Can We Make Small Molecules Lean? Optimization of a Highly Lipophilic TarO Inhibitor

Article abstract of DOI:10.1021/acs.jmedchem.7b00113

We describe our optimization efforts to improve the physicochemical properties, solubility, and off-target profile of 1, an inhibitor of TarO, an early stage enzyme in the biosynthetic pathway for wall teichoic acid (WTA) synthesis. Compound 1 displayed a TarO IC₅₀ of 125 nM in an enzyme assay and possessed very high lipophilicity (clogP = 7.1) with no measurable solubility in PBS buffer. Structure-activity relationship (SAR) studies resulted in a series of compounds with improved lipophilic ligand efficiency (LLE) consistent with the reduction of clogP. From these efforts, analog 9 was selected for our initial in vivo study, which in combination with subefficacious dose of imipenem (IPM) robustly lowered the bacterial burden in a neutropenic Staphylococci murine infection model. Concurrent with our in vivo optimization effort using 9, we further improved LLE as exemplified by a much more druglike analog 26.

Full text of DOI:10.1021/acs.jmedchem.7b00113

Subscriber access provided by UNIV OF CALIFORNIA SAN DIEGO LIBRARIES
Article
Can We Make Small Molecules Lean?-
Optimization of a Highly Lipophilic TarO Inhibitor
Mihir Baran Mandal, Zheng Tan, Christina Madsen-Duggan, Alexei V. Buevich, John P. Caldwell,
Reynalda Dejesus, Amy Flattery, Charles G. Garlisi, Charles Gill, Sookhee Nicole Ha, Ginny Ho,
Sandra Koseoglu, Marc Labroli, Kallol Basu, Sang Ho Lee, Lianzhu Liang, Jenny Liu, Todd W.
Mayhood, Debra McGuinness, David G. McLaren, Xiujuan Wen, Emma R. Parmee, Diane Rindgen,
Terry Roemer, Payal R Sheth, Paul Tawa, James R. Tata, Christine Yang, Shu-Wei Yang, Li
Xiao, Hao Wang, Christopher Tan, Haifeng Tang, Paul Walsh, Erika Walsh, Jin Wu, and Jing Su
J. Med. Chem., Just Accepted Manuscript • Publication Date (Web): 21 Mar 2017
Downloaded from http://pubs.acs.org on March 21, 2017
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 52
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
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 52
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
Can We Make Small Molecules Lean?ꢀ Optimization of a
Highly Lipophilic TarO Inhibitor
*‡
‡
‡
‡
Mihirbaran Mandal, Zheng Tan, Christina MadsenꢀDuggan, Alexei V. Buevich, John P.
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
‡
‡
Δ
∥
∥
Caldwell, Reynalda Dejesus, Amy Flattery, Charles G. Garlisi, Charles Gill, Sookhee Nicole
‡
‡
Δ
‡
‡
§
∥
Ha, Ginny Ho, Sandra Koseoglu, Marc Labroli, Kallol Basu, Sang Ho Lee, Lianzhu Liang,
Δ
Δ
Δ
Δ
∥
Jenny Liu, Todd Mayhood, Debra McGuinness, David G. McLaren, Xiujuan Wen, Emma
‡
#
§
Δ
Δ
‡
Parmee, Diane Rindgen, Terry Roemer, Payal Sheth, Paul Tawa, James Tata, Christine
‡
‡
‡
§
§
‡
⊥
Yang, ShuꢀWei Yang, Li Xiao, Hao Wang, Christopher Tan, Haifeng Tang, Paul Walsh,
#
*‡
⊥
Erika Walsh, Jin Wu, and Jing Su
‡
#
§
⊥
Global Chemistry, Early Discovery Bacteriology, Discovery Pharmaceutical Sciences,
ꢁ
∥
Pharmacokinetics, Pharmacodynamics, and Drug Metabolism, In Vivo Pharmacology, In vitro
Pharmacology
MRL, Merck & Co., Inc., Kenilworth, NJ 07033 USA
Abstract
We describe our optimization efforts to improve the physicochemical properties, solubility, and
offꢀtarget profile of 1, an inhibitor of TarO ꢀ an early stage enzyme in the biosynthetic pathway
for wall teichoic acid (WTA) synthesis. Compound 1 displayed a TarO IC of 125 nM in an
5
0
enzyme assay, and possessed very high lipophilicity (clogP 7.1) with no measureable solubility
in PBS buffer. Structure activity relationship (SAR) studies resulted in a series of compounds
with improved lipophilic ligand efficiency (LLE) consistent with the reduction of clogP. From
ACS Paragon Plus Environment

Page 3 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
these efforts, analog 9 was selected for our initial in vivo study, which in combination with subꢀ
efficacious dose of imipenem (IPM), robustly lowered the bacterial burden in a neutropenic
Staphylococci murine infection model. Concurrent with our in vivo optimization effort using 9,
we further improved LLE as exemplified by a much more drugꢀlike analog 26.
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
Introduction
After the discovery of penicillin by Alexander Fleming from Penicillium notatum in
1
1
929, history witnessed tremendous efforts poured into the antibacterial research, which led to
2
the golden age of antibiotic discovery (1940sꢀ1960s), and revolutionized the clinical practice.
3
4
Although several important classes of new drugs such as carbapenem, oxazolidinone,
5
6
monobactam and daptomycin were discovered during the 1970ꢀ1990 period, the next 25 years
of antibacterial research has been, in general, disappointing in delivering novel antibiotics for
7
8
various reasons. Additionally, antibiotic resistance, warned by Fleming in his Nobel Prize
lecture in 1945, has been rising steadily, including the most recent astonishing reports on colistin
9
resistance in animals as well as in humans. The emergence of bacterial resistance has been
considered as one of the top threats to the global public health, and has prompted governments
1
0
and World Health Organization to take actions to address this potential global crisis.
8
,11
The resistance issue has been studied extensively and reviewed.
Resistance
mechanisms range from endogenous mechanisms (such as cognate on target mutations, reduced
uptake by loss of porins, overexpression of βꢀlactamases and efflux pumps) to exogenous
1
2
mechanisms such as horizontal gene transfer. To overcome drug resistance by combining two
drugs with different mechanisms of action have been extensively applied for the treatment of
gramꢀnegative infection, and the βꢀlactam/βꢀlactamase combination therapy as exemplified by
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 52
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
amoxicillin/clavulanate developed by the Beecham Research Laboratories to address the stability
13
issue of amoxicillin represents a seminal achievement in this field. The most recent successful
1
4
15
examples are the FDA approved ceftazidime/avibactam and ceftolozane/tazobactam.
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
Combination therapy other than βꢀlactam/βꢀlactamase therapies are also being investigated at
1
1a
clinical or preꢀclinical stages.
For example, Cadazolid, a hybrid of fluoroquinolone and
oxazolidinone in one molecule, is in Phase III clinical trials, and has received “qualified
1
6
infectious disease product” (QIDP) and “Fast Track” development status from the FDA.
In our own research on methicillin resistant staphylococcus aureus (MRSA) resistance to
βꢀlactam antibiotics, an antisenseꢀbased screening of 245 MRSA essential genes revealed that
certain genes, such as TarL (an essential enzyme involved in wall teichoic acid (WTA)
1
7
biosynthesis), when partially depleted, led to resensitization of MRSA to existing βꢀlactams. In
accordance with the importance of WTA pathway in resensitizing MRSA to βꢀlactams, Campbell
et al. reported that the deletion of TarO (the first step in the WTA pathway) also resensitized
1
8
MRSA to βꢀlactams. Our subsequent WTA pathwayꢀbased whole cell screening resulted in the
19
discovery of several inhibitors targeting the WTA transporter protein TarG. Resistant mutant
selection in MRSA and MRSE and subsequent wholeꢀgenome sequence analysis indicated that in
addition to nonꢀsynonymous mutations that confer resistance due to the loss of binding of TarG
inhibitors to its target, loss of function mutations mapping to TarB, TarD, TarI, as well as earlyꢀ
2
0
stage genes TarO and TarA were identified. Loss of function mutations in TarO or TarA do
2
0
not play a role in the viability of Staphylococci since WTA is a nonꢀessential polymer.
However, the absence of WTA greatly enhances the susceptibility of MRSA to βꢀlactam
2
1
antibiotics in vitro as well as in vivo.
Analogous to the βꢀlactam/βꢀlactamase combination
therapy, as described above, finding an inhibitor that targets a nonꢀessential pathway, yet is
ACS Paragon Plus Environment

Page 5 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
synergistic with βꢀlactam antibiotics, could be a viable approach to address drug resistance for
2
1
Gramꢀpositive MRSA infection. This synergistic characteristic of the WTA pathway with βꢀ
lactam antibiotics inspired us to design a targeted phenotypic screening to identify nonꢀbioactive
agents that specifically inhibit the early steps in the WTA biosynthetic pathway, such as TarO or
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
2
TarA.
Thorough interrogation of our small molecule library resulted in the discovery of
Tarocin (1) and Tarocin B (structure in the supporting information) as specific inhibitors of TarO
21
in MRSA.
Figure 1. The physicochemical properties of hit 1 and its profile. The structure of imipenem
IPM), one of the putative antibiotics partner. MITC95 – synergistic minimum inhibitory
threshold concentration to inhibit 95% growth of methicillinꢀresistant Staphylococcal aureus
MRSA) COL in presence of 4 µg/mL of IPM. MITC95 – intrinsic minimum inhibitory
threshold concentration to inhibit 95% growth of MRSA COL. LLE: lipophilic ligand
efficiency.
S
(
i
(
Unlike tunicamycin, a known inhibitor of both TarO and MraY, which binds in the polar
substrateꢀbinding site in the intracellular domain, compound 1 and Tarocin B were predicted to
bind in the extracellularly exposed hydrophobic pocket, the putative binding domain of
2
2
hydrophobic bactoprenol substrate.
These hits possessed no intrinsic antibacterial activity
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 52
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
i
(
MITC >200 µM, minimal concentration to inhibit 95% of the growth as measured by OD ),
9
5
600
but rescued the bactericidal activity of IPM (MIC 32 µg/mL) to its clinical break point
concentration (4 µg/mL) against IPM resistant MRSA COL at concentrations of 0.9 µM (0.4
µg/mL) and 1.7 µM (0.9 µg/mL), respectively. Unfortunately, both compound 1 and Tarocin B
were highly lipophilic, and would require significant optimization to identify a tool molecule for
in vivo efficacy study. Medicinal chemistry efforts were directed to improve the LLE with
concomitant improvement in physicochemical properties while maintaining the synergistic
antibacterial effect observed when combined with IPM at its breakpoint concentration.
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
We have recently communicated our efforts on the progression of Tarocin B towards
23
lead like molecules.
Herein, we describe our optimization efforts to improve the
2
4
physicochemical properties and aqueous solubility of 1 utilizing a “hydrogenꢀbond hypothesis”
that resulted in a series of molecules with improved LLE, and a tool compound 9 that
demonstrated in vivo efficacy in a mouse infection model. Subsequent medicinal chemistry
efforts led to a more polar compound 26 (clogP 1.9). This example of turning a highly lipophilic
compound to an antibiotic like lead by lowering clogP by 5 log units demonstrated the power of
the hypothesis driven approach. Details of this medicinal chemistry effort and the in vivo results
associated with 9 will be the central focus in this report.
Chemistry
The synthesis of 1, 2, and analogs 3ꢀ5 with additional methylene spacers between the
exocyclic carbonyl and the naphthalene moiety is described in Scheme 1. Condensation of the
2
5
key intermediate 30, prepared following a known protocol, with 2ꢀ(naphthalenꢀ1ꢀyl)acetic acid
using 1,1'ꢀcarbonyldiimidazole as the activating agent afforded 1. Analogs 2ꢀ5 were prepared
accordingly from the corresponding acetic acid derivatives.
ACS Paragon Plus Environment

Page 7 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Scheme 1. Synthesis of Analogs 1ꢀ5 with an Exocyclic Amide Linkage
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) iꢀPrMgCl, THF, ꢀ10 °C, 95%; (b) toluene, 2ꢀpropanol, Al(OiPr)3,
5
0 °C, then KOH, H O, rt, 91%; (c) 2ꢀ(naphthalenꢀ1ꢀyl)acetic acid, 1,1'ꢀcarbonyldiimidazole,
2
DBU, MeCN, rt, 12ꢀ56%.
The synthesis of urea derivatives 6ꢀ11 from the corresponding oxazolidinꢀ2ꢀones (30ꢀ33)
is shown in Scheme 2. As a representative example, treatment of oxazolidinone 30 with nꢀBuLi
at ꢀ78 °C followed by quenching with trichloromethyl carbonochloridate generated the
corresponding oxazolidinoneꢀ3ꢀcarbonyl chloride, which upon treatment with isoquinolinꢀ4ꢀ
ylmethanamine afforded 6 in 20% yield. The other analogs 7ꢀ11 were prepared using the similar
protocol from the corresponding oxazolidinones (31ꢀ33) and the appropriate amines. While this
strategy proved general for most analogs, alkoxy substituted derivatives such as 14 required the
development of a slightly different synthetic strategy as illustrated in Scheme 3. Bisꢀbromide 35,
prepared readily from commercially available 2,6ꢀdichloroꢀ4ꢀ(trifluoromethyl)pyridine (34), was
converted to monoꢀalkoxy derivative 36 in 70% yield. Metalꢀhalogen exchange followed by
treatment with Weinreb amide 28 afforded a ketone, which upon stereocontrolled reduction
employing MeerweinꢀPonndorfꢀVerley conditions led to the formation of alcohol 37 in 53%
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 52
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
yield. Oxazolidinone formation employing KHMDS followed by several functional group
manipulations afforded the key intermediate 42 in 22% yield. Installation of urea fragment
employing similar protocol shown in Scheme 2 produced the desired analog 14 in 18% yield.
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
Scheme 2. Synthesis of Urea Derivatives 6ꢀ11
Reagents and conditions: (a) nꢀBuLi, THF, trichloromethyl carbonochloridate, ꢀ78 °C; (b)
amine, Et N, DCM, rt, 20ꢀ50%.
3
Scheme 3. Synthesis of Analog 14
ACS Paragon Plus Environment

Page 9 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Reagents and conditions: (a) 30% HBr in acetic acid, 100 °C, 70%; (b) propꢀ2ꢀenꢀ1ꢀol, NaH,
THF, 0 °C, 81%; (c) nꢀBuLi, THF, ꢀ78 °C, 53%; (d) toluene, 2ꢀpropanol, Al(OiPr) , 50 °C,
3
1
00%; (e) KHMDS, THF, ꢀ20 °C, 47%; (f) (Boc) O, DMAP (cat), rt, 98%; (g) Pd(PPh ) , 1,3ꢀ
2 3 4
dimethylpyrimidineꢀ2,4,6(1H,3H,5H)ꢀtrione, DCM, rt, 97%; (h) CSF, DMF, iodomethane, rt,
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
50%; (i) CF CO H, DCM, rt, (j) nꢀBuLi, THF, trichloromethyl carbonochloridate, ꢀ78 °C; (k)
3
2
(
2,6ꢀnaphthyridinꢀ4ꢀyl)methanamine, Et N, DCM, rt, 18%.
3
Urea derivatives 18ꢀ22 possessing a hydroxymethyl or a carboxyl group at the Cꢀ4
1
position (Scheme 4, generalized structure, R = hydroxymethyl or carboxyl group) were
synthesized from the commercially available Nꢀ((benzyloxy)carbonyl)ꢀLꢀserine 43 following the
steps shown in Scheme 4. Ruthenium (III) chloride mediated oxidation of alcohol functionality
of 21 afforded carboxylic acid derivative 22. Utilizing OꢀbenzylꢀNꢀ(tertꢀbutoxycarbonyl)ꢀLꢀ
homoserine 50, one carbon homologation was achieved (Scheme 5) for the preparation of
analogs 23 and 24 following the similar synthetic sequence described in Scheme 4. The analogs
25 and 26 were prepared from 2ꢀbromoꢀ6ꢀmethoxyꢀ4ꢀ(trifluoromethyl)pyridine following the
sequences of reaction described in Scheme 3, and details of the synthesis were incorporated into
supporting information.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 52
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
Scheme 4. Synthesis of Analogs 18ꢀ22
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
.
Reagents and conditions: (a) HNMe(OMe) HCl, NMM, EDCI, DCM, ꢀ5 °C, 95%; (b)
TBDMSCl, imidazole, DCM, rt, 95%; (c) i-PrMgCl, THF, ꢀ5 °C, 93%; (d) toluene, 2ꢀpropanol,
Al(OiPr) , 50 °C, 15%; (e) KOH, H O,, rt, 87%; (f) nꢀBuLi, THF, trichloromethyl
3
2
carbonochloridate, ꢀ78 °C; (g) amine, Et N, DCM, rt, 20ꢀ50%; (h) AcCl, MeOH, 71ꢀ74%; (i)
3
.
3
RuCl H O, NaIO , MeCN, EtOAc, H O, 6%.
2
4
2
ACS Paragon Plus Environment

Page 11 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Scheme 5. Synthesis of Analogs 23 and 24
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) HNMe(OMe) HCl, HOBt, EDCI, DIEA, THF, 84%; (b) 3ꢀBromoꢀ
5
ꢀfluorobenzotrifluoride, i-PrMgCl , 87%; (c) toluene, 2ꢀpropanol, Al(OiPr) , 50 °C, then KOH,
8%; (d) nꢀBuLi, THF, trichloromethyl carbonochloridate, ꢀ78 °C; (e) isoquinolinꢀ4ꢀ
ylmethanamine, Et N, DCM, rt, 37%; (f) pentamethylbenzene, BCl , DCM, 46%; (g)
RuCl .H O, NaIO , MeCN, EtOAc, H O, 24%.
3
7
3
3
3
2
4
2
Results and discussion
TarO antagonism of all the final compounds were evaluated by the liquid
chromatography–mass spectrometry (LCꢀMS) based assay using membrane bound TarO
enzyme. Intrinsic antibacterial activity of all compounds was assessed using MRSA COL as the
i
primary strain, and reported as MITC – intrinsic minimum inhibitory threshold concentration
9
5
to inhibit 95% growth based on OD600 reading. Their ability to potentiate βꢀlactam based
s
antibiotics was measured in the presence of 4 µg/mL of IPM, and reported here as MITC –
9
5
synergistic minimum inhibitory threshold concentration to inhibit 95% growth of methicillinꢀ
resistant Staphylococcal aureus (MRSA) COL.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 52
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
The compounds were also evaluated in secondary assays such as the late stage targocil
suppression assayꢀ the ability of such compounds to reverse the TarG inhibition, and the Phage K
2
2
protection assayꢀ recognition of the TarO enzyme by Phage K to lyse cells. The toxicity of the
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
26
compounds was routinely profiled using the ClickꢀiT EdU assay.
The results from the
secondary assays and in vitro toxicity data were reported in the supporting information.
After the confirmation that 1 was indeed the superior isomer among its
27
enantiomeric/diastereomeric pairs, chemistry efforts were initiated to lower the clogP to align
with our objective to administer the TarO inhibitor via intravenous route in combination with the
antibiotics partner. Additionally, analog 1 displayed poor pharmacokinetic properties. Thus, to
improve the PK profile and to lower clogP, our SAR studies were pursued by replacing the
naphthalene ring with polar heteroaromatics as well as by substituting the bisꢀ
trifluoromethylphenyl with less lipophilic aromatic moieties. Unfortunately, these efforts
resulted in compounds with significant loss of potencies. Only the isoquinoline moiety
demonstrated to be weakly tolerated with 15ꢀfold drop in TarO IC relative to analog 1. To
5
0
overcome this steep SAR issue, several approaches were undertaken, including a linker insertion
and a linker expansion strategy (Table 1). From these efforts, analog 3 with a twoꢀmethylene
spacer between the exocyclic amide carbonyl and naphthalene emerged with 19ꢀfold
improvement in TarO IC , and further lengthening the linker resulted in a decrease in potency
5
0
(analog 4). With this optimized linker, the SAR at both the naphthalene and bisꢀ
trifluoromethylphenyl regions was revisited. Soon it was discovered that either the insertion of a
polar atom into the naphthalene region or the replacement of the highly lipophilic bisꢀ
trifluoromethylphenyl with less hydrophobic aromatic moieties were tolerated in terms of TarO
potency, but concurrent reduction in polarity at the both sides led to loss in potency. From this
ACS Paragon Plus Environment

Page 13 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
exercise, analog 5 with an isoquinoline moiety replacing the naphthalene was identified with an
s
MITC95 of 23 nM, and was considered as a potential tool compound for an in vivo study.
Unfortunately, analog 5 had poor pharmacokinetic properties with AUC0ꢀ24h of 1.7 µM.hr at 100
mg/kg [formulated with Imwitor 742/tween 80 (50:50)] oral dose in mice. In addition, adequate
solubility under variety of formulation conditions was not achieved for either subcutaneous or
intravenous dosing despite the improvement in LLE (1.64 relative to ꢀ0.38 for analog 1). Thus,
analog 5 was deemed unsuitable as a tool compound for in vivo efficacy studies.
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 14 of 52
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 1. Initial SAR Investigation with a Linker Expansion Strategy
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
s
i
1
MW
TarO
COL MITC95 COL MITC95
Cpd
1
R
n
1
1
2
3
LLE
ꢀ0.38
ꢀ0.05
0.57
ꢀ0.3
(
clogP) IC (nM)
(nM)
(µM)
5
0
481
7.1)
171
2500
9
900
>200
(
482
2
ND
15
>200
>200
>200
(5.6)
495
3
(7.4)
5
09
4
28
125
(7.8)
4
96
5
2
24
1.64
23
>200
(5.9)
S
MITC95 – minimum concentration to inhibit 95% growth of MRSA COL in presence of 4
i
µg/mL of IPM. MITC – minimum concentration to inhibit 95% growth of MRSA COL. LLE:
9
5
lipophilic ligand efficiency.
ACS Paragon Plus Environment

Page 15 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Atomic charges
X = CH2
ꢀ0.61
Atom
O1
X = NH
ꢀ0.61
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
O2
ꢀ0.62
ꢀ0.65
O3
ꢀ0.61
ꢀ0.69
Figure 2. ESP derived atomic charges on oxygen atoms designated as O1, O2 and O3 for both
amide and urea analogs.
To expand SAR while improving LLE, previous SAR was carefully analyzed, and three
key findings were noted: a) substitution of oxygen (designated as O1) of the oxazolidinone 5
s
with ‘NH’ (O1 →NH, X = CH ) resulted in complete loss of potency ( MITC > 200 µM); b)
2
95
replacement of O2 with ‘S’ (O2 →S, X = CH ); and, c) removal of O3 (C=O3 →CH , X= CH )
2
2
2
led to a 150ꢀ and 4ꢀfold drop in TarO enzymatic activity, respectively. Based on these
observations, we speculated that O1, O2 and O3 could act as hydrogen bond acceptors in the
binding site.
To enhance the strength of the postulated hydrogen bonds by increasing the net negative
charge on these oxygens, we proposed to incorporate nitrogen at 2 position of the linker by
converting the exocyclic amide to urea (5 → 6) with minimum perturbation of the overall
structure. Electrostatic potential (ESP) derived atomic charges showed that such modifications
2
8
enhanced the net negative charge on oxygen O3 (Figure 2). To test this hypothesis, analog 6
was synthesized for in vitro studies. Gratifyingly, analog 6 displayed a TarO IC of 29 nM with
5
0
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 52
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
an improved LLE of 2.07 with Phage K IC50 and Targocil reversal IC50 of 94 and 191 nM
respectively (Table 1 supporting information). While the improvement in LLE was modest
(either due to enthalpy–entropy compensation or due to an entropic penalty because of
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
desolvation of the newly introduced heteroatom), this modification demonstrated that an
additional heteroatom could be incorporated without deterioration of the potency. Furthermore,
this urea forming reaction enabled a rapid exploration of the naphthalene/ isoquinolineꢀbinding
pocket with a variety of easily accessible substituted aryl methylamines (Table 2). From such
exploration, analog 7 with 1,6ꢀnaphthyridine replacing the isoquinoline moiety with LLE of 3.30
was identified. Concurrently, we investigated the bisꢀtrifluorophenyl region using moieties with
decreasing lipophilicity while holding the isoquinoline constant on the right hand side.
Replacement of one of the trifluoromethyl groups with a fluorine led to no change in TarO IC ,
5
0
but removal of a trifluoromethyl group resulted in a 6ꢀfold drop in potency ( 6 vs. 9).
Incorporation of a nitrogen into the trifluoromethyl phenyl group of 9 resulted in a significant
drop in potency (9 vs. 11) relative to incorporation of additional nitrogen into the isoquinoline
moiety (9 vs. 10). From these SAR efforts, analog 9 was quickly selected as a candidate for in
s
vivo studies based on the consideration of its in vitro potency (COL MITC = 250 nM), LLE
95
and cell toxicity (Table 1S in supporting information) in conjunction with its favorable
preliminary pharmacokinetic profile (vide infra) while we continued medicinal chemistry efforts
to further improve physicochemical properties, such as LLE and solubility.
ACS Paragon Plus Environment

Page 17 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 2. SAR Studies Using a Urea Linker.
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
i
1
2
MW
(clogP)
TarO IC50
(nM)
COL MITC95 COL MITC95
Cpd
R
R
LLE
(nM)
(µM)
>200
>200
>200
>200
>200
>200
>200
4
97
6
7
8
9
29
12
2.07
3.34
2.88
2.21
2.7
31
(5.5)
498
(4.6)
125
125
250
970
6250
375
4
47
25
(4.7)
4
29
162
402
5300
260
(4.6)
430
3.7)
1
1
0
1
(
5
44
2.05
3.47
(3.2)
4
61
14
(3.1)
S
MITC – minimum concentration to inhibit 95% growth of MRSA COL in the presence of 4
µg/mL of IPM. MITC – minimum concentration to inhibit 95% growth of MRSA COL
9
5
i
9
5
Seeking an opportunity to replace the 3,5ꢀbis(trifluoromethyl)phenyl with
heteroaromatics, we recognized that while analog 11 displayed IC of 5.3 µM, it possessed an
5
0
LLE of 2.05 and a reasonably low clogP of 3.2. Therefore to improve potency without
deteriorating clogP, selective modification of the 4ꢀtrifluoropyridine of 11 was undertaken based
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 52
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
on the previous SAR finding that the introduction of a methoxy group at the 6ꢀ position of the
pyridine of analog 12 resulted in significant improvement of potency (Figure 3).
TarO IC50
nM)
16000
COL
Cpd
R
H
s
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
(
MITC (nM)
9
5
12
>200000
13 OMe
235
320
Figure 3. Analysis of previous SAR that inspired the modification of 4ꢀ(trifluoromethyl)pyridine
of 11.
Taking these observations into account, 6ꢀmethoxyꢀ4ꢀ(trifluoromethyl)pyridine was
designed to restore the potency, and analog 14 with clogP of 3.2 was prepared and shown to have
a TarO IC of 260 nM with a significant improved LLE of 3.47 (Table 2) with Phage K IC
50
5
0
and Targocil reversal IC of 600 and 200 nM respectively (Table 1S supporting information).
50
In addition, analog 14 has a solubility of 32 µM in PBS buffer. While this improvement in
solubility relative to that of lead 1 (32 µM for 14 vs. <1 µM for 1) was a significant achievement,
2
9
the unusually high dosing regimen for the antibacterial treatment necessitated further
improvement in solubility.
ACS Paragon Plus Environment

Page 19 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
O
N
O
N
O
O
O
NH
O
O
N
N
F3C
F3C
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
CF3
CF3
15
16
TarO enzyme IC >200 µM,
TarO enzyme IC >200 µM,
50
50
s
s
MITC + IPM >200 µM
MITC + IPM >200 µM
95
9
5
Figure 4. Modification of the linker with functionalities amenable to rapid SAR exploration.
Before we continued to use this ‘urea formation strategy’ to further expand SAR, we
retrospectively contemplated whether such a strategy could have been envisioned by simply
looking for operationally straightforward transformations amenable to rapid SAR exploration
without careful consideration of the impact of strengthening the putative hydrogen bond. As a
follow up, we designed and synthesized analogs 15 and 16 by modifying the linker, as the
representative examples for rapid SAR exploration. Such simple changes led to a precipitous
drop in potency (Figure 4). Thus, in the present case scenario, prioritization of our efforts based
on our guiding hypothesis following SAR analysis signified the importance of this strategy to
finding a tool molecule expeditiously.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 52
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
TarO IC50
Cpd
X
(nM)
3
CO
9
17
CH
2
38
(a)
(b)
Figure 5. (a) Comparison of TarO enzyme IC s between analog 3 and its decarbonylated analog
7. (b) Conformational analysis of 3. Conformer 3c with the cisꢀcarbonyl orientation is 4.9
kcal/mol higher in energy than the conformer 3t with transꢀcarbonyl orientation. The carbonyl
of 3t is approximately oriented towards the vicinity of 4ꢀmethyl of oxazolidinone.
5
0
1
Having confirmed that only the urea linkage was tolerated, we investigated in incorporating the
polar solubilizing moiety into the molecule to further improve the physicochemical properties.
Among the many possibilities that could potentially be explored by a purely empirically
approach, modification at the 4ꢀposition of oxazolidinone was prioritized based on the
observation from the computational work that the exocyclic carbonyl was preferentially oriented
towards the region occupied by the methyl group from the 4ꢀ position of the oxazolidinone
(
Figure 5), and the presence of such carbonyl offered a 4ꢀfold improvement in TarO potency
ACS Paragon Plus Environment

Page 21 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
relative to decarbonylated analog 17. Thus, it was conceivable that the binding pocket occupied
by the exocyclic carbonyl was polar, and may be polar enough to accommodate solubilizing
functional groups projecting from the 4ꢀ position of the oxazolidinone without a major impact on
potency. To probe this hypothesis, a series of analogs with a variety of solubilizing groups
replacing the 4ꢀmethyl were prepared. As shown in Table 3, hydroxymethylated analog 18
displayed a TarO IC of 170 nM with Phage K IC and Targocil reversal IC of 1263 and 496
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
0
50
50
nM respectively (Table 1 supporting information) The incorporation of an additional nitrogen
into the isoquinoline moiety of 18 or replacement of one of the trifluoromethyl moieties either
with fluorine or with hydrogen resulted in minimal perturbation in TarO potency (analogs 19, 20
and 21). However, conversion of the hydroxymethyl group of analog 21 to carboxylic acid
resulted in a complete loss of potency (analog 22). Replacement of hydroxymethyl of 20 with
hydroxyl ethyl afforded 22ꢀfold loss in potency (analog 23), and conversion of this functionality
to its corresponding acid led to a complete loss in potency (analog 24). With the identification
that the hydroxymethyl moiety conferred minimum perturbation on the enzyme potency, we
sought to replace the 1,3ꢀbis(trifluoromethyl)benzene moiety with our previously validated 4ꢀ
trifluoromethylꢀ6ꢀmethoxy pyridine moiety to further lower clogP and improve solubility. From
these efforts analogs 25 and 26 were identified with TarO IC of 80 and 880 nM, and LLE of
5
0
4.34 and 4.10, respectively. The solubility of these hydroxymethylated analogs were generally
higher than their methylated counterparts as expected with analog 26 displaying 94 µM in PBS
buffer.
Selected analogs were profiled against a diverse panel of 116 various targets to assess
their offꢀtarget liabilities. In general, a compound’s offꢀtarget liability was reduced with the
30
improvement of LLE. For example, more than 50% inhibition of six offꢀtargets at 10 µM
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 52
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
concentration were observed with analog 5 (with clogP of 5.9 and LLE of 1.64), whereas, no offꢀ
target liabilities were observed with analog 26 (with clogP of 1.9 and LLE of 4.10).
Table 3. SAR by Incorporating Solubilizing Groups at 4ꢀPosition at the Oxazolidinone
Moiety.
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
COL
MITC95
COL
MITC95
(µM)
1
2
3
MW
clogP)
TarO IC50
(nM)
s
i
Cpd
R
R
R
LLE
2.55
(
(nM)
ꢀ
CH OH
513
174
142
220
390
320
650
>200
>200
>200
>200
>200
>200
>200
>200
>200
2
18
(4.2)
1
9
ꢀCH
ꢀCH
ꢀCH
2
2
2
OH
OH
OH
514
3.52
3.21
3.20
NA
(3.3)
20
21
463
210
(3.4)
445
(3.3)
300
2
2
2
2
2
2
3
4
5
6
ꢀCO
2
H
459
(4.1)
>200000
4700
>200000
80
>200000
18000
>200000
360
ꢀ(CH
)
2
OH
477
1.59
NA
2
(3.7)
ꢀCH CO H
2 2
491
(3.9)
ꢀCH
ꢀCH
OH
OH
476
4.34
4.10
2
(2.7)
477
880
900
2
(1.9)
ACS Paragon Plus Environment

Page 23 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
i
S
MITC – minimum concentration to inhibit 95% growth of MRSA COL. MITC – minimum
9
5
95
concentration to inhibit 95% growth of MRSA COL in the presence of 4 µg/mL of IPM.
Table 4. Pharmacokinetic parameters of 9 and 26 in mouse and rat.
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
In vivo
Cl
mL/min/kg)
V ,
d ss
(L/kg
)
IV Dose AUC(0ꢀ∞)
t1/2
(hr)
Cpd Species
(mg/kg)
(µM.hr)
(
9
9
rat
mouse
rat
1
0.42
92.7
6.60
1.70
2
1
1.26
0.95
61.4
36.8
13.05 6.40
1.2 0.48
26
Table 5. Tissue distribution of 9 in mouse blood, lung, liver, spleen and kidney after a 50
mg/kg subcutaneous (S.C.) dose, formulated in 30% Captisol .
®
9
@ 50 mg/kg
S.C.
Blood
9.9
Lung
Liver
Spleen
Kidney
AUC0ꢀx (µM•h)
Cmax (µM)
Tmax (h)
122.8
43.7
0.7
322
101
1.0
75
29
1.0
336
95
1.0
4
0
.9
.5
We evaluated key pharmacokinetic parameters of 9 for an in vivo efficacy study. With a
mg/kg intravenous (IV) dose, analog9 formulated in 20%DMSO/60%PEG400/20%water
2
displayed an AUC of 1.26 µM·hr in mice. It possessed a high in vivo clearance of 61
mL/min/kg, a high volume of distribution of 13.5 L/kg and a moderate terminal halfꢀlife (t1/2
6.4 h) in mice. Analog 9 displayed high clearance (93 mL/min/kg), high volume of distribution
6.6 L/kg), and a short halfꢀlife (t1/2 = 1.7 h) in rat (Table 4). The distribution of 9 in various
=
(
tissues such as lungs, liver, spleen, and kidney as well as in blood was analyzed with a solubility
limiting 50 mg/kg subcutaneous dose (S.C.) in mice. The AUC and Cmax were high in excretory
organs such as kidney, liver and lungs, and were 6ꢀ8 times higher in the spleen relative to blood
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 52
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 5). With no major toxicological findings, the 50 mg/kg S.C. dose was selected as the
highest dose for an in vivo study. In the preparation of the in vivo study, female balb/c mice
were rendered neutropenic with a single intraperitoneal dose of cyclophosphamide (250 mg/kg
on day ꢀ4.). MRSA COL (MB 5393) was grown for 16 h in trypticase soy broth at 37 °C, and
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
4
0.5 mL of diluted culture (~1·10 CFU) was administered intraperitoneally. Vehicle, 9 alone at
1
2.5, 25 and 50 mg/kg or in combination with 5 mg/kg IPM were administered at 2, 5 and 8 h
after the challenge (subcutaneously, three times day). 24 hours after the challenge, the mice
were euthanized, kidneys were harvested, weighed, homogenized, and plated to determine the
CFU remaining in the compound treated groups as compared to the vehicle group (statistical
analysis was determined by oneꢀway ANOVA). All animal procedures were performed
according to the protocol reviewed and approved by the institutional Animal Care and Use
Committee of Merck & Co., Inc., Kenilworth, NJ, USA.
Figure 6. Assessment of efficacy of 9 with doses of 12.5 mg/kg, 25 mg/kg and 50 mg/kg alone
and in combination with 5 mg/kg subꢀefficacious dose of IPM against MRSA COL. 5 mg/kg
IPM dose alone was ineffective in lowering bacterial burden. Cilastatin, an inhibitor of human
dehydropeptidase, was added to prevent degradation of IPM. *P < 0.05, oneꢀway ANOVA
ACS Paragon Plus Environment

Page 25 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
compared to the IPMꢀalone group. **P < 0.05, oneꢀway ANOVA compared to the vehicleꢀalone
group.
Due to the lack of intrinsic antibacterial activity of 9, no reduction in bacterial burden
was observed with 9 treated animals at any dose relative to vehicle treated animals. However,
the combination of 9 with a subꢀefficacious dose of IPM (5 mg/kg) demonstrated dose dependent
lowering of bacterial burden. While the 12.5 mg/kg dose was ineffective, increased doses of 25
and 50 mg/kg resulted in 2 and 2.5 log unit reduction in bacterial burden, respectively (Figure
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
6
). This synergy of subꢀefficacious dose of IPM with compound 9 was comparable to the
efficacy observed with 10 mg/kg of linezolid.
To compare the exposures of 9 and IPM among the threeꢀarm study with 50 mg/kg dose
of 9 and 5 mg/kg IPM that effected 2.5 log unit reduction in bacterial burden, satellite mice (n =
3) were dosed following the same protocol described above for the main study animals, and
blood samples were collected at the indicated time points and analyzed by LCꢀMS (Table 2S in
supporting information). Analog 9 showed similar concentrations at each time point and similar
AUC0ꢀx between 9 only treated animals and combination treatment group, but the exposure of
IPM in the IPMꢀcilastatin treatment group was on average 2ꢀ to 4ꢀ fold higher than that in the
combination treatment group. This change in pharmacokinetic profile of IPM in the presence of
9
was not clearly understood. The synergistic pharmacological effect observed with IPM in the
combination group provided strong support that the alternation of early stage WTA biosynthetic
pathway is a viable alternative for the potentiation of βꢀlactam based antibiotics.
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 52
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
Conclusions
Continual discovery of novel antibiotics is vital for the survival of humankind in our
battle against bacteria. Recent decades have witnessed both the rise in power of due to their
resistance to many once effective antibiotics as well as the decline in discovery of novel
antibiotics. If these trends continue, experts predict that by 2050, 10 million deaths will occur
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
3
1
annually from bacterial infections.
In our continuous efforts to discover novel antibacterial
mechanisms; we performed a phenotypic screening to identify WTA early stage inhibitors to
intervene WTA polymer biosynthesis for revitalization of the important class of βꢀlactam based
antibiotics. These efforts identified analogs with ‘undrugꢀlike’ properties (for example lead 1,
clogP = 7.1 with unmeasurable solubility in PBS buffer), while invigorating our continued
debate on whether the compound collection should be enriched with only ‘drugꢀlike’ molecules.
Clearly, in this case, identification of 1 would have been missed had we decided to screen a
collection of molecules with only ‘drugꢀlike’ properties. Furthermore, our optimization of the
lead 1 to improve its physicochemical properties for in vivo study proved to be very challenging.
Conventional wisdom suggested that it was much easier for medicinal chemists to add lipophilic
moieties to a lead compound to achieve optimized potency, and it was much more difficult to
reduce molecular weight and lipophilicity while maintaining potency. To further improve clogP,
the aryl group was replaced with a heteroaryl group at the 5ꢀ position of the oxazolidinone and a
hydroxyl functionality was introduced into the methyl group at the 4ꢀ position with concomitant
improvement in LLE and solubility. Ultimately, through a hypothesisꢀdriven SAR study, we
successfully developed analogs with much improved physicochemical properties such as clogP.
From these efforts, analog 9 was selected for in vivo studies, and demonstrated to provide a 2.5
log unit reduction of bacterial burden while combining a subꢀefficacious dose of IPM. These
ACS Paragon Plus Environment

Page 27 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
findings provided strong support that potentiation of betaꢀlactam based antibiotics by inhibiting
TarO and therefore preventing WTA polymer biosynthesis could be an effective alternative
treatment option for our battle against MRSA infection. These improvements, led to 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
s
discovery of many drugꢀlike molecules with favorable MITC and represented a significant
9
5
achievement in our efforts to introduce polarity into ligands that bind into the cognate
hydrophobic binding pocket of TarO. Lessons learned from such SAR studies will be very
useful in our future SAR studies and beneficial to the medicinal chemistry community.
EXPERIMENTAL SECTION
The in vivo efficacy study and the βꢀlactam potentiation assay were conducted following
2
1
the protocol described recently.
The samples from the satellite PK study were analyzed to
determine the concentration of IPM and 9.
Unless otherwise mentioned, all the reagents and chemicals were purchased from
commercial sources and used directly. Progress of the reaction was monitored either by thin
layer chromatography or by LCꢀMS. Air sensitive reactions were carried out under the
atmosphere of nitrogen. All the final compounds were purified to >95% purity by either silica
gel chromatography or using reverse phase C18 column. 400 MHz and 500 MHz NMR
spectrometers were used to collect NMR spectra, and all the peaks were reported in ppm relative
to the residual solvent peak, which served as the calibrant. The number of protons and coupling
constants were reported parenthetically after each peak. Ultraꢀhigh performance liquid
chromatography equipped with highꢀresolution tandem mass spectrometry (UHPLCꢀHRMS/MS)
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 52
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
and C18 reverse phase column with the gradient of 5% to 95% MeCN in water with 0.05% TFA
as the mobile phase was used to obtain the mass spectra and to assess the purity of the samples.
Electrostatic potential derived atomic charges were calculated at the B3LYP/6ꢀ31+G(d,p) level
of theory. Conformational analysis of 3 was done at the B3LYP/6ꢀ31+G(d,p) level of theory
with inclusion of Polarizable Continuum Model (PCM) for water (scrf=(solvent=water)). All
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
3
2
calculations were carried out with the Gaussian09 software package. The chemical structures
of all compounds in the manuscript were compared to the list of reported PAINS structures, and
no match was found.
(4S,5R)ꢀ5ꢀ(3,5ꢀBis(trifluoromethyl)phenyl)ꢀ4ꢀmethylꢀ3ꢀ(2ꢀ(naphthalenꢀ1ꢀ
yl)acetyl)oxazolidinꢀ2ꢀone (1). To a mixture of 27 (10.0 g, 34.0 mmol) and 28 (10.8 g, 40.0
mmol) in THF (100 mL) was added drop wise a solution of iꢀPrMgCl (85 mL, 1 M in THF, 85.0
mmol) at ꢀ10 °C. After the addition, the reaction mixture was warmed to rt and stirred for 12 h.
Then the reaction mixture was added slowly into a solution of 1 N HCl (190 mL) at 0 °C. The
resulting solution was extracted with EtOAc (150 mL x 3), and the combined EtOAc layers were
dried with MgSO , filtered and concentrated to afford 29, which was dissolved in a mixture of iꢀ
4
PrOH (27 mL) and toluene (60 mL) and added Al(OiPr) (6.2 g, 34.0 mmol). The resulting
3
mixture was heated at 50 °C for 12 h, and then cooled to rt. To this solution was added a
solution of KOH in water (3.9 g of KOH in 4 mL H O) and the resulting mixture was stirred at
2
20 °C for 3 h, then quenched with brine (100 mL) and extracted with EtOAc (100 mL x 2). The
EtOAc layers were combined and washed with brine (100 mL x 4), dried with MgSO , filtered,
4
concentrated, and the crude product was purified by silica gel chromatography (0% to 100%
EtOAc in hexane) to provide 30 (9.6 g, 91% yield).
ACS Paragon Plus Environment

Page 29 of 52
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
To a solution of 1,1'ꢀcarbonyldiimidazole (36.0 mg, 0.22 mmol) in acetonitrile (1 mL)
was added 2ꢀ(naphthalenꢀ1ꢀyl)acetic acid (41.0 g, 0.22 mmol), and the reaction mixture was
stirred at rt for 1 h. To this resulting mixture was added 30 (50.0 mg, 0.16 mmol) followed by
DBU (34.0 mg, 0.22 mmol), and the reaction mixture was stirred at rt for 12 h. The resulting
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
mixture was concentrated, and the crude product was purified by reverse phase HPLC to provide
1
1
(70.0 mg, 91% yield). H NMR (500 MHz, CDCl ) δ 7.97 (s, 1H), 7.95 – 7.90 (m, 2H), 7.87
3
(
dd, J = 7.3, 2.3 Hz, 1H), 7.84 (s, 2H), 7.56 (m, 2H), 7.51 – 7.45 (m, 2H), 5.84 (d, J = 7.3 Hz,
1H), 4.94 (p, J = 6.7 Hz, 1H), 4.83 (d, J = 17.2 Hz, 1H), 4.77 (d, J = 17.2 Hz, 1H), 0.97 (d, J =
.6 Hz, 3H). m/z: 482.2.
6
(4S,5R)ꢀ5ꢀ(3,5ꢀBis(trifluoromethyl)phenyl)ꢀ3ꢀ(2ꢀ(isoquinolinꢀ4ꢀyl)acetyl)ꢀ4ꢀ
methyloxazolidinꢀ2ꢀone (2). Compound 2 was prepared from 2ꢀ(isoquinolinꢀ4ꢀyl)acetic acid
1
and intermediate 30 following the acylation protocol used to prepare 1 in 40% yield. H NMR
(500 MHz, CDCl ) δ 9.62 (s, 1H), 8.66 (s, 1H), 8.35 (d, J = 8.1 Hz, 1H), 8.16 (m, 2H), 7.97 (m,
3
2H), 7.85 (m, 2H), 5.97 (d, J = 7.2 Hz, 1H), 5.03 – 4.96 (m, 1H), 4.96 – 4.84 (m, 2H), 0.97 (d, J
=
6.4 Hz, 3H). m/z: 482.9.
(4S,5R)ꢀ5ꢀ(3,5ꢀBis(trifluoromethyl)phenyl)ꢀ4ꢀmethylꢀ3ꢀ(3ꢀ(naphthalenꢀ1ꢀ
yl)propanoyl)oxazolidinꢀ2ꢀone (3).
Compound 3 was prepared from 3ꢀ(naphthalenꢀ1ꢀ
yl)propanoic acid and intermediate 30 following the acylation protocol used to prepare 1 in 17%
1
yield. H NMR (500 MHz, CDCl ) δ 8.13 (d, J = 8.4 Hz, 1H), 7.96 (s, 1H), 7.91 (d, J = 8.1 Hz,
3
1H), 7.84 – 7.77 (m, 3H), 7.59 (dd, J = 8.4, 6.9, 1H), 7.54 (dd, J = 7.4, 6.7, 1H), 7.48 – 7.42 (m,
ACS Paragon Plus Environment