” by
the basic piperazine moiety and on the “rightꢀside” by the benzoxazole.
Figure 2. Nomenclature for the three heterocyclic constituents of 4
ACS Paragon Plus Environment

Page 11 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
2
4
Initial gains in potency were achieved by modifying the rightꢀside heterocycle. Replacing the
benzoxazole ring with phenyl (25) resulted in loss of activity, but phenyl rings substituted with
electron donating groups (26 and 27) resulted in increased potency in the minigene assay (Table
2
). Compound 27 showed a modest but clearly measurable 1.3ꢀfold increase in the level of SMN
protein; however, no additional improvement could be achieved with other substituted phenyl
groups (data not shown). We then focused our attention toward replacing the benzoxazole with
novel heterocycles. Improvement of potency in the minigene assay was achieved by replacing
the benzoxazole with benzothiazole (28). Further improvement was achieved with benzimidazole
2
9 and constitutionally isomeric imidazopyridine 30. Replacing the pyridine ring in
imidazopyridine 30 with a thiazole ring gave the more potent imidazothiazole 31, with an EC1.5X
of 0.11 µM. Compound 31 was the first compound identified to achieve a >1.5 fold increase
RNA
in SMN protein as measured in SMA patient−derived fibroblasts (EC1.5X PRO = 0.36 ꢂM).
Table 2. In vitro activity of 3ꢀsubstituted coumarins
RNA minigene
SMN protein
EC1.5X PRO
Compd
R
EC1.5X RNA
(ꢀM)
Max fold
(ꢀM)
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
8.7
>32
2.0
>10
>10
>10
>10
>10
1.0
1.0
1.1
1.3
1.0
2
5
6
7
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
2
0.62
1.0
28
29
0.43
0.38
0.11
>10
>10
0.36
1.2
1.3
1.5
30
31
We next extended our SAR investigation by evaluating the effect of substitution on the
imidazopyridine ring in compound 30. The EC1.5X RNA of 380 nM for 30 was improved ~3ꢀfold
by appending a methyl group to the 6ꢀposition of the imidazopyridine in 32 (Table 3).
Substitution with the larger 6ꢀethyl group (33), however, was detrimental to potency. A modest
5
ꢀfold gain in potency could also be achieved by replacing C7 of the heterocycle with a nitrogen
atom to afford the imidazopyrazine 34. Combining the 6ꢀmethyl substituent with the Cꢀ7
nitrogen of the imidazopyrazine (35) resulted in a dramatic >100ꢀfold improvement in potency
(
EC1.5X RNA = 2 nM), demonstrating an effect often encountered in drug discovery whereby two
small structural modifications, each alone producing a minor improvement in activity, but
together imparting a profound effect on activity. This cooperative enhancement was observed
when making modifications to the imidazopyridine series (Figure 3). The dramatic improvement
ACS Paragon Plus Environment

Page 13 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
in potency observed in the minigene assay in 35 (EC1.5X RNA = 2 nM) was also observed in SMA
patient−derived fibroblasts (EC1.5X PRO = 5 nM). Further investigation revealed that the 8ꢀposition
of the imidazopyrazine could tolerate methyl substitution (36), but larger groups such as ethyl
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(
37) led to diminished potency. The 8ꢀMe modification was incorporated into the structure of
many subsequent molecules, despite the 3ꢀfold loss in potency, due to a beneficial gain in plasma
and tissue exposure in mice and rats (further described in the pharmacokinetics section).
Table 3. In vitro activity of 3ꢀsubstituted coumarins (imidazoꢀheterocycles)
RNA minigene
EC1.5X RNA (nM)
SMN protein
EC1.5X PRO (nM)
Compd
R
30
32
33
380
140
>10,000
160
2700
>10,000
3
4
70
2
120
5
35
3
6
7
2
15
3
10
210
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 3. Cooperative potency enhancement of the nitrogenꢀmethyl pair
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Additional heterocyclicꢀcontaining coumarin analogs having the Me/N pairing found in the most
active imidazopyrazines were synthesized (Table 4). Several of the compounds were very active
(38ꢀ40, EC1.5X PRO <100 nM) while others were nearly inactive (41ꢀ42, EC1.5X PRO >10,000 nM).
It became apparent that the relative orientation of the Me/N pharmacophore with respect to the
coumarin is important (Figure 4). For example, when the relative energy for rotation about the
coumarinꢀimidazopyrazine bond (e.g., 36) is calculated, the lowest energy conformation gives a
dihedral angle of 180° (see supplemental Graph 1). In this planar orientation the key nitrogen is
located above the key methyl as drawn in Figure 4. When the heterocycle is modified to
pyrazolopyrazine 41, the 180° dihedral angle is still highly favored, but the key nitrogen is
located below the key methyl when drawn in a similar planar fashion. Compounds containing
heterocycles that favor (or do not strongly disfavor) the “nitrogen above” orientation are highly
potent, while compounds that highly favor the “nitrogen below” orientation are nearly inactive.
Table 4. In vitro activity of 3ꢀsubstituted coumarins
ACS Paragon Plus Environment

Page 15 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
RNA minigene
EC1.5X RNA (nM)
SMN protein
EC1.5X PRO (nM)
Compd
R
3
6
8
2
4
15
16
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
39
4
5
29
66
40
41*
>10,000
>10,000
42
1500
>10,000
*Leftꢀside heterocycle is NꢀMeꢀpiperidine, instead of piperazine (see Table 5 for activity comparison).
Figure 4. Orientation of N/Me pair in select heterocycles
After identifying the optimized rightꢀside heterocycles, we focused our attention on the leftꢀside
heterocycle. Earlier investigation revealed the importance of a basic amine, but did not address
the optimization of amine location, pK or neighboring group effect. We decided to use the
a
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
highly potent unsubstituted piperazine 36 (EC1.5X PRO = 15 nM) as a starting point for
comparison. Methyl substitution at the 3ꢀposition of piperazine (e.g., 43 and 8) typically
provided a modest boost to potency regardless of the absolute configuration (Table 5). A
decrease in potency was observed with groups larger than methyl at the 3ꢀposition (e.g.,
isopropyl 44). Alkylation of the piperazine nitrogen (49ꢀ51) diminished potency with increasing
substituent size. Several additional modifications to the piperazine ring, where the basic amine
functionality was retained, were investigated. The ringꢀfused piperazine 52, the 4’ꢀ
dimethylaminopiperidine 53, and the Nꢀmethylpiperidine 54 had comparable potency in SMA
patient−derived fibroblasts (EC1.5X PRO = 50−66 nM), being only slightly less potent than the Nꢀ
methylpiperazine 49. Acyclic amines were generally less potent than cyclic amines (cf.
compounds 54 vs. 55). Some cyclic amine configurations were detrimental to activity, as
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
evidenced by comparing the pyrrolidinopyrrolidines 45 vs. 46. In general, pK of the basic amine
a
above 7 is required to maintain activity, as demonstrated by the diminished activity of
2
5
fluorinated analog 47 (pK = 6.3) compared to that of 48 (pK = 7.5). Although our
a
a
investigation of numerous heterocyclic amineꢀcontaining moieties revealed the difficulty in
identifying compounds with greater potency than that of those containing an unsubstituted
piperazine, e.g., 36, our understanding of the tolerated modifications was important in optimizing
the pharmacokinetic properties of the compounds.
Table 5. In vitro activity of 7ꢀsubstituted coumarins
ACS Paragon Plus Environment

Page 17 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
EC1.5X RNA
EC1.5X PRO
EC1.5X RNA
EC1.5X PRO
Compd
R
Compd
R
(
nM)
(nM)
(nM)
(nM)
3
6
3
2
15
13
49
18
10
31
62
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4
3
1
50
51
8
5
59
21
11
>10,000
66
4
4
5
23
1
76
13
52
53
4
61
46
28
640
54
55
11
58
50
4
7
8
37
19
450
60
220
4
In parallel with our efforts to optimize the rightꢀ and leftꢀsides of the coumarin lead, we began
exploring alternative cores. Our initial objectives were to identify cores with improved potency
26ꢀ28
and to seek cores lacking the potential liabilities inherent in coumarins.
Our earlier work
indicated that the rightꢀ and leftꢀside moieties had to extend from the 3ꢀ and 7ꢀpositions of the
coumarin, respectively. We attempted to maintain the same spatial orientation when replacing
the coumarin with other 6,6ꢀbicyclic heterocycles. To simplify the comparison, the rightꢀside
was fixed as 3,4ꢀdimethoxyphenyl, since this moiety showed promising activity in the coumarin
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
series (Table 2) and would be easily installed synthetically onto various cores. The leftꢀside was
fixed as piperazine.
We generated a variety of 6,6ꢀheterocycles that incorporated nitrogen atoms in the core and
contained a carbonyl group in the same position found in coumarin 27, but these analogs resulted
in decreased activity (Table 6, compounds 56ꢀ58). Improvement could be achieved, however, by
shifting the carbonyl group one position on the core as demonstrated by isocoumarin 59 (EC1.5X
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
=
80 nM). One aza variation of this theme, pyridopyrimidinone 60 (EC1.5X RNA = 120 nM),
RNA
also showed improved activity compared to that of 27 (EC1.5X RNA = 620 nM). Additional aza
variations were less promising (61ꢀ63), as was pyridopyrimidinone 64, in which the orientations
of the piperazine and dimethoxyphenyl moieties around the pyridopyrimidinone were altered.
Overall, the treatment of SMA patient−derived fibroblasts with isocoumarin 59 and
pyridopyrimidinone 60 resulted in increased production of SMN protein with improved potency
EC1.5X PRO = 290 and 830 nM, respectively) compared to that of coumarin 27 (EC1.5X PRO
( =
>10,000 nM).
Table 6. In vitro activity of analogs with modified cores
RNA minigene
EC1.5X RNA (nM)
SMN protein
Compd
EC1.5X PRO (nM)
Max fold
27
620
>10,000
1.3
ACS Paragon Plus Environment

Page 19 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
5
6
7
8800
>10,000
>10,000
1.0
1.0
5
>10,000
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
5
8
2200
80
>10,000
290
1.1
1.6
1.6
1.2
1.6
1.0
1.0
59
6
0
1
120
830
6
1400
920
>10,000
2500
62
6
3
4
>10,000
3200
>10,000
>10,000
6
Based on the activity of the isocoumarin 59 and the pyridopyrimidinone 60 we decided to
synthesize additional analogs in these series incorporating several of the rightꢀside heterocycles
that were investigated in the coumarin series (Table 7). The imidazopyrazine moiety that was
optimal in the coumarin series was also highly active in the isocoumarin series (65, EC1.5X PRO
39 nM). When the imidazopyrazine was incorporated into pyridopyrimidinone 66, greatly
diminished activity (EC1.5X PRO = 430 nM) was observed compared to that of coumarin 36
=
(
EC1.5X PRO = 15 nM). In contrast, pyridopyrimidinone 67, which contains the pyrazolopyrazine
rightꢀside moiety, was highly active (EC1.5X PRO = 31 nM) in comparison to the corresponding
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
coumarin 41 (EC1.5X PRO = >10,000 nM). This observation is consistent with the hypothesis
regarding conformation of the N/Me pair as presented in Figure 4. The lowest energy
conformation gives pyrazolopyrazine 67 the “nitrogen above” orientation and imidazopyrazine
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
66 the “nitrogen below” orientation, a trend opposite of the one seen in the coumarin series.
When the rightꢀside heterocycle is imidazopyridine (e.g., 40 and 68) there is no preference for
the “nitrogen above” or “nitrogen below” orientation. In this case the coumarin 40 (EC1.5X PRO
=
66 nM) and pyridopyrimidinone 68 (EC1.5X PRO = 60 nM) have a similar level of activity.
Table 7. In vitro activity of isocoumarins and pyridopyrimidinones
EC1.5X RNA
EC1.5X PRO
(nM)
Compd
R
(
nM)
36
2
15
41*
>10,000
5
>10,000
40
66
39
6
5
6
8
6
180
6
430
31
67
68
48
60
ACS Paragon Plus Environment

Page 21 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
*
Leftꢀside heterocycle is NꢀMeꢀpiperidine, instead of piperazine (see Table 5 for activity comparison).
The optimal pyridopyrimidinone 67 was intriguing to us because it contained both core and
rightꢀside pieces that were structurally distinct from the optimal coumarins. We synthesized
various analogs of 67 that contained alternative leftꢀside heterocycles to determine if the same
trends found in the coumarin series applied to the pyridopyrimidinone series (Table 8). Similar to
the coumarins, Nꢀalkylation of the piperazine reduced activity in the pyridopyrimidinones, with
larger alkyl groups causing greater loss of activity (e.g., 20 and 69). Also, similar to the
coumarins, piperidine 70, ringꢀfused piperazine 71, and 4’ꢀdimethylaminopiperidine 72 had
comparable activity (EC1.5X PRO = 130−210 nM) to Nꢀmethylpiperazine 20 (EC1.5X PRO = 170
nM). In contrast to the coumarin series, incorporating 3ꢀmethyl substitution (R or S) into the
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
piperazine ring of 67 resulted in reduced activity, with 3ꢀ(S)ꢀmethyl piperazine 74 (EC1.5X PRO
60 nM) showing a greater loss than 3ꢀ(R)ꢀmethyl piperazine 73 (EC1.5X PRO = 70 nM).
=
1
Table 8. In vitro activity of 7ꢀsubstituted pyridopyrimidinones
EC1.5X RNA
EC1.5X PRO
(nM)
EC1.5X RNA
(nM)
EC1.5X PRO
(nM)
Compd
R
Compd
R
(
nM)
67
6
31
71
120
20
150
130
20
35
170
72
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
69
380
53
690
210
73
74
11
19
71
70
160
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Pharmacokinetics. The goal of this program was to identify orally deliverable small molecule
therapeutics that correct alternative splicing of the SMN2 gene exon 7 to address the underlying
cause of SMA. To identify which compounds warranted further advancement, we evaluated
compounds with sufficient in vitro potency for their pharmacokinetic profile in rats. We
measured plasma drug levels of test compounds at several time points up to 6 h post dose and
total brain concentration at 6 h after oral administration of a single 10 mg/kg dose. This provided
the plasma AUC0ꢀ6h and the brain/plasma ratio (B/P) at 6 h, a measure of CNS penetrability.
Compounds 36, 65 and 67, from the coumarin, isocoumarin and pyridopyrimidinone series
respectively, showed excellent in vitro activity, but all had moderateꢀtoꢀpoor brain and/or plasma
exposure when administered orally to rats (Table 9). We determined that judicious placement of
small alkyl groups resulted in a minor reduction in activity but improved both plasma and brain
exposure. The observed improvements did not seem to be related to permeability or efflux, since
these properties were largely unaffected by the additional alkyl groups as determined by
measuring bidirectional membrane permeability across a Cacoꢀ2 cell monolayer. In general, the
compounds exhibited modest permeability with minimal efflux. In the coumarin series,
administration of the Nꢀmethyl analog 75 resulted in undetectable plasma concentrations. When
R and R were both methyl (49), the 2ꢀfold reduction in activity vs. 36 was compensated for by
1
2
a significant improvement in plasma exposure, with a brain/plasma ratio of 0.7. The addition of
ACS Paragon Plus Environment

Page 23 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
another methyl group at R (76) resulted in a 2ꢀfold decrease in activity compared to that of 49,
3
but greatly increased the brain/plasma ratio (B/P). Replacing the methyl group at R in 76 with
1
ethyl (9) achieved improvements in activity and plasma exposure. In the isocoumarin series
methylation at R and R (14) was sufficient to improve plasma exposure and achieve a B/P ratio
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
2
of 11. Similarly, in the pyridopyrimidinone series, methylation at R and R (20) resulted in
1
2
improved exposure. Pyridopyrimidinone 20 had comparable brain and plasma exposures (B/P =
.6), suggesting a more favorable distribution than observed for coumarin 9 (B/P = 19) and
1
isocoumarin 14 (B/P = 11). Additionally, pyridopyrimidinone 20 has the advantage of lower
plasma protein binding (80%) compared to that of coumarin 9 (96%) and isocoumarin 14 (93%).
Table 9. Effect of small alkyl group substitution on activity and pharmacokinetics
a
Rat PK
Cacoꢁ2
(10 mg/kg, PO)
PPBH
(%
EC1.5X PRO
Compd R1 R2
R3
Papp
(
nM)
ꢁ6 BꢁA / bound)
AꢁB
AUC0ꢁ6h
(ꢀg—h/mL) (6 h)
B/P
(
x 10
cm/s)
3
7
4
7
6
5
9
6
H
Me
H
H
H
H
15
69
31
—
—
0.74
1.1
—
97
97
0.34
<0.01
0.53
6.5
b
Me
2.7
1.4
n.d.
0.7
Me Me
Me Me (S)ꢁMe
Et Me (S)ꢁMe
63
44
39
—
—
—
0.26
1.1
19
19
9
1.2
1.9
96
65
H
Me
H
—
—
4.4
—
2.6
—
90
—
0.28
0.6
b
7
7
4
Me
120
120
<0.01
n.d.
1
Me Me
—
1.2
1.3
93
0.79
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
6
7
H
Me
H
—
—
31
14
—
2.8
—
70
—
0.38
<0.1
b
78
Me
300
<0.01
n.d.
1.6
20
Me Me
—
160
7.5
1.7
80
2.1
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Dosed PO as a suspension at 10 mg/kg in 0.5% hydroxypropyl methylcellulose and 0.1% Tweenꢀ80 to male
SpragueꢀDawley rats.
b
Not determined due to low compound concentration in the plasma.
Pharmacodynamic response. Coumarin 9, isocoumarin 14 and pyridopyrimidinone 20
demonstrated oral and CNS exposure in rats sufficient to warrant further evaluation in the ꢁ7
mouse model of SMA. The ꢁ7 transgenic mouse produces low levels of human SMN protein and
2
9
exhibits a severe SMA disease phenotype. The median survival time (MST) of untreated ꢁ7
mice is <20 days. Based on the in vitro data, we anticipated that our compounds would increase
the level of SMN protein in ꢁ7 mouse tissues by modifying SMN2 alternative splicing, leading to
a survival benefit when compared to untreated mice. The early manifestation of disease in ꢁ7
mice necessitates the administration of compound at the neonatal developmental stage. We found
that intraperitoneal dosing (IP) to neonates produced more consistent results than oral dosing.
Prior to dosing ꢁ7 mice, we determined plasma and brain exposure levels of 9, 14 and 20 after a
single 10 mg/kg IP dose to neonatal mice administered on postnatal day 9 (PND9) (Table 10).
Each compound showed sufficient plasma and brain exposure (AUC0ꢀ24h) to conduct the
pharmacodynamic and survival studies in ꢁ7 mice. The compounds were administered as a
DMSO solution once daily (QD) by IP injection. After 7 days of dosing, brain, spinal cord and
quadriceps muscle were collected. All three compounds at a dose as low as 0.1 mg/kg
demonstrated a dose dependent increase of SMN protein in all tissues analyzed (Table 10).
Table 10: Pharmacokinetic and pharmacodynamic data for 9, 14 and 20
ACS Paragon Plus Environment

Page 25 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Neonatal mouse PK
a
a
ꢂ7 mouse PD, IP, QD (PND3ꢁPND9)
IP, 10 mg/kg
EC1.5X PRO
Compd
c
(
nM)
SMN % increase above vehicle
AUC0ꢁ24h
Dose
mg/kg (N)
B/P
18
b
(
ꢀg—h/mL)
brain
spinal cord
quad
0.03 (7)
ꢁ10±11
115±49**
194±37***
7±20
92±34**
164±27***
ꢁ3±29
154±70***
147±36***
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
9
44
3.6
0.1 (6)
0.3 (6)
0
0
.1 (7)
.3 (5)
66±28**
114±49**
345±63***
395±31***
35±18*
98±49***
203±48***
284±96***
14±38
83±83
343±101***
517±218***
1
4
0
120
160
5.0
6.9
17
1
3
(10)
(11)
0
0
.1 (6)
.3 (7)
43±32
61±23*
149±57***
230±65***
26±19
6±16
33±18
129±33***
218±75***
67±22***
123±23***
189±30***
2
2.9
1
3
(7)
(7)
Abbreviations: B/P = brain to plasma ratio; PND = postnatal day; quad = quadriceps; IP = intraperitoneal; and QD =
once daily.
Dosed IP as a solution in 100% DMSO (2.5 mL/kg).
a
b
Number of mice per dose group
c
SMN % increase ± SD
* p<0.05, **p<0.01, ***p<0.001
To determine if the administration of compounds 9, 14 and 20 to ꢁ7 mice would produce a
survival benefit (an increase in median survival time), we chose doses that would elicit at least
1
00% increase in SMN protein in the brain. Coumarin 9, isocoumarin 14, and
pyridopyrimidinone 20 were dosed IP as low as 0.1, 0.1, and 1 mg/kg daily, respectively. We
switched to oral dosing twice a day at PND24 and increased the daily dose 6ꢀfold above the IP
dose to maintain the same drug levels in the plasma and tissue throughout the course of the
study. All three compounds doseꢀdependently improved survival with higher doses resulting in
3
0ꢀ32
longꢀterm survival (Table 11).
Overall, our data show that treatment of ꢁ7 mice with
compounds from multiple chemotypes (9, 14, and 20) results in increased SMN protein
production in the CNS and peripheral tissues in a dose dependent manner and that these
increases in SMN protein correlate with improved survival outcome.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Table 11: ꢁ7 mouse survival data for 9, 14 and 20
a
b
IP dose
PO dose
Survival at study
conclusion (PND
c
Compd
mg/kg/day
(PND3ꢁ23)
mg/kg/day
(PND24→)
MST (d)
concluded)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Vehicle
—
0.1
0.3
—
0.6
2
14
34
0/12 (21)
2/12 (55)
8/14 (55)
9
9
>55
9
1
0.1
0.3
1
6
0.6
2
>150
>100
>100
45
8/15 (150)
7/11 (100)
7/12 (100)
7/15 (90)
14
14
20
20
6
3
20
>190
7/11 (190)
Abbreviations: PND = postnatal day, MST = median survival time
a
Dosed as a solution in DMSO (2.5 mL/kg).
Dosed as a suspension in 0.5% hydroxypropyl methylcellulose and 0.1% Tweenꢀ80.
Number of mice surviving/number of mice enrolled
b
c
CONCLUSION
Through the discovery and activityꢀbased optimization of 7ꢀpiperazinyl coumarins, we have
identified three novel classes of molecules that potently shift splicing of SMN2 preꢀmRNA to
favor the production of SMN2ꢀFL mRNA over SMN2ꢀꢁ7 mRNA. This shift causes cells that are
deficient in SMN protein (due to the lack of the SMN1 gene) to produce additional and fully
functional SMN protein, which is critical for the proper function of cells, particularly motor
neurons. Compounds that exhibited potent in vitro activity in SMA patient−derived fibroblasts
were further optimized for oral delivery to disease relevant tissues in rats and mice. We
demonstrated that the levels of SMN protein could be increased in a dose dependent manner in
brain, spinal cord and muscle tissue in ꢁ7 mice after the administration of compounds derived
ACS Paragon Plus Environment

Page 27 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
from all three classes of molecules. Daily administration of these compounds to ꢁ7 mice resulted
in a reduction of disease manifestations and a significant increase in median survival time. These
findings support the development of orally administered small molecules for the treatment of
patients with SMA.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
EXPERIMENTAL METHODS
Starting materials and other reagents were purchased from commercial suppliers and were used
without further purification unless otherwise indicated. The syntheses of intermediates utilized
here (1b and 1c) are described in the Supporting Information. Air or moisture sensitive reactions
were performed under either a nitrogen or argon atmosphere. Flash chromatography was
performed using silica gel with standard techniques or with silica gel cartridges on an ISCO
1
Combiflash chromatography instrument. H NMR spectra were recorded at 500 MHz on a
Bruker NMR spectrometer. The chemical shifts are given in ppm referenced to the deuterated
solvent signal. Coupling constants (J) are recorded in hertz. LCꢀMS analyses were performed on
a Waters Acquity UPLC/MS system with an analytical C18 column and compounds were
detected by UV absorption at 254 nm. All final compounds with reported biological data were
1
determined to be >95% pure as determined by LC/MS and H NMR unless otherwise noted.
Ethyl 2ꢀ(benzo[d]oxazolꢀ2ꢀyl)acetate (1a). A mixture of 2ꢀaminophenol (5.0 g, 46 mmol) and
ethyl 3ꢀethoxyꢀ3ꢀiminopropanoate hydrochloride (11.7 g, 60 mmol) in EtOH (200 mL) was
stirred at 80 °C for 16 h. The mixture was partitioned between EtOAc (200 mL) and H O (200
2
mL). The organic layer was washed with brine and concentrated. The residue was
chromatographed on silica gel (10% EtOAc in hexanes) to yield 1a (9.0 g, 95%) as a pale yellow
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 54
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
oil. H NMR (DMSOꢀd ) δ 7.74 (m, 2H), 7.44ꢀ7.37 (m, 2H), 4.22 (s, 2H), 4.17 (q, J = 7.1 Hz,
6
2H), 1.22 (t, J = 7.1 Hz, 3H).
tertꢀButyl 4ꢀ(4ꢀformylꢀ3ꢀhydroxyphenyl)piperazineꢀ1ꢀcarboxylate (2). A mixture of 4ꢀ
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
fluorosalicylaldehyde (2.80 g, 20 mmol) and 1ꢀBocꢀpiperazine (4.47 g, 24 mmol) in DMSO (20
mL) was heated at 100 °C for 16 h. The mixture was diluted with H O (100 mL) and extracted
2
with EtOAc (100 mL). The organic layer was washed with brine, dried over Na SO , filtered and
2
4
concentrated. The residue was chromatographed on silica gel (0ꢀ50% EtOAc in hexanes) to yield
+
1
2
(4.6 g, 75%) as a tan solid. LC−MS: m/z = 307.2 [M+H] . H NMR (500 MHz, CDCl ) δ 11.39
3
(
br s, 1H), 9.55 (s, 1H), 7.31 (d, J = 8.8 Hz, 1 H), 6.47 (dd, J = 8.8, 2.5 Hz, 1H), 6.26 (d, J = 2.2
Hz, 1H), 3.55 (m, 4H), 3.35 (m, 4H), 1.42 (s, 9H).
tertꢀButyl 4ꢀ(3ꢀ(benzo[d]oxazolꢀ2ꢀyl)ꢀ2ꢀoxoꢀ2Hꢀchromenꢀ7ꢀyl)piperazineꢀ1ꢀcarboxylate (3).
Compound 1a (100 mg, 0.43 mmol), compound 2 (138 mg, 0.45 mmol), AcOH (15 µL, 0.25
mmol) and piperidine (44 µL, 0.45 mmol) were combined in EtOH (2 mL). The mixture was
stirred in a sealed tube at 90 °C for 24 h. After cooling to rt, the mixture was filtered. The solid
+
was washed with Et O and dried, yielding 3 (180 mg, 93%). LC−MS: m/z = 448.2 [M+H] .
2
3
ꢀ(Benzo[d]oxazolꢀ2ꢀyl)ꢀ7ꢀ(piperazinꢀ1ꢀyl)ꢀ2Hꢀchromenꢀ2ꢀone hydrochloride (4). Compound
3
(90 mg, 0.2 mmol) was dissolved in 4 N HCl in 1,4ꢀdioxane (2 mL). The mixture was stirred at
rt for 3 h. The solid was collected and dried, yielding 4 (66 mg, 85%) as a yellow powder.
+
1
LC−MS: m/z = 348.2 [M+H] . H NMR (500 MHz, CDCl ) δ 9.31 (br s, 2H), 8.92 (s, 1H), 7.84ꢀ
3
7
.76 (3H), 7.44 (m, 2H), 7.12 (dd, J = 9.0, 2.2 Hz, 1H), 7.01 (d, J = 2.2 Hz, 1H), 3.75 (m, 4H),
3.23 (m, 4H).
ACS Paragon Plus Environment

Page 29 of 54
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
3
ꢀAcetylꢀ7ꢀfluoroꢀ2Hꢀchromenꢀ2ꢀone (5). Piperidine (100 µL, 1 mmol) was added to a mixture
of 4ꢀfluoroꢀ2ꢀhydroxybenzaldehyde (1.4 g, 10 mmol) and ethyl 3ꢀoxobutanoate (1.3 g, 10
mmol). The mixture was stirred at rt for 10 min. The precipitate that formed was collected by
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
vacuum filtration. The solid was washed with ethanol and aqueous HCl (1 N), filtered and dried
1
to give 5 (1.96 g, 95%) as a pale yellow solid. H NMR (500 MHz, CDCl ) δ 8.51 (s, 1H), 7.68
3
(m, 1H), 7.13ꢀ7.07 (m, 2H), 2.73 (s, 3H).
3ꢀ(2ꢀBromoacetyl)ꢀ7ꢀfluoroꢀ2Hꢀchromenꢀ2ꢀone (6). A solution of bromine (1.6 g, 10 mmol) in
CHCl (10 mL) was added dropwise to a solution of 5 (1.96 g, 9.5 mmol) in CHCl (20 mL). The
3
3
mixture was stirred at rt for 1 h. The solid material that was present in the mixture was collected
by vacuum filtration, washed with CHCl and dried to give 6 (1.96 g, 72%) as a pale yellow
3
1
solid. H NMR (500 MHz, CDCl ) δ 8.63 (s, 1H), 7.72 (m, 1H), 7.17ꢀ7.10 (m, 2H), 4.73 (s, 2H).
3
3ꢀ(6,8ꢀDimethylimidazo[1,2ꢀa]pyrazinꢀ2ꢀyl)ꢀ7ꢀfluoroꢀ2Hꢀchromenꢀ2ꢀone hydrobromide (7).
A mixture of 6 (0.684 g, 2.4 mmol) and 3,5ꢀdimethylpyrazinꢀ2ꢀamine (0.246 g, 2.0 mmol) in
CH CN (10 mL) was stirred at 120 °C in a sealed tube for 20 min. The mixture was cooled to rt
3
and diluted with Et O to produce a precipitate. The solid was collected by vacuum filtration,
2
+
washed with Et O and dried to give 7 (0.7 g, 90%) as a tan solid. LC−MS: m/z = 310.1 [M+H] .
2
1
H NMR (500 MHz, DMSOꢀd ) δ 8.92 (s, 1H), 8.84 (s, 1H), 8.58 (s, 1H), 8.10 (dd, J = 6.3, 8.8
6
Hz, 1H), 7.48 (dd, J = 2.3, 9.6 Hz, 1H), 7.33 (dt, J = 2.5, 8.7 Hz, 1H), 2.91 (s, 3H), 2.46 (s, 3H).
(
S)ꢀ3ꢀ(6,8ꢀDimethylimidazo[1,2ꢀa]pyrazinꢀ2ꢀyl)ꢀ7ꢀ(3ꢀmethylpiperazinꢀ1ꢀyl)ꢀ2Hꢀchromenꢀ2ꢀ
one (8). Compound 7 (100 mg, 0.25 mmol) was combined with (S)ꢀ2ꢀmethylpiperazine (52 mg,
.52 mmol) and K CO (0.14 g, 1.0 mmol) in DMSO (0.5 mL). The mixture was stirred at 120
0
2
3
°
C for 2 h. The mixture was cooled to rt and diluted with water to produce a precipitate. The
ACS Paragon Plus Environment