Disubstituted 1-aryl-4-aminopiperidine library synthesis using computational drug design and high-throughput batch and flow technologies

Article abstract of DOI:10.1021/co400078r

A platform that incorporates computational library design, parallel solution-phase synthesis, continuous flow hydrogenation, and automated high throughput purification and reformatting technologies was applied to the production of a 120-member library of 1-aryl-4-aminopiperidine analogues for drug discovery screening. The application described herein demonstrates the advantages of computational library design coupled with a flexible, modular approach to library synthesis. The enabling technologies described can be readily adopted by the traditional medicinal chemist without extensive training and lengthy process development times.

Full text of DOI:10.1021/co400078r

Subscriber access provided by UNIV FED DO PARANA UFPR
Article
Disubstituted 1-Aryl-4-Aminopiperidine Library
Synthesis Using Computational Drug Design and
High-Throughput Batch and Flow Technologies
Marian C. Bryan, Christopher D. Hein, Hua Gao, Xiaoyang Xia, Heather
Eastwood, Bernd A. Bruenner, Steven W. Louie, and Elizabeth M Doherty
ACS Comb. Sci., Just Accepted Manuscript • DOI: 10.1021/co400078r • Publication Date (Web): 08 Aug 2013
Downloaded from http://pubs.acs.org on August 13, 2013
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.
ACS Combinatorial Science 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 23
ACS Combinatorial Science
1
2
3
4
5
6
Disubstituted 1-Aryl-4-Aminopiperidine Library Synthesis Using Computational
Drug Design and High-Throughput Batch and Flow Technologies
7
8
9
Marian C. Bryan1, Christopher D. Hein1, Hua Gao2, Xiaoyang Xia2, Heather Eastwood2, Bernd A.
Bruenner3, Steven W. Louie3, Elizabeth M. Doherty*,1
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
1Medicinal Chemistry Research Technologies, 2Molecular Structure and Characterization,
3Pharmacokinetics and Drug Metabolism, Therapeutic Discovery, Amgen Inc.
One Amgen Center Drive, Thousand Oaks, CA 91320
Abstract: A platform that incorporates computational library design, parallel solution-phase
synthesis, continuous flow hydrogenation, and automated high throughput purification and
reformatting technologies was applied to the production of a 120-member library of 1-aryl-4-
aminopiperidine analogs for drug discovery screening. The application described herein demonstrates
the advantages of computational library design coupled with a flexible, modular approach to library
synthesis. The enabling technologies described can be readily adopted by the traditional medicinal
chemist without extensive training and lengthy process development times.
Graphical abstract:
Keywords: drug discovery, drug-likeness, lead-likeness, computational library design, library
synthesis, continuous flow hydrogenation, enabling technologies, aminopiperidines
ACS Paragon Plus Environment

ACS Combinatorial Science
Page 2 of 23
1
2
3
4
Introduction
5
6
7
8
9
The costs associated with drug research continue to rise and many US and EU-based
pharmaceutical companies are responding by creating greater efficiencies through the use of predictive
computational tools and innovative enabling technologies for the design and production of drug
candidates. Many of these tools and technologies originated in the combinatorial chemistry world of
the 1990s1 with the more recent addition of the highly popular microwave and, less commonly used,
continuous flow reactors.2 In some larger organizations, dedicated high-throughput (HT) synthesis
groups tasked with creating libraries to support drug discovery are often the specialist users of these
technologies. Nevertheless, whether an organization has the capacity to maintain a dedicated library
synthesis team or not, adoption of these computational tools and enabling technologies by traditional
medicinal chemists is highly useful.
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 the state of the art advances, ingenious library synthesis platforms have emerged that
integrate multiple synthetic steps,3 integrate synthesis with purification,4 and even integrate synthesis
with biological evaluation.5 The siren call of the fully integrated synthesis, purification, characterization,
and biological evaluation platform continues to drive many research efforts. In contrast to an integrated
approach, however, our group places an emphasis on identifying enabling technologies that are modular
in design and can be incorporated into a library synthesis platform accessible to every medicinal
chemist. Techniques that will readily become a part of the day-to-day toolkit of the medicinal chemist
are robust, intuitive, and require minimal specialized training. Because the library synthesis platform is
modular, each tool can be utilized independently with no delays or disruptions to the overall workflow.
The platform consists of five main features:
(1) a web-based computational library design tool using molecular structure-based calculated
properties coupled with a predictive algorithm to select for “drug-like”6 and “lead-like”7 molecules,
ACS Paragon Plus Environment

Page 3 of 23
ACS Combinatorial Science
1
2
3
4
(2) a custom-weighed reagent service coupled with a proprietary and commercial reagent collection,
5
6
7
(3) a suite of parallel and HT-synthesis and purification tools, each of which is supported by a dedicated
8
9
“power-user” who provides technical assistance and training on an ad hoc basis,
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(4) a high-throughput mass-triggered HPLC purification and reformatting process (HTPP) that is
supported by an internal team of HT-purification specialists, and
(5) an electronic laboratory notebook-based (ELN) compound tracking and record keeping system with a
batch compound registration process that correlates sample numbers and analytical data to products in
bar-coded vials or 96-well plates.
To demonstrate an application of the platform, an aminopiperidine-based 120-member library
synthesis and evaluation is described herein. The library production incorporates various commercially
available synthesis-enabling tools, including parallel reactor blocks, automated chromatography
systems, and a flow hydrogenator. The library was evaluated in vitro for key indicators of drug-likeness
(solubility, microsomal stability, cell permeability) and the results compared to those predicted using our
predictive algorithm.
Results and Discussion
Library Design. Because of its representation in many biologically active compounds, the 4-
aminopiperidine substructure may qualify as a “privileged” scaffold.8 As a subset of this class, 1-aryl-4-
aminopiperidines appear in a variety of therapeutic agents, for example the cardiac stimulant
Buquineran,9 and the antidiabetic ABT-27910 (Figure 1).
Figure 1. The 1-aryl-4-aminopiperidine scaffold represented in bioactive molecules.
ACS Paragon Plus Environment

ACS Combinatorial Science
Page 4 of 23
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
For the purposes of library synthesis, the 4-aminopiperidine scaffold provides a convenient
opportunity for diversification at two sites, a primary and a secondary amine, which we proposed to
differentiate by means of a suitable protecting group strategy. Thus our synthetic strategy consisted of
three steps (Scheme 1): (1) arylation of the piperidinyl amine via thermally induced SNAr, (2)
deprotection of the primary amine, and (3) reaction of the primary amine with an isocyanate, sulfonyl
chloride, or acid chloride.
Scheme 1. Proposed synthesis of 1-aryl-4-aminopiperidine library.
For the SNAr step, a minimal set of eight aryl and heteroaryl fluorides (1{1-8}, Figure 2) was
selected. The set included aromatic rings with ortho- or para-activating groups to enhance reactivity
and provide an orthogonal functional handle for potential future elaboration (acyl 1{1}, nitrile 1{2-3},
ester 1{4-5}) thus affording “lead-like” structures. Reactive heteroaromatics (2-pyridine 1{6}, 2-
pyrimidine 1{7}) were included as well as an example incorporating a kinase binding moiety, 2-
aminopyrimidine 1{8}.
ACS Paragon Plus Environment

Page 5 of 23
ACS Combinatorial Science
1
2
3
4
Figure 2. Aryl and heteroaryl fluorides selected for SNAr step.
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
A set of putative R-groups for the second point of attachment (urea, sulfonamide, amide) was
selected from reagents available in Amgen’s proprietary reagent collection, which is inventoried and
managed by Frontier Scientific Services, Inc. (FSSI). Located in Newark, Delaware, FSSI also maintains
their own commercial reagent collection and custom weighing service available to the public. The
reagents can be accessed on demand and delivered to the chemist’s bench, within 2-3 days of request,
custom-weighed in the requisite reaction vessel. In the selection process, the inventory of potential
reagents was searched via a custom web interface using a structural filter based on required functional
groups (acid chloride, isocyanate, sulfonyl chloride) and undesired groups (aldehydes, nitro groups,
carboxylic acids, primary amines), thus eliminating obvious pharmacokinetic liabilities and competing
reactivity. The reduced list (301 acid chlorides, 127 isocyanates, 134 sulfonyl chlorides) in combination
with the 4-aminopiperidine core and aryl groups in Figure 2 was used to enumerate a virtual library of
4496 disubstituted 4-aminopiperidines. A script written using Pipeline Pilot™ software11 simultaneously
generated the structures and associated structure-based calculated properties. The calculated
properties were used to refine the list further based on molecular weight (<400 Da), clogP (0-4), polar
surface area (45-125 Å), total hydrogen bond donors and acceptors (<7), and number of aromatic rings
(1-3). While consideration of prior SAR, Topliss analysis, and virtual docking programs are routinely used
to design libraries at Amgen, for the exercise described herein a simple visual inspection of the
ACS Paragon Plus Environment

ACS Combinatorial Science
Page 6 of 23
1
2
3
4
truncated list led to the selection of five each from the acid chlorides, isocyanates, and sulfonyl chlorides
5
6
to apply to library production (2{1-15}, Figure 3).
7
8
9
Figure 3. Acid chloride, isocyanate, and sulfonyl chloride reagents.
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
Library Synthesis. The synthesis commenced with SNAr arylation of a protected 4-aminopiperidine. It
was quickly determined that N-carbamate protecting groups (Boc or Cbz) were unsuitable due to
thermal instability of the carbamate under the SNAr reaction conditions (DMSO, K2CO3, 90 °C), yielding
by-products resulting from decomposition to the isocyanate. We then considered using an azide as an
amine precursor which offers the attraction of atom-efficiency, thus reducing the mass required in the
first steps of the library synthesis. While many organic azides are known to be thermally unstable, it has
been reported that azides may be stable when the number of nitrogen atoms does not exceed the
number of carbon atoms.12 The 4-azidopiperidine (4) was prepared from the commercially available
Boc-protected precursor (3) by acid deprotection (Table 1). Without further purification, the resulting
hydrochloride salt was used directly in the arylation step (1{1-8}, DMSO, K2CO3, 90 °C) under which
ACS Paragon Plus Environment

Page 7 of 23
ACS Combinatorial Science
1
2
3
4
5
6
7
8
conditions the azide proved thermally stable. The crude products were readily purified by silica gel
chromatography using an automated flash silica gel chromatography system. The overall process
afforded the desired 1-aryl-4-azidopiperidines 5{1-8} in 23-84% isolated yield over two steps.
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. Generation of 1-aryl-4-azidopiperidines 5{1-8} from 4-azidopiperidine.
Entry
ArF
1{1}
1{2}
1{3}
1{4}
Product
5{1}
Yield (%)a
Entry
ArF
1{5}
1{6}
1{7}
1{8}
Product
5{5}
Yield (%)
1
2
3
4
61
67
84
23
5
6
7
8
47
47
72
70
5{2}
5{6}
5{3}
5{7}
5{4}
5{8}
aisolated yield after purification by silica gel chromatography.
While the azide may be reduced by chemical means (Ph3P or NaBH4), catalytic hydrogenation
was preferred due to the relative convenience of product isolation via filtration. Furthermore, by using
an H-Cube® continuous flow reactor (ThalesNano, Budapest, Hungary) the time taken from reaction
screening to scale-up was minimized. The reactor introduces hydrogen into a flowing stream of
substrate in carrier solvent which is then passed through a stainless steel cartridge (CatCart®) containing
solid catalyst. The product is isolated in high yield from the eluent after evaporation of solvent. The
conversion of azide 5{1} to amine 6{1} (Scheme 2) was first examined on a screening scale (0.08 mmol)
to optimize conversion and minimize the formation of the over-reduced side product 7. For screening
reaction conditions, the H-Cube® was equipped with a syringe injector port and a 2 mL sample loop,
previous studies having shown that this configuration provides the best correlation between small-scale
ACS Paragon Plus Environment

ACS Combinatorial Science
Page 8 of 23
1
2
screening results and larger continuous-flow preparative scale results.13 Commercial 10% Pd/C, 5%
Pd/C, 20% Pd(OH)2, and 10% Pt/C CatCarts® were evaluated under standard hydrogenation conditions (1
mL/min flow rate, 30 °C, “full” hydrogen mode) and the effect of MeOH vs. THF was compared (0.04 M
substrate concentration).14 All three Pd catalysts afforded the undesired alcohol 7 when the reduction
was performed in MeOH, though this over-reduction appeared to be mitigated to some extent in THF.
In comparison, 10% Pt/C afforded excellent conversion (>90%) and selectivity in both MeOH and THF for
all of the azides 5{1-8} on a 2 mL screening scale. Therefore, 10% Pt/C was used for all preparative scale
azide reductions.
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
Scheme 2. Reduction of azide 5{1} to amine 6{1} and over-reduced product 7.
Initially, in the course of screening and scale-up, preservative (BHT)-free THF was used so that
the presence of preservative would not interfere with GC analysis. Using this solvent, however, led to
varying amounts of a side product with a mass consistent with n-butanol addition. In the case of the
reduction of azide 5{2} with a 10% Pt/C CatCart®, this side product was isolated and identified as alcohol
8 (Scheme 3). The addition of n-butanol to amines in the presence of Pd has been reported,15 though
not to our knowledge in the presence of Pt. To prevent the formation of this by-product, MeOH could
be substituted, otherwise the use of preservative (BHT)-containing THF appeared to mitigate the issue.
ACS Paragon Plus Environment

Page 9 of 23
ACS Combinatorial Science
1
2
3
4
Scheme 3. Alkylation of amine under H-Cube® reduction conditions in preservative-free THF.
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
To reconfigure the H-Cube® for preparative scale reductions (>2 mL), the syringe injector port
was bypassed and azide solution was introduced directly into the reactor via the solvent inlet.14 As
shown in Table 2, the mass recovery using MeOH or THF was high for all preparative scale (1.1-5.5
mmol) reductions with the exception of the 2-aminopyrimidine analog (entry 8) in which case recovery
was improved using 0.02 M AcOH/MeOH as the carrier solvent. The same 10% Pt/C CatCart® was re-
used for all sequential reductions until conversion was observed to deteriorate, presumably due to
catalyst poisoning. Performance was regained by heating the CatCart® to 50 °C during the reduction or
by replacing with a fresh CatCart®. When incomplete reduction occurred, the recovered product
mixture could be resubjected to the reaction conditions to provide desired product without over-
reduction. The arylpiperidinyl amines were isolated by evaporation of the solvent in vacuo and the
products carried through the subsequent library synthesis step without purification.
Table 2. Reduction of azides 5{1-8} to amines 6{1-8} on preparative scale (1.1-5.5 mmol).
ACS Paragon Plus Environment

ACS Combinatorial Science
Page 10 of 23
1
2
3
4
5
6
7
8
9
Entry
Azide Scale (mmol)
Solvent
MeOH
MeOH
MeOH
MeOH
MeOH
MeOH
THF
Temp (°C)
Product
6{1}
Yield (%)a
Purity (%)b
1
2
3
4
5
6
7
5{1}
5{2}
5{3}
5{4}
5{5}
5{6}
5{7}
3.3
2.9
2.6
2.5
2.8
1.1
4.4
50
50
50
50
30
30
30
50
96
97
96
90
98
97
96
89
92
6{2}
6{3}
93
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{4}
>99
>99
90
6{5}
6{6}
6{7}
95c
0.02 M
8
5{8}
5.5
6{8}
84d
86
AcOH/MeOH
aisolated yield after removal of solvent unless otherwise indicated; bpurity by reverse phase HPLC at 254 nm unless
otherwise indicated; cpurity by GC; disolated yield after aqueous workup described fully in Experimental Section.
For the final N-functionalization of amines 6{1-8}, the reactions were performed on a 50 mg
scale in parallel in 1-dram glass vials held in 24-position reactor blocks. The amines were added as
solutions to the acid chlorides 2{1-5}, isocyanates 2{6-10}, sulfonyl chlorides 2{11-15} which had been
pre-weighed in the glass vials by FSSI, Inc. After sealing the vials, the blocks were agitated at ambient
temperature on an orbital shaker. Workup consisted of simply quenching excess reagent by the
addition of methanol and the reaction mixtures were concentrated to dryness in the reactor block using
a centrifugal evaporator. In most cases, the crude products could be purified using Amgen’s dedicated
high-throughput mass-triggered reverse-phase HPLC purification and reformatting platform (HTPP).16
The HTPP at Amgen is the only component of the library production platform that requires a dedicated
specialist. The flexibility of our modular library production platform is highlighted by the fact that those
compounds that failed to progress through HTPP due to insufficient solubility or because of poor
ionization by MS were simply redirected to the high-throughput flash silica gel chromatography systems.
ACS Paragon Plus Environment

Page 11 of 23
ACS Combinatorial Science
1
2
3
4
5
6
7
8
9
The final purified compounds could then be directed back again to HTPP for reformatting into the
appropriate vials for storage or further evaluation. Examples purified by each of these purification
methods are shown in Table 3 (see Supporting Information for full set of final yields). Each library
member was analyzed by LC-MS and HT-NMR.17 Typically, library purification, reformatting, LC-MS and
1H NMR analysis, and registration of library members for submission to the screening collection are
completed within 6-10 business days. A 100% success rate was achieved for this library, with all 120
compounds obtained in sufficient quantity and purity (>90% by LC at 215 nm) for submission to the
compound screening collection for further testing.
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 3. Yield and purity of a representative chemset 9{1-2,1-15} from the final library.
N-functionalizing
Purification
Methoda
Isolated
Yield (%)
Entry
Amine
Product
Purity (%)b
reagent
1
2
3
4
5
6
7
6{1}
6{1}
6{1}
6{1}
6{1}
6{1}
6{1}
2{1}
2{2}
2{3}
2{4}
2{5}
2{6}
2{7}
9{1,1}
9{1,2}
9{1,3}
9{1,4}
9{1,5}
9{1,6}
9{1,7}
A
A
A
A
A
A
A
>99
98
27
60
38
45
49
30
26
>99
>99
>99
>99
94
ACS Paragon Plus Environment

ACS Combinatorial Science
Page 12 of 23
1
2
3
4
5
6
7
8
9
8
6{1}
6{1}
6{1}
6{1}
6{1}
6{1}
6{1}
6{1}
6{2}
6{2}
6{2}
6{2}
6{2}
6{2}
6{2}
6{2}
6{2}
6{2}
6{2}
6{2}
6{2}
6{2}
6{2}
2{8}
9{1,8}
9{1,9}
A
>99
93
20
23
11
34
32
10
35
32
84
82
94
100
92
59
59
23
64
57
88
79
44
89
89
9
2{9}
2{10}
2{11}
2{12}
2{13}
2{14}
2{15}
2{1}
A
A
A
A
A/B
A
A
B
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
9{1,10}
9{1,11}
9{1,12}
9{1,13}
9{1,14}
9{1,15}
9{2,1}
>99
>99
>99
>99
>99
>99
>99
97
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}
9{2,2}
B
2{3}
9{2,3}
B
96
2{4}
9{2,4}
B
97
2{5}
9{2,5}
B
>99
>99
98
2{6}
9{2,6}
B
2{7}
9{2,7}
B
2{8}
9{2,8}
B
>99
>99
>99
>99
98
2{9}
9{2,9}
B
2{10}
2{11}
2{12}
2{13}
2{14}
2{15}
9{2,10}
9{2,11}
9{2,12}
9{2,13}
9{2,14}
9{2,15}
B
B
B
B
>99
>99
>99
B
B
aPurification method A = HTPP; B = silica gel chromatography; A/B = HTPP followed by silica gel chromatography.
bPurity by HPLC at 215 nm.
ACS Paragon Plus Environment

Page 13 of 23
ACS Combinatorial Science
1
2
3
4
5
6
7
8
9
Library Analysis. The library members were tested for rat and human liver microsomal stability (RLM,
HLM) and the results compared with predicted. The in silico models18 were constructed and validated
based on in house data generated from over 40,000 proprietary compounds. The molecular descriptors
used in the models were selected from a pool of 820 different molecular descriptors covering charged
surface areas, shape and Chi molecular indices, hydrophobicity, solvation properties, E-state indices,
intermolecular H-bond features, and types of H-bond counts. The most important molecular properties
with regard to microsomal stability were: (1) hydrophobicity: compounds with higher clogP in general
have higher clearance; (2) number of acid groups: acidic groups like COOH decrease clearance; (3) H
bonds: increasing number of H bonds decreases clearance; and (4) aromaticity: aromatic compounds in
general have higher clearance. Regarding the performance of the models used, typically ~70% of new
compounds are predicted within two folds and ~80% of new compounds are within three folds of
experimental values. Of course, the predictive performance of the in silico model is highly dependent
on the chemical space coverage of new compounds, that is, properties for new compounds similar to
compounds already represented in the model are better predicted.
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 expected, for the array described herein the accuracy of the prediction varied based on
structural type (see Supplemental Information for full data set). For example, amides (60% within 2-
folds) and sulfonamides (61% within 2-folds) were significantly better predicted than ureas (21% within
2-folds) in the RLM assay, while amides (68% within 2-folds) and ureas (77% within 2-folds) were better
predicted than sulfonamides (28% within 2-folds) in the HLM assay. Overall, the aminopyrimidines 9{8,
1-15} (Table 4) were the best predicted structural sub-type and exhibited the most desirable drug-like
ACS Paragon Plus Environment

ACS Combinatorial Science
Page 14 of 23
1
2
3
4
properties of the array, with 100% of these compounds demonstrating microsomal stability <100
5
6
ꢀL/min/mg and 87% with aqueous solubility >100 ꢀM.
7
8
9
This subset was further evaluated for cell permeability and efflux in LLC-PK1 cells stably over-
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
expressing human P-glycoproteins. Cell permeability (Papp) from the apical to basolateral surface (A2B)
as well as efflux ratio (B2A/A2B) is shown in Table 4. While within the predicted range of values, the
experimentally determined permeability was generally higher than predicted. The measured efflux
ratios were in many cases much lower than predicted due to the higher actual A2B permeability. Based
on these Papp values and efflux ratios, as well as high solubility, this chemset would be expected to have
good intestinal absorption.
Table 4. Properties of aminopyrimidine chemset 9{8, 1-15}.
Predicted Observed
Observed
Observed
RLM CLe
L/min/mg
Observed
HLM CLe
L/min/mg L/min/mg
ꢀ
Predicted
RLM CL
Predicted
HLM CL
ꢀ
Observed
ERg
Predicted
ER
f
PSAa
Å
Ar Solubilityd
Papp
Papp
cm/s
ꢀ
MW
Chemset g/mol clogPa
HBAb HBDc rings
M
L/min/mg
ꢀ
cm/s
ꢀ
ꢀ
ꢀ
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
10.1
5.3
ndh
9{8,1}
332.8
0.00
97.0
4
2
2
500
50
<14
63
<14
14.3
3.0
9{8,2}
9{8,3}
9{8,4}
9{8,5}
9{8,6}
9{8,7}
9{8,8}
9{8,9}
353.4
301.4
357.5
291.4
278.4
318.4
318.4
340.4
2.02
0.40
84.1
84.1
3
3
4
3
3
3
3
3
3
4
4
5
5
4
2
2
2
2
3
3
3
3
3
2
2
2
2
2
3
1
2
1
1
2
1
2
2
2
2
2
2
2
11
94
29
15
26
29
95
28
35
44
41
31
118
31
27
18
<14
<14
<14
<14
16
41
24
15
20
17
52
15
21
28
29
17
17
16
23
<14
<14
<14
<14
<14
<14
<14
<14
<14
<14
<14
<14
<14
<14
3.5
9.7
12.6
11.7
1.1
1.7
7.6
1.5
9.0
3.0
6.9
2.3
1.2
8.2
ndh
28.3
38
11.3
7.5
500
43
2
0.59 102.0
8.5
1.6
1.2
18.8
6.5
2.4
1.8
1.5
3.8
2.5
4.9
6.2
1.6
0.71
0.22
0.90
1.42
1.46
1.49
84.1
96.2
96.2
96.2
96.2
96.2
500
500
107
497
500
166
500
500
500
206
500
31.9
1.1
6.0
21.2
31.6
6.7
7.5
<14
<14
26
14.3
10.2
35.5
9.9
34.5
12.3
29.3
11.0
34.5
36.9
8.6
9{8,10} 346.8
9{8,11} 333.4
9{8,12} 347.4
9{8,13} 368.5
9{8,14} 358.4
9{8,15} 351.4
0.85 101.2
0.71 101.2
1.31 114.1
0.99 125.0
1.30 101.2
61
21
12.1
11.5
7.3
28
22
56
27.5
aPartition coefficients (clogP) and polar surface area (PSA) were calculated using Daylight (V481) toolkit module:
Daylight Chemical Information Systems, Inc. bHydrogen bond acceptors. cHydrogen bond donors. dSolubility in pH
7.4 phosphate buffer. eRat liver microsomal (RLM) and human liver microsomal (HLM) intrinsic clearance measured
in vitro in liver homogenate. fApparent cell permeability (A2B) in LLC-PK1 cells stably over-expressing human P-
glycoprotein is measured by LC-MS/MS. gEfflux ratio = B2A/A2B. not determined due to low recovery in cell assay.
h
Conclusion
ACS Paragon Plus Environment

Page 15 of 23
ACS Combinatorial Science
1
2
3
4
5
6
7
8
9
By allowing the chemistry to guide the selection of appropriate accelerating technologies, as
opposed to forcing the chemistry to work on an existing fully integrated platform, a 100% success rate in
the preparation of all library members can be achieved with minimal process development time. This
library was added to the corporate compound collection for further high-throughput screening.
Computational tools utilized in the design of this library proved useful in determining the relative “drug-
likeness” of the target compounds. The techniques and computational tools described herein have been
adopted by medicinal chemists across Amgen’s Drug Discovery organization.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Experimental Section
Caution: While the authors experienced no adverse events in the preparation of the aliphatic azides
described herein, we advise the use of appropriate personal protective equipment and blast-shielding
when repeating these procedures.
General remarks. All reagents were obtained from commercial suppliers and used without further
purification. Solvents were obtained from EM Science and used directly. Silica gel chromatography was
performed on an Isolera-4 flash purification system with prepacked silica gel cartridges (Biotage,
Uppsala, Sweden). The H-Cube® and CatCarts® were purchased from ThalesNano (Budapest, Hungary).
Pre-weighed reagents for final N-functionalization were obtained in 1-dram glass screw-cap vials from
Frontier Scientific Services (Newark, DE). OptiBlock 24-position reactor blocks were obtained from
ChemGlass (Vineland, NJ). Reactor blocks were agitated on a MaxQ 2000 orbital shaker (J-Kem
Scientific, St. Louis, MO). Polypropylene 24-well plates were obtained from Thomson Instrument Co.
(Oceanside, CA) and 24-well filter plates were obtained from The Automation Partnership
(Hertfordshire, UK). Samples in 24-well plates were concentrated in vacuo on an EZ-2 Evaporator
(Genevac Inc., NY). All final compounds were purified to >90% purity (at 215 nm) as determined by high-
ACS Paragon Plus Environment

ACS Combinatorial Science
Page 16 of 23
1
2
3
4
5
6
7
8
9
performance liquid chromatography (HPLC) on an Agilent 1100 series analytical system with tandem LC-
MSD quadrupole ESI-MS. NMR spectra were determined with a Bruker DRX 400 MHz spectrometer.
4-Azidopiperidine hydrochloride (4). A solution of tert-butyl 4-azidopiperidine-1-carboxylate (3, 1.2 g,
5.2 mmol) in 1,4-dioxane (10 mL) was treated with 1 M HCl in 1,4-dioxane (10 mL) and stirred at 60 °C.
After deprotection was complete, the reaction was allowed to cool to 23 °C and concentrated in vacuo
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
to give the desired product as the HCl salt. H NMR (300 MHz, DMSO-d6) δ 3.80 - 3.91 (m, 1 H), 3.15 -
3.25 (m, 2 H), 2.97 (ddd, J = 3.43, 9.68, 13.04 Hz, 2 H), 1.96 - 2.09 (m, 2 H), 1.61 - 1.76 (m, 2 H). MS (ESI,
pos. ion) m/z: 127.1 [M+H].
General procedure for the preparation of 1-aryl 4-azidopiperidines 5{1-8}. A mixture of 4 (0.5 M in
DMSO), aryl halide 1{1-8} (1 equiv), and K2CO3 (2.5 equiv) was heated at 90 °C with stirring. After 18 h,
the reaction was allowed to cool to 23 °C, diluted with water, and extracted with DCM (2x). The
combined organic layers were dried over Na2SO4, filtered, and concentrated in vacuo to give the crude
product. The crude material was adsorbed onto a plug of silica gel and purified by silica gel
chromatography, eluting with a gradient of EtOAc in hexanes to provide the desired products.
5{1} 1-(2-(4-Azidopiperidin-1-yl)phenyl)ethanone. Purification gradient: 5-50% EtOAc in hexanes; 61%
1
yield. H NMR (300 MHz, CDCl3-d) δ 7.38 - 7.47 (m, 2H ), 7.06 - 7.15 (m, 2 H), 3.55 - 3.67 (m, 1 H), 3.18 -
3.30 (m, 2 H), 2.92 (t, J = 9.87 Hz, 2 H), 2.66 (s, 3 H), 2.01 - 2.14 (m, 2 H), 1.75 - 1.90 (m, 2 H). MS (ESI,
pos. ion) m/z: 245.0 [M+H].
5{2} 2-(4-Azidopiperidin-1-yl)benzonitrile. Purification gradient: 0-30% EtOAc in hexanes; 67% yield.
1H NMR (300 MHz, CDCl3-d) δ 7.58 (dd, J = 1.61, 8.18 Hz, 1 H), 7.45 - 7.53 (m, 1 H), 6.99 - 7.06 (m, 2 H),
3.66 (tt, J = 4.07, 8.28 Hz, 1 H), 3.40 - 3.51 (m, 2 H), 3.06 (ddd, J = 3.14, 8.62, 12.06 Hz, 2 H), 2.11 (tdd, J =
3.34, 6.50, 12.83 Hz, 2 H), 1.84 - 1.99 (m, 2 H). MS (ESI, pos. ion) m/z: 228.0 [M+H].
ACS Paragon Plus Environment

Page 17 of 23
ACS Combinatorial Science
1
2
3
4
5
6
7
8
5{3} 4-(4-Azidopiperidin-1-yl)benzonitrile. Purification gradient: 10-45% EtOAc in hexanes; 84% yield.
1H NMR (300 MHz, CDCl3-d) δ 7.49 - 7.55 (m, 2 H), 6.93 (d, J = 9.06 Hz, 2 H), 3.62 - 3.76 (m, 3 H), 3.13 -
3.23 (m, 2 H), 2.00 - 2.12 (m, 2 H), 1.70 - 1.83 (m, 2 H). MS (ESI, pos. ion) m/z: 228.1 [M+H].
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
5{4} Methyl 2-(4-azidopiperidin-1-yl)benzoate. Purification gradient: 5-50% EtOAc in hexanes; 23%
1
yield. H NMR (300 MHz, CDCl3-d) δ 7.75 (dd, J = 1.61, 7.75 Hz, 1 H), 7.38 - 7.46 (m, 1 H), 6.97 - 7.10 (m,
2 H), 3.91 (s, 3 H), 3.58 (qd, J = 4.40, 8.73 Hz, 1 H), 3.22 - 3.35 (m, 2 H), 2.85 - 2.96 (m, 2 H), 2.06 (d, J =
10.67 Hz, 2 H), 1.78 - 1.92 (m, 2 H). MS (ESI, pos. ion) m/z: 261.0 [M+H].
5{5} Methyl 4-(4-azidopiperidin-1-yl)benzoate. Purification gradient: 5-30% EtOAc in hexanes; 47%
1
yield. H NMR (300 MHz, CDCl3-d) δ 7.89 - 7.97 (m, 2 H), 6.91 (d, J = 8.33 Hz, 2 H), 3.88 (s, 3H), 3.62 -
3.76 (m, 3 H), 3.15 (ddd, J = 3.22, 9.32, 12.90 Hz, 2 H), 2.04 (br. s., 2 H), 1.69 - 1.84 (m, 2 H). MS (ESI, pos.
ion) m/z: 261.0 [M+H].
5{6} 2-(4-Azidopiperidin-1-yl)pyridine. Purification gradient: 5-50% EtOAc in hexanes; 47% yield. 1H
NMR (300 MHz, CDCl3-d) δ 8.19 (td, J = 1.06, 4.02 Hz, 1 H), 7.48 (ddd, J = 1.97, 7.02, 8.70 Hz, 1 H), 6.68
(d, J = 8.62 Hz, 1 H), 6.62 (ddd, J = 0.66, 4.93, 7.13 Hz, 1 H), 4.05 (td, J = 4.28, 13.52 Hz, 2 H), 3.60 - 3.72
(m, 1 H), 3.21 (ddd, J = 3.22, 9.79, 13.30 Hz, 2 H), 1.93 - 2.06 (m, 2 H), 1.61 - 1.77 (m, 2 H). MS (ESI, pos.
ion) m/z: 204.0 [M+H].
5{7} 2-(4-azidopiperidin-1-yl)pyrimidine. Purification gradient: 5-50% EtOAc in hexanes; 72% yield. 1H
NMR (300 MHz, CDCl3-d) δ 8.33 (d, J = 4.82 Hz, 2 H), 6.52 (t, J = 4.82 Hz, 1 H), 4.39 (td, J = 4.71, 13.67 Hz,
2 H), 3.65 - 3.76 (m, 1 H), 3.38 - 3.50 (m, 2 H), 1.92 - 2.04 (m, 2 H), 1.58 - 1.73 (m, 2 H). MS (ESI, pos. ion)
m/z: 205.0 [M+H].
5{8} 4-(4-Azidopiperidin-1-yl)pyrimidin-2-amine. Purification gradient: 90-100% EtOAc in hexanes;
70% yield. 1H NMR (300 MHz, DMSO-d6) δ 12.03 (br. s., 1 H), 7.64 - 7.86 (m, 3 H), 6.56 (d, J = 7.60 Hz, 1
ACS Paragon Plus Environment

ACS Combinatorial Science
Page 18 of 23
1
2
3
4
5
6
H), 3.94 (dt, J = 4.31, 8.66 Hz, 2 H), 3.46 (t, J = 10.38 Hz, 2 H), 1.84 - 2.08 (m, 2 H), 1.42 - 1.60 (m, 2 H). MS
(ESI, pos. ion) m/z: 219.9 [M+H].
7
8
9
Screening-scale reduction of azides 5{1-8}. An H-Cube® hydrogenation flow reactor (ThalesNano
Technology, Inc., Budapest, Hungary) was adapted to allow for fixed-loop injections (2 mL) as previously
described.14
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
For each reaction screen, a 30 x 4 mm catalyst cartridge (CatCart®, ThalesNano, Inc.) was placed into the
CatCart® holder. The temperature was set to 30 °C, flow rate to 1 mL/min, and hydrogen production to
full H2 mode. The catalyst was fully reduced and the system primed by running under stable H2 in full
mode for 10 min with solvent alone. With the system stabilized, azide solution (2 mL, 0.04 M) was
introduced via the manual injector. The eluent was collected over 10 min and product purity was
determined by HPLC.
Preparative-scale reduction of azides 5{1-8} to amines 6{1-8}. The injector port was bypassed on the H-
Cube® for the preparative scale reductions. A 10% Pt/C CatCart® (30 x 4mm, ThalesNano, Inc.) was
placed into the CatCart® holder. The temperature was set to 30 °C, flow rate to 1 mL/min, and the
hydrogen to full H2 mode. A solution of azide 5 in solvent (0.04 M) was then introduced directly into the
H-Cube® through the Knauer pump via the solvent inlet. Once the volume of azide solution was reduced
to 1-2 mL, additional solvent was added to the flask while pumping was continued for 10 min. The
combined eluent was concentrated in vacuo to give the products 6{1-7} (90-98% recovered yield), which
were derivatized without further purification. In the case of 6{8}, after concentrating in vacuo, the crude
product was dissolved in water and washed with DCM (2x). The aqueous phase was basified with 5 N
NaOH and extracted with DCM. The DCM extract was concentrated in vacuo to afford 6{8} in 84% yield.
2-(4-((4-Hydroxybutyl)amino)piperidin-1-yl)benzonitrile (8). Upon reduction of 5{2} in preservative-
free THF using a 10% Pt/C CatCart®, the crude material was purified by reverse-phase preparative HPLC
ACS Paragon Plus Environment

Page 19 of 23
ACS Combinatorial Science
1
2
3
4
using a Phenomenex Gemini-NX, 10ꢀm, C-18, 110 Å, 100x50 mm column, A = 0.1% TFA in H2O, B = 0.1%
5
6
1
TFA in CH3CN, gradient 10% to 90% B over 15 min to provide 8. H NMR (300 MHz, CDCl3) δ 9.02 (br. s.,
7
8
9
2 H), 7.42 - 7.61 (m, 2 H), 6.91 - 7.11 (m, 2 H), 5.58 (s, 4 H), 3.68 (t, J = 5.41 Hz, 1 H), 3.60 (d, J = 12.28 Hz,
2 H), 3.20 (br. s., 1 H), 3.02 - 3.14 (m, 2 H), 2.86 (t, J = 11.55 Hz, 2 H), 2.22 (d, J = 10.96 Hz, 2 H), 1.80 -
2.00 (m, 4 H), 1.69 (quin, J = 5.88 Hz, 2 H). 13C NMR (300 MHz, CDCl3) δ 155.19, 134.10, 134.02, 122.43,
119.12, 118.08, 106.14, 61.37, 55.15, 50.22, 45.09, 29.32, 28.73, 23.86. MS (ESI, pos. ion) m/z: 274.1
[M+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
General procedure for N-functionalization. A stock solution of each amine 6{1-8} (0.1 M in DCM) was
aliquotted (50 mg per reaction) into 1 dram screw-cap glass vials containing reagents 2{1-15} (1.5
equiv). Triethylamine (3 equiv) was added to each vial via micropipette, the vials were tightly capped,
placed in a 24-well block, and agitated on an orbital shaker at ambient temperature. Reaction progress
was monitored by LC-MS. Upon completion, a small amount of methanol was added to each vial to
quench excess reagent and dissolve any precipitate that may have formed. The solutions were then
transferred to a 24-well polypropylene plate and solvents removed in vacuo. The residues were re-
dissolved in DMSO (1-2 mL) and filtered through a 24-well filter plate prior to HTPP. For the crude
products that were poorly soluble in DMSO, purification by silica gel column chromatography was
performed, eluting with a gradient of 0-5% or 0-6% A to B where A = 2.0 M NH3 in MeOH and B = DCM.
Solubility Assay. Compounds as 10 mM DMSO stock solutions were charged to a 96 deep-well plate at a
volume of 10
ꢀL per well. The DMSO was removed in vacuo (Genevac Evaporator, 2.5 h, 40 °C, under
full vacuum). After dry down was complete, a Tecan Evo Liquid Handler was used to transfer 200
ꢀL PBS
(1x), pH 7.4 to each well, with a comparison solution in DMSO prepared for each compound. The plates
were sealed and centrifuged at 1,000 rpm for 1 minute to push all liquid from the walls to the bottom of
the wells. The plates were shaken at 1,500 rpm on a 3 mm radius orbital shaker for 1 h. The samples
were equilibrated at room temperature for 72 h. The plates were centrifuged at 4,000 rpm for 30 min.
ACS Paragon Plus Environment

ACS Combinatorial Science
Page 20 of 23
1
2
3
4
The supernatant was analyzed by LC (215 nm, 2
ꢀL injection volume). Peak area in PBS was compared to
5
6
DMSO standard to determine solubility, accurate within the range of 1-500 ꢀM.
7
8
9
Microsomal clearance assay. Pooled human or rat liver microsomes (0.25 mg/mL) were mixed with test
compounds (1 µM) in phosphate buffer (67 mM, pH 7.4) and pre-incubated for 5 minutes at 37 °C.
Subsequently, the test samples were treated with or without 1 mM NADPH and incubated at 37 °C for
30 minutes in a total volume of 0.2 mL. Reactions were stopped by the addition of 200 µL of ice-cold
acetonitrile containing 0.5% formic acid and temazepam (500 ng/mL) as the internal standard. The
quenched test samples were centrifuged at approximately 1,650 x g for 20 minutes, and the
supernatants were analyzed directly for unchanged test compound using liquid chromatography with
tandem mass spectrometric detection (LC-MS/MS). Intrinsic clearance was calculated based on
substrate disappearance rate assuming first order elimination of compound over the 30 minute
incubation.
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
Cell permeability assay. The parental LLC-PK1 cell line (porcine renal epithelial cells, LLC-PK1) was
purchased from American Type Culture Collection (ATCC, Manassass, VA). Transfections of LLC-PK1 cells
with human MDR1 gene (hMDR1-LLC-PK1) were generated. Cells were grown as previously described19
in Medium 199 (HyClone, Logan, UT) supplemented with 2 mM L-glutamine, 50 units/mL penicillin,
50 µg/mL streptomycin, and 10% (v/v) fetal bovine serum (Invitrogen, Grand Island, NY). Cells were
seeded onto transwell filter membranes at a density of 90,000 cells/well in the presence of 1 µg/mL
human fibronectin (BD Biosciences, Bedford, MA). Media change was performed on day 3. Compound
incubations were performed 6 days post seeding. All cultures were incubated at 37 °C in a humidified
(95% relative humidity) atmosphere of 5% CO2 / 95% air.
Prior to the transport experiment, culture medium was aspirated from both apical and basolateral wells,
cells were rinsed with warmed (37 °C) Hank’s balanced salt solution supplemented with 10 mM Hepes at
pH 7.4 (HHBSS, Invitrogen) as previously described.20 HHBSS was removed from wells prior to dosing
ACS Paragon Plus Environment

Page 21 of 23
ACS Combinatorial Science
1
2
3
4
5
6
7
8
9
with test drugs at 5 µM in transport buffer (HHBSS containing 0.1% bovine serum albumin). One
hundred-fifty microliters of transport buffer were added to receiver chambers prior to dosing in
triplicate to apical or basolateral chambers. Transport studies were incubated for two hours at 37 oC on
a shaking platform. At the end of the incubation period, 100 µl samples were collected from receiver
reservoirs and analyzed by LC-MS/MS on an API4000 (Applied Biosystem, Foster City, CA) triple
quadruple mass spectrometer interfaced with turbo IonSpray operated in positive mode using Analyst
1.6.1 software.
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 apparent permeability coefficient (Papp) of all tested agents was estimated from the slope of a plot
of cumulative amount of the agent versus time based on the following equation:
Papp = (dQ / dt) / (A * C0)
where dQ / dt is the penetration rate of the agent (ng/s), A is the surface area of the cell layer on the
Transwell (0.11 cm2), and C0 is the initial concentration of the test compound (ng/ml). Efflux ratio (ER)
was calculated from the basolateral-to-apical permeability divided by the apical-to-basolateral
permeability: ER = Papp B2A/ Papp A2B.
Associated Content
Supporting Information. Tables of calculated physical properties, predicted RLM and HLM clearance,
experimental RLM/HLM clearance, experimental solubility data, final yields and characterization data for
all library members can be found in the Supporting Information.
Author Information
Corresponding author. *E-mail: edoherty@amgen.com.
Acknowledgements
The authors thank John Eschelbach, James Petersen, Olindo Lammawin Jr., Kyung Gahm, and Dean
Hickman for technical support and guidance. We also thank Les Miranda and Randy Hungate for
supporting this research.
ACS Paragon Plus Environment

ACS Combinatorial Science
Page 22 of 23
1
2
3
4
5
6
7
References
8
9
1. (a) Wildonger, R. A.; Deegan, T. L.; Lee, J. W. J. Autom. Methods Manag. Chem. 2003, 25, 57–61. (b)
Kennedy, J. P.; Williams, L.; Bridges, T. M.; Daniels, R. N.; Weaver, D.; Lindsley, C. W. J. Combi. Chem.
2008, 10, 345–354.
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. (a) Sadler, S.; Moeller, A. R.; Jones, G. B. Expert Opin. Drug Discov. 2012, 7(12), 1107-1128. (b) Malet-
Sanz, L.; Susanne, F. J. Med. Chem. 2012, 55, 4062–4098.
3. (a) Baumann, M.; Baxendale, I. R.; Kuratli, C.; Ley, S. V.; Martin, R. E.; Schneider, J. ACS Combi. Sci.
2011, 13, 405–413. (b) Battilocchio, C.; Baumann, M.; Baxendale, I. R.; Biava, M.; Kitching, M. O.; Ley, S.
V.; Martin, R. E.; Ohnmacht, S. A.; Tappin, N. D. C. Synthesis 2012, 635–647. (c) Lange, P.; James, K. ACS
Combi. Sci. 2012, 14, 570–578.
4. Hochlowski, J. E.; Searle, P. A.; Tu, N. P.; Pan, J. Y.; Spanton, S. G.; Djuric, S. W. J. Flow Chem. 2011, 1,
56–61.
5. (a) Guetzoyan, L.; Nikbin, N.; Baxendale, I. R.; Ley, S. V. Chem. Sci. 2013, 4, 764–769. (b) Desai, B.;
Dixon, K.; Farrant, E.; Feng, Q.; Gibson, K. R.; van Hoorn, W. P.; Mills, J.; Morgan, T.; Parry, D. M.;
Ramjee, M. K.; Selway, C. N.; Tarver, G. J.; Whitlock, G.; Wright, A. G. J. Med. Chem. 2013, 56 (7), 3033-
3047. (c) Czechtizky, W.; Dedio, J.; Desai, B.; Dixon, K.; Farrant, E.; Feng, Q.; Morgan, T.; Parry, D. M.; K.
Ramjee, M. K.; Selway, C. N.; Schmidt, T.; Tarver, G. J.; Wright, A. G. ACS Med. Chem. Lett. Articles ASAP,
DOI: 10.1021/ml400171b.
6. (a) Lipinski, C. A.; Lombardo, F.; Dominy, B. W.; Feeney, P. J. Adv. Drug Deliv. Rev. 2001, 46, 3–26. (b)
Veber, D. F.; Johnson, S. R.; Cheng, H. Y.; Smith, B. R.; Ward, K. W.; Kopple, K. D. J. Med. Chem. 2002, 45,
2615–2623.
7. Nadin, A.; Hattotuwagama, C.; Churcher, I. Angew. Chem. Int. Ed. Engl. 2012, 51, 1114–1122.
8. Welsch, M. E.; Snyder, S. A.; Stockwell, B. R. Curr. Opin. Chem. Bio. 2010, 14, 347–361.
9. Alabaster, C. T.; Blackburn, K. J.; Joice, J. R.; Massingham, R.; Scholfield, P. C. Br. J. Pharmacol. 1977,
60(2), 284-285.
10. Madar, D. J.; Kopecka, H.; Pireh, D.; Yong, H.; Pei, Z.; Li, X.; Wiedeman, P. E.; Djuric, S. W.; Von
Geldern, T. W.; Fickes, M. G. J. Med. Chem. 2006, 49, 6416–6420.
11. Pipeline Pilot Version 7.5.2, Accelrys, Inc. San Diego, CA.
12. Bräse, S.; Gil, C.; Knepper, K.; Zimmermann, V. Angew. Chem. Int. Ed. Engl. 2005, 44, 5188–5240.
13. Eschelbach, J. W.; Wernick, D.; Bryan, M. C.; Doherty, E. M. Aus. J. Chem. 2012, 66(2), 165-171 .
ACS Paragon Plus Environment

Page 23 of 23
ACS Combinatorial Science
1
2
3
4
5
6
14. Bryan, M. C.; Wernick, D.; Hein, C. D.; Petersen, J. V.; Eschelbach, J. W.; Doherty, E. M. Beilstein J.
Org. Chem. 2011, 7, 1141–1149.
15. Russell, H. F.; Bremner, J. B.; Bushelle-Edghill, J.; Lewis, M. R.; Thomas, S. R.; Bates, F. Tetrahedron
Lett. 2007, 48, 1637–1639.
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
16. Thomas, S.; Notari, S.; Semin, D.; Cheetham, J.; Woo, G.; Bence, J.; Schulz, C.; Provchy, J. J. Liq.
Chromatogr. Relat. Technol. 2006, 29, 699-715.
17. For HT-NMR, samples were dissolved in d6-DMSO in 96-well plates and transferred to 1.7 mm
capillary NMR tubes using a Gilson® automated liquid handler. NMR was performed on a Bruker® 400
MHz NMR spectrometer with SampleJet sample changer.
18. Lee, P.H.; Cucurull-Sanchez L.; Lu, J.; Du Y.J J. Comput. Aided Mol. Des. 2007, 21, 665-673.
19. Schinkel, A. H.; Wagenaar, E.; van Deemter, L.; Mol, C. A. A. M.; Borst, P. J. Clin. Invest. 1995, 96,
1698-1705.
20. Booth-Genthe, C.L.; Louie, S.W.; Carlini, E.J.; Li, B.; Leake, B.F.; Eisenhandler, R.; Hochman, J.H.; Mei,
Q.; Kim, R.B.; Rushmore, T.H.; Yamazaki, M. J. Pharmacol. Tox. Methods 2006, 54, 78-89.
ACS Paragon Plus Environment