Ugi Reaction-Derived α-Acyl Aminocarboxamides Bind to Phosphatidylinositol 3-Kinase-Related Kinases, Inhibit HSF1-Dependent Heat Shock Response, and Induce Apoptosis in Multiple Myeloma Cells

Article abstract of DOI:10.1021/acs.jmedchem.6b01613

Heat shock transcription factor 1 (HSF1) has been identified as a therapeutic target for pharmacological treatment of multiple myeloma (MM). However, direct therapeutic targeting of HSF1 function seems to be difficult due to the shortage of clinically suitable pharmacological inhibitors. We utilized the Ugi multicomponent reaction to create a small but smart library of α-acyl aminocarboxamides and evaluated their ability to suppress heat shock response (HSR) in MM cells. Using the INA-6 cell line as the MM model and the strictly HSF1-dependent HSP72 induction as a HSR model, we identified potential HSF1 inhibitors. Mass spectrometry-based affinity capture experiments with biotin-linked derivatives revealed a number of target proteins and complexes, which exhibit an armadillo domain. Also, four members of the tumor-promoting and HSF1-associated phosphatidylinositol 3-kinase-related kinase (PIKK) family were identified. The antitumor activity was evaluated, showing that treatment with the anti-HSF1 compounds strongly induced apoptotic cell death in MM cells.

Full text of DOI:10.1021/acs.jmedchem.6b01613

Subscriber access provided by UB + Fachbibliothek Chemie | (FU-Bibliothekssystem)
Article
Ugi reaction-derived #-acyl aminocarboxamides bind to phosphatidylinositol
3-kinase-related kinases, inhibit HSF1-dependent heat shock
response, and induce apoptosis in multiple myeloma cells
Matthias Bach, Anna Lehmann, Daniela Brünnert, Jens T. Vanselow, Andreas
Hartung, Ralf C. Bargou, Ulrike Holzgrabe, Andreas Schlosser, and Manik Chatterjee
J. Med. Chem., Just Accepted Manuscript • Publication Date (Web): 28 Apr 2017
Downloaded from http://pubs.acs.org on April 29, 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 46
Journal of Medicinal Chemistry
1
2
3
4
5
Ugi reaction-derived α−acyl aminocarboxamides
6
bind
to
phosphatidylinositol
3-kinase-related
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
kinases, inhibit HSF1-dependent heat shock
response, and induce apoptosis in multiple myeloma
cells
Matthias Bach4, ¥, Anna Lehmann1, ¥, Daniela Brünnert2, ¥, Jens T. Vanselow4,
Andreas Hartung1, Ralf C. Bargou3, Ulrike Holzgrabe1*, Andreas Schlosser4, §, Manik
Chatterjee2, §
.
1 Institute of Pharmacy and Food Chemistry, University of Würzburg, Am Hubland,
97074 Würzburg, Germany
2 Department of Internal Medicine II, Translational Oncology, University Hospital of
Würzburg, Versbacher Str. 5, 97078 Würzburg, Germany.
3 Comprehensive Cancer Center Mainfranken, University Hospital of Würzburg,
Versbacher Str. 5, 97080 Würzburg, Germany.
4 Rudolf Virchow Center for Experimental Biomedicine, University of Würzburg, Josefꢀ
SchneiderꢀStr. 2, 97080 Würzburg, Germany.
*Corresponding author: ulrike.holzgrabe@uni-wuerzburg.de (Ulrike Holzgrabe)
Tel.: +49ꢀ931 31ꢀ85460; Fax: +49ꢀ931 31ꢀ85494.
¥ These authors contributed equally
§ Both authors contributed equally
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 46
1
2
3
4
Abstract
5
6
7
8
9
The heat shock transcription factor 1 (HSF1) has been identified as a therapeutic
target for pharmacological treatment of multiple myeloma (MM) However, direct
therapeutic targeting of HSF1 function seems to be difficult due to the shortage of
clinically suitable pharmacological inhibitors. We utilized the Ugi multicomponent
reaction to create a small, but smart library of αꢀacyl aminocarboxamides, and
evaluated their ability to suppress heat shock response (HSR) in MM cells. Using the
INAꢀ6 cell line as MM model, and the strictly HSF1ꢀdependent HSP72 induction as a
HSR model, we identified potential HSF1 inhibitors. Mass spectrometryꢀbased affinity
capture experiments with biotinꢀlinked derivatives revealed a number of target
proteins and complexes, which exhibit an armadillo domain. Also, four members of
the tumorꢀpromoting and HSF1ꢀassociated phosphatidylinositol 3ꢀkinaseꢀrelated
kinase (PIKK) family were identified. The antitumor activity was evaluated, showing
that treatment with the antiꢀHSF1 compounds strongly induced apoptotic cell death in
MM 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
Keywords
Multiple myeloma, HSF1, Ugi reaction, PIKKs, DNAꢀPK, affinity capture, quantitative
mass spectrometry, armadillo domain
2
ACS Paragon Plus Environment

Page 3 of 46
Journal of Medicinal Chemistry
1
2
3
4
Introduction
5
6
7
8
9
The evolutionarily highly conserved heat shock response (HSR) buffers fatal
proteotoxic effects in acutely stressed cells and thus prevents induction of cell death.1
As a central HSR regulator, the heat shock transcription factor 1 (HSF1) controls
expression of multiple heat shock proteins (HSPs) which are essential for cellular
growth at physiologically relevant conditions by supporting correct protein folding and
protein translocation.2 Previous investigations demonstrated that the HSF1/HSP
transcriptome is critically involved in malignant transformation, inhibition of stressꢀ
mediated programmed cell death (apoptosis) and malignant growth of human
cancer.3ꢀ4 These findings might explain the observed correlation between increased
HSP expression and resistance to chemotherapeutic treatment.5ꢀ6 HSF1 as well as
HSPs like HSP90 or HSP70 have therefore been intensely studied as potential antiꢀ
cancer targets.3,7ꢀ8
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
Multiple myeloma (MM) remains a largely incurable B cell neoplasm mainly due to
the development of resistance against initially effective therapeutic agents.9ꢀ10
Increasing experimental and clinical evidence suggests that vulnerability of MM cells
to proteotoxic stress represents a therapeutic target, as evidenced by the successful
clinical establishment of pharmacological proteasome inhibition. Accordingly, HSPs
have been shown to critically contribute to the malignant phenotype of MM.11ꢀ15 In
addition, we could recently demonstrate a key role of HSF1ꢀdependent HSP
induction in response to treatment of MM cells with proteasome or HSP90
inhibitors.16 This finding indicates that the development of combination approaches
including HSF1 inhibition might represent a promising therapeutic strategy
overcoming HSRꢀmediated drug resistance in MM. However, the shortage of
clinically suitable pharmacological HSF1 inhibitors currently impairs further clinical
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 46
1
2
3
translation.
4
5
6
7
8
Previously, a number of structurally diverse compounds were found to putatively
inhibit the HSF1ꢀregulated heat shock response.17ꢀ18 The Workman group identified a
4,6ꢀsubstituted pyrimidine (1)19 as well as the bisamide CCT251236 (2)20 (Figure 1)
as strong inhibitors of HSFꢀ1 stress pathway. Those compounds have the chance to
be tested in preclinical and clinical trials. The chemical scaffold for a variety of natural
compounds ranges from the flavonoid quercetin (3)21 to the benzylidene lactame
KNK437 (4),22 natural products like stresgenin B (5)23 produced by a Streptomyces
sp. ASꢀ9 strain and triptolide (6)24 isolated from Tripterygium wilfordii (Figure 1). A
recent screening of over 80,000 natural and synthetic compounds, evaluated by
Santagata and colleagues,25 revealed that diverse chemical classes of natural
products were successfully tested for their anticancer activity by targeting the HSF1ꢀ
dependent stress response. Analysing the structural similarities between the
compounds dehydrocurvularin (7),25 withaferin A (8),25 colletofragarone A2 (9)25 and
the previous mentioned 3-6, it becomes apparent that all these substances share an
α,βꢀunsaturated carbonyl moiety (Figure 1).
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
4
ACS Paragon Plus Environment

Page 5 of 46
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
Figure 1: The identified compounds 1–919ꢀ25 block the HSF1ꢀregulated heat shock
response.
Here, we aimed to investigate the importance of this α,βꢀunsaturated carbonyl moiety
for HSF1ꢀrelated activity (structureꢀactivity relationship) and synthesized a library of
diversely substituted αꢀacyl aminocarboxamides performing Ugi multicomponent
reaction (Uꢀ4CR) out of four starting materials in a oneꢀpot reaction.26 In order to
identify cellular targets, we synthesized biotinꢀlinked versions of the most potent antiꢀ
HSF1 compounds and performed mass spectrometryꢀbased affinity capture
experiments.
5
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 46
1
2
3
4
Results and Discussion
5
6
7
Compound Synthesis
8
9
As can be seen in the Figure 1, currently reported HSF1 inhibitors of natural origin
consist of a highly reactive Michael system which belongs to the so called panꢀassay
interference compounds (PAINS).27ꢀ28 Those compounds are prone to react
covalently with proteins. However, covalent binding to a certain target in cancer
therapy might be eligible. E.g. the kinase inhibitors afatinib29 and ibrutinib30 were
recently launched to the market. Thus, HSF1 inhibitors containing an α,βꢀunsaturated
carbonyl moiety might be good starting point. Furthermore, studying the chemical
scaffold of aforementioned HSF1ꢀrelated compounds 1–921ꢀ25 it becomes apparent
that these substances possess a certain degree of bulkiness due to aromatic and
saturated rings and diverse residual groups. Hence, we used the Ugi fourꢀcomponent
reaction (Uꢀ4CR)31 to establish a diversely substituted library. Beside 2,5ꢀ
diketopiperazines,32 this reaction gives easy access to broad range of
αꢀacyl aminocarboxamides (10 or 11) in a singleꢀstep conversion using an aldehyde
(12), and amine (13), an isocyanide (14) and a carboxylic acid (15 or 16) as reaction
participants (Scheme 1).31 Beside some polar functional groups being crucial for a
balanced lipophilicity, various aliphatic substituted reactants were chosen for Uꢀ4CR
to be able to increase the stereoscopic specificity and therefore the associated
selectivity.33 For that purpose, starting materials with aromatic moieties (phenyl,
benzyl), aliphatic residues (methyl, propyl, butyl, pentyl, hexyl) and polar functional
groups (furyl, morpholinyl, hydroxyl ether, methyl ether) were applied for the
synthesis approach (Table 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
6
ACS Paragon Plus Environment

Page 7 of 46
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
Scheme 1: The Uꢀ4CR gives access to
aminocarboxamides (10 or 11) with and without an α,βꢀunsaturated carbonyl moiety
(shown in bold, respectively) in only one synthesis step.
Reagents and conditions: 12 (1.0 equiv), 13 (1.0 equiv), 14 (1.2ꢀ1.3 equiv), 15 or 16
(1.0ꢀ1.3 equiv), abs. CH3OH, r.t., 2 hꢀ15 d, 36ꢀ81%.
a
great variability of αꢀacyl
At first the corresponding aldehydes (12) and amines (13) were preꢀcondensated in
absolute CH3OH. The concentrations of 12 and 13 ranged between 0.4 M and 1.2 M
as it was reported that Uꢀ4CR ensure more effectiveness if the reactants are
available in high concentration.26 Corresponding isocyanides (14) and carboxylic
acids (15 or 16) were added subsequently in a slight excess of 1.0 to 1.3 equiv and
reacted to the αꢀacylaminocarboxamides 10a-m and 11a-g at r.t (Table 1). Some
products (e.g. 10a, 10b, 10c, 10e, 11b, 11c) precipitated after short reaction time and
were isolated by filtration and recrystallization in good yields. The products 11a and
11d-g were purified by means of medium pressure liquid chromatography (MPLC)
and were obtained in moderate yields.
After preliminary evaluation of the inhibitory effect of these compounds (data not
shown), the two most potent compounds 10b and 10j were structurally modified to
analyse the effect of the bromoꢀsubstituent R1 at the phenyl residue on the one hand
and the effect of the highly reactive enone function on the other hand (Scheme 1).
10l and 10m are the bromoless analogues of 10b and 10j, respectively, and were
synthesized by using the benzaldehyde instead of 2ꢀbromoꢀbenzaldehyde.
Compounds 11a and 11b are analogous to 10l and 10m and were obtained by
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 46
1
2
3
4
5
means of hydrocinnamic acid and 3ꢀ(2,4ꢀdichlorophenyl)propionic acid instead of the
corresponding unsaturated derivatives, respectively (Table 1).
6
7
8
9
Table 1: Synthesised series of αꢀacyl aminocarboxamides with (10a-m) and without
(11a-g) α,βꢀunsaturated carbonyl moiety.
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
product
(±) 10a
(±) 10b
(±) 10c
(±) 10d
(±) 10e
(±) 10f
(±) 10g
(±) 10h
(±) 10i
(±) 10j
(±) 10k
(±) 10l
(±) 10m
(±) 11a
(±) 11b
(±) 11c
(±) 11d
R1
2ꢀBr
2ꢀBr
2ꢀBr
2ꢀBr
2ꢀBr
2ꢀBr
2ꢀBr
2ꢀBr
2ꢀBr
2ꢀBr
2ꢀBr
H
R2
Me
R3
R4
[%]
77
57
75
60
72
81
61
69
41
36
37
73
53
60
62
66
80
67
Bn
Ph
cꢀPropyl
Bn
Ph
Bn
Bn
Ph
4ꢀMeOꢀBn
nꢀPropyl
Furanꢀ2ꢀylꢀmethyl
3,4ꢀ(MeO)2ꢀphenethyl
2ꢀClꢀBn
Bn
4ꢀMeOꢀPh
cꢀHexyl
4ꢀMeOꢀPh
cꢀHexyl
Me
nꢀPentyl
H
tꢀButyl
Me
2ꢀ(2ꢀHydroxyethoxy)ethyl
Bn
tꢀButyl
H
2ꢀMorpholinoethyl
2,4ꢀ(Cl)2ꢀPh
nꢀPropyl
cꢀPropyl
2ꢀMorpholinoethyl
Ph
Bn
Ph
H
Bn
2ꢀMorpholinoethyl
2,4ꢀ(Cl)2ꢀPh
2ꢀBr
2ꢀBr
4ꢀBr
4ꢀOH
cꢀPropyl
Bn
Ph
Bn
2ꢀMorpholinoethyl
2,4ꢀ(Cl)2ꢀPh
cꢀPropyl
Bn
Bn
Bn
Bn
Bn
Ph
Ph
Ph
cꢀPropyl
(±) 11e 4ꢀ(2ꢀBromoethoxy)
(±) 11f 4ꢀBr
(±) 11g 4ꢀ(2ꢀBromoethoxy)
cꢀPropyl
cꢀPropyl
4ꢀHydroxyphenyl 77
4ꢀHydroxyphenyl 73
cꢀPropyl
For the identification of cellular drug targets by means of affinity capture experiments,
three different biotinꢀlabelled compounds (20, 22, 24) were prepared (Scheme 2 and
3). First of all, 4ꢀ(2ꢀbromoethoxy)benzaldehyde (17) was prepared according to Jiang
8
ACS Paragon Plus Environment

Page 9 of 46
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
and colleagues.34 Condensation of 17, cꢀpropylamine, benzyl isocyanide and
3ꢀ(4ꢀhydroxyphenyl)propionic acid gave 11g, which was reacted with
2,2´ꢀ(ethylenedioxy)bis(ethylamine) and basic K2CO3 to give 21. Subsequently, 21
was biotinylated to achieve 22 which was purified by means of reversed phase
MPLC. The synthesis of compound 20 is analogous.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Scheme 2: Synthesis route to the biotinylated compounds 20 and 22 which were
used for the affinity capture experiment
Reagents and conditions: a) abs. CH3OH, r.t., 16 h; b) 2,2´ꢀ
(ethylenedioxy)bis(ethylamine), K2CO3, abs. CH3CN, 80 °C; 3ꢀ6 h; c) NHSꢀBiotin,
DIPEA, abs. DMF, r.t., 16ꢀ24 h.
The synthesis of biotinꢀlabelled compound 24 started off by substitution of 11f
(Scheme 3) with 1,4ꢀdibromobutane in presence of K2CO3. The obtained compound
18 was converted with 2,2´ꢀ(ethylenedioxy)bis(ethylamine) including K2CO3 and
biotinylated by means of N,NꢀDiisopropylehtylamine and NHSꢀBiotin in absolute
DMF. Purification by reversed phase MPLC gave compound 24.
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 46
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
Scheme 3: Synthetic route of the biotinylated compound 24 for the affinity capture
experiment.
Reagents and conditions: a) abs. CH3OH, r.t.; b) 1,4ꢀdibromobutane, K2CO3,
abs. CH3CN, 100 °C, 3 h; c) 2,2´ꢀ(ethylenedioxy)bis(ethylamine), K2CO3, abs.
CH3CN, 100 °C, 4 h; d) NHSꢀBiotin, DIPEA, abs. DMF, r.t., 22 h.
Structure-HSF1-dependent HSR activity relationship analysis
The isolated αꢀacyl aminocarboxamide products were tested in a recently established
biological HSR model assay for their ability to inhibit the HSF1ꢀmediated upregulation
of the heat shock proteins HSP72 and HSP27 upon cellular stress.16 In brief, INAꢀ6
MM cells were preꢀtreated with the compounds 10aꢀm, 11aꢀg and 18ꢀ24 (used
concentrations: 23 12.5 ꢀM, 19 25 ꢀM, and all others 50 ꢀM), respectively, prior to
incubation with the pharmacological HSP90 inhibitor and HSR inducer NVPꢀAUY922
(25),35 harvested and analysed for HSP72, HSP27, HSF1 or phosphoꢀHSF1
expression by Western blot (Figure 2). Of note, this cellular assay requires that a
tested compound can pass the cell membrane. For detection of HSP72 protein, the
main HSR readout marker, two complementary assays were employed using the
same protein lysate: Western blot analysis as a semiꢀquantitative method (Figure 2),
and ELISA for quantitative analysis of the HSP72 protein expression (Figure 3). In
addition, HSP72 mRNA level were determined for selected active or inactive
10
ACS Paragon Plus Environment

Page 11 of 46
Journal of Medicinal Chemistry
1
2
compounds to investigate whether HSP72 induction is mediated by HSF1ꢀmediated
transcription (Figure 4).
3
4
5
6
7
8
9
O
N
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
HO
O
OH
O
N
NH
)
25
NVPꢀAUY922 (
Figure 2: Western blot analyses of HSF1 expression or HSF1 phosphorylation at the
Serine residue 326, or of HSF1ꢀdependent induction of the heat shock proteins
HSP72 or HSP27. INAꢀ6 MM cells were incubated with the compounds 10a-m, 11a-g
or 18-24 for 4 hours, and concomitantly treated with the compound 25 (to induce
HSF1 activity by inhibition of HSP90) for additional 4 h. INAꢀ6 cells treated only with
DMSO or with the compound 25 served as negative or positive controls, respectively.
Staining of βꢀactin served as a control for equal loading.
Figure 3: Quantitative analysis of HSP90 inhibitionꢀinduced, HSF1ꢀdependent
HSP72 expression using an ELISA assay. INAꢀ6 cells were incubated with the
compounds 10a-m, 11a-g or 18-24 for 4 h, followed by HSF1 stimulation through coꢀ
treatment with the HSP90 inhibitor compound 25 for additional 4 h. INA6 cells treated
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 46
1
2
3
with DMSO or with compound 25 alone served as negative or positive controls,
4
5
respectively.
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
12
ACS Paragon Plus Environment

Page 13 of 46
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
Figure 4: Transcriptional changes in the expression of HSP72 and HSF1. INAꢀ6 cells
were treated with the compounds 19 (25 ꢁM), 23 (12.5 ꢁM), 10d (50 ꢁM) and 11d (50
ꢁM) for 4 h and were coꢀtreated with 10 nM of 25 for an additional 4 h. Afterwards,
total RNA was isolated and gene expression of HSP27, HSP72 and HSF1 was
measured by realꢀtime PCR. Data are normalized against αꢀtubulin. Shown are
means and standard deviations of three independent experiments.
With regard to the initially synthesized Michaelꢀacceptor compounds 10a–m the
induction of HSP72 or HSP27 expression by 25 was completely abrogated by 10b,
10j and 10m and strongly suppressed by 10a, 10e, 10g, 10h and 10l, whereas 10c,
10d, 10f, 10i and 10k had only moderate or no effect. Perfectly in line with the
HSP72 Western blot results, the findings from the ELISAs illustrate that preꢀtreatment
with compounds 10c, 10d, 10f, 10i and 10k show no or only minor effects on the
amount on induced HSP72. In contrast, treatment with the αꢀacyl
aminocarboxamides 10a, 10b, 10e, 10g, 10h, 10j, 10l and 10m significantly
prevented induction of HSP72 or HSP27. The strongest inhibitory effect was
observed for compounds 10b, 10j, 10l and 10m which show a HSP72 or HSP27 level
comparable to unstressed cells indicating a sufficient suppression of HSF1 function.
As 10l and 10m are the bromoless derivatives of 10b and 10j, respectively, the
bromoꢀsubstituent R1 in the ortho position is not crucial for the inhibitory activity.
However, among the compounds 10aꢀm the potency to inhibit the HSF1ꢀdependent
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 46
1
2
3
4
5
HSP72/HSP27ꢀupregulation was unaffected by their respective structural alteration,
and thus no obvious structureꢀHSF1 activity relationship could be derived.
6
7
8
9
For the investigation of the effect of the Michael system the compounds 10b and 10j
were compared to the saturated analogues 11a and 11b. Western blot and ELISA
analysis show similar induction of HSP72 expression so that the effect of the
α,βꢀunsaturated carbonyl moiety and the related covalent binding to the target are
unlikely. The compounds 11c and 11d differ in the para substituent R1 with 11d
carrying an OH group. ELISA analysis show strong induction effect of 11c but no
effect for 11d, indicating that a hydrogen bond donor at this molecule residue is not
crucial for the inhibitory effect. Concurrently, a hydrogen bond acceptor with a short
linker at R1 is tolerable as the compound 11e shows moderate activity. At the residue
R4 compound 11f carries an additional hydrogen bond donor in comparison to 11c.
Both compounds show the same activity pointing to the unimportance of the phenolic
OH group. However, replacement of the OH substituent with a bromobutoxy group at
R4 reduced the inhibitory effect. Of note the compound 11g consisting of the
bromoethoxy group in the position R1 (cf. 11e) and the phenolic OH group in the
position R4 (cf. 11f) belongs to the most active substances of the library.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
The biotinylated compounds 20, 22 and 24 and their intermediates 19, 21 and 23
were tested for the ability to inhibit the HSF1ꢀmediated upregulation of HSP72 or
HSP27, as well. Because the substances 19-24 are quite large molecules with
molecular weight up to 940 g/mol and many hydrophilic residues, we supposed that
their membrane permeability to the cell cytosol might be hindered. With the exception
of substance 19, this hypothesis was confirmed by Western blots as well as by ELISA
analyses.
In order to analyze potential mechanisms underlying the suppression of HSP72 and
HSP27, we additionally determined HSF1 and phosphoꢀHSF1 protein expression by
14
ACS Paragon Plus Environment

Page 15 of 46
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Western blot, and HSF1 mRNA level by RTꢀPCR. Western blot analyses revealed
downregulation of both HSF1 and phosphoꢀHSF1 protein level only upon treatment
with the antiꢀHSRꢀactive compounds (Figure 2), whereas HSF1 mRNA, as detected
by RTꢀPCR, remained unaffected for all compounds tested (Figure 4). This result
suggests that impairment of HSF1 protein stabilization or HSF1 mRNA translation
rather than inhibition of HSF1 transcription or HSF1 protein activation are potential
mechanisms underlying the observed antiꢀHSR effects of the active compounds. Our
finding is in a good accordance with a previous report showing downregulation of
HSF1 after PI3 kinase inhibition.13
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
Drug Target Identification
In order to identify cellular drug targets of compounds that efficiently inhibit HSF1ꢀ
dependent heat shock response, the biotinꢀlabelled compounds 20, 22 and 24
(Scheme 2 and 3) were applied for affinity capture experiments using streptavidin
magnetic beads in combination with quantitative mass spectrometry. To distinguish
between specific and unspecific binders, empty streptavidin beads were used as a
control. Compoundꢀloaded beads and control beads were incubated with INAꢀ6 cell
lysate under the same conditions. After washing, proteins were eluted, digested with
trypsin and analysed by nanoLCꢀMS/MS. Protein intensity ratios between inhibitor
and control samples were calculated for all identified proteins by labelꢀfree
quantification (Figure 5).
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 46
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
Figure 5: Workflow of the affinity capture experiment. Biotinylated inhibitor was
loaded on streptavidin beads; empty beads were used as control. Both beads were
incubated with INAꢀ6 cell lysate for 3 h at 4 °C, washed, and captured proteins were
eluted with LDS sample buffer. Proteins were digested with trypsin and analyzed by
nanoLCꢀMS/MS. Labelꢀfree quantification (LFQ) of both samples was performed for
all identified proteins.
The result of the analysis of five biological replicates for compound 22 is shown in
Figure 6A. All quantified proteins are listed in Supplemental Table S1. In summary,
we identified 68 proteins that are significantly enriched with inhibitorꢀloaded beads
over the control. The enriched proteins are a mixture of direct drug targets and
proteinꢀprotein interaction partners thereof, and it is not possible from the present
data to distinguish between those. A number of protein complexes are specifically
enriched with the compound 22: the MICOS complex, which regulates mitochondria
morphology and protein transport,36 four subunits of the SMN complex, which has an
essential role in the assembly of the splicesosomal small nuclear ribonucleoproteins
(snRNPs),37 and six subunits of the CCR4ꢀNOT complex, which is among other
things involved in the DNA damage response.38 In addition, we identified a number of
16
ACS Paragon Plus Environment

Page 17 of 46
Journal of Medicinal Chemistry
1
2
peroxisomal proteins; proteins associated with apoptosis, such as MAP3K5,
MAP3K6, PDCD6 and PEF1, five subunits of phosphatase 6, an enzyme involved in
mitosis and cell cycle progression.39 Interestingly, we identified four members of the
phosphatidylinositol 3ꢀkinaseꢀrelated kinases (PIKK), namely DNAꢀPK (PRKDC),
ATM and ATR which are involved in the DNA damage response (DDR) and interact
with each other,40 and mTOR which regulates cell growth by modulating many
processes including protein synthesis, ribosome biogenesis and autophagy.41ꢀ42
Beside its role in the DDR, DNAꢀPK interacts with HSF143 and it was shown that the
DNAꢀPK is important for the HSF1ꢀmediated upregulation of HSP72 during the heat
shock response (HSR).44
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
Figure 6: A) Identification of potential drug targets from affinity capture experiments
with 22 and INAꢀ6 cell lysate. Proteins significantly enriched (BenjaminiꢀHochberg
adjusted limma pꢀvalue < 0.02) over control are represented by closed circles.
Protein ratios and intensities are median values from five biological replicates (n
replicates > 1). The size of the circles correlates with the number of identified razor
and unique peptides. Proteins from the same complex, the same protein family or
with a similar function are marked in the same color. Carboxylases were enriched
specifically with the streptavidin control beads as they have Biotin as prosthetic
group. B) Significantly enriched, nonꢀredundant InterPro domains (Benjaminiꢀ
Hochberg adjusted Fisher´s exact test pꢀvalue < 0.01). Blue bars show fraction of
terms present in significantly enriched proteins (BenjaminiꢀHochberg adjusted limma
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 46
1
2
3
4
5
pꢀvalue < 0.02, n replicates > 1 and log2 ratio > 0) (drug targets) or in not significantly
enriched proteins (grey bars, unspecific proteins).
6
7
8
9
In order to find a common theme among these various proteins with diverse
biological functions, we performed an enrichment analysis for protein domains
(Figure 6B). The most common protein domain, found in more than 17% of the
potential drug targets, is the armadillo domain (ARM). Within the 68 potential drug
targets we identified 13 proteins (PRKDC, ATM, ATR, MTOR, XPOT, PP6R1,
PP6R2, MON2, CIP2A, CNOT1, RCD1, GCN1, BIG2) containing the armadillo
domain which is composed of several variable helical repeats forming a superhelix
that is important for proteinꢀprotein interaction.45
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
These 13 armadillo domainꢀcontaining proteins ꢀ together with their high confident
interaction partners listed in the HitPredict database46 ꢀ cover most of the 68 potential
drug targets identified by our affinity capture approach (Supplemental Table S1).
Thus, it is tempting to speculate that the armadillo domain is the primary target of the
investigated αꢀacyl aminocarboxamides. Although this domain differs from protein to
protein in repeatꢀlength, it was shown by Dahlstrom and coworkers that the peptide
binding site contains binding pockets with highly conserved residues.47
In addition to affinity capture experiments with compound 22, we analyzed the
compound with a different linker position 24 (Scheme 2 and 3) with the same
strategy. The potential cellular targets identified from these two compounds largely
overlap (Supplemental Figure S6). Especially, the PIKKs are identified in both
experiments as drug targets indicating that neither the linker position of compound 22
nor the linker position of compound 24 is essential for binding to PIKKs. However, for
a few protein complexes the linker position seems to interfere with protein binding.
E.g. the linker position of 24 seems to interfere with binding of CCR4ꢀNOT complex,
and the linker position of 22 seems to interfere with binding of the COG complex
18
ACS Paragon Plus Environment

Page 19 of 46
Journal of Medicinal Chemistry
1
2
(Supplemental Figure S6). Consistently, affinity capture experiments with compound
20 indicated that the absence of the phenolic hydroxyl moiety of 22 does not
significantly influence drug target binding (data not shown).
3
4
5
6
7
8
9
In order to validate some of the MS data, we performed Western blot analyses of the
INAꢀ6 MM cell eluates using specific antibodies against the PIKKs DNAꢀPK, ATM,
ATR or mTOR. As compared to the control approach, all PIKKs were significantly
higher detectable in the cell eluate that had been incubated with the beadsꢀlabelled
compound 22 (Figure 7).
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 7: Shown are Western blot analyses of the PI3 kinaseꢀrelated kinases
(PIKKs) DNAꢀPK ATM, ATR and mTOR upon incubation of INAꢀ6 cell lysates either
with compound 22ꢀcoupled beads or with beads only as a control. Staining against
H2AX or Ponceau S served as loading controls.
Anti-tumor activity
Next, we aimed to evaluate potential antiꢀtumor effects of the antiꢀHSF1ꢀsuppressive
compounds, and analyzed tumor cell survival in the MM cell line model INAꢀ6. Cells
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 46
1
2
were treated either with DMSO as a solvent control or with increasing concentrations
of the compounds 19, 23 or 10d, respectively, for three days prior to viability
assessment by annexin VꢀFITC/propidium iodide staining and measurement by flow
cytometry. Based on the detected viable cell fractions, which are negative for both
annexin V and propidium iodide, doseꢀresponse curves of three independent
experiments were calculated using GraphPad Prism software. As shown in Figure 8,
treatment with the both antiꢀHSRꢀactive compounds 19 and 23 strongly reduced
INAꢀ6 cell viability by induction of apoptotic cell death within a low micromolar
concentration range (8.6 ꢁM and 6.3 ꢁM, respectively) (Figure 8A and B), whereas
the antiꢀHSRꢀinactive compound 10d had no effect on cellular viability (Figure 8C).
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
Figure 8: Viability analysis of INAꢀ6 MM cells upon treatment with the compounds 19
(A), 23 (B) or 10d (C). INAꢀ6 cells were incubated either with DMSO as a solvent
control or with increasing concentrations of the compounds 19, 23 or 10d for three
days prior to viability assessment by annexin VꢀFITC/ propidium iodide staining and
measurement by flow cytometry. Shown are mean values and standard deviations of
three independent experiments.
In addition, we aimed to evaluate potential synergistic effects of an antiꢀHSRꢀactive
compound and the HSP90 inhibitor 25, and performed a combination experiment.
INAꢀ6 cells were incubated with subꢀmaximally effective concentrations either of the
compound 19 (for 72 h), the HSP90 inhibitor 25 (for 68 h), or a combination of both
compounds prior to viability analysis. As shown in Figure 9, we observed a strong
synergistic apoptotic effect of the combination approach, which strengthens our
20
ACS Paragon Plus Environment

Page 21 of 46
Journal of Medicinal Chemistry
1
2
hypothesis that concomitant inhibition of the HSR might trigger apoptotic effects
HSP90 inhibitors.
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
Figure 9: Combined inhibition of HSP90 and HSF1ꢀdependent heat shock response
(HSR). INAꢀ6 cells were incubated with subꢀmaximally effective concentrations either
of the compound 19 (for 72 h), the HSP90 inhibitor 25 (for 68 h), or a combination of
both compounds. 4 h after the treatment start with either with DMSO as a solvent
control or compound 19, either DMSO or 10 nM of 25 were added for an additional
68 h followed by flow cytometryꢀbased viability analysis (annexin VꢀFITC/ propidium
iodide staining). Data show the means and the standard deviations of the viable cell
fractions based on three independent experiments. Percentages were calculated
relative to the respective DMSOꢀtreated controls.
Since MSꢀbased target search revealed several PIKKs as potential targets of our
active compounds, we finally wanted to evaluate the role of DNAꢀPK and mTOR, two
PIKK family members, for HSF1ꢀmediated HSR using the dual DNAꢀPK/mTOR
inhibitor NU7441 (26).48 INAꢀ6 cells were treated with the dual DNAꢀPK/mTOR
inhibitor 26 (10 ꢁM) for 4 h and were coꢀtreated with 10 nM of 25 for an additional 4 h
prior to analyses of HSF1, phosphoꢀHSF1, HSP72 or HSP27 proteins by Western
blot (Figure 10A), or the mRNA expression level of HSF1 or HSP72 by RTꢀPCR
(Figure 10B). Both analyses show similar results regarding the HSF1ꢀdependent
HSR suppression as compared to our antiꢀHSR active compounds underscoring their
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 46
1
2
proposed mode of action, namely blockage of HSP90 inhibitorꢀdependent HSP72
induction on mRNA and protein level as well as HSF1 and phosphoꢀHSF1
downregulation. The similar observation of HSF1 downregulation upon both
treatment approaches, either with the dual DNAꢀPK/mTOR inhibitor or with our active
compounds, indicates involvement of common mechanisms affecting translation of
mRNA into protein or protein stability but not transcriptional or phosphorylationꢀ
mediated activation of HSF1.
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
S
O
O
N
O
26
)
NU7441 (
Figure 10: Dual DNAꢀPK/mTOR inhibition blocks HSF1ꢀdependent HSR upon
HSP90 inhibition. INAꢀ6 cells were treated with the dual DNAꢀPK/mTOR inhibitor 26
(10 ꢁM) for 4 h and were coꢀtreated with 10 nM of 25 for an additional 4 h.
Afterwards, protein as well as total mRNA were isolated. Expression level of HSF1,
phosphoꢀHSF1, HSP72 or HSP27 were analyzed by Western blot. Staining of aꢀ
tubulin served as a loading control (A). The mRNA expression level of HSF1 or
HSP72 were determined by RTꢀPCR. Data were normalized against αꢀtubulin.
Results are means and standard deviations of three independent experiments (B).
22
ACS Paragon Plus Environment

Page 23 of 46
Journal of Medicinal Chemistry
1
2
3
4
Conclusion
5
6
7
8
9
Experimental data suggests that HSF1 is critically involved in pathological processes,
e.g. of neurodegenerative diseases or cancer.3ꢀ4,49 We have recently established the
critical role of HSF1 for the protective upꢀregulation of HSP72 expression in MM, in
particular upon exposure to therapeutic agents that increase proteotoxic stress, like
the HSP90 inhibitor 25 or the proteasome inhibitor bortezomib. As this mechanism
distinctly contributes to the survival of MM cells, it is regarded to play an essential
role in the observed drugꢀresistance in the treatment of Multiple Myeloma.16
However, from a clinical perspective, the translation of direct therapeutic targeting of
HSF1 as a therapeutic principle into clinical trials is currently hampered, mainly due
to the shortage of clinically suitable pharmacological inhibitors. Therefore, the aim of
this study was the development of potent and selective inhibitors of the HSF1ꢀ
dependent HSR (for combined approaches with stressꢀinducing agents).
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
Utilizing the wellꢀknown Ugi multicomponent reaction a smart library of structurally
diverse compounds was synthesized and their potential to inhibit the HSF1 activity
was tested. Monitoring the HSF1ꢀdependent induction of HSP72 and HSP27
expression using the HSP90 inhibitor 25 revealed two very potent inhibitors bearing a
highly reactive Michael system and a bromophenyl ring. Omission of both moieties
resulted in equally potent HSFꢀ1 inhibitors indicating that both groups are not
necessary for inhibitory activity. Thus, the compounds 11a and 11b as well as 19 can
be regarded as excellent lead structures for further optimization.
Additional functional analyses revealed that the observed antiꢀHSR effect of these
active compounds is neither the result of a functional inactivation by inhibition of
HSF1 phosphorylation, nor caused by transcriptional downregulation. In fact,
treatment with the active compounds resulted in a loss of total HSF1 protein
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 46
1
2
3
4
5
expression indicating destabilization of HSF1 protein or impaired translation of HSF1
mRNA into HSF1 protein as potential underlying mechanisms.
6
7
8
9
We identified members of the PIKK family, namely DNAꢀPK, ATM, ATR and mTOR
as strong candidates for primary and therapeutically relevant targets. Beside their
role in the DNA damage response,40 it has been reported that DNAꢀPK is involved in
the HSF1ꢀmediated upregulation of HSP7244 and might therefore represent a
significant therapeutic point of attack. In line with this assumption, we here could
mimic the antiꢀHSR effects of our active compounds using a dual DNAꢀPK/mTOR
inhibitor.
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
Yet, HSF1 remains an important therapeutic target for drug development as recent
HSF1 phenotypic screens by the Workman group led to potent novel inhibitors
additionally targeting CDK9 and Pirin.19ꢀ20
24
ACS Paragon Plus Environment

Page 25 of 46
Journal of Medicinal Chemistry
1
2
3
4
Experimental
5
6
7
Chemistry: General Methods for Synthesis:
8
9
Melting points were determined on a Stuart melting point apparatus SMP11 (Bibby
Scientific, UK) and Melting point meter MPMꢀH2 (Schorpp Geraetetechnik,
Ueberlingen, Germany) and were not corrected. IR spectra were obtained using a
JASCO FT/IRꢀ6100 spectrometer (JASCO, GrossꢀUmstadt, Germany). TLC was
performed on preꢀcoated aluminium sheets with silica gel 60 F254 and preꢀcoated
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
aluminium sheets with silica gel C18 F254 (MachereyꢀNagel, Dueren, Germany). H
(400 MHz) and 13C (100 MHz) NMR spectra were recorded on a Bruker AV 400
instrument (Bruker Biospin, Ettlingen, Germany). Chemical shifts are given in ppm
1
and were calibrated on residual solvent peaks as internal standard (C6D6: H
7.15 ppm, 13C 128.0 ppm; CDCl3: 1H 7.26 ppm, 13C 77.1 ppm, CD3CN: 1H 1.94 ppm,
13C 1.24 ppm50). NMR signals were specified as s (singlet), d (doublet), t (triplet), m
(multiplet), bs (broad singlet), dd (doublet of doublets), dt (doublet of triplets).
Coupling constants
J are given in Hz. Medium pressure liquid chromatography
(MPLC) was performed on puriFlash®430 system (Interchim, Montluçon, France)
using preꢀpacked silica gel 50ꢁ columns from Interchim (Montluçon, France).
Reversed phase MPLC was performed using preꢀpacked C18ꢀHQ silica gel 15ꢁ
columns from Interchim (Montluçon, France) and HPLC grade solvents from Sigmaꢀ
Aldrich Chemie GmbH (Munich, Germany). MS data were obtained using an Agilent
1100 Series LC/MSD Trap (Agilent Technologies, Boeblingen, Germany). Purity of
substances 10a-m, 11aꢀg, 18, 19 and 21ꢀ23 was determined by a chromatographic
method (HPLC) and found to be > 95%. The purity of substances 20 and 24 was
determined by absolute qNMR according to ACS instruction to purity determination
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 46
1
2
3
4
5
and found to be > 91% and >88%, respectively. Commercial available chemicals
were used without further purification.
6
7
8
General procedure for the synthesis of the α-acyl aminocarboxamides 10a-m, 11a-g.
9
To a solution of the respective benzaldehyde (1.0 equiv) in absolute CH3OH the
corresponding amine (1.0 equiv) was added under Argon atmosphere and stirred for
20 min at r.t. Then the isocyanide compound (1.2ꢀ1.3 equiv) and the corresponding
acid derivative (1.0ꢀ1.3 equiv) were added. The solution stirred until the reaction,
controlled by means of TLC, was finished. The particular compounds were purified
either by filtration and crystallization or by means of chromatographic purification
methods. Detailed synthesis procedures and structure characterization data of 10a-m
and 11a-g are described in the Supplemental Information.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
4-(2-bromoethoxy)benzaldehyde (17) was prepared according to Jiang and
colleagues34: 4ꢀhydroxybenzaldehyde (1.99 g, 16.2 mmol) was dissolved in absolute
CH3CN (50 mL). K2CO3 (4.50 g, 32.5 mmol) and 1,2ꢀdibromoethane (14.1 mL, 164
mmol) were added and the reaction mixture refluxed 24 h at 90 °C. The reaction
mixture was filtered and the filtrate was concentrated in vacuo. After the purification
by MPLC (petroleum ether/EtOAc 4/1 to 1/1) 17 was obtained as colourless solid
1
(3.26 g, 87%). H NMR (CDCl3, 400 MHz): δ 9.90 (s, 1H), 7.85 (d,
J
=8.8 Hz, 2H),
7.02 (d,
J
=8.7 Hz, 2H), 4.38 (t,
J
= 6.2 Hz, 2H), 3.67 (t,
J
=6.2 Hz, 2H); 13C NMR
(CDCl3, 100 MHz): δ 190.7, 163.0, 132.1, 130.5, 114.9, 68.0, 28.5; mp 52ꢀ53 °C.
N-(2-(benzylamino)-1-(4-bromophenyl)-2-oxoethyl)-3-(4-(4-bromobutoxy)phenyl)-N-
cyclopropylpropanamide (18): 11f (300 mg, 0.59 mmol) was dissolved in absolute
CH3CN (6.0 mL) under Argon atmosphere. K2CO3 (166 mg, 1.20 mmol) and
1,4ꢀdibromobutane (700 ꢁL, 5.83 mmol) were added and the reaction mixture
refluxed 3 h at 100 °C. Afterwards the suspension was filtered, the filtrate
26
ACS Paragon Plus Environment

Page 27 of 46
Journal of Medicinal Chemistry
1
2
3
concentrated in vacuo and purified by MPLC (petroleum ether/EtOAc 4/1 to 1/4) to
4
5
6
1
obtain 18 as colourless solid (261 mg, 70%). H NMR (CDCl3, 400 MHz): δ 7.43 (m,
7
8
2H), 7.35ꢀ7.25 (m, 5H), 7.16ꢀ7.12 (m, 4H), 6.81 (m, 2H), 6.51 (t,
J=5.6 Hz, 1H, NH),
9
5.59 (s, 1H), 4.46 (m, 2H), 3.99 (t, =6.0 Hz, 2H), 3.51 (t, =6.6 Hz, 2H), 2.97ꢀ2.84
J
J
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(m, 4H), 2.51ꢀ2.46 (m, 1H), 2.13ꢀ2.06 (m, 2H), 1.99ꢀ1.92 (m, 2H), 0.95ꢀ0.90 (m, 1H),
0.86ꢀ0.80 (m, 1H), 0.71ꢀ0.60 (m, 2H); 13C NMR (CDCl3, 100 MHz): δ 176.3, 169.7,
157.4, 138.1, 134.8, 133.3, 131.6 (2C), 130.8 (2C), 129.5 (2C), 128.7, 127.6, 127.4,
122.1, 114.5 (2C), 66.9, 65.8, 43.7, 36.5, 33.5, 31.2, 30.4, 29.5, 28.0, 10.6, 9.3; IR ꢀȬ
3252, 3065, 2925, 1642, 1512, 1397, 1246, 1071, 1011, 822, 700 cmꢀ1; ESIꢀMS: m/z
642.9 [M+H+]; mp. 112ꢀ113 °C.
N-(1-(4-(2-((2-(2-(2-aminoethoxy)ethoxy)ethyl)amino)ethoxy)phenyl)-2-
(benzylamino)-2-oxoethyl)-N-cyclopropyl-3-(4-hydroxyphenyl)propanamide (21): 11g
(496
mg,
0.90
mmol),
K2CO3
(149
mg,
1.08
mmol)
and
2,2´ꢀ(ethylenedioxy)bis(ethylamine) (660 ꢁL, 4.50 mmol) were suspended in absolute
CH3CN (4.5 mL) under Argon atmosphere. The reaction mixture refluxed 3 h at
80 °C. After the suspension was filtered, the filtrate was concentrated in vacuo and
the crude product was purified by MPLC (EA/EtOH/NEt3 1/1/0.1) to isolate 21 as
1
yellowish oil (490 mg, 88%). H NMR (CDCl3, 400 MHz): δ 7.29ꢀ7.19 (m, 5H),
7.03ꢀ7.01 (m, 2H), 6.92ꢀ6.90 (m, 2H), 6.75ꢀ6.73 (m, 2H), 6.46ꢀ6.60 (m, 3H, NH), 5.43
(s, 1H), 4.48ꢀ4.36 (m, 2H), 4.10ꢀ4.01 (m, 2H), 3.63ꢀ3.58 (m, 6H), 3.53ꢀ3.51 (m, 2H),
3.00 (t, J= 5.13 Hz, 2H), 2.91ꢀ2.79 (m, 8H), 2.34ꢀ2.29 (m, 1H), 0.94ꢀ0.73 (m, 2H),
0.69ꢀ0.58 (m, 2H); 13C NMR (CDCl3, 100 MHz): δ 176.3, 170.5, 158.4, 155.7, 138.3,
131.5, 130.6, 129.6, 128.6, 127.9, 127.6, 127.3, 115.4, 114.4, 72.3, 70.4, 70.2, 70.1,
67.2, 66.3, 49.1, 48.5, 43.7, 41.2, 36.0, 31.3, 30.5, 10.3, 9.7; IR ꢀȬ 3300, 2865, 1646,
1510, 1453, 1240, 1100, 1047, 828, 555 cmꢀ1; ESIꢀMS: m/z 619.2 [M+H+].
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 46
1
2
3
N-(2-(2-(2-((2-(4-(2-(benzylamino)-1-(N-cyclopropyl-3-(4-
4
5
6
7
8
hydroxyphenyl)propanamido)-2-oxoethyl)phenoxy)ethyl)amino)ethoxy)ethoxy)ethyl)-
5-((3aS,4S,6aR)-2-oxohexahydro-1H-thieno[3,4-d]imidazol-4-yl)pentanamide (22):
NHSꢀBiotin (40.0 mg, 0.12 mmol) was dissolved in absolute DMF (1.5 mL) under
Argon atmosphere. Solution of 21 (87.0 mg, 0.14 mmol) and DIPEA (51.0 ꢁL, 0.28
mmol) in absolute DMF (1.5 mL) was added dropwise and the reaction mixture stirred
24 h at r.t. After the solvent was removed in vacuo the crude product was directly
purified by reversed phase MPLC (H2O/ CH3OH 8/2 to 1/9) to isolate 22 as a
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
1
colourless solid (45.0 mg, 38%). H NMR (CDCl3, 400 MHz): δ 7.30ꢀ7.19 (m, 5H),
7.07ꢀ7.05 (m, 2H), 6.94ꢀ6.84 (m, 4H, 2 NH), 6.76ꢀ6.74 (m, 2H), 6.67ꢀ6.65 (m, 2H),
6.36 (d,
3H), 4.20ꢀ4.16 (m, 1H), 4.11ꢀ4.02 (m, 2H), 3.63ꢀ3.48 (m, 8H), 3.40ꢀ3.36 (m, 2H),
3.06ꢀ2.78 (m, 10H), 2.65ꢀ2.62 (m, 1H), 2.36ꢀ2.31 (m, 1H), 2.16 (t, =7.4 Hz, 2H),
J=5.1 Hz, 1H, NH), 5.55 (s, 1H, NH), 5.48 (d, J=3.0 Hz, 1H), 4.49ꢀ4.37 (m,
J
1.68ꢀ1.53 (m, 4H), 1.41ꢀ1.31 (m, 2H), 0.97ꢀ0.88 (m, 1H), 0.80ꢀ0.73 (m, 1H), 0.69ꢀ0.55
(m, 2H); 13C NMR (CDCl3, 100 MHz): δ 176.4, 173.8, 170.5, 164.0, 158.3, 155.4,
138.5, 131.8, 130.7, 129.6, 128.6, 128.1, 127.7, 127.3, 115.45, 114.4, 70.2 (2C),
70.1, 70.0, 67.2, 66.3, 61.9, 60.2, 55.6, 49.1, 48.6, 43.7, 40.5, 39.3, 36.1, 35.9, 31.2,
30.5, 28.3, 28.1, 25.6, 10.3, 9.8; IR ꢀȬ 3288, 2924, 1686, 1644, 1511, 1453, 1238,
1100, 1029, 828, 576 cmꢀ1; ESIꢀMS: m/z 845.2 [M+H+]; mp. 67ꢀ69 °C.
3-(4-(4-((2-(2-(2-aminoethoxy)ethoxy)ethyl)amino)butoxy)phenyl)-N-(2-
(benzylamino)-1-(4-bromophenyl)-2-oxoethyl)-N-cyclopropylpropanamide (23): 18
(600
mg,
0.93
mmol),
K2CO3
(250
mg,
1.80
mmol)
and
2,2´ꢀ(ethylenedioxy)bis(ethylamine) (675 ꢁL, 4.59 mmol) were suspended in absolute
CH3CN (10 mL) under Argon atmosphere. The reaction mixture refluxed 4 h at
100 °C. Afterwards the suspension was filtered, the filtrate concentrated in vacuo and
the crude product purified by MPLC (EtOAc/EtOH/NEt3 1/1/0.1) to isolate 23 as
28
ACS Paragon Plus Environment

Page 29 of 46
Journal of Medicinal Chemistry
1
2
3
4
1
yellowish oil (465 mg, 71%). H NMR (CDCl3, 400 MHz): δ 7.42ꢀ7.38 (m, 2H), 7.32ꢀ
5
6
7
8
7.21 (m, 5H), 7.13ꢀ7.08 (m, 4H), 6.79ꢀ6.77 (m, 2H), 6.45 (t,
J
=5.6 Hz, 1H, NH), 5.53
=6.4 Hz, 2H), 3.61ꢀ3.59 (m, 6H), 3.50 (t, =5.2
=5.2 Hz, 2H), 2.69 (t, =7.2 Hz, 2H), 2.49ꢀ2.43
(s, 1H), 4.49ꢀ4.37 (m, 2H), 3.94 (t,
Hz, 2H), 2.93ꢀ2.84 (m, 6H), 2.81 (t,
J
J
9
J
J
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
(m, 1H), 1.85ꢀ1.78 (m, 2H), 1.71ꢀ1.64 (m, 2H), 0.93ꢀ0.76 (m, 2H), 0.69ꢀ0.58 (m, 2H);
13C NMR (CDCl3, 100 MHz): δ 176.4, 169.7, 157.6, 138.1, 134.8, 133.1, 131.7, 130.8,
129.5, 128.7, 127.6, 127.5, 122.1, 114.6, 73.4, 70.6, 70.4, 70.3, 67.8, 66.0, 49.6,
49.3, 43.8, 41.8, 36.5, 31.3, 30.4, 27.2, 26.7, 10.7, 9.3; IR ꢀȬ 3312, 2863, 1647, 1509,
1395, 1239, 1106, 1010, 824, 699, 550 cmꢀ1; ESIꢀMS: m/2z 355.6 [M+2H+].
N-(2-(2-(2-((4-(4-(3-((2-(benzylamino)-1-(4-bromophenyl)-2-
oxoethyl)(cyclopropyl)amino)-3-oxopropyl)phenoxy)butyl)amino)ethoxy)ethoxy)ethyl)-
5-((3aS,4S,6aR)-2-oxohexahydro-1H-thieno[3,4-d]imidazol-4-yl)pentanamide
(24):
NHSꢀBiotin (81.0 mg, 0.23 mmol) was dissolved in absolute DMF (1.5 mL) under
Argon atmosphere. Solution of 23 (200 mg, 0.28 mmol) and DIPEA (100 ꢁL, 0.58
mmol) in absolute DMF (1.5 mL) was added dropwise and the reaction mixture stirred
22 h at r.t. After the solvent was removed in vacuo the crude product was directly
purified by reversed phase MPLC (H2O/ CH3OH 8/2 to 1/9) to isolate 24 as
1
hygroscopic colourless solid (149 mg, 53%). H NMR (CDCl3, 400 MHz): δ 7.41ꢀ7.39
(m, 2H), 7.32ꢀ7.22 (m, 5H), 7.13ꢀ7.08 (m, 4H), 6.82ꢀ6.77 (m, 3H, NH), 6.63 (dt,
17.7 Hz, 1H, NH), 6.27 (d, =8.8 Hz, 1H, NH), 5.58 (s, 1H), 5.38 (d, =7.7 Hz, 1H,
NH), 4.94ꢀ4.37 (m, 3H), 4.26ꢀ4.23 (m, 1H), 3.94 (t, =6.3 Hz, 2H), 3.60ꢀ3.58 (m, 6H),
3.54 (t, =4.9 Hz, 2H), 3.43ꢀ3.38 (m, 2H), 3.12ꢀ3.08 (m, 1H), 2.93ꢀ2.79 (m, 8H), 2.71ꢀ
J=5.8,
J
J
J
J
2.66 (m, 3H), 2.46ꢀ2.41 (m, 1H), 2.23ꢀ2.16 (m, 2H), 1.85ꢀ1.78 (m, 2H), 1.74ꢀ1.60 (m,
6H), 1.45ꢀ1.37 (m, 2H), 0.94ꢀ0.85 (m, 1H), 0.83ꢀ0.76 (m, 1H), 0.70ꢀ0.57 (m, 2H);
13C NMR (CDCl3, 100 MHz): δ 176.4, 173.3, 169.8, 163.8, 157.5, 138.2, 134.9, 133.2,
29
ACS Paragon Plus Environment