Structural Optimization and Pharmacological Evaluation of Inhibitors Targeting Dual-Specificity Tyrosine Phosphorylation-Regulated Kinases (DYRK) and CDC-like kinases (CLK) in Glioblastoma

Article abstract of DOI:10.1021/acs.jmedchem.6b01840

The DYRK family contains kinases that are up-regulated in malignancy and control several cancer hallmarks. To assess the anticancer potential of inhibitors targeting DYRK kinases, we developed a series of novel DYRK inhibitors based on the 7-azaindole scaffold. All compounds were tested for their ability to inhibit DYRK1A, DYRK1B, DYRK2, and the structurally related CLK1. The library was screened for anticancer efficacy in established and stem cell-like glioblastoma cell lines. The most potent inhibitors (IC₅₀ ≤ 50 nM) significantly decreased viability, clonogenic survival, migration, and invasion of glioblastoma cells. Target engagement was confirmed with genetic knockdown and the cellular thermal shift assay. We demonstrate that DYRK1A’s thermal stability in cells is increased upon compound treatment, confirming binding in cells. In summary, we present synthesis, structure-activity relationship, and efficacy in glioblastoma-relevant models for a library of novel 7-azaindoles.

Full text of DOI:10.1021/acs.jmedchem.6b01840

Subscriber access provided by University of Newcastle, Australia
Article
Structural optimization and pharmacological evaluation of inhibitors
targeting dual-specificity tyrosine phosphorylation-regulated
kinases (DYRK) and CDC-like kinases (CLK) in glioblastoma
Qing-Qing Zhou, Athena F. Phoa, Ramzi H. Abbassi, Monira Hoque, Tristan A. Reekie, Josep S. Font,
Renae M. Ryan, Brett W. Stringer, Bryan W. Day, Terrance G Johns, Lenka Munoz, and Michael Kassiou
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/acs.jmedchem.6b01840 • Publication Date (Web): 16 Feb 2017
Downloaded from http://pubs.acs.org on February 17, 2017
Just Accepted
“Just Accepted” manuscripts have been peer-reviewed and accepted for publication. They are posted
online prior to technical editing, formatting for publication and author proofing. The American Chemical
Society provides “Just Accepted” as a free service to the research community to expedite the
dissemination of scientific material as soon as possible after acceptance. “Just Accepted” manuscripts
appear in full in PDF format accompanied by an HTML abstract. “Just Accepted” manuscripts have been
fully peer reviewed, but should not be considered the official version of record. They are accessible to all
readers and citable by the Digital Object Identifier (DOI®). “Just Accepted” is an optional service offered
to authors. Therefore, the “Just Accepted” Web site may not include all articles that will be published
in the journal. After a manuscript is technically edited and formatted, it will be removed from the “Just
Accepted” Web site and published as an ASAP article. Note that technical editing may introduce minor
changes to the manuscript text and/or graphics which could affect content, and all legal disclaimers
and ethical guidelines that apply to the journal pertain. ACS cannot be held responsible for errors
or consequences arising from the use of information contained in these “Just Accepted” manuscripts.
Journal of Medicinal Chemistry is published by the American Chemical Society. 1155
Sixteenth Street N.W., Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the course
of their duties.

Page 1 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Structural optimization and pharmacological evaluation of inhibitors targeting dualꢀ
specificity tyrosine phosphorylationꢀregulated kinases (DYRK) and CDCꢀlike kinases (CLK)
in glioblastoma
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
2
2
1
Qingqing Zhou , Athena F. Phoa , Ramzi H. Abbassi , Monira Hoque , Tristan A. Reekie , Josep S.
3
3
4
4
5
2*
Font , Renae M. Ryan , Brett W. Stringer , Bryan W. Day , Terrance G. Johns , Lenka Munoz and
1
*
Michael Kassiou
1
School of Chemistry, The University of Sydney, New South Wales 2006, Australia
2
School of Medical Sciences, Discipline of Pathology and Charles Perkins Centre, The University
of Sydney, New South Wales 2006, Australia
3
School of Medical Sciences, Discipline of Pharmacology, The University of Sydney, New South
Wales 2006, Australia
4
QIMR Berghofer Medical Research Institute, 300 Herston Road, Herston, Queensland 4006,
Australia
5
Oncogenic Signaling Laboratory, Centre for Cancer Research, Hudson Institute of Medical
Research, 27 Wright Street, Clayton, Victoria 3168, Australia
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 65
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 (150 words)
The DYRK family contains kinases that are upꢀregulated in malignancy and control several cancer
hallmarks. To assess the antiꢀcancer potential of inhibitors targeting DYRK kinases, we developed a
series of novel DYRK inhibitors based on the 7ꢀazaindole scaffold. All compounds were tested for
their ability to inhibit DYRK1A, DYRK1B, DYRK2 and the structurally related CLK1. The library
was screened for antiꢀcancer efficacy in established and stem cellꢀlike glioblastoma cell lines. The
most potent inhibitors (IC50 ≤ 50 nM) significantly decreased viability, clonogenic survival,
migration and invasion of glioblastoma cells. Target engagement was confirmed with genetic
knockdown and the cellular thermal shift assay. We demonstrate that DYRK1A’s thermal stability
in cells is increased upon compound treatment, confirming binding in cells. In summary, we present
synthesis, structureꢀactivity relationship and efficacy in glioblastomaꢀrelevant models for a library
of novel 7ꢀazaindoles.
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
ACS Paragon Plus Environment

Page 3 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
INTRODUCTION
Dualꢀspecificity tyrosine phosphorylationꢀregulated kinases (DYRK) and CDCꢀlike kinases (CLKs)
are members of the CMGC group of prolineꢀdirected kinases, which also includes cyclinꢀdependent
kinases (CDKs), mitogenꢀactivated protein kinases (MAPKs) and glycogen synthase kinases
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
(
GSKs).
While the CDK, GSK and MAPK signaling pathways have been extensively studied
with numerous inhibitors developed; less in known about DYRK and CLK signaling with the
development of inhibitors only becoming of interest recently.
The DYRK family consist of 5 kinases (DYRK1A, DYRK1B, DYRK2, DYRK3, DYRK4)
with DYRK1A being the most studied subtype. DYRK1A is produced as a constitutively active
kinase and its activity is controlled by subtle changes in the expression levels, in contrast to the
more paradigmatic phosphorylationꢀregulated activation of MAPKs, CDKs and CLKs. Increased
DYRK1A expression induces phosphorylation of substrates at serine/threonine residues which
3, 4
modulates various pathological processes. Most notably, the DYRK1A gene maps to chromosome
2
1 and the extra gene copy in Down syndrome individuals results in 1.5ꢀfold increased DYRK1A
expression. This alteration in DYRK1A expression levels has been correlated with structural
5
changes and cognitive impairment in Down syndrome population. Furthermore, trisomy 21ꢀ
mediated DYRK1A upꢀregulation has been linked to increased risk of childhood leukaemia and
6
early onset of Alzheimer’s disease.
Potent ATPꢀcompetitive DYRK1A inhibitors have emerged from natural sources, such as
7
8
meridianines, leucettine L41 and the alkaloid harmine (1; Figure 1), which is the most
experimentally used DYRK1A inhibitor. This small βꢀcarboline alkaloid (1) inhibits DYRK1A
9ꢀ11
(
IC50 = 33ꢀ80 nM)
and other kinases of the DYRK family; IC50 values of 0.16 ꢁM (DYRK1B), 2
9
, 10
ꢁ
M (DYRK2), 0.41 ꢁM (DYRK3) and 80 ꢁM (DYRK4) have been reported
analogues of the marine sponge leucettine, leucettine L41 with potent inhibition of DYRK1A (IC50
55 nM) and CLK1 (IC50 = 71 nM) was developed.
. From a series of
=
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 65
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
Heterocyclic DYRK1A inhibitors not derived from natural products are represented by the
1
2
13
benzothiazole (INDY) (2a) and 7ꢀazaindole (DANDY) (3a) (Figure 1). 2a is a derivative of the
1
4, 15
CLK inhibitor TG003 (2b)
that inhibits DYRK1A with IC50 = 240 nM, although is equipotent
at inhibiting other DYRK and CLK subtypes. The 3aꢀtype inhibitors remain one of the most potent
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
DYRK1A inhibitors to date. The lead molecules in this series inhibit DYRK1A and CLK1 at low
12, 13
nanomolar concentrations (IC = 3 ꢀ 20 nM).
As all DYRK1A inhibitors inhibit related kinases
5
0
of the CMGC group (e.g. DYRK1B, DYRK2, CLK1, CLK4, CDK5), efforts have been directed
16, 17
towards the development of selective DYRK1A inhibitors.
For example, quinoloneꢀ6ꢀ
carboxylic acid based inhibitor resulted in a DYRK1A selective inhibitor which displayed a 100ꢀ
fold selectivity for DYRK1A (IC50 = 6 nM) over DYRK1B (IC50 = 600 nM) and CLK1 (IC50 = 500
17
nM). More recent efforts developed quinazolineꢀ2ꢀcarbimidates selective for DYRK1A and
1
8
DYRK1B. Selective inhibitors are valuable tools to probe DYRK1A pathways and could serve as
leads for the development of new drugs to ameliorate the cognitive deficits in the population
affected by Down syndrome.
Cancer, however, is a complex disease driven by deregulation of numerous pathways and
molecularly targeted cancer therapy has greatly benefited from kinase inhibitors targeting multiple
kinases. Increasing evidence suggests that DYRK and CLK kinases are overꢀexpressed in
1
9ꢀ21
tumours
mechanisms.
degradation of the epidermal growth factor receptor (EGFR)
where they have been linked to the deregulation of cell cycle through various
22ꢀ24,25, 26
For example, DYRK1Aꢀdependent phosphorylation of Sprouty2 blocks
2
7, 28ꢀ29, 30
and favors recycling of EGFR
to the cell surface to amplify cancer cell proliferation. Furthermore, DYRK1A and CLK1 are prime
regulators of alternative splicing. Through this mechanism, DYRK1A and CLK1 not only increase
31
32
L
cell proliferation, but also increase expression of antiꢀapoptotic proteins BclꢀX and Mclꢀ1 ,
resulting in resistance to apoptotic stimuli. DYRK1B induce reversible cell arrest in a quiescent
state and inhibition of DYRK1B has been suggested to renew the cell cycle and sensitize cancer
25, 33ꢀ36
cells to chemotherapy and molecularly targeted therapy.
Finally, DYRK kinases
4
ACS Paragon Plus Environment

Page 5 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
3
7ꢀ39
phosphorylate actin and tubulin, which are cytoskeletal proteins crucial for cell migration.
Thus,
inhibition of this kinase family might be a beneficial strategy to combat proliferation and migration
of cancer cells.
Herein, we describe the synthesis of novel 7ꢀazaindole based inhibitors and pharmacological
evaluation against DYRK1A, DYRK1B, DYRK2 and CLK1. We extended the structureꢀactivity
relationship (SAR) of this series of compounds by employing primarily a bioisosteric approach
where we explored substitution of the phenol group at the C3 or C5 position of compound 3a
independently. Furthermore, we have chosen to evaluate these novel inhibitors in a glioblastomaꢀ
relevant model as DYRK kinases are overꢀexpressed in glioblastoma tumors and genetic knockꢀ
down of DYRK1A reduced tumor growth in vivo, suggesting that DYRK inhibitors could have
therapeutic potential.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 1. Selected DYRK1A inhibitors
RESULTS AND DISCUSSION
SYNTHESIS
13
The synthesis of 3a followed the published method starting from 5ꢀbromoꢀ7ꢀazaindole (4a)
(
Scheme 1). Regioselective iodination at the 3ꢀposition with Nꢀiodosuccinimide afforded 5a
followed by pꢀtoluenesulfonyl (tosyl) protection to give 6a. Suzuki crossꢀcoupling at both halide
positions afforded the diaryl compound 7a, which could then be deprotected, first by removing the
tosyl group to give 8a, followed by demethylation affording the lead compound 3a. We then
generated a library of 3a analogues. Initial investigations focused on the 7ꢀazaindole core and what
changes could be tolerated. The indole and indazole based compounds, 3b and 3c respectively, were
prepared in an analogous manner to 3a (Scheme 1).
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 65
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 1. Synthetic route of diphenolꢀazaindole derivatives
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: (i) NIS (1.0 eq), KOH (0.5 eq), CH
NCl (0.02 eq), CH Cl
steps; (iii) arylboronic acid (2.0 eq), Pd(PPh (4 mol%), K CO
00 °C, 3.5 h, 84ꢀ87%; (iv) KOH (5.0 eq), MeOH, 80 °C, 2 h, 91ꢀ95%; (v) BBr
°C to RT, 15 h, 58ꢀ65%. NIS = Nꢀiodosuccinimide.
2
Cl
, 0ꢀRT, 30 min; 80ꢀ89% over two
2 M (4.0 eq), toluene/EtOH 3:1,
(6.0 eq), CH Cl
2
, RT, 10 h; (ii) NaH (60% in
mineral oil, 3.0 eq), TsCl (1.2 eq), BnEt
3
2
2
3
)
4
2
3
1
0
3
2
2
,
Next, we focused on the modification of the phenol group at the C3 or C5 position of the azaindole
40, 41
scaffold, utilizing the key intermediate 6a. Sequential Suzuki crossꢀcoupling reactions
were
employed to first afford 7dꢀg, then the desired diarylꢀsubstituted compounds 8dꢀi (Scheme 2).
Hydrolysis of the tosyl group under basic conditions resulted in compounds 9aꢀc and 9eꢀg. The
42
desired compounds 8jꢀk were realized by cyclization of the benzonitriles 8f or 8i
,
respectively,
with TMSN under microwave conditions followed by hydrolysis of the tosyl group. The nitroꢀ
3
substituted compounds 9a and 9e were reduced in the presence of iron powder under acidic
conditions to afford anilines 9d and 9h
,
respectively. Further modifications were performed on the
6
ACS Paragon Plus Environment

Page 7 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
4
3
44
newly formed aniline 9h by acetylation or sulfonylamide formation with treatment of acetic
anhydride in acetic acid or with methanesulfonyl chloride in H O to afford 10a and 10b,
2
respectively. The compound 10c was also obtained in 70% yield from urea and catalyzed by iodine
45
and citric acid in mannitol.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
a
Scheme 2. Synthetic route of arylꢀazaindole derivatives
HO
HO
I
R1
R
1
R
1
Br
i
ii
iv
Br
N
N
N
N
N
N
N
Ts
N
H
Ts
Ts
Ph
1
1
9d R1 = 4-NH Ph
7
7
7
7
d R = 4-NO
2
Ph
8d R = 4-NO
2
6a
2
v
1
1
1
e R
=
A
8e R
8f R
=
=
A
9a R = 4-NO Ph
2
1
1
9b R1
f R = 4-CNPh
g R = 4-OHPh
4-CNPh
= A
iii
1
1
9c R1 = B
8j R = B
NH
ii
A =
OH
OH
OH
3
R HN
R2
iv
R
2
vi or vii or viii
N
HN
N
N
N
N
N
B =
H
N
N
N
Ts
H
R2 = 4-NO
2
9h
9e
R
R
2
2
= 4-NH Ph
= 4-NO Ph
A
= B
10a R
10b R
10c R
3
3
3
= MeCO
=
MeSO2
= NH CO
2
8
8
8
8
g
h
i
2
Ph
2
v
R
=
A
2
2
9f
R
2
=
R
= 4-CNPh
iii
2
2
9g R
k R = B
a
Reagents and conditions: (i) arylboronic acid, Pd(PPh ) (2 mol%), K CO 2 M (2.0 eq),
3
4
2
3
toluene/EtOH 3:1, 90 °C, 3ꢀ5 h, 78ꢀ85%; (ii) arylboronic acid, Pd(PPh
2.0 eq), toluene/EtOH 3:1, 110 °C, 3ꢀ5 h, 80ꢀ93%; (iii) TBAF, TMSN
CuSO •5H O, tBuOH/H O, 1:1 (V/V), MW (100 w, 80 °C), 2 h, 81ꢀ83%; (iv) KOH, MeOH, 80 °C,
2 h, 90ꢀ95%; (v) Fe powder (10.0 eq), HCl (3 N solution in water, 0.03 M), 110 °C, 3 h, 70ꢀ84%;
vi) Ac O (1.2 eq), AcOH, 110 °C, 5 h, 77%; (vii) methanesulfonyl chloride (1.5 eq), H O, RT, 10
h, 64%; (viii) urea (15.0 eq), I (0.1 eq), citric acid, mannitol, 80 °C, 6 h, 70%.
3
)
4
(2 mol%), K
2 3
CO 2 M
(
3
, sodium ascorbate,
4
2
2
(
2
2
2
To investigate diverse skeletons with different geometries and increased spatial flexibility, we
added a linker between the 7ꢀazaꢀindole core and phenol group and varied the phenol substitution
resulting in compounds 17aꢀb and 21aꢀb (Scheme 3). The 5ꢀcarboxamide 17aꢀb could be easily
prepared from 4a, the same starting material used previously, which was firstly protected with the
tosyl group, followed by the formation of the 5ꢀcarboxylic acid 12 by palladiumꢀcatalyzed
4
6
carboxylation using freshly made aceticꢀformic anhydride. The resulting acid underwent HBTU
amide coupling with aniline derivative 13a or 13b to afford compounds 14aꢀb in comparable
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
7
48
yield. Concomitant deprotection of the tosyl and TBS groups gave compounds 15aꢀb, which
4
9
2
followed by iodination at the C3 position with I and KOH, gave compounds 16aꢀb. The final
compounds 17aꢀb could then be easily obtained by a standard Suzuki crossꢀcoupling reaction.
Similarly, compounds 21aꢀb could be obtained by again starting with 4a (Scheme 3), firstly by
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
50
acylation with trichloroacetyl chloride followed by hydrolysis to afford carboxylic acid 18.
Subsequent amide formation with 13aꢀb afforded 19aꢀb, which, followed by Suzukiꢀcoupling gave
20aꢀb. Final compounds 21aꢀb were obtained in high yield after the TBAF induced deprotection of
51
the TBS group.
a
Scheme 3. Further modification for lead compound 3a
OTBS
OTBS
OH
O
HO
O
COOH
NH
NH
HO
O
Br
Br
NH
iii
vi
ix
N
N
N
N
N
H
H
N
H
N
N
H
19a 4-OTBS
9b 3-OTBS
2
0a 4-OTBS
21a 4-OH
21b 3-OH
18
1
20b 3-OTBS
vii, viii
TBSO
HO
Br
HOOC
ii
iii
iv
O
O
N
R
N
Ts
N
N
N
N
H
H
4
1
a R = H
1 R = Ts
12
N
N
N
N
i
H
Ts
1
1
4a 4-OTBS
4b 3-OTBS
15a 4-OH
15b 3-OH
OH
v
TBSO
HO
HO
O
vi
O
I
NH
2
N
N
H
H
1
3a 4-OTBS
13b 3-OTBS
N
N
N
H
N
H
1
1
7a 4-OH
7b 3-OH
16a 4-OH
16b 3-OH
a
Reagents and conditions: (i) tosyl chloride (1.2 eq), NaH (1.5 eq), THF, 0 °C, 1 h, 90%; (ii)
Pd(OAc) (5 mol%), dppf (10 mol%), aceticꢀformic anhydride (5.0 eq), DIPEA (5.0 eq), DMF, 110
C, 5 h, 83%; (iii) HBTU (1.2 eq), DIPEA (2.5 eq), aniline 13a or 13b (1.1 eq), DMSO, 0~RT, 2 h,
2
°
6
1ꢀ79%; (iv) KOH (5.0 eq), MeOH, 80 °C, 1 h, 94ꢀ96%; (v) I (1.2 eq), KOH (3.0 eq), DMF, RT, 8
2
h, 80ꢀ86%; (vi) arylboronic acid, Pd(PPh
3 4 2 3
) (2 mol%), K CO 2 M (2.0 eq), toluene/EtOH 3:1, 100
°
C, 3ꢀ5 h, 60ꢀ65%; (vii) trichloroacetyl chloride (1.2 eq), AlCl
3
(2.5 eq), 0 °C, 3 h; (viii) NaOH (3
8
ACS Paragon Plus Environment

Page 9 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
M), RT, 5 h, 70% over two steps; (ix) TBAF (1.5 eq), THF, RT, 30 min, 93ꢀ95%. DIPEA = N,Nꢀ
diisopropylethylamine, dppf =1,1’ꢀferrocenediylꢀbis(diphenylphosphine), HBTU = (2ꢀ(1Hꢀ
benzotriazolꢀ1ꢀyl)ꢀ1,1,3,3ꢀtetramethyluronium
hexafluorophosphate),
TBAF
=
tetraꢀnꢀ
butylammonium fluoride, OTBS = tertꢀbutyldimethylsilyl ether.
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
STRUCTUREꢀACTVITY RELATIONSHIP
New derivatives were tested for their potency to inhibit DYRK1A, DYRK1B, DYRK2 and CLK1.
Kinase assays were performed with Woodtide as a substrate for DYRK kinases and RS peptide as a
substrate for CLK1. ATP concentration (100 ꢁM) was kept constant to allow comparison of
potencies across the kinases. In our hands, harmine (1) and the lead analogue 3a inhibited
DYRK1A with IC50 of 360 and 14 nM, respectively. Inhibitor 3a showed comparable potencies
against DYRK1B (15 nM) and CLK1 (48 nM), but it was 10ꢀfold less potent at inhibiting DYRK2
(166 nM, Table 1). Replacement of the azaindole with indole (3b) or benzimidazole (3c) led to the
loss of kinase inhibition (> 1 ꢁM).
Table 1. Kinase inhibition and cellular efficacy of 3aꢀc and C3ꢀmodified analogues 9bꢀd.
Kinase inhibition assay (IC , nM)
DYRK1A DYRK1B DYRK2 CLK1
Cell viability assay (EC , ꢁM)
50
5
0
1
X, Y and R
RN1
WK1
JK2
SJH1
3a
14 ± 2
15 ± 5
166 ±
48 ± 7
2.0 ±
0.4
1.1 ±
0.1
5.1 ±
0.6
4.1 ±
1.0
2
2
3b
>1,000
>1,000
>1,000 >1,000
>1,000 >1,000
>
>
50
50
> 50
> 50
> 50
> 50
> 50
> 50
3
c
>1,000
>1,000
9
b
169 ± 52 83 ± 25
280 ± 53 172 ± 53
571 ±
49
117 ±
35
5.7 ±
1.3
6.2 ±
2.3
6.3 ±
1.7
2.2 ±
0.5
1
9
c
416 ±
85 ± 5
> 50
> 50
> 50
> 50
4
6
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 65
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
9d
>1,000
>1,000
>1,000 >1,000
> 50
> 50
> 50
> 50
Next, we tested how modification of the phenol at the C3 position affects kinase inhibition.
Bioisosteric replacement of the phenol moiety with indole (9b) resulted in 10ꢀfold weaker
DYRK1A inhibition (169 nM) and comparable loss of activity was observed against DYRK1B (83
nM), DYRK2 (571 nM) and CLK1 (117 nM). Substitution of phenol for a phenyltetrazole moiety
at C3 as seen in 9c further reduced kinase inhibition activity (85 ꢀ 280 nM) and the aniline analogue
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
(9d) resulted in complete loss of kinase inhibition (> 1 ꢁM).
Based on this SAR, we retained the 7ꢀazaindole core and C3 phenol, but diversified
substitution on C5 (Table 2). 4ꢀNitrophenyl (9e) and indole (9f) analogues showed weaker
DYRK1A inhibition (212 and 204 nM, respectively). Interestingly, indoleꢀanalogue 9f was more
potent at inhibiting DYRK1B (51 nM) and CLK1 (26 nM). Analogues with phenyltetrazole (9g),
aniline (9h), sulfonamide (10b) or urea (10c) in the paraꢀposition of the C5 phenyl inhibited
DYRK1A with IC50 in doubleꢀdigit nanomolar potency (31 ꢀ 86 nM). Comparable nanomolar
potency was determined for these inhibitors against DYRK1B and CLK1, and approximately 10ꢀ
fold higher IC50 were calculated from the DYRK2 inhibition assay. The introduction of
phenylacetamide at C5 (10a) led to the strongest DYRK1A inhibition (6.6 nM) with 5ꢀfold
selectivity over DYRK1B (38 nM) and CLK1 (35 nM), and 80ꢀfold selectivity over DYRK2 (536
nM).
1
0
ACS Paragon Plus Environment

Page 11 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Table 2. Kinase inhibition and cellular efficacy of C5 analogues.
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
Kinase inhibition assay (IC50, nM)
Cell viability assay (EC50, ꢁM)
R
DYRK1A DYRK1B DYRK2 CLK1
RN1
WK1
JK2
SJH1
1
2
87 ±
7
9
e
212 ± 36 198 ± 45 >1,000
> 50
> 50
> 50
> 50
6
44 ±
9f
204 ± 46 51 ± 8
26 ± 6 > 50
70 ± 9 > 50
> 50
> 50
> 50
> 50
30 ± 5
> 50
99
4
6
5
08 ±
9g
86 ± 17
43 ± 9
42 ± 6
0
67 ±
1.1 ±
34 ± 8
0.9 ±
0.3
8.0 ±
0.2
3.2 ±
0.3
9h
32 ± 16
168
0.3
5
1
36 ±
08
35 ±
3.5
2.1 ±
0.3
0.9 ±
0.2
8.6 ±
1.7
5.1 ±
0.8
10a
6.6 ± 1.2 38 ± 15
724 ±
153
10b
56 ± 17
31 ± 10
130 ± 37
17 ± 8
44 ± 14 14 ± 4 42 ± 5 13 ± 2 42 ± 6
4
6
76 ±
0
9.6 ±
1.5
10c
22 ± 4 >50
> 50
28 ± 8
1
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 65
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. Kinase inhibition and cellular efficacy of analogues with an amide linker.
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
Kinase inhibition assay (IC50, nM)
DYRK1A DYRK1B DYRK2 CLK1
Cell viability assay (EC50, ꢁM)
RN1 WK1 JK2 SJH1
44 ± 84 128 ± 31 >1,000 12 ± 3 12.5 ± 21.0 ± 37.8 ± 5.2 ±
1
2
R and R
3
17a
2.9
4.7
3.4
2.2
1
27 ± 54 138 ± 51 847 ± 34 ± 8 13.8 ±
13.2 ±
1.2
38.9 ±
0.1
5.7 ±
0.1
17b
50
1.2
>
1,000
>1,000
>1,000
>1,000 >1,000
> 50
> 50
> 50
> 50
> 50
> 50
2
1a
1b
>1,000
>1,000 >1,000
> 50
> 50
2
In the final series (Table 3), we explored the effect of an amide linker at position C3 or C5
on kinase inhibition. In line with results obtained with C3 modified analogues 9bꢀd, the
introduction of an amide linker into C3 position (21a and 21b) resulted in loss of kinase inhibition
(all IC50 > 1 ꢁM). The same modification in the position C5 was better tolerated, however
analogues 17a and 17b inhibited DYRK1A with higher IC values (344 and 127 nM, respectively).
50
The finding that C3ꢀmodified analogues failed to inhibit all kinases underlines the importance of the
C3 phenolic group attached to the 7ꢀazaindole core. The C5 position, however, offered scope for
improvement of activity and yielded analogues 9h and 10a with low nanomolar potency.
1
2
ACS Paragon Plus Environment

Page 13 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
CELLꢀBASED SCREENING
Glioblastoma is a heterogeneous cancer with great variability between patients and even within the
same patient. Glioblastoma tumors have been subꢀclassified into four molecular subtypes ꢀ
52
classical, proneural, neural and mesenchymal. Each subtype is characterized by distinct molecular
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
traits; however, the established glioblastoma cell lines inadequately mirror genotypes and
phenotypes seen in glioblastoma patients. Accordingly, we observed that the established U251 and
A172 glioblastoma cell lines express higher quantities of DYRK1A than the patientꢀderived
glioblastoma cell lines RN1, JK2, WK1 and SJH1 (Figure 1A). As DYRK1A activity correlates
with the protein quantity, higher DYRK1A quantity (thus higher DYRK1A activity) in the
established glioblastoma cells could render these cells more sensitive to DYRK1A inhibition,
however, would not be clinically relevant. Furthermore, DYRK1B, DYRK2, CLK1 expression
pattern in patientꢀderived glioblastoma cells also differs from the profile of U251 and A172 cells
(Figure 1A). With the exception of SJH1 cells that express only DYRK1A; the RN1, WK1 and JK2
cells express also DYRK1B, DYRK2 and CLK1 ꢀ albeit to a lesser degree than DYRK1A
expression. We have therefore chosen to screen all inhibitors in patientꢀderived glioblastoma cells.
These cells were grown as stem cells under defined conditions in order to maintain the phenotype
53,54, 55
and genotype of the primary resected tumors (Figure 1B).
Viability of cells treated with test
compounds was determined in the Alamar blue assay (Table 1 ꢀ 3).
A
B
DYRK1A
Cell line
Subtype
Genotype
RN1
Classical
p53wt, CDKN2A null,
DYRK1B
DYRK2
CLK1
PTENwt
p53 mut, CDKN2A null,
PTENwt
JK2
Proneural
WK1
SJH1
Mesenchymal p53 wt, CDKN2A null,
PTEN null
Neural
p53 mut, CDKN2A null,
PTENwt, NF1 mut
βꢀactin
1
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Figure 1. Expression of DYRK/CLK1 kinases and genotype of patientꢀderived glioblastoma
cell lines. (A) Cell lysate of untreated cells were analyzed with Western blotting using indicated
antibodies (representative image of 2 independent blots). (B) Subtype and genotype of patientꢀ
54, 55
derived glioblastoma cell line RN1, JK2, WK1 and SJH1 as previously described
.
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 lead compound 3a reduced viability of RN1 cells with EC50 of 2.0 ꢁM (Table 1). The
two most potent kinase inhibitors developed in this study, compounds 9h and 10a reduced RN1
viability with 1.1 ꢁM and 2.1 ꢁM potency, respectively (Table 2). Compounds 10b and 10c,
inhibiting DYRK1A, DYRK1B, DYRK2 and CLK1 with potencies (9.6 – 724 nM) comparable to
9h were 13ꢀfold less efficacious in the cell viability assays (EC > 13 ꢁM) compared to 9h (EC =
50
50
1
.1 ꢁM). This reduction in efficacy could arise from the increased polarity and reduced cell
membrane permeability, however, offꢀtarget effects could also be playing a role. For the remaining
compounds, cellular efficacy correlated with the potency in the kinase inhibition assay.
For example, compounds 10b and 17b with nanomolar kinase inhibition activity (IC50 = 26 ꢀ
6
44 nM) reduced RN1 viability with midꢀrange micromolar potency (EC50 > 10 ꢁM), whereas
compounds lacking kinase inhibition activity, e.g. 3b, 9d, 21b (IC50 > 1 ꢁM) were not cytotoxic to
RN1 cancer cells up to 50 ꢁM (Table 1 and Table 3).
This trend in SAR was also observed in WK1, JK2 and SJH1 cells, and DYRK1A inhibition
activity correlated with the cellular efficacies (R = 0.71, 0.63, 0.60 and 0.69 for DYRK1A (IC50) vs
RN1, WK1, JK and SJH1 (EC ), respectively; Spearman correlation). The most efficacious
50
compounds across all cell lines were compounds 9h and 10a, with the classical RN1 and
mesenchymal WK1 cells being most responsive to these inhibitors (EC50 = 0.9 – 2.1 ꢁM). Viability
of the neural SJH1 and proneural JK2 cell lines was reduced by 9h and 10a with the EC50 values
ranging 3.2 ꢀ 8.6 ꢁM.
CELLꢀBASED FUNCTIONAL EVALUATION
To evaluate the most potent inhibitors further, we tested analogues 9h and 10a in several cellꢀbased
assays. As a negative control, we employed analogue 9d that lacks the kinase inhibition activity and
1
4
ACS Paragon Plus Environment

Page 15 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
was not cytotoxic to primary cells (Table 1). Accordingly, analogue 9d did not reduce viability of
established A172 and U251 glioblastoma cell lines (EC50 > 50 ꢁM), whereas 9h and 10a decreased
A172 and U251 viability with efficacies ranging 1.6 – 4.6 ꢁM.
Cellular thermal shift assay (CETSA). To demonstrate target engagement in cells, compound 9h
was assessed for its ability to penetrate cells and bind to DYRK1A in a manner that was
independent of a phenotypic end point (i.e. reduced cell viability). For this, we used a cellular
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
6, 57
thermal shift assay (CETSA)
to assess DYRK1A engagement in U251 cells as this cell line
expresses high quantities of DYRK1A. In brief, U251 cells were treated with analogue 9h (10 ꢁM)
and cell suspensions were heated to different temperatures to induce protein denaturation. Soluble
proteins were extracted with an aqueous buffer and analyzed with Western blotting. Inhibitors that
bind to their cognate target in cells would yield a stabilized protein, which will be detected at higher
temperatures compared to the denaturation profile in cells without drug treatment. Indeed, 9h
efficiently stabilized DYRK1A, but not p38α MAPK (Figure 2AꢀB), validating that this inhibitor
binds to cellular DYRK1A and most likely the related DYRK1B, DYRK2 and CLK1 kinases;
whereas the stability of the structurally unrelated p38α MAPK was not affected by 9h. To
investigate 9h concentration effects, we derived the isothermal doseꢀresponse fingerprint
(ITDRFCETSA), which is a characteristic ligandꢀinduced protein stabilization at a constant
temperature. In this experiment, the DYRK1A biochemical potency of 9h (IC50 = 43 nM) translated
well into its cellular target engagement (ITDFRCETSA = 711 nM; Figure 2CꢀD).
1
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 65
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
B
Temp (°C) 40 43 46 49 52 55 58 61 64 67
DYRK1A
1
20
Ctr
9h
DMSO
100
8
0
DYRK1A
9h (10 µM)
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
6
0
4
0
Temp (°C) 40 43 46 49 52 55 58 61 64 67
p38α
20
0
4
0
45 50 55 60 65 70
Temperature(°C)
DMSO
p38α
9h (10 µM)
6
C
D
5
4
3
2
1
0
9h (µM)
0
.01 .1 .5
1
5 10 20
DYRK1A
p38α
00 101 102 103 104 105
1
Compound 9h (nM)
Figure 2. Treatment of U251 cells with inhibitor 9h stabilizes DYRK1A and decreases cell
viability. CETSA melt curves (40 – 67 °C) for DYRK1A and p38α (A) and quantification (mean ±
SEM) of 3 independent experiments (B) were determined in intact U251 cells treated with
compound 9h (1 h, 10 ꢁM). Representative image (C) and quantification (mean ± SEM) of 2
independent experiments (D) of ITDRFCETSA in intact U251 cell treated with compound 9h (0 ꢀ 20
ꢁ
M, 1 h) at 54 °C.
EGFR degradation. To further confirm DYRK1A inhibition in cells treated with 9h, we performed
a EGFR degradation assay. EGF binding to EGFR induces dimerization and mutual
phosphorylation. The phosphorylated EGFR dimer activates signaling pathways to stimulate cell
proliferation. EGFR dimerization also triggers endocytosis and lysosomal degradation of EGFR.
DYRK1A overꢀexpression has been shown to block this pathway of EGFR degradation, resulting in
1
6
ACS Paragon Plus Environment

Page 17 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
2
7
longer halfꢀlife of EGFR protein. Thus, in cells where DYRK1A is inhibited, EGFR is expected to
degrade at a faster rate compared to untreated cells. We therefore tested 9h for its ability to increase
EGFR degradation in U251 cells. EGFR protein translation was blocked with cyclohexamide and
EGFR degradation was induced with EGF (100 ng/mL). Cell lysates were collected at indicated
time points and the amount of nonꢀdegraded EGFR quantified by Western blotting (Figure 3A).
Basal EGFR halfꢀlife (t ) in untreated U251 cells was 25.6 ± 7.6 min. Preꢀtreatment of cells with
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
9h (2.5 ꢁM) reduced EGFR halfꢀlife to 11.3 ± 2.5 min (Figure 3B). Importantly, a similar effect
was observed with genetic knockꢀdown of DYRK1A. Using siRNA approach, we achieved on
average 80% reduction in DYRK1A expression (Figure 3C) and this DYRK1A downꢀregulation
reduced EGFR halfꢀlife to 3.3 ± 2.5 min (Figure 3D).
A
B
Ctr (t1/2 = 25.6 ± 7.6 min)
h (t1/2 = 11.3 ± 2.5 min)
120
Ctr
15 30 60 120
Compound 9h
9
100
EGF (min)
EGFR
0
5
0
5
15 30 60 120
80
60
βꢀactin
40
20
0
5
15 30
60
EGF (min)
120
C
D
Ctr siRNA
DYRK1A siRNA
Ctr siRNA (t1/2 = 9.4 ± 1.1 min)
EGF (min)
EGFR
0
5
15 30 60 120
0
5
15 30 60 120
120
100
80
DYRK1A siRNA (t1/2 = 3.3 ± 2.5 min)
DYRK1A
long exposure
short exposure
6
4
0
0
DYRK1A
βꢀactin
20
0
5 15 30
60
120
EGF (min)
Figure 3. DYRK1A inhibition induces EGFR degradation. (A) U251 cells were treated with DMSO
Ctr) or compound 9h (2.5 ꢁM) for 4 h. Cells were incubated with cycloheximide (30 ꢁg/mL; 1 h)
(
and EGF (100 ng/mL) was added for indicated time points. (B) U251 cells were treated with
scramble (Ctr) or DYRK1Aꢀtargeting siRNa for 24 h. Cells were incubated with cycloheximide (30
ꢁ
g/mL; 1 h) and EGF (100 ng/mL) was added for indicated time points. Cell lysates were analyzed
1
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 65
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
by Western Blotting using indicated antibodies. (C ꢀ D) EGFR levels were normalized to loading
control (βꢀactin) and are expressed as percentage of untreated cells (0 min). Halfꢀlife (t1/2) values
were calculated by nonꢀlinear regression analysis and represent mean ± SEM from 3 independent
experiments.
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
Clonogenic survival. Based on kinase inhibition in vitro and in cells, we evaluated compounds 9h
and 10a for their ability to reduce longꢀterm survival of glioblastoma cells. Compounds 9h and 10a
doseꢀdependently inhibited clonogenic survival of RN1 with EC50 values of 0.8 and 2.1 ꢁM (Figure
4
AꢀB), respectively; mirroring the EC50 values obtained in the shortꢀterm viability assay (Table 2).
Compound 9h was more efficacious at reducing the survival of the established U251 cell line (EC50
0.18 ꢁM, Figure 4C).
As numerous kinase inhibitors, including inhibitors targeting DYRK1B , directly bind
=
58
tubulin and act as tubulin inhibitors, we tested compounds 9h and 10a together with the inactive
analogue 9d for their potential effect on tubulin polymerization in vitro. None of the tested
compounds altered the rate of tubulin polymerization, whereas the tubulinꢀtargeting agents
paclitaxel and vinblastine changed the rate of tubulin polymerization in agreement with their
established mechanism of action (data not shown).
A
B
C
EC50 = 0.8 ± 0.2 ꢀM
EC50 = 2.1 ± 0.1 ꢀM
EC50 = 0.18 ± 0.06 ꢀM
120
100
80
1
120
100
40
1
20
100
*
***
80
60
40
*
8
0
60
60
40
****
4
2
0
0
***
*
***
20
*
***
20
0
1
0
****
1
0
****
****
0
0.01 0.1 0.5
1
5
0
0.01 0.1 0.5
1
5
0
0.01 0.1 0.5
1
5
compound 9h (ꢀM)
compound 10a (ꢀM)
compound 9h (ꢀM)
Figure 4. Clonogenic survival following drug treatment. Patientꢀderived RN1 glioblastoma cells
were treated with 9h (A) or 10a (B) for 10 days. Established U251 glioblastoma cells were treated
with 9h (C) for 10 days. Colonies were fixed, stained with Toluidine Blue and counted using the
ImageJ software. Data were normalized to vehicleꢀtreated controls (set as 100% survival) and EC50
1
8
ACS Paragon Plus Environment

Page 19 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
values calculated by nonꢀlinear regression analysis. EC50 values represent mean ± SEM from 3
independent experiments performed in duplicate. Oneꢀway ANOVA, followed by Dunnett’s
multiple comparison test was used to determine statistical significance. (*P < 0.05, *** P < 0.001,
*
***P < 0.0001 in relation to vehicleꢀtreated cells).
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
Migration and invasion. We assessed 9h and 10a, together with the inactive structural analogue
9d using in vitro assays of cell migration and invasion (Table 4). In the migration assay, the
movement across the wound is measured, whereas in the invasion assay the motility through the
wound coated with Matrigel is measured. We employed U251, A172 and RN1 cell lines, all of
which show high basal velocity (υ) of migration (υ > 2,000 ꢀm/h) and invasion (υ > 1,000 ꢀm/h).
The inhibitor 9h potently reduced migration of A172, U251 and RN1 cells, with more than 75%
inhibition of migration reached at 5 ꢁM concentration (Table 4). This inhibitor also blocked
invasiveness of A172 and U251 cells in a doseꢀdependent manner (Table 4). Similar effects were
observed across all assays with the analogue 10a, whereas the inactive analogue 9d that lacks
DYRK/CLK kinase inhibition activity (Table 1) was ineffective in the migration and invasion
assays (Table 4).
1
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 65
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 4. Inhibition of migration and invasion of glioblastoma cells by CMGC inhibitors 9h and
10a, and inactive analogue 9d. Values represent mean ± SEM from 3 independent experiments.
U251
migration
(ύ, ꢁm/h)
U251
invasion
(ύ, ꢁm/h)
RN1
migration
(ύ, ꢁm/h)
Conc.
ꢁM)
A172 migration
A172 invasion
(
(ύ, ꢁm/h)
(ύ, ꢁm/h)
0
1
5
.0
.0
.0
2,163 ± 133
1,247 ± 176
127 ± 56
1,337 ± 93
1,139 ± 112
234 ± 22
3,123 ± 103
2,828 ± 271
728 ± 308
2,659 ± 674
2,675 ± 475
1,538 ± 696
714 ± 585
1,174 ± 158
355 ± 28
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
9
h
118 ± 43
10.0
129 ± 52
ꢀ149 ± 12
1,710 ± 97
736 ± 117
34 ± 153
131 ± 129
208 ±77
0
.0
.0
.0
2,159 ± 48
1,125 ± 93
173 ± 38
2,739 ± 428
2,891 ± 147
2,296 ± 564
3,047 ± 506
2,731 ± 431
2,027 ± 419
1,115 ± 110
332 ± 33
1
10a
5
210 ± 64
1
0.0
111 ± 1
ꢀ118 ± 37
1,151 ± 133
1,451 ± 69
651 ± 300
2,650 ± 675
3,123 ± 745
1,718 ± 517
2,990 ± 997
2,411 ± 241
196 ± 47
937 ± 17
0
.0
2,653 ± 395
2,331 ± 143
1
.0
1,070 ± 246
9d
5.0
10.0
2,370 ± 39
1,481 ± 181
1,380 ± 73
3,065 ± 718
2,827 ± 709
2,232 ± 498
2,707 ± 491
929 ± 253
791 ± 236
2,702 ± 239
CONCLUSION
DYRK1B and DYRK2 have been recently implicated in the progression of various cancers,
33, 37
however the therapeutic potential of simultaneously targeting DYRK kinases has not been
evaluated. The hypothesis that inhibition of DYRK protein family would have antiꢀcancer efficacy
was tested in this study, with glioblastoma being an example of a tumor type where a strong
response might be expected. Glioblastoma cells upꢀregulate DYRK1A, and the link between
3, 27
DYRK1A overꢀexpression and glioblastoma pathogenesis has been established in vivo.
To
address the therapeutic potential of simultaneously targeting DYRK kinases, all of which are
variously expressed in glioblastoma subtypes, a series of potent and cellꢀactive DYRK inhibitors
have been developed and profiled in biochemical and glioblastoma cellꢀbased models.
2
0
ACS Paragon Plus Environment

Page 21 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
1
7, 18
While others have concentrated on the development of selective DYRK1A inhibitors,
our goal was to understand SAR across the DYRK family. Using compounds within and across
series that exhibit a full spectrum of potencies, we demonstrate that the SAR established for
inhibition of DYRK1A is nearly identical to that for DYRK1B inhibition. This observation is not
surprising as these two DYRK isoforms share high degree of homology, with only one amino acid
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
9
difference in the ATP binding pocket. However, while the 7ꢀazaindoleꢀbased scaffold explored in
this study offers little opportunity for developing selective DYRK1A inhibitors, quinoloneꢀ6ꢀ
17
carboxylic acid derivatives have been developed into highly selective DYRK1A inhibitors,
demonstrating that within the CMGC family of protein kinases selective inhibition of DYRK1A is
possible. Numerous DYRK1A inhibitors potently bind to and inhibit the related CLK family
13, 60
kinases
and this has been also demonstrated in our study. The potency of tested analogues to
inhibit DYRK1A and DYRK1B activity was comparable to the potency of inhibitors against CLK1;
with exception of compound 10c which inhibited CLK1 with >2ꢀfold higher potency than
DYRK1A/B. Nevertheless, CLK1 inhibition offers a therapeutic advantage as alternative splicing
32
driven by excessive CLK1 activity drives numerous hallmarks of cancer. Finally, all inhibitors
inhibited DYRK2 with >10ꢀfold higher IC values.
5
0
To address the clinical potential of these novel DYRK inhibitors, their cellular efficacy was
assessed in patientꢀderived glioblastoma cells. Our screening platform considers both the diversity
between patients and the heterogeneity of cells within a patient, and allowed us to identify not only
the most efficacious inhibitors from the series, but also the glioblastoma subtype most sensitive to
the inhibition of DYRK and CLK family of kinases. We established a correlation between
biochemical and cellular activities across the series, and confirmed that the antiꢀproliferative
activity is not a result of a tubulin targeting effect. However, the role of other kinases and nonꢀ
kinase targets cannot be completely discounted, as the proteome profile of the lead inhibitors is
unknown at this stage of the project (?). Nevertheless, the most potent DYRK/CLK inhibitors
attenuated viability of glioblastoma cells with low micromolar potency; and RN1 and WK1 cells
2
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 65
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
were most sensitive to inhibitors 9h and 10a,. These two cell lines represent the classical (60%) and
mesenchymal (13%) glioblastoma tumors, respectively; thus, our data suggests that DYRK/CLKꢀ
targeted therapy could be beneficial for majority of patients diagnosed with glioblastoma.
To further demonstrate the antiꢀcancer potential, inhibitor 9h was evaluated in the EGFR
degradation, clonogenic cell survival, migration and invasion assays. In all assays, compound 9h
was efficacious in the low micromolar range. Although the cellular efficacy and biochemical IC50
values differ, several factors must be taken into consideration when translating the potency values
obtained in kinase inhibition assays to the activity of inhibitors in cells. Firstly, kinase inhibition
assays are performed with recombinant, preꢀactivated enzymes and a large excess of the
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
6
1
phosphoacceptor, which renders kinases more sensitive to the inhibitor. Secondly, for the ATP
competitive kinase inhibitors (such as those developed in this study), the IC value depends on the
50
competition from ATP under the assay conditions. We performed all kinase assays at 100 ꢁM ATP,
based on the K (ATP) value for DYRK1A (30 ꢁM). However, when the ATP concentration
exceeds the K (ATP), the IC50 value increases as well. As the ATP concentration in cells is in the
M
M
1ꢀ5 mM range, an inhibitor concentration that is higher than the IC50 value determined from
biochemical kinase inhibition assays will be required to inhibit the targeted kinase in the cell. It is
suggested that an ATPꢀcompetitive kinase inhibitor should be active in cells at a concentration
approximately 10 to 100ꢀfold above its IC50 value as determined in biochemical assays using the
62
M
K (ATP) concentration. In the cellular assays, efficacy of 9h was significant in the 0.8 – 2.5 ꢁM
range, which is 20 to 70ꢀfold above the biochemical IC50 values for the inhibition of DYRK1A,
DYRK1B and CLK1 (34 – 43 nM).
In addition to using an appropriate inhibitor concentration in the functional studies and
CETSA assays, we demonstrate that genetic knockꢀdown of DYRK1A induced EGFR degradation
comparable to the degradation kinetics determined upon pharmacological DYRK1A inhibition with
compound 9h. Together, these data imply DYRK1A as a kinase responsible for EGFR stabilization.
However, as DYRK1A knockꢀdown had only mild effects on cell viability (data not shown) we
2
2
ACS Paragon Plus Environment

Page 23 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
conclude that the antiꢀproliferative efficacy of inhibitor 9h is due to simultaneously targeting the
DYRKꢀCLK family of kinases, though other offꢀtarget effects could also be a factor. In summary,
the protein kinases of the DYRK family are emerging as potential targets for small molecule cancer
therapy. We present a library of novel potent and cellꢀactive DYRK kinase inhibitors to further
explore the role of the DYRK kinases in an in vivo glioblastoma model setting.
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
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 65
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
EXPERIMENTAL SECTION
General chemical synthesis details
Unless noted otherwise, commercially obtained reagents were used as purchased without further
purification. Solvents for flash chromatography were distilled prior to use, or used as purchased for
HPLC grade, with the eluent mixture reported as the volume/volume ratio (v/v). Flash
chromatography was performed using Merck Kieselgel 60 (230–400 mesh) silica gel. Analytical
thinꢀlayer chromatography was performed using Merck aluminumꢀbacked silica gel 60 F254 (0.2
mm) plates (Merck, Darmstadt, Germany), which were visualized using shortwave (254 nm)
ultraviolet fluorescence. Melting points were measured with a rate of 6 °C/min and are uncorrected.
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
–1
Infrared absorption spectra were reported as vibrational frequency (cm ). Nuclear magnetic
resonance spectra were recorded at 300 K using a 200 MHz, 300 MHz, 400 MHz or a 500 MHz
spectrometer. The data are reported as chemical shift (δ ppm) relative to the residual protonated
solvent resonance, relative integral, multiplicity (s = singlet, br.s = broad singlet, d = doublet, dd =
doublet of doublets, t = triplet, m = multiplet, etc.) and coupling constants (J Hz). Low resolution
mass spectra (LRMS) was obtained from a ThermoQuest Finnigan LCQ Deca ion trap mass
spectrometer with electroꢀspray ionization in either positive (+ESI) or negative (ꢀESI) mode. Data
is expressed as observed mass (m/z), assignment (M = molecular ion), and relative intensity (%).
High resolution mass spectra was performed on a Bruker Apex Qe 7T Fourier Transform Ion
Cyclotron Resonance mass spectrometer equipped with an Apollo Π ESI dual source. Samples were
run with syringe infusion at 150 ꢀL/hr on a Cole Palmer syringe pump into electrospray ionization
(ESI). High performance liquid chromatography (HPLC) analysis of organic purity was conducted
TM
on a Waters Alliance 2695 instrument using a SunFire C18 column (5 ꢀm, 2.1 x 150 mm) and
detected using a Waters 2996 photodiode array (PDA) detector set at 254 nm. Separation was
achieved using water (solvent A) and acetonitrile (solvent B) at flow rate of 0.2 mL/min and a
gradient of 0% B to 100% or 40% over 30 min. HPLC data is reported as percentage purity (≥95%)
and retention time (RT) in minutes.
2
4
ACS Paragon Plus Environment

Page 25 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
General procedure A: synthesis of 5ꢀbromoꢀ3ꢀiodoꢀ1ꢀtosylꢀbased derivatives
13
The compounds ware prepared according to the literature. To a solution of indoleꢀbased
compound 4aꢀc (1.00 equiv) in CH Cl (0.05 M) was added KOH (0.50 equiv) at RT. After 30 min,
Nꢀiodosuccinimide (1.00 equiv) was added, the mixture was stirred for 10 h, quenched with a
saturated solution of Na and extracted with CH Cl . The combined organic layers were dried
2
2
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
S
2 2
O
3
2
2
(MgSO ) and concentrated in vacuo. The desired product was used in the next step without further
4
purification.
2 2
To a solution of 5aꢀc (1.00 equiv) in CH Cl (0.05 M) were added sodium hydride 60% (3.00
equiv) and benzyltriethylammonium chloride (0.02 equiv) under argon at 0 °C. After 30 min, 4ꢀ
methylbenzenesulfonyl chloride (1.20 equiv) was added at 0 °C, and the mixture was stirred at RT.
After 2 h, water was added and extracted with CH Cl . The combined organic layers were dried
2
2
(
MgSO ) and concentrated in vacuo.
4
General procedure B for diꢀSuzuki Coupling Reaction: To a solution of 6aꢀc (1.00 equiv) in
toluene/ethanol 3:1 (0.02 M) were added (4ꢀmethoxyphenyl)boronic acid (2.00 equiv), K CO (2 M
2
3
solution in water, 4.00 equiv), and Pd(PPh ) (4 mol %), and the reaction was heated to 110 °C for 5
3 4
h under argon. The reaction mixture was cooled to room temperature, concentrated in vacuo, and
then partitioned between water and ethyl acetate. The aqueous layer was extracted with ethyl
acetate, and the combined organic layers were dried (MgSO ) and concentrated in vacuo. The
4
residue was purified using flash chromatography (hexane/ethyl acetate 5:1 → 3:1) to give the
product.
General procedure C for monoꢀSuzuki Coupling Reaction: To solution of 6a or 7dꢀg or 16aꢀb
or 19aꢀb (1.00 equiv) in toluene/ethanol 3:1 (0.02 M) were added arylboronic acid (1.00 equiv),
K CO (2 M solution in water, 2.00 equiv), and Pd(PPh ) (2 mol %), and the reaction was heated to
2
3
3 4
9
0 °C for 3 h under argon. The reaction mixture was cooled to room temperature, solvent was
2
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 65
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
removed under reduce pressure, and then partitioned between water and ethyl acetate. The aqueous
layer was extracted with ethyl acetate, and the combined organic layers were dried (MgSO ) and
4
concentrated in vacuo.
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
General procedure D for the synthesis of tetrazole: To a solution of cyanated compound 8f or 8i
(
1.00 equiv) in tꢀBuOH and H
2
O (1:1 v/v, 0.125 M), were added TBAF (1 M in THF, 1.00 equiv),
•5H O (0.01 equiv) and sodium ascorbate (0.03 equiv),
trimethylsilyl azide (10.00 equiv), CuSO
4
2
and the mixture was stirred at 80 °C under microwave irradiation (100 W) for 2 h. The mixture was
quenched with water and extracted with ethyl acetate (3 x 10 ml), the combined organic layers were
dried (MgSO ) and concentrated under reduced pressure.
4
General procedure E for deprotection of tosylꢀgroup: To a solution of 7aꢀc, 8dꢀk or 14aꢀb (1.00
equiv) in MeOH (0.02 M) was added KOH (5.00 equiv), and the reaction was heated to 80 °C for 2
h. After completion, the solvent was removed under reduced pressure, and then partitioned between
water and ethyl acetate. The aqueous layer was extracted with ethyl acetate, and the combined
organic layers were dried (MgSO ) and concentrated in vacuo.
4
General procedure F for demethylation: To a solution of 8aꢀc (1.00 equiv) in dry CH Cl (0.05
2
2
2 2
M) was added BBr (1M solution in CH Cl , 6.00 equiv) under nitrogen. The reaction mixture was
3
stirred at room temperature for 15 h, then quenched at 0 °C with MeOH and concentrated in vacuo.
General procedure G for reduction of nitro compounds. To a solution of 9a or 9e (1.00 equiv)
in HCl (3N solution in water, 0.03 M) was added iron powder (10.00 equiv), and the reaction was
heated to 110 °C for 3 h. After completion and cooling down, the mixture was neutralized with
NaHCO (aq), and extracted with ethyl acetate, and the combined organic layers were dried
3
(
MgSO ) and concentrated in vacuo.
4
2
6
ACS Paragon Plus Environment

Page 27 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
General procedure H for the protection of 4ꢀ or 3ꢀhydroxyaniline with tertꢀButyldimethyl
6
3
group : To a solution of 4ꢀhydroxyaniline or 3ꢀhydroxyaniline (1.00 equiv, 5 mmol) in dry DMF
(
0.05 M) was added imidazole (2.50 equiv, 12.5 mmol) at 0 °C, followed by the addition of TBSCl
1.50 equiv, 7.5 mmol). The reaction mixture was stirred at RT for 12 h, after completion monitored
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
(
by TLC, the mixture was extracted with ethyl acetate (3 x 30 mL) and water (50 mL), the organic
layers were dried (MgSO ) and concentrated in vacuo.
4
General procedure I for amide coupling reaction of aromatic acid and substituted aniline: To
a solution of acid 12 or 18 (1.00 equiv) in DMSO (0.05 M) was added HBTU (1.20 equiv) under
nitrogen at 0 °C, followed by the addition of DIPEA (2.50 equiv) and aniline 13a or aniline 13b
(
1.10 equiv). The reaction mixture was stirred at RT for 2 h. after completion monitored by TLC,
the mixture was extracted with ethyl acetate and H O, the organic layers were dried (MgSO ) and
concentrated in vacuo. The residue was purified by flash chromatography (hexane/ethyl acetate 5:1
3:1) to give the product.
2
4
→
4
,4'ꢀ(1Hꢀpyrrolo[2,3ꢀb]pyridineꢀ3,5ꢀdiyl)diphenol (3a). This compound was prepared according
to General procedure F. The residue was purified by flash chromatography (CH
2
Cl
2
/MeOH 100:1
1
→
50:1) to give the product as a pale yellow solid (65% yield). R (CH Cl
f
2
2
/MeOH 20:1): 0.45; H
NMR (500 MHz, DMSOꢀd): δ 11.83 (1H, s), 9.69ꢀ9.23 (1H, br s), 8.47 (1H, d, J = 2.0 Hz), 8.30
1H, d, J = 2.0 Hz), 7.71 (1H, d, J = 2.5 Hz), 7.56 (4H, dd, J = 0.9 Hz, J = 8.6 Hz), 6.87 (4H, dq, J
(
13
= 2.0 Hz, J = 8.6 Hz), one OH signal not observed; C NMR (125 MHz, DMSOꢀd): δ 156.8,
1
55.7, 147.2, 140.5, 129.5, 128.7, 128.2, 127.7, 125.6, 125.3, 123.2, 117.9, 115.8, 115.7, 115.0;
+
ꢀ1
HRMS (ESI+) calcd for C19
H
14
N
2
O
2
[M + H] 303.1128, found 303.1130. IR (neat, cm ):
ν
%
3107, 2943, 1611, 1249. HPLC: 98.9%, RT: 16.4 mins.
2
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 65
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
4
,4'ꢀ(1Hꢀindoleꢀ3,5ꢀdiyl)diphenol (3b). This compound was prepared according to General
procedure F. The residue was purified by flash chromatography (CH
2
Cl
2
/MeOH 100:1 → 50:1) to
1
give the product as a white solid (60% yield). m.p. 246ꢀ248 °C; R (CH
f
2
Cl
2
/MeOH 20:1): 0.50; H
NMR (500 MHz, MeOD): δ 7.90 (1H, d, J = 1.0 Hz), 7.48 (4H, dd, J = 8.6 Hz, J = 23.2 Hz), 7.42
1H, d, J = 8.5 Hz), 7.35ꢀ7.33 (2H, m), 6.86 (4H, dd, J = 8.5 Hz, J = 14.8 Hz), NH and OH signals
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
(
13
not observed; C NMR (125 MHz, MeOD): δ 157.2, 156.5, 137.6, 135.8, 134.2, 129.4, 129.1,
129.0, 127.6, 123.1, 122.0, 118.6, 117.9, 116.5, 116.4, 112.6; HRMS (ESI+) calcd for C20 15NO
2
H
+
ꢀ1
[
M + Na] 324.0995, found 324.0998. IR (neat, cm ):
ν
%
3402, 3022, 1514, 1219, 1170, 798.
HPLC: >99.9%, RT: 21.7 mins.
4
,4'ꢀ(1Hꢀindazoleꢀ3,5ꢀdiyl)diphenol (3c). This compound was prepared according to General
procedure F. The residue was purified by flash chromatography (CH
2
Cl
2
/MeOH 100:1 → 50:1) to
1
give the product as a white solid (58% yield). m.p. 251ꢀ252 °C; R (CH
f
2
Cl
2
/MeOH 20:1): 0.50; H
NMR (500 MHz, MeOD): δ 8.11 (1H, s), 7.97 (1H, d, J = 8.7 Hz), 7.82 (2H, d, J = 7.4 Hz), 7.73
(1H, d, J = 8.7 Hz), 7.51 (2H, d, J = 7.4 Hz), 7.06 (2H, d, J = 7.4 Hz), 6.89 (2H, d, J = 7.2 Hz), NH
13
and OH signals not observed; C NMR (125 MHz, MeOD): δ 161.6, 158.7, 144.6, 141.4, 139.2,
1
33.2, 132.6, 131.1, 129.5, 120.3, 119.6, 119.5, 117.7, 116.9, 112.9; HRMS (ESI+) calcd for
+
ꢀ1
C H N O [M + H] 303.1128, found 303.1131. IR (neat, cm ):
1
ν
%
3343, 2980, 1609, 1497,
9
14
2
2
1
263, 1182, 809, 507. HPLC: 98.8%, RT: 19.9 mins.
5ꢀbromoꢀ3ꢀiodoꢀ1ꢀtosylꢀ1Hꢀpyrrolo[2,3ꢀb]pyridine (6a). This compound was prepared according
to general procedure A. The residue was purified using flash chromatography (hexane/ethyl
acetate 15:1 → 10:1) to give the product as a white solid (81% yield over two steps). m.p. 186ꢀ187
1
°C; R
f
(hexane/ethyl acetate 10:1): 0.65; H NMR (500 MHz, CDCl
3
): δ 8.44 (1H, d, J = 2.1 Hz),
13
8.06ꢀ8.04 (2H, m), 7.86 (1H, s), 7.78 (1H, d, J = 2.1 Hz), 7.29 (2H, d, J = 8.1 Hz), 2.37 (3H, s); C
NMR (125 MHz, CDCl ): δ 146.7, 146.0, 144.7, 134.7, 132.5, 131.4, 130.0, 128.4, 126.7, 116.0,
3
2
8
ACS Paragon Plus Environment

Page 29 of 65
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
+
ꢀ1
6
0.4, 21.8; LRMS (+ESI): m/z 499/501 (100/97, [M + Na] ). IR (neat, cm ):
ν
%
3138, 1619, 1371,
41
1
167, 1140, 1015, 527. The spectroscopic data matched that reported in the literature .
5ꢀbromoꢀ3ꢀiodoꢀ1ꢀtosylꢀ1Hꢀindole (6b). This compound was prepared according to general
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
procedure A. The residue was purified using flash chromatography (hexane/ethyl acetate 20:1 →
1
1
0:1) to give the product as a white solid (89% yield over two steps). H NMR (200 MHz, CDCl
3
):
δ 7.83 (1H, d, J = 8.7 Hz), 7.75 (2H, d, J = 8.4 Hz), 7.68 (1H, s), 7.51 (1H, d, J = 1.7 Hz), 7.45 (1H,
1
3
dd, J = 1.9 Hz, J = 8.7 Hz), 7.25 (2H, d, J = 8.0 Hz), 2.36 (3H, s); C NMR (50 MHz, CDCl
3
): δ
145.8, 134.8, 134.4, 133.3, 131.1, 130.3, 128.8, 127.1, 125.0, 117.7, 115.0, 65.5, 21.8. The
64
spectroscopic data matched that reported in the literature .
5
ꢀbromoꢀ3ꢀiodoꢀ1ꢀtosylꢀ1Hꢀindazole (6c). This compound was prepared according to general
procedure A. The residue was purified using flash chromatography (hexane/ethyl acetate 20:1 →
10:1) to give the product as an offꢀwhite solid (80% yield over two steps). m.p. 160ꢀ161 °C; R
f
1
(
hexane/ethyl acetate 12:1): 0.50; H NMR (500 MHz, CDCl ): δ 8.05 (1H, dd, J = 0.5 Hz, J = 8.9
3
Hz), 7.86 (2H, dd, J = 1.7 Hz, J = 6.7 Hz), 7.69 (1H, dd, J = 1.8 Hz, J = 8.9 Hz), 7.60 (1H, d, J =
13
1
.6 Hz), 7.27 (2H, d, J = 8.1 Hz), 2.38 (3H, s); C NMR (125 MHz, CDCl
3
): δ 146.2, 139.2, 134.1,
133.5, 131.9, 130.2, 127.9, 125.0, 118.2, 114.8, 102.4, 21.8; LRMS (+ESI): m/z 499/501 (100/86,
+
ꢀ1
[
ν
M + Na] ). IR (neat, cm ): 3102, 3071, 2923, 1593, 1373, 1238, 663, 531.
%
3,5ꢀbis(4ꢀmethoxyphenyl)ꢀ1ꢀtosylꢀ1Hꢀpyrrolo[2,3ꢀb]pyridine (7a). This compound was prepared
according to general procedure B to give the product as a white solid (87% yield). m.p. 158ꢀ160
1
°
C; R (hexane/ethyl acetate 2:1): 0.45; H NMR (400 MHz, CDCl ): δ 8.64 (1H, d, J = 2.1 Hz),
f
3
8
.14 (2H, d, J = 2.0 Hz), 8.12 (1H, d, J = 1.8 Hz), 7.81 (1H, s), 7.54 (2H, dd, J = 2.2 Hz, J = 6.7
Hz), 7.49 (2H, dd, J = 2.2 Hz, J = 6.7 Hz), 7.29 (2H, dd, J = 0.6 Hz, J = 8.6 Hz), 7.01 (4H, dt, J =
13
2
3
.2 Hz, J = 8.8 Hz), 3.87 (3H, s), 3.85 (3H, s), 2.38 (3H, s); C NMR (100 MHz, CDCl ): δ 159.6,
2
9
ACS Paragon Plus Environment