Development of Small Molecules that Specifically Inhibit the D-loop Activity of RAD51

Article abstract of DOI:10.1021/acs.jmedchem.5b01762

RAD51 is the central protein in homologous recombination (HR) DNA repair and represents a therapeutic target in oncology. Herein we report a novel class of RAD51 inhibitors that were identified by high throughput screening. In contrast to many previously

Full text of DOI:10.1021/acs.jmedchem.5b01762

Subscriber access provided by Weizmann Institute of Science
Article
Development of Small Molecules that
Specifically Inhibit the D-loop Activity of RAD51
Wei Lv, Brian Budke, Michal Pawlowski, Philip P. Connell, and Alan P. Kozikowski
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.5b01762 • Publication Date (Web): 06 Apr 2016
Downloaded from http://pubs.acs.org on April 7, 2016
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 59
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
Development of Small Molecules that Specifically Inhibit the D-loop Activity
of RAD51
†§
‡§
†
‡
†
∥
∥
Wei Lv, Brian Budke, Michal Pawlowski, Philip P. Connell, and Alan P. Kozikowski
†
Department of Medicinal Chemistry and Pharmacognosy, Drug Discovery Program, University
of Illinois at Chicago, Chicago, Illinois 60612, United States
‡
Department of Radiation and Cellular Oncology, University of Chicago, Chicago, Illinois
60637, United States
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 59
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
Abstract
RAD51 is the central protein in homologous recombination (HR) DNA repair and represents
a therapeutic target in oncology. Herein we report a novel class of RAD51 inhibitors that were
identified by high throughput screening. In contrast to many previously reported RAD51
inhibitors, our lead compound 1 is capable of blocking RAD51ꢀmediated Dꢀloop formation (IC50
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
21.3 ± 7.8 µM) at concentrations that do not influence RAD51 binding to ssDNA. In human cells,
1 inhibits HR (IC50 13.1 ± 1.6 µM) without blocking RAD51’s ability to assemble into subꢀ
nuclear foci at sites of DNA damage. We determined that the active constituent of 1 is actually
an oxidized derivative (termed RI(dl)-1 or 8) of the original screening compound. Our SAR
campaign also yielded RI(dl)-2 (hereafter termed 9h), which effectively blocks RAD51’s Dꢀloop
activity in biochemical systems (IC50 11.1 ± 1.3 µM) and inhibits HR activity in human cells
(
IC 3.0 ± 1.8 ꢁM).
50
2
ACS Paragon Plus Environment

Page 3 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Introduction
Homologous recombination (HR) is an evolutionarily conserved DNA repair process, which
repairs DNA double strand breaks (DSBs) and promotes tolerance of interꢀstrand DNA crossꢀ
links (ICLs). Unlike the errorꢀprone nonꢀhomologous endꢀjoining (NHEJ) pathway of DSB
rejoining, HR faithfully repairs DNA damage by utilizing an undamaged homologous DNA
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
1
, 2
template to guide the repair process. An initial step of HR involves nuclease processing that
generates a 3' singleꢀstranded DNA (ssDNA) tail at the site of DNA damage (Figure 1). The
ssDNA is then coated with RAD51 protein to form a helical nucleoprotein filament. This
resulting nucleoprotein filament subsequently searches for a homologous DNA sequence and
invades it to form a joint molecule intermediate, termed a Dꢀloop. Finally, with the assistance of
other related HR proteins, accurate DNA synthesis is performed using the undamaged
homologous sequence as a template.
Since HR facilitates cellular recovery from harmful DNA lesions, cells with deficient HR
functionality are especially vulnerable to radiotherapy and chemotherapeutic agents that generate
2
ꢀ4
DSBs or DNA replicationꢀblocking lesions. RAD51, the central protein of HR, is overꢀ
5
, 6
expressed in various types of human cancer cells. These high expression levels of RAD51 can
elevate HR efficiency in cancer cells, thereby inducing cellular resistance to DNAꢀdamaging
7
ꢀ10
chemotherapy and radiotherapy.
Inhibition of RAD51 by antisense RNAs or RAD51
inhibitors are reported to enhance the sensitivity of cancer cells to chemotherapy and
1
1ꢀ13
radiotherapy.
the development of small molecule therapeutics to sensitize tumors.
molecule RAD51 inhibitors have been developed, most of these inhibitors act by preventing the
These observations support the concept of RAD51 as a therapeutic target and
1
4ꢀ16
Although several small
1
3, 14, 16ꢀ19
formation of RAD51ꢀssDNA nucleoprotein filaments.
One possible exception is
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 59
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
20
halenaquinone, a chemical purported to specifically inhibit Dꢀloop formation by RAD51.
However, halenaquinone also reduces the subꢀnuclear appearance of RAD51 foci at DSB sites in
cells, which raises questions as to its mechanistic interaction with RAD51 in cells.
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
In addition to its function in HR, RAD51 filaments also serve an important function at
2
1, 22
stalled DNA replication forks.
These RAD51 nucleoprotein filaments protect the ssDNA
from extensive nuclease processing, thereby aiding in replication and preserving genomic
integrity. Since most RAD51 inhibitors act by preventing the formation of RAD51 nucleoprotein
filaments on ssDNA, they are expected to inhibit both HR and replication fork stabilization
activities of RAD51. As such, their use is predicted to generate unintended toxicity to normal
cells by interrupting this centrallyꢀimportant function in normal replication.
We reasoned that specifically inhibiting RAD51ꢀmediated Dꢀloop formation while
preserving RAD51’s ability to form nucleoprotein filaments may target DNA repair while
minimizing replicationꢀassociated toxicity in normal tissue. Therefore, we carried out a twoꢀpart
highꢀthroughput (HT) screen of the ASDI diversity library and LOPAC1280 compound library
for drug candidates that selectively target RAD51’s Dꢀloop function. The preferred lead
compound resulting from this screen, compound 1, inhibits RAD51’s Dꢀloop activity while
having minimal effects on RAD51’s ssDNA binding activity. Compound 1 is a commercially
obtained sample labeled as 4ꢀ(7ꢀmethoxyꢀ4,5ꢀdihydropyrrolo[1,2ꢀa]quinoxalinꢀ4ꢀyl)ꢀN,Nꢀ
dimethylaniline, but its structure was found to be inconsistent with the named compound. This
report details our HT screen and subsequent structureꢀactivity relationship (SAR) optimizations
performed to improve the potency and selectivity of 1.
4
ACS Paragon Plus Environment

Page 5 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Results and Discussion
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
Screen for RAD51 D-loop inhibitors. We carried out two parallel HT screens to identify
small molecules that inhibit RAD51 Dꢀloop activity without affecting its ssDNA binding activity.
A previouslyꢀdescribed fluorescence polarizationꢀbased microplate assay was used to assess the
2
3
compounds’ effects on RAD51ꢀssDNA binding (Figure 2a). In order to enable a screen for Dꢀ
loop specific RAD51 inhibitors, we previously developed a microplateꢀbased assay that can
2
4
accurately quantify Dꢀloop formation in a highꢀthroughput fashion (Figure 2b). This method
makes use of two substrates: a closed linear doubleꢀhairpin dsDNA, and an ssDNA fused with a
black hole quencher. The sequence homology of the two substrates enables RAD51 to catalyze
the formation of Dꢀloops, which is quantified as a function of quenched fluorescence. Among the
screening compounds tested, we focused on those that prevented RAD51 Dꢀloop formation while
exerting little effect on RAD51ꢀssDNA binding (Figure 2c). Potential hits were confirmed using
secondary lowꢀthroughput reꢀtesting, which involved a gelꢀbased Dꢀloop assay to exclude the
possibility that the observed inhibition of Dꢀloop activity was due to artefactual interference by
compound autoꢀfluorescence or fluorescence quenching.
Using this screening platform, the ASDI diversity library (6800 compounds of diverse
chemical structures) and the LOPAC library of known drugs (library of 1280 pharmacologically
active compounds) were screened, leading to the identification of 10 hit compounds that were
capable of inhibiting Dꢀloop activities with minimal effects on ssDNA binding. The best of these
10 compounds is shown in Figure 3 and hereon referred to as 1. Compound 1 inhibits the Dꢀloop
activity of RAD51 by up to 74%, while exerting minimal doseꢀdependent inhibition of RAD51
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 59
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
filament assembly in the same concentration range (Figure 3). As expected, parallel
experiments demonstrate that the more generalized inhibitors of RAD51 (like RIꢀ1) do generate
doseꢀdependent inhibition of RAD51 filament assembly using identical assay conditions (see
Figure 7b for reference).
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
Synthesis and activities of the reduced and oxidized forms of 1. As compound 1 was
identified from a commercial compound library, we decided to first confirm its chemical
structure by carrying out our own independent synthesis as detailed in Scheme 1. The starting
material 2a was first condensed with dimethoxytetrahydrofuran (3) in refluxing acetic acid to
29
29
afford 4a. The nitro group of 4a was then reduced, and the aniline 5a was reacted with
benzaldehyde 6a in the presence of 1,2,3ꢀbenzotriazole and AlCl to provide the desired product
3
7 in moderate yield.
a
Scheme 1. Synthesis of 7
a
Reagents and conditions: (a) AcOH, reflux, 3 h, 90%; (b) H , Pd/C, MeOH, 5 h, 97%; (c) 1,2,3ꢀ
2
benzotriazole, AlCl , THF, 4 h, 32%.
3
The newly synthesized compound 7 was then subjected to biological evaluation as shown in
Table 1. Unexpectedly, compound 7 displayed a much weaker Dꢀloop inhibitory activity in
comparison to the compound 1 that we obtained from the commercial source (maximum
inhibition of 17% for 7 vs 74% for 1). Moreover, compound 7 also showed significant
6
ACS Paragon Plus Environment

Page 7 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
stimulatory activity in RAD51ꢀssDNA binding when compared to 1. These discrepancies
suggested that compound 7 and 1 might be different substances. Interestingly, we observed that
compound 7 was not stable when exposed to air at room temperature; instead, it slowly oxidized
over time to compound 8 (Scheme 2). Additionally, we found during the preparation of
compound 7, some of the oxidized product 8 was always produced. These observations raised the
possibility that the commercially obtained material 1 might have undergone oxidation during its
storage, and that the oxidized form 8 might be the active constituent responsible for the Dꢀloop
inhibitory activity of 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
To confirm this hypothesis, the structural identity of compound 1 was determined.
1
According to the H NMR spectrum, the commercially obtained material 1 was impure, with the
oxidized compound 8 as the major component (Figure S1). The presence of compound 7 was not
observed. The mass spectrum also confirmed the presence of compound 8 rather than compound
7
. To further confirm the Dꢀloop inhibitory activity of this oxidized compound, compound 8 was
synthesized by following the route as shown in Scheme 2. The aniline 5a and benzaldehyde 6a
were reacted in the presence of 1,2,3ꢀbenzotriazole and AlCl at room temperature until the
3
conversion was complete, which resulted in a mixture of the reduced (7) and oxidized (8) forms.
This crude mixture was then heated in acetone with MnO to induce further oxidization, thereby
2
yielding the oxidized form 8 as the exclusive product. Subsequent testing in our Dꢀloop assay
(Table 1) showed that compound 8 exhibits similar activity to 1 in terms of maximal inhibition
(
67% vs 74%, respectively) and IC values (13.0 vs 21.3 ꢁM, respectively). These results thus
50
confirm that the oxidized form 8 is the actual substance that generates the Dꢀloop inhibitory
activity of the commercially obtained compound 1. Therefore, we have named compound 8 as
“RI(dl)-1”, for Dꢀloop activity of RAD51 Inhibitor #1.
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 59
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
a
Scheme 2. Synthesis of 8
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Reagents and conditions: (a) 1,2,3ꢀbenzotriazole, AlCl , THF, overnight; (b) MnO , acetone, 55
3
2
o
C, 24 h, 59% in two steps.
Table 1. The RAD51 Dꢀloop inhibitory and RAD51ꢀssDNA binding stimulatory/inhibitory
activities of 1, 7 and 8.
DꢀLoops
ssDNA binding
Max
inhibition
effect
74±2%
17±3%
67±11%
Cpd
1
IC (ꢁM)
Max effect IC (ꢁM)
5
0
50
21.3±7.8
>100
33±10%↓
47±33%↑
40±28%↑
>100
Stim
Stim
7
8
13.0±4.7
“↑” Indicates that the compound stimulates RAD51ꢀssDNA binding. “↓” Compound inhibits
RAD51ꢀssDNA binding. “Stim” Compound is a stimulator.
Structure optimization of 8. To improve upon the Dꢀloop inhibitory activity, medicinal
chemical studies were undertaken to optimize 8. An overview of the optimization strategy is
summarized in Figure 4. First, different functional groups were appended to rings A and B as
substituents, including alkyl groups of different sizes, electron withdrawing/donating groups, and
hydrogen bond donor/acceptor groups. Next, replacement of the phenyl ring B with different
heterocycles or other saturated/unsaturated rings was investigated in order to probe its role. For
8
ACS Paragon Plus Environment

Page 9 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
each of the prepared compounds, we used the gelꢀbased Dꢀloop assay and the microplateꢀbased
DNA binding assay to characterize the biochemical activities.
Optimization of ring B. To synthesize analogues with different substituents in the Bꢀring,
the aniline 5e was treated with different benzaldehydes 6a-h in the presence of 1,2,3ꢀ
benzotriazole and AlCl (in Scheme 3). Then the crude products were oxidized with MnO to
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
3
2
provide 9a-j in good yield. Interestingly, when benzaldehyde 6f was reacted with the aniline 5e
under these conditions, partial deꢀethylation was observed and the monoꢀethylated product 9h
was isolated as the major product. The analogues 9c and 9g were obtained by reducing the nitro
31
group of 9b and 9f via hydrogenation. To synthesize analogues in which the Bꢀring was
replaced with other aromatic or saturated rings, the aniline 5e was reacted with different
3
1-33
aldehydes 10a-d under similar conditions (in Scheme 4) to afford the products 11a-d
in good
yield.
a
Scheme 3. Synthesis of Analogues 9a-j
a
Reagents and conditions: (a) 1,2,3ꢀbenzotriazole, AlCl , THF, overnight; (b) MnO , acetone, 55
3
2
o
C, 24 h, 19ꢀ70% in two steps; (c) H , Pd/C, methanol, 5h, 57ꢀ87%.
2
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 59
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
a
Scheme 4. Synthesis of Analogues 11a-d
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Reagents and conditions: (a) 1,2,3ꢀbenzotriazole, AlCl , THF, overnight; (b) MnO , acetone, 55
3
2
o
C, 24 h, 35ꢀ71% in two steps.
The Dꢀloop inhibitory activities measured for compounds 9a-j and 11a-d are shown in
Table 2 (see Dꢀloop gels in Figure S2). Results indicate that a dimethylamino group (9a), amino
group (9g), ethylamino group (9h) or morpholine ring (9j) in the para position of the Bꢀring
allow for Dꢀloop inhibition, and that 9g and 9h display potencies that are similar to that of
compound 8. By contrast, introducing a hydroxyl group (9d), trifluoromethoxyl group (9e), nitro
group (9f), or pyrrolidine group (9i) in the paraꢀposition significantly decreased the Dꢀloop
inhibitory activities. Moving the amino group from the paraꢀ (9g) to the metaꢀposition (9c) also
decreased the Dꢀloop inhibitory activity. Replacing the Bꢀring with other aromatic or saturated
rings only resulted in moderate (11a, 11b) or weak (11c, 11d) Dꢀloop inhibitors. The ability of
these analogs to modulate RAD51’s ssDNA binding was also evaluated as shown in Table 2.
Many of these compounds had only a weak influence on RAD51ꢀssDNA binding. On the other
hand, compounds 9a, 9b, 9i, 11c, and 11d had a modest influence on RAD51ꢀssDNA binding.
1
0
ACS Paragon Plus Environment

Page 11 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 2. The RAD51 Dꢀloop inhibitory and RAD51ꢀssDNA binding stimulatory/inhibitory
activities of 9a-j and 11a-d.
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
DꢀLoops
Max
ssDNA binding
inhibition
effect
Cpd
R
IC (ꢁM)
Max effect IC (ꢁM)
5
0
50
9a
4ꢀN(CH
3
)
2
61±4%
20±8%
79±8%
38±3%
22±8%
2±21%
100±2%
91±1%
27.6±6.0
>100
88±22%↑
52±14%↓
11±9%↓
Stim
>100
>100
>100
Stim
>100
>100
>100
9
b
3ꢀNO
3ꢀNH
2
9c
2
91.1±10.3
>100
9
d
4ꢀOH
14±16%↓
30±21%↑
16±13%↓
18±11%↓
2±24%↑
9
e
4ꢀOCF
3
>100
9
f
4ꢀNO
2
>100
9
g
4ꢀNH
2
22.1±2.4
11.1±1.3
9
h
4ꢀNHEt
9i
9j
8±7%
>100
69±35%↑
11±19%↓
Stim
>100
68±7%
27.2±3.0
1
1a
70±13%
63±4%
5±11%
71.7±18.1
79.4±6.0
>100
6±10%↓
7±38%↓
42±17%↑
>100
>100
Stim
1
1b
11c
1
1d
42±11%
>100
56±31%↑
Stim
“
↑” Compound stimulates RAD51ꢀssDNA binding. “↓”Compound inhibits RAD51ꢀssDNA
binding. “Stim” Compound is a stimulator.
Optimization of ring A. To synthesize analogues with different substituents in the Aꢀring,
the substituted anilines 5b-d were first prepared according to the route shown in Scheme 5. The
starting materials 2b-d were condensed with dimethoxytetrahydrofuran (3) in refluxing acetic
1
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 59
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
30
acid to afford 4b-d. Then, the nitro group was reduced to provide the products 5b-d.
Subsequently, the anilines 5a-d were reacted with benzaldehydes (6a, 6c and 6e) to provide the
analogues 12a-m in good yield (in Scheme 6). For analogues with an amino group (12c, 12f, 12i
and 12l), compounds were prepared from the corresponding nitro analogues (12b, 12e, 12h, 12k)
via hydrogenation. Two analogues (13a-b) with a 3ꢀpyridine ring were also prepared from
anilines 5a-b and aldehyde 10a as shown in Scheme 7.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
a
Scheme 5. Synthesis of 5b-d
Reagents and conditions: (a) AcOH, reflux, 3 h, 45ꢀ63%; (b) H , Pd/C, MeOH, 5 h, 37ꢀ97%.
2
a
Scheme 6. Synthesis of Analogues 12a-m
1
2
ACS Paragon Plus Environment

Page 13 of 59
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
Reagents and conditions: (a) 1,2,3ꢀbenzotriazole, AlCl , THF, overnight; (b) MnO , acetone, 55
3
2
o
C, 24 h, 7ꢀ77% in two steps; (c) H , Pd/C, MeOH, 5 h, 50ꢀ93%.
2
a
Scheme 7. Synthesis of Analogues 13a-b
Reagents and conditions: (a) 1,2,3ꢀbenzotriazole, AlCl , THF, overnight; (b) MnO , acetone, 55
3
2
o
C, 24 h, 52ꢀ78% in two steps.
1
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 59
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
To explore the effects of replacing the hydroxyl group of compound 12f with different
alkyl groups, compound 12e was first treated with allyl bromide to introduce the allyl group
(14a), and then the nitro group was reduced by hydrogenation (in Scheme 8). The side chain
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
double bond was also reduced under these conditions to afford the nꢀpropoxy product 15a. The
fluoroethyl compound 15b was prepared in a similar fashion but employed the Mitsunobu
2
5, 26
reaction to introduce the side chain (in Scheme 8).
An analogue 16 with a 2ꢀ
methylpropionamide side chain was also prepared as shown in Scheme 9.
a
Scheme 8. Synthesis of Analogues 15a-b
o
Reagents and conditions: (a) allyl bromide, Cs CO , DMF, microwave, 80 C, 30 min, 58%; (b)
2
3
o
FCH CH OH, DEAD, Ph P, THF, microwave, 60 C, 40 min; (c) H , Pd/C, MeOH, overnight,
2
2
3
2
36ꢀ45%.
a
Scheme 9. Synthesis of Analogues 16
o
Reagents and conditions: (a) Br(CH ) CHCONH , NaOH, DMA, 50 C, 3 h, 41%.
3
2
2
1
4
ACS Paragon Plus Environment

Page 15 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The Dꢀloop inhibitory activities for compounds 12a-m, 13a-b, 15a-b and 16 are shown in
Table 3. These results indicate that a methoxyl group in the 7 position of Aꢀring improves its Dꢀ
loop inhibitory activity, which can be seen by comparing compound 12c (IC 13.2 µM) against
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
9g (IC 22.1 µM). A similar improvement was observed when a hydroxyl group was introduced
5
0
at the same position (compare 12f, IC 8.4 µM and 12g, IC 9.8 µM against 9g, IC 22.1 µM
5
0
50
50
and 9a, IC 27.6 µM). In contrast, the introduction of a fluorine atom in the 7 position or two
5
0
fluorine atoms in the 7 and 8 positions of the Aꢀring yielded compounds whose Dꢀloop inhibitory
potential was moderately diminished (compound 12i IC50 25.0 µM and 12l IC50 34.0 µM) or
nearly absent (compound 12j and 12m). For compounds with 3ꢀpyridyl substituents (compound
13a and 13b), the incorporation of a methoxy or hydroxyl group in the Aꢀring yielded a slight
improvement in Dꢀloop inhibition compared with 11a. The appendage of larger alkoxy groups at
the 7 position of the Aꢀring (compounds 15a-b and 16) did not provide more potent Dꢀloop
inhibitors (except for compound 15b), when compared to 9g.
The influence of these alterations on RAD51ꢀssDNA binding activity was also evaluated
for these analogs (Table 3). Most of the compounds influenced RAD51ꢀssDNA binding weakly
or modestly, except for compounds 12b, 15a, 15b, and 16, which were stimulators of RAD51ꢀ
ssDNA binding, and 12a, which showed relatively strong ssDNA binding inhibition. Among the
most potent Dꢀloop inhibitors in this series, 12c and 12g only weakly influenced ssDNAꢀRAD51
binding; 12f and 15b exerted modest effects on ssDNAꢀRAD51 binding.
1
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 59
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 3. The RAD51 Dꢀloop inhibitory and RAD51ꢀssDNA binding stimulatory/inhibitory
activities of 12a-m, 13a-b, 15a-b and 16.
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
DꢀLoops
Max
ssDNA binding
inhibition
effect
Cpd
2a
2b
2c
2d
2e
2f
2g
2i
12j
2l
12m
3a
13b
5a
R1
7ꢀOCH
7ꢀOCH
7ꢀOCH
7ꢀOH
7ꢀOH
7ꢀOH
7ꢀOH
7ꢀF
R2
IC (ꢁM)
Max effect IC (ꢁM)
50
50
1
3
3
3
ꢀOH
13±12%
17±7%
100±4%
97±8%
13±10%
100±3%
91±8%
95±6%
31±4%
59±4%
19±15%
74±5%
53±30%
75±7%
>100
>100
100±18%↓ 35.0±14.0
1
ꢀNO
2
105±47%↑
29±8%↓
Stim
>100
1
ꢀNH
2
13.2±2.8
27.6±4.9
>100
1
ꢀOH
72±11%↓
92±21%↓
>100
1
ꢀNO
2
57.8±3.0
1
ꢀNH
2
8.4±3.4
9.8±3.5
25.0±5.1
>100
82±23%↓ 96.9±44.9
34±32%↑ Stim
60±18%↓ 81.5±16.6
1
ꢀN(CH
3
3
3
)
2
)
2
)
2
1
ꢀNH
2
7ꢀF
ꢀN(CH
57±28%↑
42±8%↓
Stim
>100
Stim
>100
>100
Stim
1
7,8ꢀdiF
ꢀNH
2
34.0±2.9
>100
7,8ꢀdiF ꢀN(CH
65±22%↑
16±28%↓
87±10%↓
117±18%↑
1
7ꢀOCH
3
38.6±6.2
64.9±31.6
46.1±2.8
7ꢀOH
1
ꢀNH
2
15b
ꢀNH
2
98±5%
20.0±4.1
79±23%↑
Stim
16
ꢀN(CH
3
)
2
38±4%
>100
131±28%↑
Stim
“↑” Compound stimulates RAD51ꢀssDNA binding. “↓”Compound inhibits RAD51ꢀssDNA
binding. “Stim” Compound is a stimulator.
1
6
ACS Paragon Plus Environment

Page 17 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Cellular effects of 1 and 9h. We used the DRꢀGFP assay to test the ability of select
2
7
compounds to inhibit HR in human cells. Briefly, cells containing the chromosomal DRꢀGFP
reporter were transfected with a plasmid that expresses IꢀSceI, a rareꢀcutting endonuclease that
makes a DSB within the DRꢀGFP cassette. Compounds were added at the time of transfection
and the cells were outgrown for 24 hours. We noted that for longer outgrowth periods such as 48
hours, cells regained some HR activity (data not shown). This suggests the possibility of
compensatory responses that oppose RAD51 inhibition by compounds; these effects will be
detailed in a separate report. To overcome this technical limitation, we selected the relatively
early time point of 24 hours to describe the in vivo effects of these compounds.
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
We demonstrated that 1 is able to inhibit HR activity in human cells (Figure 5A).
Importantly, 1 yields 50% inhibition of cellular HR activity at 13.1 µM, which is comparable to
its observed activity in biochemical assays. The HR inhibitory activity of 1 reaches 83% at
concentrations of 40 µM, which is a dose where the RAD51ꢀssDNA binding activity remains
unaffected in biochemical assays. Next, we compared 1 against its pure oxidized and pure
reduced forms. In contrast to our results using biochemical assays, the inhibitory activity of 7 in
the cellꢀbased HR assay was comparable to that of 1 and 8. As noted earlier, 7 spontaneously
oxidizes to 8. As such, this observed inhibitory activity of 7 is likely caused by its oxidation into
the active constituent (8) during the HR assay; such would be expected given the relatively long
(24ꢀhour) incubation of 7 in an aerated tissue culture medium.
Next we used this cellꢀbased HR quantification method to evaluate analogs that exhibited
IC values less than 20 µM in biochemical Dꢀloop assays. While compounds 12f and 12g had
50
the lowest IC50 values in the Dꢀloop assay, they both yielded comparable or slightly less
inhibitory activity in the DRꢀGFP assay relative to 1 and 8. The most potent inhibitor of cellular
1
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 59
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
HR was 9h, having an IC50 of 3.0 ± 1.8 ꢁM, which is over fourꢀfold lower than the IC50
concentrations of the starting lead compound 1.
Compound 9h was tested further using cells containing the chromosomal DRꢀSSA
reporter, which quantifies the ability of the cells to perform singleꢀstrand annealing (SSA), as a
means to repair IꢀSceI induced DSBs. SSA is an alternative method of reꢀligating DNA when
HR cannot be completed. SSA efficiency is known to be elevated in situations where RAD51
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
2
7, 28
function has been generally disrupted
. Similarly, we previously demonstrated that both RIꢀ1
1
9
and RIꢀ2 induce elevated levels of SSA activity . By contrast, we found that 9h does not
stimulate SSA, using a relevant concentration range that inhibits HR (Figure 5b). This suggests
that 9h stabilizes nucleoprotein filaments in a nonꢀfunctional state, which are incapable of Dꢀ
loop activity and simultaneously shielded from related (e.g. RAD52ꢀmediated) pathways that
promote SSA.
We tested whether 9h permits the timely assembly of RAD51 into subꢀnuclear foci in
response to DNA damage. As expected, 9h does not inhibit the appearance of RAD51 into subꢀ
nuclear foci in response to radiation (Figure 6a), even at concentrations of 9h that had inhibited
cellular HR by 75% in prior DRꢀGFP assays (Figure 5b). Importantly, 9h does not affect the coꢀ
localization of RAD51 foci with RPA, an HR associated protein that binds to ssDNA prior to the
recruitment of RAD51 (Figure 6b). We observed 57% coꢀlocalization in 9h treated cells and
60% coꢀlocalization in vehicleꢀtreated cells. This important control confirms that our imaged
subꢀnuclear RAD51 puncta represent true DSBꢀassociated foci, rather than offꢀtarget
accumulations of RAD51 protein. Additionally, 9h does not affect RAD51 protein levels
(cytosolic, nuclear, or chromatinꢀbound fractions) at comparable time points (Figure S3). Our
interpretation is further substantiated by quantitative gelꢀshift assays (Figures 7a-c, S4). We also
1
8
ACS Paragon Plus Environment

Page 19 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
tested the selectivity of 9h as a specific inhibitor of human RAD51 by comparing its ability to
inhibit the Dꢀloop activity of human RAD51 versus the distantlyꢀrelated S. cerevisiae Rad51 and
E. coli RecA recombinase proteins and found no significant inhibition of these proteins by 9h
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(
Figure 7d-f). Unlike the generalized RAD51 inhibitor RIꢀ1, 9h does not modulate the affinity
of RAD51 to ssDNA or the stability of preꢀformed RAD51ꢀssDNA complexes when challenged
with high concentrations of salt.
Since 9h inhibits HRꢀmediated repair of DSBs, we next investigated whether 9h could
sensitize cancer cells to radiation damage. Indeed, we observed that 9h significantly sensitizes
three different cancer cell lines, when 9h was administered immediately after following
irradiation and allowed to incubate with cells for the duration of clonogenic outgrowth thereafter
(Figure 8). Given the substantial improvements of 9h over the starting compound identified in
our high throughput screen, we renamed compound 9h as RI(dl)-2, for Dꢀloop activity of
RAD51 Inhibitor #2.
Conclusion
A common hallmark of tumor cells is the overꢀexpression of the RAD51 recombinase
protein. Numerous lines of evidence indicate that tumor cells are especially dependent upon the
11, 12, 29
HR function of RAD51.
This wellꢀestablished feature of cancer cells has made RAD51 an
attractive drug target, and several small molecules that inhibit cellular RAD51 function have
16
been described as reviewed by Huang and Mazin . In addition to its classical role in catalyzing
homologous strand exchange, RAD51 additionally plays a central role in protecting stalled
2
1, 22
replication forks from nucleolytic degradation.
This protective function in replication
requires that RAD51 is able to bind ssDNA, however it may not require the strand invasion
1
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 59
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
activity of RAD51. Therefore, we carried out a drug development campaign to inhibit RAD51’s
Dꢀloop activity while sparing its ssDNA binding activity. Compound 1 was identified as a
promising lead compound, and subsequent SAR efforts yielded compound 9h with improved
biochemical and cellꢀbased activities.
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 believe that this series of compounds represents a major step forward in the
development of RAD51ꢀtargeting small molecules. To our current knowledge, these are the only
known compounds that specifically inhibit RAD51’s Dꢀloop activity without interfering with its
ability to bind ssDNA both in vitro and in cells. Halenaquinone was previously reported to
strongly inhibit RAD51’s Dꢀloop activity while sparing its ssDNA binding activity in
2
0
biochemical assays. However, Halenaquinone was shown to inhibit the formation of RAD51
foci in human cells following irradiation, suggesting that in a cellular context this compound
may prevent sufficient loading of RAD51 onto damaged DNA at the site of radiationꢀinduced
DSBs to form visible foci. Other characterized RAD51 inhibitors including our previouslyꢀ
1
3,
described inhibitors RIꢀ1 and RIꢀ2 act by preventing RAD51 from loading onto damaged DNA.
19
Compound 1 and its analogs represent potential cancer therapeutics aimed at sensitizing
tumors to DNAꢀdamaging therapies. Considering its high potency for inhibiting both Dꢀloop
formation and HR, compound 9h provides an important candidate for further investigation.
Meanwhile, since compound 9h has almost no impact on RAD51ꢀssDNA binding, it also
presents a novel tool for basic scientists who wish to study the different functions of RAD51.
2
0
ACS Paragon Plus Environment

Page 21 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Experimental Section
1
13
General. The nuclear magnetic resonance ( H and C NMR) spectra were obtained using
a Bruker spectrometer with TMS as an internal standard. Automated column chromatography
was performed using the CombiFlash Rf apparatus loaded with Merck silica gel (40−60 mesh).
Preparative HPLC was carried out using a Shimadzu preparative liquid chromatograph with a
column from ACE 5 AQ (150 mm × 21.2 mm) with 5 ꢁm particle size. All solvents used in
preparative HPLC were spiked with 0.05% TFA. The purities of biologically important
compounds are ≥ 95% as determined by analytical HPLC (ACE 3AQ C18 column (150 mm ×
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
.6 mm, particle size 3 ꢁM), 0.05% TFA in H O/0.05% TFA in MeOH gradient eluting system).
2
General procedure for preparation of 4a-d. A solution of 2a-d (20 mmol) and 3 (20
mmol) in acetic acid (30 mL) was heated to reflux for 3 h. After cooling down, the mixture was
poured into water (150 mL) and extracted with diethyl ether (150 mL X 3). The organic layers
were combined and dried over Na SO . Remove solvent provided the crude products of 4a-d.
2
4
30
1-(4-Methoxy-2-nitrophenyl)-1H-pyrrole (4a). The crude product of 4a was further
purified by flash silica gel column chromatography (ethyl acetateꢀhexanes, gradient up to 1:4) to
1
provide the product as orange solid (90%). H NMR (400 MHz, CDCl ) δ 7.41ꢀ7.38 (m, 2 H),
3
1
3
7
.19ꢀ7.16 (m, 1 H), 6.76 (s, 2 H), 6.35 (s, 2 H), 3.92 (s, 3 H); C NMR (100 MHz, CDCl ) δ
3
158.3, 145.6, 129.0, 126.9, 121.4, 118.8, 110.0, 109.1, 55.8.
31
3-Nitro-4-(1H-pyrrol-1-yl)phenol (4b). The crude product of 4b was further purified by
flash silica gel column chromatography (ethyl acetateꢀhexanes, gradient up to 1:2) to provide the
1
product as orange solid (59%). H NMR (400 MHz, methanolꢀd and CDCl ) δ 7.25ꢀ7.22 (m, 2
4
3
13
H), 7.05ꢀ7.02 (m, 1 H), 6.68 (s, 2 H), 6.25 (s, 2 H); C NMR (100 MHz, methanolꢀd and
4
CDCl ) δ 156.5, 145.6, 129.0, 125.6, 121.5, 119.8, 110.7, 109.5.
3
2
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
-(4-Fluoro-2-nitrophenyl)-1H-pyrrole (4c). The crude product of 4c was further
purified by flash silica gel column chromatography (ethyl acetateꢀhexanes, gradient up to 1:7) to
1
provide the product as yellow oil (63%). H NMR (400 MHz, CDCl ) δ 7.64ꢀ7.61 (m, 1 H), 7.52ꢀ
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
7
.48 (m, 1 H), 7.43ꢀ7.39 (m, 1 H), 6.80ꢀ6.78 (m, 2 H), 6.40ꢀ6.38 (m, 2 H).
1-(4,5-Difluoro-2-nitrophenyl)-1H-pyrrole (4d). The crude product of 4d was further
purified by flash silica gel column chromatography (ethyl acetateꢀhexanes, gradient up to 1:9) to
1
provide the product as yellow oil (45%). H NMR (400 MHz, CDCl ) δ 7.85ꢀ7.80 (m, 1 H), 7.36ꢀ
3
7.32 (m, 1 H), 6.78ꢀ6.76 (m, 2 H), 6.40ꢀ6.38 (m, 2 H).
General procedure for preparation of 5a-d. A suspension of 4a-d (3 mmol) and
palladium on charcoal (10%, 60 mg) in methanol (10 mL) was stirred under a hydrogen balloon
for 5 h. The black solid was removed by filtration and the filtrate was concentrated to provide the
crude product of 5a-d.
30
5
-Methoxy-2-(1H-pyrrol-1-yl)aniline (5a). The crude product of 5a was further
purified by flash silica gel column chromatography (ethyl acetateꢀhexanes, gradient up to 1:4) to
1
provide the product as pale yellow oil (97%). H NMR (400 MHz, CDCl ) δ 7.13ꢀ7.09 (m, 1 H),
3
1
3
6.83ꢀ6.82 (m, 2 H), 6.40ꢀ6.37 (m, 4 H), 3.84 (s, 3 H); C NMR (100 MHz, CDCl ) δ 159.9,
3
1
43.4, 128.2, 122.1, 121.2, 109.2, 103.6, 101.1, 55.4.
3-Amino-4-(1H-pyrrol-1-yl)phenol (5b). The crude product of 5b was further purified by
flash silica gel column chromatography (ethyl acetateꢀhexanes, gradient up to 1:2) to provide the
1
product as yellow solid (37%). H NMR (400 MHz, methanolꢀd and CDCl ) δ 6.93ꢀ6.90 (m, 1
4
3
1
3
H), 6.71ꢀ6.69 (m, 2 H), 6.26ꢀ6.24 (m, 3 H), 6.22ꢀ6.19 (m, 1 H); C NMR (100 MHz, methanolꢀ
d and CDCl ) δ 156.7, 142.9, 127.8, 121.8, 120.1, 108.4, 105.1, 102.0.
4
3
2
2
ACS Paragon Plus Environment

Page 23 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
5
-Fluoro-2-(1H-pyrrol-1-yl)aniline (5c). The crude product of 5c was further purified by
flash silica gel column chromatography (ethyl acetateꢀhexanes, gradient up to 1:7) to provide the
1
product as white solid (97%). H NMR (400 MHz, CDCl ) δ 7.16ꢀ7.12 (m, 1 H), 6.85ꢀ6.83 (m, 2
3
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
H), 6.55ꢀ6.50 (m, 2 H), 6.42ꢀ6.41 (m, 2 H).
4,5-Difluoro-2-(1H-pyrrol-1-yl)aniline (5d). The crude product 5d was further purified
by flash silica gel column chromatography (ethyl acetateꢀhexanes, gradient up to 1:7) to provide
1
the product as white solid (87%). H NMR (400 MHz, CDCl ) δ 7.06ꢀ7.00 (m, 1 H), 6.82ꢀ6.80
3
(m, 2 H), 6.62ꢀ6.56 (m, 1 H), 6.39ꢀ6.38 (m, 2 H).
4-(7-Methoxy-4,5-dihydropyrrolo[1,2-a][13]quinoxalin-4-yl)-N,N-dimethylaniline (7).
A mixture of the aniline 5a (94 mg, 0.50 mmol), the benzaldehydes 6a (75 mg, 0.50 mmol),
1,2,3ꢀbenzotriazole (59 mg, 0.50 mmol) and AlCl (18 mg, 0.13 mmol) in THF (4 mL) was
3
stirred at room temperature for 4 h. The mixture was diluted with water (15 mL) and extracted
with ethyl acetate (15 mL X 3). The organic layers were combined, washed with saturated
NaHSO solutions (15 mL) three times, and then washed with 2 N NaOH aqueous solution (15
3
mL) and brine (15 mL) sequentially, dried over Na SO . The solvent was evaporated, the residue
2
4
was further purified by flash silica gel column chromatography (ethyl acetateꢀhexanes, gradient
up to 1:3). The purified product was recrystallized in methanolꢀacetonitrile (1:1, 3 mL) solutions
1
to provide the pure product of 7 as white solid (52 mg, 32%). H NMR (400 MHz, CDCl ) δ
3
7.35ꢀ7.33 (m, 2 H), 7.27ꢀ7.24 (m, 1 H), 7.15ꢀ7.13 (m, 1 H), 6.76ꢀ6.73 (m, 2 H), 6.42ꢀ6.38 (m, 1
H), 6.31ꢀ6.30 (m, 1 H), 6.25ꢀ6.23 (m, 1 H), 5.62ꢀ5.60 (m, 1 H), 5.44 (s, 1 H), 4.11 (s, 1 H), 3.79
1
3
(
s, 3 H), 2.98 (s, 6 H); C NMR (100 MHz, CDCl ) δ 156.6, 150.2, 137.2, 129.8, 128.6, 128.4,
3
119.5, 115.0, 113.4, 112.0, 109.1, 104.8, 103.7, 100.8, 55.4, 55.0, 40.2; HRESIMS m/z cacld for
2
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
+
C H N O (MH ) 320.1763, found 320.1768. HPLC purity 98.3%, 98.9% (Cꢀ18 reverse phase,
20
22
3
MeOHꢀH O).
2
General procedure for preparation of 8, 9a-j, 11a-d, 12a-m and 13a-b. A mixture 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
the aniline (5aꢀd, 1 mmol), the aldehyde (6a-h or 10a-d, 1 mmol), 1,2,3ꢀbenzotriazole (1 mmol)
and AlCl (0.2 mmol) in THF (4 mL) was stirred at room temperature overnight. The mixture
3
was diluted with water (15 mL) and extracted with ethyl acetate (15 mL X 3). The organic layers
were combined, washed with 2 N NaOH solution (15 mL) and brine (15 mL) sequentially, and
dried over Na SO (for preparation of compounds with phenolic hydroxyl groups, including 9d,
2
4
1
2a, 12d-g and 13b, the organic layers were not washed with NaOH solution, only washed with
brine). The solvent was evaporated, the residue was combined with MnO (10 mmol) and
2
o
dissolved with acetone (5 mL). The suspension was heated to 55 C for 24 h. MnO was removed
2
by filtration, and the solvent was evaporated to provide the crude products of 8, 9a-j, 11a-d, 12a-
m and 13a-b.
4
-(7-Methoxypyrrolo[1,2-a]quinoxalin-4-yl)-N,N-dimethylaniline (8). The aniline 5a
and benzaldehyde 6a were reacted according to the general procedure, and the crude product was
first purified by flash silica gel column chromatography (ethyl acetateꢀdichloromethane, gradient
up to 3:17), and then further purified by preparative HPLC (with methanolꢀwater, method 1) to
1
afford the products 8 as TFA salt, orange solid (59%). H NMR (400 MHz, methanolꢀd and
4
CDCl ) δ 8.50ꢀ8.49 (m, 1 H), 8.06ꢀ8.03 (m, 1 H), 7.86ꢀ7.83 (m, 2 H), 7.57ꢀ7.55 (m, 1 H), 7.43ꢀ
3
13
7
.42 (m, 1 H), 7.21ꢀ7.16 (m, 2 H), 6.83ꢀ6.80 (m, 2 H), 3.88 (s, 3 H), 3.12 (s, 6 H); C NMR (100
MHz, methanolꢀd and CDCl ) δ 158.3, 153.7, 150.0, 131.0, 126.4, 122.2, 121.9, 119.2, 119.0,
4
3
+
1
17.9, 117.2, 116.0, 114.3, 111.4, 102.3, 55.2, 39.1; HRESIMS m/z cacld for C H N O (MH )
20 20 3
3
18.1606, found 318.1598. HPLC purity 98.8%, 99.0% (Cꢀ18 reverse phase, MeOHꢀH O).
2
2
4
ACS Paragon Plus Environment

Page 25 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
N,N-dimethyl-4-(pyrrolo[1,2-a]quinoxalin-4-yl)aniline (9a). The aniline 5e and
benzaldehyde 6a were reacted according to the general procedure, and the crude product was
purified by flash silica gel column chromatography (ethyl acetateꢀdichloromethane, gradient up
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 1:9) to provide the product 9a as yellow solid (28%). H NMR (400 MHz, methanolꢀd ) δ
4
8.06ꢀ8.04 (m, 1 H), 7.89ꢀ7.87 (m, 1 H), 7.82ꢀ7.80 (m, 1 H), 7.76ꢀ7.73 (m, 2 H), 7.40ꢀ7.33 (m, 2
1
3
H), 6.94ꢀ6.92 (m, 1 H), 6.79ꢀ6.76 (m, 3 H), 2.96 (s, 6 H); C NMR (100 MHz, methanolꢀd ) δ
4
1
54.4, 151.9, 135.4, 129.5, 127.9, 126.7, 126.6, 125.1, 124.8, 115.3, 113.7, 111.3, 109.4, 39.0;
+
HRESIMS m/z cacld for C H N (MH ) 288.1501, found 288.1506. HPLC purity 98.1%,
19
18
3
9
8.4% (Cꢀ18 reverse phase, MeOHꢀH O).
2
4
-(3-Nitrophenyl)pyrrolo[1,2-a]quinoxaline (9b). The aniline 5e and benzaldehyde 6b
were reacted according to the general procedure, and the crude product was purified by flash
silica gel column chromatography (ethyl acetateꢀdichloromethane, gradient up to 1:19) to
1
provide the product 9b as light yellow solid (65%). H NMR (400 MHz, CDCl ) δ 8.93ꢀ8.91 (m,
3
1
H), 8.41ꢀ8.38 (m, 2 H), 8.07ꢀ8.04 (m, 2 H), 7.94ꢀ7.91 (m, 1 H), 7.76ꢀ7.72 (m, 1 H), 7.61ꢀ7.57
13
(
m, 1 H), 7.53ꢀ7.49 (m, 1 H), 7.02ꢀ7.00 (m, 1 H), 6.98ꢀ6.95 (m, 1 H); C NMR (100 MHz,
CDCl ) δ 151.1, 148.1, 139.7, 135.5, 134.1, 130.0, 129.3, 127.9, 126.8, 125.2, 124.3, 124.1,
3
+
1
23.3, 114.8, 114.1, 113.4, 107.8; HRESIMS m/z cacld for C H N O (MH ) 290.0930, found
1
7
12
3
2
2
90.0918. HPLC purity 99.5%, 99.4% (Cꢀ18 reverse phase, MeOHꢀH O).
2
4-(Pyrrolo[1,2-a]quinoxalin-4-yl)phenol (9d). The aniline 5e and benzaldehyde 6c were
reacted according to the general procedure, and the crude product was first purified by flash
silica gel column chromatography (ethyl acetateꢀdichloromethane, gradient up to 2:3), and then
further purified by preparative HPLC (with methanolꢀwater, method 1) to afford the products 9d
1
as TFA salt, yellow solid (57%). H NMR (400 MHz, methanolꢀd ) δ 8.87 (s, 1 H), 8.40ꢀ8.37 (m,
4
2
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 59
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
1
H), 8.03ꢀ8.01 (m, 1 H), 7.96ꢀ7.94 (m, 2 H), 7.82ꢀ7.78 (m, 1 H), 7.73ꢀ7.67 (m, 2 H), 7.35ꢀ7.33
13
(m, 1 H), 7.15ꢀ7.12 (m, 2 H); C NMR (100 MHz, methanolꢀd ) δ 163.0, 151.5, 131.6, 129.6,
4
1
27.3, 126.0, 125.8, 124.1, 123.1, 121.0, 120.3, 120.0, 118.6, 116.2, 115.5; HRESIMS m/z cacld
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
+
for C H N O (MH ) 261.1028, found 261.1025. HPLC purity 99.5%, 99.5% (Cꢀ18 reverse
1
7
13
2
phase, MeOHꢀH O).
2
4-(4-(Trifluoromethoxy)phenyl)pyrrolo[1,2-a]quinoxaline (9e). The aniline 5e and
benzaldehyde 6d were reacted according to the general procedure, and the crude product was
purified by flash silica gel column chromatography (ethyl acetateꢀhexanes, gradient up to 1:9) to
1
provide the product 9e as yellow solid (70%). H NMR (400 MHz, CDCl ) δ 8.08ꢀ8.03 (m, 3 H),
3
1
3
7
.97 (s, 1 H), 7.85ꢀ7.82 (m, 1 H), 7.53ꢀ7.40 (m, 4 H), 6.97ꢀ6.96 (m, 1 H), 6.90ꢀ6.89 (m, 1 H); C
NMR (100 MHz, CDCl ) δ 152.7, 150.3, 137.1, 136.1, 130.2, 130.1, 129.9, 127.7, 127.1, 125.3,
3
125.0, 124.4, 121.0, 120.5 (q, CF ), 116.7, 114.8, 114.1, 113.6, 108.4; HRESIMS m/z cacld for
3
+
C H N OF (MH ) 329.0902, found 329.0908. HPLC purity 97.3%, 97.1% (Cꢀ18 reverse
18 12
2
3
phase, MeOHꢀH O).
2
32
4
-(4-Nitrophenyl)pyrrolo[1,2-a]quinoxaline (9f). The aniline 5e and benzaldehyde 6e
were reacted according to the general procedure, and the crude product was purified by flash
silica gel column chromatography (ethyl acetateꢀdichloromethane, gradient up to 1:9) to provide
1
the product 9f as yellow solid (51%). H NMR (400 MHz, CDCl ) δ 8.43ꢀ8.40 (m, 2 H), 8.22ꢀ
3
8.19 (m, 2 H), 8.07ꢀ8.05 (m, 2 H), 7.94ꢀ7.92 (m, 1 H), 7.62ꢀ7.57 (m, 1 H), 7.53ꢀ7.49 (m, 1 H),
13
6
1
.97 (s, 2 H); C NMR (100 MHz, CDCl ) δ 151.4, 148.2, 144.0, 135.5, 130.1, 129.3, 128.0,
3
26.8, 125.3, 124.4, 123.4, 114.8, 114.1, 113.4, 107.9; HRESIMS m/z cacld for C H N O
2
17 12
3
+
(MH ) 290.0930, found 290.0922. HPLC purity 99.0%, 98.1% (Cꢀ18 reverse phase, MeOHꢀ
H O).
2
2
6
ACS Paragon Plus Environment

Page 27 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
N-ethyl-4-(pyrrolo[1,2-a]quinoxalin-4-yl)aniline (9h). The aniline 5e and benzaldehyde
f were reacted according to the general procedure, and the crude product was first purified by
6
flash silica gel column chromatography (ethyl acetateꢀhexanes, gradient up to 1:3), and then
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
further purified by preparative HPLC (with methanolꢀwater, method 1) to afford the products 9h
1
as TFA salt, red oil (23%). H NMR (400 MHz, methanolꢀd ) δ 8.64 (s, 1 H), 8.19ꢀ8.16 (m, 1 H),
4
7.89ꢀ7.87 (m, 1 H), 7.80ꢀ7.78 (m, 2 H), 7.65ꢀ7.61 (m, 2 H), 7.56ꢀ7.52 (m, 1 H), 7.21ꢀ7.18 (m, 1
1
3
H), 6.80ꢀ6.77 (m, 2 H), 3.25 (q, J = 7.2 Hz, 2 H), 1.31 (t, J = 7.2 Hz, 3 H); C NMR (100 MHz,
methanolꢀd ) δ 154.1, 150.6, 131.6, 128.7, 127.0, 125.5, 125.3, 123.3, 122.4, 120.2, 119.9, 118.0,
4
+
1
15.2, 114.7, 112.0, 37.1, 13.0; HRESIMS m/z cacld for C H N (MH ) 288.1501, found
1
9
18
3
2
88.1504. HPLC purity 98.5%, 98.4% (Cꢀ18 reverse phase, MeOHꢀH O).
2
4-(4-(Pyrrolidin-1-yl)phenyl)pyrrolo[1,2-a]quinoxaline (9i). The aniline 5e and
benzaldehyde 6g were reacted according to the general procedure, and the crude product was
first purified by flash silica gel column chromatography (ethyl acetateꢀhexanes, gradient up to
1:3), and then further purified by preparative HPLC (with methanolꢀwater, method 1) to afford
1
the products 9i as TFA salt, orange solid (61%). H NMR (400 MHz, methanolꢀd ) δ 8.63 (s, 1
4
H), 8.17ꢀ8.14 (m, 1 H), 7.88ꢀ7.85 (m, 1 H), 7.77ꢀ7.74 (m, 2 H), 7.63ꢀ7.58 (m, 2 H), 7.54ꢀ7.50 (m,
1
3
1
H), 7.20ꢀ7.17 (m, 1 H), 6.60ꢀ6.58 (m, 2 H), 3.31 (s, 4 H), 2.07 (s, 4 H); C NMR (100 MHz,
methanolꢀd ) δ 151.7, 150.3, 131.4, 128.6, 127.0, 125.4, 125.2, 123.1, 122.3, 120.2, 119.9, 118.0,
4
+
115.1 113.9, 111.7, 47.3, 24.9; HRESIMS m/z cacld for C H N (MH ) 314.1657, found
2
1
20
3
3
14.1651. HPLC purity 95.6%, 97.7% (Cꢀ18 reverse phase, MeOHꢀH O).
2
4-(4-(Pyrrolo[1,2-a]quinoxalin-4-yl)phenyl)morpholine (9j). The aniline 5e and
benzaldehyde 6h were reacted according to the general procedure. The crude product was
combined with NaBH (68mg, 1.8 mmol), dissolved with methanol (3 mL)ꢀdichloromethane (0.5
4
2
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 59
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
mL) mixture solvent, and stirred at room temperature for 3 h. The mixture was diluted with water
(10 mL) and extracted with ethyl acetate (10 mL X 3). The organic layers were combined and
dried over Na SO . Solvent was evaporated and the residues were further purified by flash silica
2
4
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
gel column chromatography (ethyl acetateꢀhexanes, gradient up to 1:2) and the product was
1
recrystallized in acetonitrile to provide the product 9j as yellow solid (19%). H NMR (400
MHz, CDCl ) δ 8.04ꢀ7.97 (m, 4 H), 7.87ꢀ7.84 (m, 1 H), 7.47ꢀ7.44 (m, 2 H), 7.06ꢀ7.04 (m, 3 H),
3
13
6
.90ꢀ6.88 (m, 1 H), 3.92ꢀ3.87 (m, 4 H), 3.32ꢀ3.27 (m, 4 H); C NMR (100 MHz, CDCl ) δ
3
153.5, 152.0, 136.1, 129.6, 129.4, 129.3, 126.6, 126.5, 125.0, 124.8, 114.5, 114.0, 113.4, 113.2,
+
1
08.1, 66.4, 48.3. HRESIMS m/z cacld for C H N O (MH ) 330.1606, found 330.1577. HPLC
2
1
20
3
purity 98.4%, 98.9% (Cꢀ18 reverse phase, MeOHꢀH O).
2
4-(Pyridin-3-yl)pyrrolo[1,2-a]quinoxaline (11a). The aniline 5e and aldehyde 10a were
reacted according to the general procedure and the crude product was purified by flash silica gel
column chromatography (hexaneꢀethyl acetate, gradient up to 2:3) to provide the product 11a as
1
white solid (64%). H NMR (400 MHz, CDCl ) δ 9.26ꢀ9.25 (m, 1 H), 8.77ꢀ8.74 (m, 1 H), 8.31ꢀ
3
8.29 (m, 1 H), 8.02ꢀ7.96 (m, 2 H), 7.84ꢀ7.81 (m, 1 H), 7.52ꢀ7.41 (m, 3 H), 6.96ꢀ6.94 (m, 1 H),
1
3
6.89ꢀ6.87 (m, 1 H); C NMR (100 MHz, CDCl ) δ 150.9, 150.2, 149.0, 135.8, 135.6, 133.8,
3
129.8, 127.6, 126.7, 125.1, 124.5, 123.2, 114.7, 113.9, 113.3, 107.9. HRESIMS m/z cacld for
+
C H N (MH ) 246.1031, found 246.1038. HPLC purity 99.1%, 98.9% (Cꢀ18 reverse phase,
16
12
3
MeOHꢀH O).
2
32
4
-(Pyridin-4-yl)pyrrolo[1,2-a]quinoxaline (11b). The aniline 5e and aldehyde 10b
were reacted according to the general procedure and the crude product was further purified by
flash silica gel column chromatography (ethyl acetateꢀdichloromethane, gradient up to 7:3) to
1
provide the product 11b as yellow solid (64%). H NMR (400 MHz, CDCl ) δ 8.83ꢀ8.81 (m, 2
3
2
8
ACS Paragon Plus Environment

Page 29 of 59
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
H), 8.05ꢀ8.02 (m, 2 H), 7.91ꢀ7.87 (m, 3 H), 7.58ꢀ7.54 (m, 1 H), 7.50ꢀ7.46 (m, 1 H), 7.00ꢀ6.98 (m,
1
3
1
1
H), 6.94ꢀ6.92 (m, 1 H); C NMR (100 MHz, CDCl ) δ 151.3, 149.9, 145.3, 135.5, 130.1,
3
27.9, 126.9, 125.1, 124.3, 122.5, 114.7, 114.0, 113.3, 107.8; HRESIMS m/z cacld for C H N
3
1
6
12
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
+
(
MH ) 246.1031, found 246.1025. HPLC purity 98.0%, 98.9% (Cꢀ18 reverse phase, MeOHꢀ
H O).
2
3
3
4
-(Cyclohex-1-en-1-yl)pyrrolo[1,2-a]quinoxaline (11c). The aniline 5e and aldehyde
1
0c were reacted according to the general procedure and the crude product was further purified
by flash silica gel column chromatography (ethyl acetateꢀhexanes, gradient up to 1:9) to provide
1
the product 11c as pale yellow oil (35%). H NMR (400 MHz, CDCl ) δ 7.97ꢀ7.94 (m, 1 H),
3
7
.92ꢀ7.91 (m, 1 H), 7.82ꢀ7.79 (m, 1 H), 7.48ꢀ7.40 (m, 2 H), 6.98ꢀ6.96 (m, 1 H), 6.85ꢀ6.83 (m, 1
H), 6.59ꢀ6.57 (m, 1 H), 2.69ꢀ2.67 (m, 2 H), 2.35ꢀ2.32 (m, 2 H), 1.90ꢀ1.86 (m, 2 H), 1.82ꢀ1.77 (m,
13
2
1
H); C NMR (100 MHz, CDCl ) δ 156.2, 136.3, 135.6, 130.8, 129.5, 126.7, 126.6, 124.8,
3
24.6, 113.8, 113.1, 113.0, 107.8, 26.6, 25.3, 22.4, 21.7; HRESIMS m/z cacld for C H N
17 17
2
+
(
MH ) 249.1392, found 249.1392. HPLC purity 98.8%, 98.9% (Cꢀ18 reverse phase, MeOHꢀ
H O).
2
3
4
4-(Naphthalen-1-yl)pyrrolo[1,2-a]quinoxaline (11d). The aniline 5e and aldehyde 10d
were reacted according to the general procedure and the crude product was first purified by flash
silica gel column chromatography (ethyl acetateꢀhexanes, gradient up to 3:17), and then further
purified by preparative HPLC (with methanolꢀwater, method 0) to afford the products 11d as
1
TFA salt, pale yellow oil (71%). H NMR (400 MHz, methanolꢀd and CDCl ) δ 8.50ꢀ8.49 (m, 1
4
3
H), 8.19ꢀ8.16 (m, 1 H), 8.13ꢀ8.11 (m, 2 H), 7.99ꢀ7.96 (m, 1 H), 7.84ꢀ7.82 (m, 1 H), 7.79ꢀ7.72 (m,
2
H), 7.67ꢀ7.61 (m, 2 H), 7.58ꢀ7.54 (m, 1 H), 7.48ꢀ7.44 (m, 1 H), 7.13ꢀ7.10 (m, 1 H), 7.03ꢀ7.02
13
(
m, 1 H); C NMR (100 MHz, methanolꢀd and CDCl ) δ 151.6, 133.3, 131.8, 130.4, 129.6,
4
3
2
9
ACS Paragon Plus Environment