Easy-To-Synthesize Spirocyclic Compounds Possess Remarkable in Vivo Activity against Mycobacterium tuberculosis

Article abstract of DOI:10.1021/acs.jmedchem.8b01533

Society urgently needs new, effective medicines for the treatment of tuberculosis. To kick-start the required hit-to-lead campaigns, the libraries of pharmaceutical companies have recently been evaluated for starting points. The GlaxoSmithKline (GSK) library yielded many high-quality hits, and the associated data were placed in the public domain to stimulate engagement by the wider community. One such series, the spiro compounds, are described here. The compounds were explored by a combination of traditional in-house research and open source methods. The series benefits from a particularly simple structure and a short associated synthetic chemistry route. Many members of the series displayed striking potency and low toxicity, and highly promising in vivo activity in a mouse model was confirmed with one of the analogues. Ultimately the series was discontinued due to concerns over safety, but the associated data remain public domain, empowering others to resume the series if the perceived deficiencies can be overcome.

Full text of DOI:10.1021/acs.jmedchem.8b01533

Subscriber access provided by University of Leicester
Article
Easy-to-Synthesize Spirocyclic Compounds Possess
Remarkable In Vivo Activity Against Mycobacterium Tuberculosis
Ana Guardia, Jessica Baiget, Monica Cacho, Arancha Pérez, Montserrat Ortega-Guerra, Winston
Nxumalo, Setshaba Khanye, Joaquín Rullas, Fatima Ortega, Elena Jimenez, E Perez-Herran,
María Teresa Fraile-Gabaldón, Jorge Esquivias, Raquel Fernández, Esther Porras-De Francisco,
Lourdes Encinas, Marta Alonso, Ilaria Giordano, Cristina Rivero, Juan Miguel-Siles, Javier Osende,
Katrina Badiola, Peter J. Rutledge, Matthew H. Todd, Modesto Remuiñán, and Carlos Alemparte
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.8b01533 • Publication Date (Web): 20 Nov 2018
Downloaded from http://pubs.acs.org on November 21, 2018
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 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 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.
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 44
Journal of Medicinal Chemistry
1
2
3
4
5
6
Remuiñán, Modesto; GlaxoSmithKline, Tres Cantos Medicines
Development Campus
Alemparte, Carlos; GlaxoSmithKline, Tres Cantos Medicines Development
Campus
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
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 44
1
2
3
4
5
6
Easy-to-Synthesize Spirocyclic Compounds Possess Remarkable In Vivo
Activity Against Mycobacterium Tuberculosis
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
Ana Guardia,† Jessica Baiget,‡ Mónica Cacho,† Arancha Pérez,† Montserrat Ortega-Guerra,† Winston
Nxumalo,† Setshaba D. Khanye,† Joaquín Rullas,† Fátima Ortega,† Elena Jiménez,† Esther Pérez-
Herrán,† María Teresa Fraile-Gabaldón,† Jorge Esquivias,† Raquel Fernández,† Esther Porras-De
Francisco,† Lourdes Encinas,† Marta Alonso,† Ilaria Giordano,† Cristina Rivero,† Juan Miguel-Siles,†
Javier G. Osende,‡ Katrina A. Badiola,‡ Peter J. Rutledge,‡ Matthew H. Todd,*‡§ Modesto Remuiñán,†
and Carlos Alemparte*†
†GlaxoSmithKline, Tres Cantos Medicines Development Campus, Severo Ochoa 2, 28760 Tres
Cantos, Madrid, Spain
‡School of Chemistry, The University of Sydney, NSW 2006, Australia
§School of Pharmacy, University College London, 29-39 Brunswick Square, London WC1N 1AX, UK
ABSTRACT: Society urgently needs new, effective medicines for the treatment of tuberculosis. To
kick-start the required hit-to-lead campaigns, the libraries of pharmaceutical companies have recently
been evaluated for starting points. The GlaxoSmithKline (GSK) library yielded many high-quality hits
and the associated data were placed in the public domain to stimulate engagement by the wider
community. One such series, the Spiro compounds, are described here. The compounds were explored
by a combination of traditional in-house research and open source methods. The series benefits from a
particularly simple structure and a short associated synthetic chemistry route. Many members of the
series displayed striking potency and low toxicity, and highly promising in vivo activity in a mouse
model was confirmed with one of the analogs. Ultimately the series was discontinued due to concerns
over safety, but the associated data remain public domain, empowering others to resume the series if
the perceived deficiencies can be overcome.
1
ACS Paragon Plus Environment

Page 3 of 44
Journal of Medicinal Chemistry
1
2
3
4
INTRODUCTION
5
6
7
8
Tuberculosis (TB) is a major public health problem, with one-third of the world's population infected
with the causative agent, Mycobacterium tuberculosis, resulting in 10 million new cases of clinical
TB in 2017 and over 4,000 deaths, daily.1 The End TB Strategy, adopted by the World Health
Assembly in 2014, aims in the next 15 years to reduce TB deaths by 90% and reduce new cases by
80%. The Strategy’s resolutions explicitly call for the development of new drugs, yet few new
treatments have been approved in recent decades,2 and the pipeline remains weak.3 The pressing need
for new small molecule therapeutics is made greater by the emergence of multi- and totally-resistant
strains,4,5 and the troubling finding that resistance can arise without fitness costs.6 Even where new
therapeutics are developed and are made available, the associated costs of access can severely limit
the numbers of people treated,7 illustrating the importance of pursuing simple, inexpensive medicines
if the burden of TB is to be genuinely reduced. New initiatives continue to pursue candidate
compounds. These range from screening libraries within pharmaceutical companies and sharing the
resulting data openly8 through to new incentive mechanisms to stimulate research and development.9
From a high-throughput screening campaign of GSK’s corporate compound collection, the spirocyclic
compound 1 was identified as a promising antitubercular hit from a cluster of seven compounds
(termed the “Spiro” series) (Table 1).10 As previously reported, treatment of Mycobacterium bovis
BCG with 1 produced accumulation of intracellular trehalose monomycolate (TMM) by interfering
with the ability of MmpL3 to act as a TMM transporter.11 MmpL3 protein is a member of the MmpL
(Mycobacterial membrane protein Large) family of transporters that has been identified as essential
for Mtb viability.12 MmpL3 has been recently identified as a promising target for TB; in fact, several
new TB inhibitors with structural diversity, including the drug candidate SQ109, may interact with
MmpL3 as a mode of action13 though there remains uncertainty over whether the mechanism of action
is via a direct inhibition or a related perturbation in the transmembrane proton gradient.14
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
Table 1. Profile of Initial Hit 1.
2
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 44
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
Parameter
1
Mtb H37Rv MIC90 (µM)
0.30
0.60
0.25
≥16
36
Mtb MIC90 (µM) (108 strains)
Intracellular H37Rv MIC80 (µM)
Antibacterial panel IC50 (µM)
Mammalian cell (HepG2) Tox50 (µM)
clogP
2.99
>30
25
CLint (mL/min·g) mouse microsomes
CLint (mL/min·g) human microsomes
Solubility CLND (µM)
266
The compound’s intra-macrophage activity, selectivity for mycobacteria vs. other bacterial species,
and potency against a broad panel of clinical isolates including MDR and XDR strains, in vitro -cidal
behavior and a low frequency of spontaneous resistance13a suggested a highly promising
antitubercular profile. The low microsomal stability was a concern and required improvement. In this
paper, we report our efforts to optimize this family of compounds through the synthesis of analogs
aimed at retaining the antitubercular potency while improving the overall profile for the series.
RESULTS AND DISCUSSION
Synthesis and In Vitro Evaluation of Analogs
Aiming to explore the chemical space of the series, a library of novel compounds was synthesized and
evaluated. The compounds were isolated either as free amines or as salts (hydrochloride or
3
ACS Paragon Plus Environment

Page 5 of 44
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
trifluoroacetate, see Experimental Section and Supporting Information), but for simplicity only the
parent compound is drawn throughout the paper. The three different regions of the molecule were
explored in turn: the northern part (the thiophene ring), the central part (methylenepiperidine moiety)
and the southern part (the aryl pendant).
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
Figure 1. Areas of variation in the synthetic chemistry towards understanding the SAR.
1) Modifications to the North. The first region to be explored for determination of the Structure-
Activity Relationship (SAR) was the northern part of the molecule, with a view to replacing the
thiophene group that had the potential to promote oxidative metabolism.15 These analogs were
obtained using an oxa-Pictet Spengler reaction on 1-benzyl-4-piperidinone with different aryl alcohols
under mild conditions (Scheme 1).16 Subsequent N-deprotection and reductive amination with 4-
benzodioxane-6-carboxaldehyde generated the corresponding compounds that were evaluated (Table
2). The thiophene isomer of 1 (i.e. 4,7-dihydro-5H-thieno[2,3-c]pyran) was found to be inactive, but
the sample displayed instability under ambient conditions (visual color degradation, and appearance
1
of new signals in the H NMR spectrum) that made its assessment challenging (Compound S1,
Supporting Information). Compounds 2 and 3 were instead found to exhibit superior stability and
confirm the strong link between thiophene orientation and biological potency. The benzopyran
analogue 4 was better tolerated than the indole analogue 6, but the addition of a methoxy group on the
phenyl ring (5) in order to block potential metabolic reactions resulted in loss of potency.
Scheme 1. Synthesis of Compounds with Modifications to the Northern part
4
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 44
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
Reagents and Conditions: a) MsOH, toluene, 80°C, 3 h; b) 1-chloroethyl chloroformate, THF, –78°C, 90 min; c)
MeOH, 65°C, 1 h; d) 1,4-benzodioxan-6-carboxaldehyde (or benzofuran-2-carbaldehyde for compounds 2 and
3), NaBH(OAc)3, DCM or DCE, 40–100ºC, 1.5 – 48 h.
Table 2. Evaluation of Analogs Modified in the Northern Part
Mtb H37Rv
Mtb H37Rv
Compound
R
Compound
R
MIC90 (µM)
MIC90 (µM)
S
1
2
0.30
2.7
4
5
0.94
>10
MeO
see Scheme 1
see Scheme 1
3
>100
6
93
N
H
2) Modifications to the Center. The second region to be examined was the central part of the
molecule, in order to verify the importance of the methylenepiperidine moiety (Scheme 2). The amine
716 was coupled with the appropriate carboxylic acid, sulfonyl chloride and isocyanate to afford the
corresponding amide 8, sulfonamide 9 and urea 10, respectively. The five-membered ring analog (12)
and the open-chain compounds (14) were synthesized from 2-(thiophen-2-yl)ethanol with the oxa-
5
ACS Paragon Plus Environment

Page 7 of 44
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Pictet Spengler reaction as before, employing the appropriate Cbz-protected ketone or aldehyde as the
coupling partner, followed by deprotection and reductive amination with 2,3-dihydro-1,4-
benzodioxin-6-carbaldehyde. The quaternary analog 15 was synthesized in a two-step sequence from
2-(4-methyl-6,7-dihydro-4H-thieno[3,2-c]pyran-4-yl)ethanol which was subjected to mesylation and
nucleophilic substitution with the relevant amine, 2,3-dihydro-1,4-benzodioxin-6-ylmethylamine.
Replacement of the methylene group between the piperidine and benzene rings with amide,
sulfonamide or urea groups generated inactive molecules (Table 3). Similar results were obtained
when the piperidine moiety was replaced with a pyrrolidine moiety. Interestingly, the acyclic
analogues, which lack the spirocyclic structure, were also inactive. This set of molecules revealed the
importance of the methyl dihydrospiropiperidine structure for retaining antitubercular 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
Scheme 2. Synthesis of Compounds with Modifications to the Central part
Reagents and Conditions: a) oxa-Pictet Spengler reaction, then deprotection.16 b) 1,4-benzodioxane-6-carboxylic
acid, HOBt, EDCI, DCM, 20 °C, O/N; c) 2,3-dihydrobenzo[b][1,4]dioxine-6-sulfonyl chloride, Et3N, DCM, 20
°C, 2 h; d) 6-isocyanato-2,3-dihydrobenzo[b][1,4]dioxine, DIPEA, DCM, 0 ºC, 2 h; e) (i) 3-pyrrolidinone,
pTsOH·H2O, Na2SO4, MeCN, 82 °C, O/N, (ii) TMSI, DCM, 20 °C, O/N; f) 2,3-dihydro-1,4-benzodioxin-6-
carbaldehyde, Biotage® MP-cyanoborohydride, AcOH, DCM, MW, 100°C, 40 min.; g) 2,3-dihydro-1,4-
6
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 44
1
2
3
4
5
6
benzodioxin-6-carbaldehyde, (Ti(OiPr)4, EtOH, rt, 1 h then NaBH4, 20 °C, 36 h. h) (i) MsCl, Et3N, DCM, 20
°C, O/N, (ii) 2,3-dihydro-1,4-benzodioxin-6-ylmethylamine, Et3N, 75°C, O/N.
7
8
9
Table 3. Evaluation of Analogs modified in the Central Part
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
S
O
X
O
O
Mtb H37Rv
Mtb H37Rv
Compound
X
Compound
X
MIC90 (µM)
MIC90 (µM)
O
O
1
0.30
12
> 5
N
N
O
O
H
8
> 80
14
> 5
N
HN
O
O
O
9
> 10
> 10
15
> 5
N
HN
S O
O
10
7
ACS Paragon Plus Environment

Page 9 of 44
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
O
O
N
N
H
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) Modifications to the South. The southern area of the lead compound was examined through the
synthesis of 25 compounds containing a modified N-substituent attached to the central 6',7'-
dihydrospiro[piperidine-4,4'-thieno[3,2-c]pyran by reductive amination (Scheme 3, Table 4). The
aldehydes employed were either commercially available or easily synthesized by adapting methods
already described in the literature.16, 17
Scheme 3. Synthesis of Compounds with Modifications to the Southern part
Reagents and Conditions: See Experimental Section and Supporting Information.
Table 4. Evaluation of Analogs modified in the Southern Part
Mtb H37Rv
CLint (mL/min·g)
Compound
R
MIC90 (µM)
human
mouse
O
O
1
7
0.30
> 5
25
nd
> 30
nd
H
16
0.23
2.9
> 30
O
8
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 44
1
2
3
4
5
6
7
8
9
17
18
0.02
> 5
22.6
nd
> 30
nd
O
O
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
N
1916
> 5
nd
nd
22.5
> 30
20
0.20
5.2
> 30
21
22
23
0.63
> 5
nd
nd
0.08
18.9
> 30
24
0.02
9.9
> 30
25
26
0.14
0.63
3.0
4.6
18.2
18.1
F5S
F3C
9
ACS Paragon Plus Environment

Page 11 of 44
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
27
28
0.05
0.12
12.0
1.9
28.8
1.1
O
HO
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
O
F3CO
2911
0.06
0.20
0.6
1.2
O
Br
30
31
1.5
4.1
0.9
N
H
> 5
28.7
NH
N
3216
3
nd
nd
F
N
33
34
3
nd
nd
nd
nd
N
14
N
10
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 44
1
2
3
4
5
6
7
8
9
N
35
36
14
nd
nd
N
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
N
3
nd
nd
MeO
37
38
0.63
4.5
> 30
N
0.94
0.31
3.1
1.3
7.0
HN
Cl
39
10.3
HN
nd: not determined.
In general, placing aromatic rings in the south provided potent compounds (vs. analog lacking
substituents as in 7). However, there is a very high level of sensitivity to the substitution pattern, with
the unsubstituted-benzyl compound 19 and the 2-methyl analogue 22 being inactive, whilst the 3- or
4-methyl isomers (20 and 21) and the 3,4-dimethyl (23) analogue were highly potent. The fused
cyclohexyl (17) analogue was found to be more potent than the parent compound 1 or the
corresponding chromane analogue 16, yet the pyridyl analogue (18) was found to be inactive. Several
substituents were well tolerated in the 4-position of the phenyl ring (24, 25), as was methyl branching
in the benzylic position (26, racemic). Substitution of the phenyl ring for heterocyclic groups had
varying effects, with the potency of the indole set (37–39) on the one hand, and the inactive pyrrole
analogue 31 on the other. (Compound 31 was the least lipophilic analogue examined to this point; in
11
ACS Paragon Plus Environment

Page 13 of 44
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
order to evaluate the influence of polarity on metabolic stability this compound was, despite being
inactive, progressed to the microsome assay but was found to be metabolically unstable.) Reasonable
levels of potency were seen with a set of N-aryl pyrrole substitutions in this position (32, 33, 36),
although the N-pyridyl variants (34, 35) were found to exhibit low potency. Biphenyl ether analogue
27 displayed outstanding activity but microsomal clearance was still high. Blocking the para-position
with hydroxymethyl, trifluoromethoxy and bromine substituents produced very potent and fairly
stable compounds (28–30).
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
Pharmacokinetics
In order to confirm translation of the microsomal data to an in vivo setting, the pharmacokinetic
profiles of the initial hit 1 and the di-aryl analogues (27–30) that gave the best results observed in
vitro were selected for in vivo PK studies following a single dose administered through oral and
intravenous routes (Table 5).
Table 5. In Vivo Pharmacokinetic Profile for Selected Compounds
iv PK mice
po PK mice
Dosea
Cl
Dose Cmax AUC
Compound
Structure
Vss
(mg/k t1/2 (h)
g)
(mL/m (mg/ (µg/m (µg∙h/ F (%)
(L/kg)
in/kg)
Kg)
L)
mL)
S
O
1
4
0.7
3.9
0.2
7.4
214.2
49.2
0.47
1.78
14.6
1.5
N
O
O
S
O
27
28
3.6
3.9
19.4
7.0
67.7
18.7
44.0
0.05
0.04
0.10
0.04
N
O
S
O
721.8
2.5
N
HOH2C
O
12
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 44
1
2
3
S
O
4
5
6
7
8
29
30
3.8
4.5
3.2
7.3
10.4
12.0
36.3
19.8
44.0
45.0
1.27
1.02
8.74
42.6
N
F3CO
O
S
9
O
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
N
17.87
84.6
Br
N
H
aFor iv route, all compounds were administered in 5% DMSO/20% Encapsin; this was identified as a suitable
vehicle to dissolve the compounds at the required dose.
In the in vitro studies, compound 27 displayed already high intrinsic clearance in liver microsomes.
Consistently, the low bioavailability observed in the oral administration profile confirmed an
extensive first-pass metabolism of the compound. Compound 28 contrasted the in vitro data and
presented high in vivo clearance and low exposure following i.v. and p.o. administration, respectively.
This result might be attributed to the enhanced intrinsic clearance likely due to phase II metabolism of
the primary alcohol. In contrast, compounds 29 and 30 presented the best bioavailability with lower
intrinsic clearance, longer half-lives and oral AUC values suitable for progression into efficacy
studies. Cross-resistance of compound 29 with resistant mutants generated against initial hit 1
confirmed that, despite the structural changes incorporated during the optimization process, the
biological target is maintained.11
In vivo Efficacy
The efficacy of compound 29, showing promising pharmacokinetic profile and high in vitro potency,
was studied in an acute TB murine infection model. The data obtained are shown in Figure 2.
Compound 29 proved to be very efficacious showing a -cidal response, 4.2 log CFU reduction
compared to the untreated control, similar to isoniazid. The good result obtained in a short assay in
which the compound is administered only for 8 days is evidence for a fast-killing profile. ED99
(Efficacious Dose that gives 2 log colony-forming units (CFU) reduction compared to the untreated
control) was 11.6 mg/kg. This encouraging value, although higher than that of isoniazid, is in the
range of most existing antitubercular dugs. A maximum recommended dose of 50 mg/kg was used
13
ACS Paragon Plus Environment

Page 15 of 44
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
since tolerability issues had been observed at higher doses. The exciting results that confirmed the
potential of this chemical series to efficiently kill Mtb in vivo prompted us to continue our medicinal
chemistry efforts towards optimization of compound 29.
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
8
Isoniazid
Compound 29
2 logCFU
reduction
6
4
2
0
Limit of detection (DL)
0.01
0.1
1
10
100
1000
Dose (mg/kg)
Figure 2. Efficacy in a mouse model of acute TB infection under oral QD dosing regimen (once a
day). C57BL/6J mice were infected with M. tuberculosis H37Rv intratracheally (~105 CFU) and were
dosed starting on the following day after infection for 8 days. Mice were sacrificed at least 24 h after
the last drug administration.
Decreasing Lipophilicity
Despite its remarkable efficacy, a number of factors precluded further progression of compound 29
(see Tables 6 and 7). The Cell Health (CH) assay is a triple-readout phenotypic assay measuring
changes in cell structure as a preliminary consequence of cytotoxic injury. The assay uses automated
imaging to measure IC50 values for the effect of compounds on human liver-derived HepG2 cells (48
h incubation) as a surrogate of hepatotoxicity. For high doses (>100 mg/day), an IC50 value lower than
80 µM in any of the readouts of the assay is considered an indication of high hepatotoxicity risk (see
Table 6 for data of compound 29). Both high doses and long treatments, considered as an additional
risk factor, are common to antitubercular drugs. Inhibition of the human ether-a-go-go-related gene
(hERG) channel was also a concern for potential cardiotoxicity (see below and Table 7 for
14
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 44
1
2
3
4
5
6
7
8
9
discussion). Additionally, the compound presented a poor solubility and high lipophilicity. We
envisioned that a decrease of the latter would be a reasonable strategy to improve the general ADMET
profile of compound 29. At this stage, we used the property forecast index (PFI), a more relevant
measure of the lipophilicity of the compounds, to guide the design of new analogs. PFI is the sum of
the Chrom LogD7.4 and the number of aromatic rings and values lower than 7 predict promising
compounds in terms of solubility, permeation, cytochrome P450s, intrinsic clearance, and hERG
binding and promiscuity.18 In an effort to reduce lipophilicity (compound 29 has a PFI value of ca.
12), several compounds lacking the thiophene ring were proposed that afforded improved predicted
PFI values (Table 6); such changes had not previously been addressed in modifying the northern part
of the molecule (Table 2).
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
Accordingly, these compounds (40–50) were synthesized (Scheme 4) by reductive amination of 4-[4-
(trifluoromethoxy)phenoxy]benzaldehyde or 4-tert-butylbenzaldehyde and the corresponding cyclic
amines. The amines employed were either commercially available or easily synthesized. In the case of
compounds 42 and 50, the spirocyclic amine (58) required was synthesized as shown in Scheme 5.
The synthesis started with the Grignard reaction of N-Boc-4-piperidone with allylmagnesium bromide
to form the corresponding alcohol 55, which was alkylated with allyl bromide to form ether 56. Ring
closing metathesis of 56 with Grubbs I catalyst formed the unsaturated spirocyclic compound 57,
which underwent hydrogenation followed by N-deprotection to form 58.
Scheme 4. Synthesis of Analogs Lacking the Thiophene Ring
Scheme 5. Synthesis of 1-Oxa-9-azaspiro[5.5]undecane
15
ACS Paragon Plus Environment

Page 17 of 44
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
Reagents and Conditions: a) Allylmagnesium bromide, Et2O, 0°C, then 20 °C, O/N; b) allylbromide, NaH,
DMF, 20 °C; c) Grubbs I, DCM; d) Pd/C, H2, MeOH, Et2O, 0°C, then 20 °C, O/N; e) 3 M HCl, MeOH, 2 h, 20
°C.
Table 6. Evaluation of Compounds Lacking the Thiophene Ring
Solubilityb
(µM)
CLint (mL/min·g)
MIC
(µM)
Cell Health
Cmpd
R
PFI
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/B/C, µM)a
Human
Mouse
R
N
F3CO
O
S
29
40
41
0.06
> 5
0.6
nd
1.2
12.6/10.7/11.0
46.3/49.5/44.2
28.3/21.8/25.9
6
12.37
9.17
O
H
H
nd
156
<1
0.10
0.5
2.8
12.66
42
43
0.31
0.65
1.7
1.2
43.7/46.3/38.7
<1
53
9.91
9.56
O
< 0.5
< 0.5
25.7/28.8/26.3
49.8/53.8/46.9
O
O
O
O
44
45
> 5
1.3
nd
nd
52
9.68
9.79
1.5
0.6
132.6/140.9/96.9
Nd
O
16
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 44
1
2
3
4
5
6
7
8
9
NH
46
47
>5
>5
nd
nd
nd
nd
5.7/6.6/4.9
117
87
6.21
O
29.2/36.3/39.4
8.91
N
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
O
48
49
>5
nd
nd
nd
nd
75.9/>200/74.1
8.0/12.2/6.9
121
205
9.24
7.57
O
N
2.5
R
N
50
51
0.21
0.63
3.2
7.3
18.1
48.0
157.7/134.6/129.0
>200/>200/>200
≥596
7.23
7.23
O
O
363
O
52
53
54
5
2.5
5
0.95
1.4
2.25
13.5
27.0
>200/>200/>200
>200/>200/>200
>200/169.8/166.0
≥399
374
6.56
7.45
8.80
O
O
O
O
4.1
194
aA: membrane permeability; B: mitochondrial potential; C: nuclear morphology. bKinetic solubility measured by
CLND (see Experimental Section). nd: not determined.
The results indicated that activity is lost when the R group is just a hydrogen atom (the non-spiro
compound 40). Cyclohexyl (41) and 2-pyranyl (42) groups, both lacking the fused thiophene ring, can
be incorporated without significant detriment to the potency or stability, and with improved PFI
17
ACS Paragon Plus Environment

Page 19 of 44
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
(under 10) in the case of 42. Other more significant alterations to the pyran ring, such as the
incorporation of a 6- or 5-membered ring acetal (43 and 44) or substitution with a 2-tetrahydrofuran
ring (45), gave improved PFI and CH but higher MIC values, with clearance rates remaining low.
Further changes incorporating amines (46, 49), an ,-unsaturated amide (47) and a lactone (48) gave
compounds with good solubilities and PFI values (particularly in the case of 46) but with poor
potency values. Compound 50, containing the 2-pyranyl moiety of 42, but with a 4-tert-butylphenyl
ring in place of the bis-phenyl ether presented a good PFI value as two aromatic rings had been
removed, showing at the same time good potency and solubility, reasonable human microsomal
stability and low cytotoxicity. These trends continued for the related compounds containing isomeric
pyrans (51, 52) or ketals (53, 54) which displayed excellent CH and PFI values alongside low
clearance rates, but with generally lower potency than compound 50.
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
As mentioned above, the cardiovascular risk associated to hERG inhibition was also a concern that
needed to be monitored so in vitro hERG data were generated for a selection of compounds (Table 7).
Table 7. In vitro Cardiovascular Data for Selected Compounds
hERG
IW IC50
(µM)
hERG PXP
Compound
Structure
PFI
In silico
IC50 (µM)a
S
O
29
12.37
10.0
1.5 (98.4)
hERG alert
N
F3CO
O
S
O
30
10.39
7.9
2.3 (93.2)
hERG alert
N
Br
N
H
18
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 44
1
2
3
4
5
O
6
7
8
9
42
9.91
2.5
nd
hERG alert
N
F3CO
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
O
O
N
50
7.23
15.8
8.1 (82.0)
No alert
aMaximum inhibition (%) indicated in brackets. nd: not determined.
The in silico prediction from the structures pointed to potential hERG channel inhibition in vitro in
mammals for compounds 29, 30 and 42. This alert describes a structure-based pharmacophore
developed primarily from compounds that have been reported to be strong inhibitors of the hERG
potassium channel.18 An ion channel inhibition in vitro assay, carried out using automated planar chip
electrophysiology platforms IonWorks (IW) and PatchExpress (PXP), confirmed the alert. Although
the less lipophilic compound 50 did not show an in silico alert, the experimental hERG inhibition was
decreased by less than one order of magnitude. We decided to assess if the cardiovascular liability of
the series was also observed in ex vivo studies (perfused rabbit ventricular wedge assay, RVW) with
compound 50 (Table 8).
Table 8. Rabbit Ventricular Wedge Data for Compound 50
QT (µM)
(% change)
QRS (µM)
(% change)
Tp-e (µM)
(% change)
Compound
50
TdP scorea @ 10µM
10 (12.2)
30 (25.4)
10 (44.4)
3.0
aTdP score > 2.5 predicts a risk of cardiovascular toxicity.
19
ACS Paragon Plus Environment

Page 21 of 44
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Compound 50 caused a concentration-dependent prolongation of QT interval and Tp-e from 10 µM
and also caused a prolongation of QRS interval from 30 µM. The maximum Torsades de Pointes
(TdP) score19 was 3.0 at 10 µM that exceed the 2.5 threshold (based on GSK criteria) indicating a risk
of developing TdP arrhythmias. These results confirm the cardiovascular risk associated to this
chemical series.
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
Furthermore, in vivo tolerability studies (data not shown) suggested that 50 presented a maximum
tolerated exposure lower than that of compound 29: AUC<6.6 µg·h/mL for 50 vs. 13.7 µg·h/mL for
29. It is evident that in this case the considerable reduction of lipophilicity (almost 5 units of PFI) and
simplification of the structure (MW reduced from 475.5 to 301.5) achieved in the optimization
process that led from compound 29 to 50 afforded only a modest improvement in hERG in vitro data
and no progress in terms of in vivo tolerability. At this point, despite the excellent efficacy observed
with this chemical class, we considered that the associated safety risk was high and decided to
discontinue our work on the series.
CONCLUSIONS
Through exploration of a class of small molecule spirocyclic compounds, made easier by the ease
with which the compounds can be accessed synthetically, we have identified highly potent
antitubercular compounds. Several compounds examined possessed good pharmacokinetic properties.
Most importantly, a representative member of the series displayed remarkable potency in an in vivo
model of the disease. These features would normally mark the series out as highly promising for
further investigation, but a significant risk of hERG inhibition and a lower-than-desired in vivo
tolerability have combined to lead us to park the series. There may be ways to engineer around these
series liabilities and members of the broader drug discovery community may discover such solutions.
To encourage collaboration with an open source research mechanism, the data and related materials
for the series are posted on a community resource created for this purpose,20 adopting principles that
have driven open source research mechanisms in other disease areas.17c Any scientist interested in
furthering this series, or adopting an open approach to other series, is encouraged to do so using this
free online infrastructure.
20
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 44
1
2
3
4
5
6
EXPERIMENTAL SECTION
7
8
9
We can confirm that all animal studies were ethically reviewed and carried out in accordance with
European Directive 2010/63/EU and the GSK Policy on the Care, Welfare and Treatment of Animals.
Compound synthesis. All commercially available reagents and solvents were used without further
purification. High performance liquid chromatography (HPLC) was used to determine the purity of
the compounds synthesized. The data confirmed that the target compounds generally had ≥95% of
purity with the exception of compounds 14 (90%), 18 (91%), 23 (90%), 36 (90%) and 47 (85%).
When reactions were performed under microwave conditions, a Biotage Initiator was used. The yields
refer to the purified products, and they were not optimized. Reactions were monitored by TLC on
silica gel with detection by UV light (254 nm) and/or stained with the relevant reagent. TLC analysis
was performed using Merck Silica Gel 60 F254 precoated aluminium plates (0.2 mm) or Polygram
precoated silica gel TLC sheets SIL G/UV254. When necessary, crude reaction products were purified
by using flash chromatography, ion exchange purifications and/or preparative HPLC. Flash
purifications were usually performed on a Biotage ISOLERA One flash system equipped with an
internal variable dual wavelength diode array detector (200−400 nm). SNAP cartridges (10−100 g,
flow rate of 10−100 mL/ min) were used. Dry sample loading was done by selfpacking samplet
cartridges using silica or Celite 545. Gradients used varied for each purification. However, typical
gradients used for normal phase were gradients of 0−100% ethyl acetate in cyclohexane or 0−15%
methanol in CH2Cl2. For Ion-exhange purifications, SOLUTE® SCX-2 Ion-exchange columns (1–50
g) were used with elutions of MeOH followed by 7 N NH3 in MeOH. Purification using prep-HPLC
was accomplished using a Gilson HPLC system with a Gemini 5 μm C18 column (150 mm × 21.2
mm i.d.). Typically, the mobile phase used for HPLC was a linear gradient of water (A) and
acetonitrile (B). The water and acetonitrile were mixed with 0.1% TFA or 0.1% HCOOH. The flow
rate was maintained at 0.8−1.2 mL/min for analytic HPLC and 20 mL/min for prep-HPLC, and the
eluent was monitored with UV detector at 220 nm. Characterization of all final compounds was
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
1
performed using H NMR spectroscopy and mass spectrometry. H NMR spectra were recorded at
300K on Bruker spectrometers: either an AVANCE 300 (1H at 300 MHz), AVANCE 200 (1H at 200
21
ACS Paragon Plus Environment

Page 23 of 44
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
MHz) or Ultrashield DPX 400 spectrometer (1H at 400 MHz). Chemical shifts (δ) are given in ppm
relative to the solvent reference as an internal standard (d6-DMSO, δ = 2.50 ppm; CDCl3, δ = 7.27
ppm; CD3OD, δ = 3.31 ppm). Data are reported as follows: chemical shift (multiplicity (s for singlet,
d for doublet, t for triplet, m for multiplet, br for broad), integration, coupling constant(s) in Hz. Low-
resolution mass spectrometry (m/z) was carried out on a Finnigan quadrupole ion trap mass
spectrometer using electrospray ionisation (ESI) or atmospheric-pressure chemical ionisation (APCI)
or Agilent LC−MS 1200 with 6110 MS detector. Positive and negative detection is indicated by the
charge of the ion, e.g. [M+H]+ indicates positive ion detection. The purities of the final compounds
were ≥95% unless stated otherwise, as determined by Agilent 1100 instrument equipped with a
Sunfire C18 column (30 mm x 2.1 mm i.d., 3.5 mm packing diameter) at 40 ºC coupled with a Waters
ZMD2000 mass spectrometer; the method of ionization was alternate-scan positive and negative
electrospray. The following section comprises the synthetic procedures for some representative
intermediates and several products reported in this publication. Complementary data for the rest of
intermediates and final compounds can be found in the Supporting Information.
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
General Procedure A. Reductive amination
A solution of secondary amine (1 equiv) and the appropriate aldehyde (1–1.1 equiv) in CH2Cl2 or
DCE (molarity 0.05–0.14 M) was stirred at rt for 1–4 h. Sodium triacetoxyborohydride or solid-
supported Amberlite® IRA-900 cyanoborohydride (1–6 equiv) was added and the reaction was stirred
at room temperature or 40–100 ºC for 1.5–48 h. After allowing to cool, the reaction was quenched by
addition of saturated NaHCO3 solution and extracted with CH2Cl2, the organic layer was dried
(Na2SO4), filtered and concentrated to afford a crude mixture that was purified by column
chromatography and/or preparative HPLC to give the corresponding tertiary amine.
General Procedure B. Reductive amination
A solution of secondary amine (1 equiv), the appropriate aldehyde (1–1.1 equiv), Biotage® MP-
Cyanoborohydride (1.7 equiv) or solid-supported Amberlite® IRA-900 cyanoborohydride (1–1.7
equiv) and acetic acid (4 equiv) in CH2Cl2 (molarity 0.06–0.16 M) was subjected to MW irradiation
for 40–180 min at 100 ºC. The reaction was filtered, quenched by addition of saturated NaHCO3
solution and extracted with CH2Cl2. The organic layer was dried (Na2SO4), filtered and concentrated
22
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 44
1
2
3
4
5
6
to afford a crude mixture that was purified by column chromatography or preparative HPLC to give
the corresponding tertiary amine.
7
8
1-(4-Phenoxybenzyl)-6',7'-dihydrospiro[piperidine-4,4'-thieno[3,2-c]pyran] (27)
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
The title compound was prepared using General Procedure A. White solid, yield 80% (600 mg, 1.532
1
mmol). H NMR (400 MHz, CDCl3) δ ppm: 7.36–7.31 (m, 4H), 7.12–7.07 (m, 2H), 7.04–6.94 (m,
4H), 6.83 (d, 1H, J=5.3), 3.95–3.93 (m, 2H), 3.54 (s, 2H), 2.85–2.82 (m, 2H), 2.76–2.73 (m, 2H),
2.43–2.38 (m, 2H), 2.03–1.95 (m, 2H), 1.89–1.84 (m, 2H). [ES+ MS] m/z 392 (M+H)+.
(4-(4-((6',7'-Dihydrospiro[piperidine-4,4'-thieno[3,2-c]pyran]-1-
yl)methyl)phenoxy)phenyl)methanol, hydrochloride (28)
The compound as a free amine was obtained using General Procedure A with 28a (see Supporting
Information) as colorless oil. The compound as a free amine (11.8 mg, 0.028 mmol) was dissolved in
CH2Cl2 (1.0 mL) and 1 M HCl in Et2O (0.034 mL, 0.034 mmol) was added. After stirring at rt for 10
min, the solvent was evaporated to obtain the hydrochloride salt of the compound. White solid, yield
10% (12.8 mg, 0.028 mmol). 1H NMR (400 MHz, CDCl3+CD3OD) δ ppm: 7.40–7.38 (m, 2H), 7.21–
7.19 (m, 2H), 6.96 (d, 1H, J=5.3), 6.86–6.84 (m, 4H), 6.75 (d, 1H, J=5.1), 4.45 (s, 2H), 4.03 (s, 2H),
3.20–3.14 (m, 4H), 3.06–3.00 (m, 2H), 2.68–2.66 (m, 2H), 2.44–2.37 (m, 2H), 1.87–1.83 (m, 2H).
[ES+ MS] m/z 422 (M+H)+.
4-Bromo-N-(4-((6',7'-dihydrospiro[piperidine-4,4'-thieno[3,2-c]pyran]-1-
yl)methyl)phenyl)aniline, hydrochloride (30)
The compound as a free amine was obtained using the general procedure B with 30b (see Supporting
Information) as colorless oil. The compound as a free amine (480 mg, 1.02 mmol) was dissolved in
CH2Cl2 (3 mL) and treated with 2 M HCl in Et2O (0.51 mL, 1.02 mmol). The precipitate was filtered
and washed with CH2Cl2, hexane and acetonitrile to afford the title compound as a pink solid, yield
1
59% (446 mg, 0.873 mmol). H NMR (400 MHz, DMSO-d6) δ ppm: 10.49 (bs, 1H), 8.59 (s, 1H),
7.47–7.36 (m, 4H), 7.36 (d, 1H, J=5.1), 7.12–7.05 (m, 4H), 6.77 (d, 1H, J=5.3), 4.24 (bd, 2H, J=5.1),
3.88 (bt, 2H, J=5.1), 3.25–3.22 (m, 2H), 3.14–3.06 (m, 2H), 2.78 (bt, 2H, J=5.1), 2.38–2.30 (m, 2H),
1.98–1.95 (m, 2H). [ES+ MS] m/z 469 (M+H)+.
23
ACS Paragon Plus Environment

Page 25 of 44
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
9-(4-(4-(Trifluoromethoxy)phenoxy)benzyl)-1-oxa-9-azaspiro[5.5]undecane, hydrochloride (42)
The compound as a free amine (91 mg, 0.218 mmol) was obtained using General Procedure A and 58
as a pale yellow oil. The compound as a free amine was dissolved in CH2Cl2 (5 mL) and 1 M HCl in
Et2O (0.218 mL, 0.218 mmol), the solvent was evaporated to obtain the hydrochloride salt of 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
1
compound. White solid, yield 47% (100 mg, 0.218 mmol). H NMR (400 MHz, DMSO-d6) δ ppm:
10.41 (bs, 1H), 7.71 (d, 2H, J=8.5), 7.50 (d, 2H, J=8.5), 7.23 (d, 2H, J=9.0), 7.19 (d, 2H, J=9.0), 4.35
(d, 2H, J=5.3), 3.64–3.62 (m, 2H), 3.24–3.19 (m, 2H), 3.10–3.01 (m, 2H), 2.16–2.13 (m, 2H) 1.81–
1.74 (m, 2H), 1.64–1.62 (m, 2H), 1.52–1.50 (m, 2H), 1.47–1.44 (m, 2H). [ES+ MS] m/z 422 (M+H)+.
9-(4-(tert-Butyl)benzyl)-1-oxa-9-azaspiro[5.5]undecane, hydrochloride (50)
The compound as a free amine (53.7 mg, 0.178 mmol) was obtained using General Procedure A with
58. The compound was dissolved in CH2Cl2 (5 mL) and 1 M HCl in Et2O (178 μL, 0.178 mmol). The
solvent was evaporated to obtain the hydrochloride salt of the compound. White solid, yield 38% (60
1
mg, 0.178 mmol). H NMR (400 MHz, DMSO-d6) δ ppm: 10.30 (bs, 1H), 7.59 (d, 2H, J=8.5), 7.52
(d, 2H, J=8.5), 4.31 (d, 2H, J=5.3), 3.63–3.61 (m, 2H), 3.21–3.17 (m, 2H), 3.09–3.00 (m, 2H), 2.15–
2.12 (m, 2H) 1.80–1.72 (m, 2H), 1.65–1.62 (m, 2H), 1.51–1.49 (m, 2H), 1.46–1.42 (m, 2H), 1.36 (s,
9H). [ES+ MS] m/z 302 (M+H)+.
tert-Butyl 4-allyl-4-hydroxypiperidine-1-carboxylate (55)
Allylmagnesium bromide in Et2O (1 M, 29.1 mL, 29.1 mmol) was slowly added to a solution of tert-
butyl 4-oxopiperidine-1-carboxylate (5.0 g, 25.1 mmol) in Et2O (80 mL) at 0ºC. The reaction mixture
was allowed to warm to rt overnight. The reaction was quenched by addition of saturated NH4Cl
solution and the organic layer was separated, dried (Na2SO4), filtered and evaporated to obtain a
yellow oil (4.9 g). A fraction of crude (2 g) was purified by flash chromatography to obtain the title
compound. Pale yellow oil, yield 45% (1.1 g, 4.56 mmol). 1H NMR (400 MHz, CDCl3) δ ppm: 5.92–
5.81 (m, 1H), 5.23–5.13 (m, 2H), 3.82–3.77 (m, 2H), 3.17–3.14 (m, 2H), 2.24 (d, 2H, J=7.5), 1.56–
1.52 (m, 4H), 1.46 (s, 9H).
tert-Butyl 4-allyl-4-(allyloxy)piperidine-1-carboxylate (56)
Allylbromide (1.082 mL, 12.43 mmol) was added to a solution of 55 (1.0 g, 4.1 mmol) in DMF (20
mL) at 0ºC. After 5 min, NaH (0.167 g, 6.63 mmol) was added. The reaction mixture was allowed to
24
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 44
1
2
3
4
5
6
7
8
warm to rt and stirred for 48 h. The reaction mixture was portioned between water and TBME. The
organic layer was separated, washed with water (3  20 mL), dried (MgSO4) and concentrated. The
crude product was purified by flash chromatography to afford the title compound. Pale yellow oil,
9
1
yield 67% (780 mg, 2.77 mmol). H NMR (400 MHz, CDCl3) δ ppm: 5.98–5.88 (m, 1H), 5.86–5.75
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
(m, 1H), 5.34–5.28 (m, 1H), 5.17–5.05 (m, 3H), 3.97–3.89 (m, 2H), 3.79–3.77 (m, 2H), 3.11–3.07 (m,
2H), 2.28 (d, 2H, J=6.9), 1.79–1.75 (m, 2H), 1.46 (s, 9H), 1.43–1.40 (m, 2H).
tert-Butyl 1-oxa-9-azaspiro[5.5]undec-3-ene-9-carboxylate (57)
A solution of Grubbs I (114 mg, 0.139 mmol) in CH2Cl2 (10 mL) under argon atmosphere was slowly
added to a solution of 56 (780 mg, 2.77 mmol) in CH2Cl2 (15 mL) at 0 ºC. The reaction mixture was
allowed to warm to rt and stirred overnight. The reaction was concentrated and the crude was purified
by flash chromatography to afford the title compound. Pale brown oil, yield 95% (670 mg, 2.64
1
mmol). H NMR (400 MHz, CDCl3) δ ppm: 5.78–5.68 (m, 2H), 4.12–4.10 (m, 2H), 3.79–3.73 (m,
2H), 3.19–3.17 (m, 2H), 1.98–1.96 (m, 2H), 1.81–1.78 (m, 2H), 1.46 (s, 9H), 1.43–1.40 (m, 2H).
1-Oxa-9-azaspiro[5.5]undecane, hydrochloride (58)
Pd/C (281 mg, 0.264 mmol) was added to a solution of 57 (670 mg, 2.64 mmol) in MeOH (15 mL).
The reaction was hydrogenated at 2.5 bar for 2 h. The reaction was filtered over Celite® and
concentrated to afford the free amine. Off-white solid, yield 97% (658 mg, 2.58 mmol). 1H NMR (400
MHz, CDCl3) δ ppm: 3.73–3.70 (m, 2H), 3.66–3.63 (m, 2H), 3.50–3.49 (m, 2H), 3.15–3.12 (m, 2H),
1.86–1.83 (m, 2H), 1.65–1.52 (m, 6H), 1.46 (s, 9H). To a solution of the free amine (658 mg, 2.58
mmol) in CH2Cl2 (15 mL) 3 M HCl in MeOH (6.87 mL, 20.61 mmol) was added. The reaction was
stirred at rt for 2 h. The crude reaction was concentrated to afford the 1-oxa-9-azaspiro[5.5]undecane,
1
hydrochloride. Off-white solid, yield 100% (514 mg, 2.58 mmol). H NMR (400 MHz, DMSO-d6) δ
ppm: 8.96–8.83 (m, 2H), 3.13–3.09 (m, 2H), 3.02–2.93 (m, 2H), 1.69–1.61 (m, 5H), 1.54–1.45 (m,
6H).
Determination of MIC for M. tuberculosis. Mycobacterium tuberculosis H37Rv was grown at 37 ºC
in Middlebrook 7H9 broth (Difco) supplemented with 0.025% Tween 80 and 10% albumin-dextrose-
catalase (ADC) or on Middlebrook 7H10 plates supplemented with 10% oleic acid-albumin-dextrose-
25
ACS Paragon Plus Environment

Page 27 of 44
Journal of Medicinal Chemistry
1
2
3
4
5
6
catalase (OADC). The antitubercular activity against extracellular or intracellular Mycobacterium
strains was performed as previously described.21
7
8
9
Microsomal Fraction Stability. Pooled mouse and human liver microsomes were purchased from
Xenotech. Microsomes (final protein concentration 0.5 mg/mL, 5 mM MgCl2) and test compound
(final substrate concentration 0.5 µM; final DMSO concentration 0.5 %) in 0.1 M phosphate buffer
pH 7.4 were pre-incubated at 37 °C prior to the addition of NADPH (final concentration 1 mM) to
initiate the reaction. The final incubation volume was 600 µL. Control samples were included for each
compound tested where 0.1 M phosphate buffer pH 7.4 was added instead of NADPH (minus
NADPH). Midazolan was included as control in every experiment. Each compound was incubated for
30 min and samples (90 µL) were taken at 0, 5, 10, 20 and 30 min. The minus NADPH control was
sampled at 0 and 30 min only. The reactions were stopped by the addition of 200 µL of
acetonitrile:methanol (3:1) containing an internal standard, followed by centrifugation at 3700 rpm for
15 min at 4 °C to precipitate the protein. Quantitative analysis was performed using specific LC-
MS/MS conditions. Data analysis: from a plot of ln peak area ratio (compound peak area/internal
standard peak area) against time, the gradient of the line was determined. Subsequently, half-life and
intrinsic clearance were calculated using the equations below:
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
Elimination rate constant (k) = (- gradient)
0.693
k
Half life (t1/2) (min) =
V 0.693
t1 / 2
Intrinsic Clearance (CLint) (mL/min/g protein) =
where V=Incubation volume mL/g microsomal protein.
Procedure for PK Studies. For pharmacokinetic studies, C57BL/6 female mice (18–20 g) were used
and compound concentrations were determined using peripheral whole blood.
All compounds were administered by intravenous route at 4 mg/kg single dose in 5% DMSO/20%
Encapsin and by oral gavage at 50 mg/kg single dose in 1% methyl cellulose (1% MC) or
PEG400/Solutol 70:30 (compound 1). For i.v. route, aliquots of 25 μL of blood were taken from the
lateral tail vein by puncture from each mouse (n =5) at 5, 15 and 30 minutes, 1, 2, 4 and 8 hours post-
26
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 44
1
2
3
4
5
6
7
8
9
dose; for oral route aliquots of 25 μL of blood were taken from the lateral tail vein by puncture from
each mouse (n = 5) at 15, 30, and 45 minutes, 1, 2, 4, 8 and 24 hours post-dose. LC-MS/MS was used
as the analytical method for the establishment of compound concentration in blood. Pharmacokinetic
analysis was performed by non-compartmental data analysis (NCA) with Phoenix WinNonlin 6.3
(Pharsight, Certara L.P) and supplementary analysis was performed with GraphPad Prism 6
(GraphPad Software, Inc).
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
Murine Model of Acute TB Infection. Specific pathogen-free, 8–10 week-old female C57BL/6 mice
were purchased from Harlan Laboratories and were allowed to acclimate for one week. The
experimental design for the acute assay has been previously described.22 In brief, mice were
intratracheally infected with 100,000 CFU/mouse of M. tuberculosis H37Rv. Compounds were
administered for 8 consecutive days starting one day after infection. One mouse was administered per
product and dose. Lungs were harvested 24 hours after the last administration and all lung lobes were
aseptically removed, homogenized and frozen. Homogenates were plated onto 10% OADC–7H11
medium supplemented with activated charcoal (0.4%) and incubated for 18 days at 37 °C. The viable
CFU were converted to Log10, which were then fitted to a sigmoidal dose/response curve (GraphPad
Prism Software).
Cell Health. The Cell Health (CH) assay is a 3-parameter automated imaging cell-based assay to
measure the cytotoxic effect of compounds in human liver derived HepG2 cells.23,24 In brief, the key
parameters measured in this assay using fluorescent staining are nuclear size, mitochondrial
membrane potential and plasma membrane permeability. HepG2 cells (Biocat #97134) were
incubated with the test compounds in 384-well plates, and after 48 h the staining cocktail was added:
Hoechst 33342 is used to stain nuclei and quantify changes in nuclear morphology; TMRM is a
cationic dye that accumulates in healthy mitochondria that maintain a mitochondrial membrane
potential. This dye leaks out of the mitochondria when the mitochondrial membrane potential is
dissipated; TOTO-3 iodide labels nuclei of permeabilized cells and is used to measure plasma
membrane permeability.
27
ACS Paragon Plus Environment

Page 29 of 44
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
Following 45 min incubation, the plate was sealed for reading on INCell 2000 (GE Healthcare). Each
parameter produced a % of cells which were 'LIVE' or 'DEAD'. Analysis of data showed the IC50, the
compound concentration that produced 50% of inhibition of growth of cells.
9
Hydrophobicity Assay (ChromlogD). 10 mL of 10 mM DMSO stock solutions were diluted to 750
mL with octanol saturated phosphate buffer pH 7.4 and 160 mL buffer saturated octanol in a 96-well
deep well block. Blocks were sealed and inverted for 3 sets of 50 inversions, then centrifuged at 300 g
for 20 min. Both phases were then quantified using generic gradient UV-HPLC.
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
Chemiluminescent Nitrogen Detection (CLND) Solubility. 5 mL of 10 mM DMSO stock solution
was diluted to 100 mL with phosphate-buffered saline, pH 7.4, equilibrated for 1 h at room
temperature, and filtered through Millipore Multiscreen HTS-PCF filter plates (MSSL BPC). The
eluent was quantified by suitably calibrated flow injection CLND.
Cardiovascular Toxicity. In silico prediction of hERG inhibition in vitro was performed with DereK
Nexus v.5.0.2 (Lhasa Ltd). Based on cardiovascular progression in GSK, ion channel inhibition assay
was carried out through automated planar chip electrophysiology platforms, IonWorks25 and Qpatch.26
of high and medium throughput, respectively. The whole cell voltage clamp technique directly
measures the effect of drugs/compounds on the hERG channel by electrophysiological method on the
IonWorks Barracuda and Qpatch platform, respectively. This assay generates a full compound dose-
response curve. CHO S1 hERG cell line (BIOCAT109040) expressing KCNH2 (hERG ether-a-go-go-
related potassium channel protein) were plated in a Patchclamp plate with a 4-fold 6 concentration
points dose response curve. Electrophysiology in an IonWorks Barracuda system (Molecular Devices)
or Qpatch 48HTX (Sophion Bioscience, Copenhagen) was measured. Data for each well for each
compound was transformed to a fraction of the baseline current and a pIC50 was obtained.
Subsequently ex vivo assay was performed to confirm the risk flagged from in vitro ion channel
assays. Surgical preparation of the rabbit left ventricular wedge has been described in detail
previously.19 Briefly, female New Zealand white rabbit was anaesthetized and the heart removed and
immediately cannulated and perfused with cold cardioplegic solution (24 mM potassium (K+),
buffered with 95% O2/5% CO2). A transmural wedge was dissected from the left ventricle. The
preparation was placed in a tissue bath and arterially perfused with Tyrode’s solution (4 mM K+,
28
ACS Paragon Plus Environment