Promiscuity and selectivity in covalent enzyme inhibition: A systematic study of electrophilic fragments

Article abstract of DOI:10.1021/jm5006918

Covalent ligand-target interactions offer significant pharmacological advantages. However, off-target reactivity of the reactive groups, which usually have electrophilic properties, must be minimized, and the selectivity of irreversible inhibitors is a crucial requirement. We therefore performed a systematic study to determine the selectivity of several electrophilic groups that can be used as building blocks for covalently binding ligands. Six reactive groups with modulated electrophilicity were combined with 11 nonreactive moieties, resulting in a small combinatorial library of 72 fragment-like compounds. These compounds were screened against a group of 11 enzyme targets to assess their selectivity and their potential for promiscuous binding to proteins. The assay results showed a considerably lower degree of promiscuity than initially expected, even for those members of the screening collection that contain supposedly highly reactive electrophiles.

Full text of DOI:10.1021/jm5006918

Subscriber access provided by Universitaetsbibliothek | Johann Christian Senckenberg
Article
Promiscuity and selectivity in covalent enzyme
inhibition: A systematic study of electrophilic fragments
Christian Jöst, Christoph Nitsche, Therese Scholz, Lionel Roux, and Christian D. Klein
J. Med. Chem., Just Accepted Manuscript • DOI: 10.1021/jm5006918 • Publication Date (Web): 22 Aug 2014
Downloaded from http://pubs.acs.org on August 25, 2014
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 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Promiscuity and selectivity in covalent enzyme
inhibition: A systematic study of electrophilic
fragments
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
Christian Jöst, Christoph Nitsche, Therese Scholz, Lionel Roux, Christian D Klein*
Medicinal Chemistry, Institute of Pharmacy and Molecular Biotechnology IPMB, Heidelberg
University, Im Neuenheimer Feld 364, Dꢀ69120 Heidelberg, Germany
* To whom correspondence should be addressed: c.klein@uniꢀheidelberg.de, phone ++49ꢀ6221ꢀ
544875
1
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 2 of 38
1
2
3
4
5
6
7
8
9
ABSTRACT: Covalent ligandꢀtarget interactions offer significant pharmacological advantages.
However, offꢀtarget reactivity of the reactive groups, which usually have electrophilic properties,
must be minimized and the selectivity of irreversible inhibitors is a crucial requirement. We
therefore performed a systematic study to determine the selectivity of several electrophilic groups
that can be used as building blocks for covalently binding ligands. Six reactive groups with
modulated electrophilicity were combined with eleven nonꢀreactive moieties, resulting in a small
combinatorial library of 72 fragmentꢀlike compounds. These compounds were screened against a
group of eleven enzyme targets to assess their selectivity and their potential for promiscuous
binding to proteins. The assay results showed a considerably lower degree of promiscuity than
initially expected, even for those members of the screening collection that contain supposedly
highly reactive electrophiles.
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
ACS Paragon Plus Environment

Page 3 of 38
Journal of Medicinal Chemistry
1
2
3
4
INTRODUCTION
5
6
7
8
Irreversible inhibitors of enzymes exhibit several properties that are beneficial for their clinical
use. Nonꢀequilibrium binding of irreversible inhibitors improves their biochemical efficiency1
and confers robustness against pharmacokinetic liabilities such as fast clearance and binding to
serum proteins.2 Side effects can be minimized, since selective irreversible inhibitors retain their
effect after elimination.3 This allows the creation of rapidly metabolized and excreted drugs that
still have high and lasting effectiveness. Covalent inhibitors can also avoid some resistance
mechanisms, especially when targeting residues of a target protein that are essential for its proper
function.4 An interruption of metabolic pathways by covalent inhibitors is difficult to overcome
by feedback mechanisms, which offers a significant advantage for antibiotic or anticancer drugs.
The outstanding pharmacological properties of irreversible enzyme inhibitors are highlighted by
such prominent examples as the βꢀlactam antibiotics,5 fosfomycin,6 and proton pump inhibitors.7
Etacrynic acid is a noteworthy example of a drug that appears to be activated by covalent binding
to cysteine and binds to its target as a cysteineꢀconjugate.8 Avibactam is a reversible, covalent βꢀ
lactamase inhibitor that binds to the active site serine of βꢀlactamases.9
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
Irreversible enzyme inhibitors contain a chemical functionality, most frequently an
electrophile, which is prone to react with (usually nucleophilic) protein residues. It is therefore of
extreme importance to tune the reactivity of the electrophilic "warhead" to the intended enzyme
target, in order to avoid offꢀtarget reactivity. In the examples given above, this requirement is
fulfilled nearly to perfection. In the case of βꢀlactam antibiotics, offꢀtarget binding is relevant for
the only major side effect, allergic reaction.
Numerous electrophiles have been employed to design covalent target binding into enzyme
inhibitors or receptor ligands. Among the more frequently used functional groups are acrylamides
and other α,βꢀunsaturated groups, boronic acids and αꢀhalogen ketones. However, the majority
3
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 4 of 38
1
2
3
4
5
6
7
8
9
of drugs with a covalent binding mode were discovered serendipitously and were only
retrospectively identified as covalent inhibitors.10 Only few drugs that contain the
abovementioned "classical" electrophiles have entered clinical practice. This prompted us to
search for asꢀyet underexplored chemical functionalities, in order to increase the number of
electrophilic warhead options available for the design of covalent target binders. Some of these
warheads are described for the first time in detail in the present study.
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 present study has two objectives: First, to study the selectivity and reactivity of an
extended group of electrophiles towards biological targets; second, to identify electrophilic
fragments that covalently modify the target enzymes currently pursued in our laboratory. We
therefore studied the binding behavior of six electrophiles (Table 1) towards eleven structurally
and functionally diverse enzyme targets (Table 3).
Some of the electrophiles primarily target cysteine residues (2ꢀ5). Chloroacetylamides (4) were
expected to be more promiscuous due to their (expectedly) high reactivity. The
chloroacetylamide group in metazachlor and related herbicides is responsible for the covalent
binding of these compounds to an active site cysteine of the very longꢀchain fatty acid
elongase.11 These herbicides react irreversibly with protein sidechains via nucleophilic
substitution of the halogen.
The dimethylsulfoniumacetylamides (3) are structurally similar but less reactive. This
electrophile was successfully used to target a cysteine in waterꢀsoluble tissue transglutaminase
inhibitors12 and modifies the cysteine sulfur by substitution of the dimethylsulfide moiety or by
methylation.
A classical motif for covalent inhibitors is the Michaelꢀacceptor acrylamide (2) that targets
nucleophiles like cysteine13 and in some cases threonine.14 Acrylamide groups are present in
4
ACS Paragon Plus Environment

Page 5 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
numerous natural products and rationally designed, targeted inhibitors and covalent receptor
ligands.
7
8
9
Bromodihydroisoxazoles (5) are derived from the natural product acivicin,15 which targets a
cysteine sidechain in γꢀglutamyltranspeptidase and has been also used to target cysteines in
cysteine peptidases.16 Their binding mode involves nucleophilic replacement of the bromine.
Two reactive groups with preference for serine residues, 2ꢀcyanoacetamides and imidazoleꢀ1ꢀ
carboxamides (carbonylimidazoles, 6, 7) were included in this study. 2ꢀCyanoacetamides (6) are
widely used in Medicinal Chemistry not only as reactive group but also as a metabolically labile
moiety to accelerate degradation. They react with oxygen and sulfur nucleophiles by means of an
addition to the triple bond with subsequent hydrolysis to the corresponding ester.17 Imidazoleꢀ1ꢀ
carboxamides (7) have been used to target serines in elastase inhibitors,18 and form carbamates
by substitution of the imidazole. Apart from that, they are rarely used in Medicinal Chemistry.
These electrophiles were linked to eleven amine and aniline scaffolds (Table 2). We chose to
link the electrophiles to these scaffolds via amide linkers because of their modulating influence
on the reactivity of the adjacent electrophile. The amides have a three orders of magnitude lower
reaction rate constant towards GSH when compared to alkyl vinyl ketones.19 The amide linkers
also increase the polarity, when compared to compounds with an allꢀcarbon backbone, which is
beneficial for their solubility and prevents aggregation. The scaffolds were chosen to be diverse
in size, substitution patterns and electronic effects. Since this work focusses on the covalent
rather than the nonꢀcovalent part of the intermolecular interaction between target and ligand,
relatively small scaffolds were chosen, resulting in fragmentꢀlike test compounds. The electronic
properties of the aromatic scaffolds range from electron withdrawing (d, e) to electron donating
(b, c) in order to modulate the reactivity of the electrophiles.
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
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 6 of 38
1
2
3
4
5
6
7
8
9
Several of the electrophiles presented here (3ꢀ5, 7) are somewhat exotic in Medicinal
Chemistry, and very few or no information on their binding behavior towards biological
nucleophiles is currently available. These "rare" electrophiles were included in the study to
specifically assess their usefulness for the design of covalent inhibitors.
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 72 fragmentꢀlike substances resulting from the combination of various electrophiles with
the nonꢀreactive moieties were assayed for inhibition of eleven enzyme targets, reactivity towards
glutathione and stability under assay conditions. Over 900 biochemical measurements were
performed by means of reproducible and robust assays that were established previously in the
context of drug discovery efforts in our laboratory.20ꢀ23 The enzymes studied in this work are
described in Table 3. They represent a wide range of structural recognition motifs, catalytical
functions and source organisms. While some similarities exist between DEN and WNV as well
as between hsMetAPꢀ1 and E. coli MetAP, all other enzymes vary considerably in their structure
and mechanism.
Irreversible inhibitors have been previously described for some of these enzymes, in particular
fosfomycin for MurA6 and fumagillin24 for the MetAPs. MurA has previously been described to
be particularly reactive towards smallꢀmolecular electrophiles;25, 26 therefore we expected this
enzyme to be inhibited readily by promiscuous covalent binders. The E. coli and human MetAPs
contain cysteine residues in the active site, which we also expected to be reactive towards the
tested electrophiles.
RESULTS
Chemistry
6
ACS Paragon Plus Environment

Page 7 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
The allylamines and anilines 1a-k were synthesized according to previously published
procedures.27ꢀ29 The amide/anilide compounds 2a-k, 3a-k, 5a-i, 7a-k, 9 and 10 were synthesized
by amide couplings (Scheme 1, for structures see Table 1, Table 2 and Table 4).
Dimethylsulfonium salts (3a-k) were obtained by treatment of bromoacetylanilides and ꢀamides
with dimethylsulfide following the procedure by Izquierdo et al.30 Bromodihydroisoxazoles (5a-
k) were obtained by cycloaddition of in situ formed bromonitriloxide to acrylanilides and ꢀ
amides (2a-k) following the procedure of Girardin et al.;31 8 was obtained from styrene in the
same fashion. Imidazoleꢀ1ꢀcarboxamides (7a-e) were synthesized by addition of imidazole to the
corresponding isocyanates. 11 and 12 were purchased. The five substituted anilines bear paraꢀ
sustituents that range from highly electronꢀdonating (methoxy, c) to highly electronꢀwithdrawing
(nitro, d) substituents, which we expected to have an influence on the reactivity. Aliphatic
amines were methylꢀ (j) butyl (i) and cyclopentylamine (h). Two amines bearing an aromatic
residue (benzylamine f and phenylethylamine g) and norleucine methyl ester (k) were also
included (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
The acrylamides (2), dimethylsulfonium salts (3) and chloroacetylamides (4) are considered to
be electrophiles that primarily react with cysteine residues, whereas the nitriles (6, 9, 10) as well
as imidazoleꢀ1ꢀcarboxamides (7) are considered to have a preference for catalytic serine residues.
Bromodihydroisoxazoles (5, 8) are expected to target cysteines, serines, lysines or similar
nucleophiles. Allylamines, anilines (1), and norleucine methyl ester (11) were included as
negative controls to identify nonꢀcovalent components of inhibitorꢀenzyme recognition. 4ꢀ
Nitroacetanilide was included in order to study the possibility of an acyl group transfer by the
nitroanilides.
7
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 8 of 38
1
2
3
4
O
Cl
H
5
6
7
8
Cl
Br
H
Cl
N
N
NH2
NH2
NH2
R
Cl
N
R
R
R
R
R
Base
Base
O
1a-e
4a-k
9
1. Di-t-butyl
dicarbonate
+
F3C-COO-
H2
N
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
O
OH
H
N
H
N
R
Br
Br
2. NaH/
allylchloride
3. TFA
R
R
1f-k
KHCO3
O
2a-k
O
5a-k
O
O
CN
H
N
H
O
Cl
N
NH2
NH2
Cl
R
R
R
Base
Base
Me2S
O
6a-k
O
2a-k
H
H
H
N
Imidazole
N
N
N
N
N
R
Br
R
S+
R
C
R
Br-
O
O
O
3a-k
O
7a-e
Scheme 1: Synthesis of the compounds. Cf. Table 2 for definitions of group R.
Biochemical data
The inhibition data of the eleven enzymes and the reactivity towards GSH for all compounds
are listed in Table 4 and Chart 1.
ESI-MS analysis of covalent inhibitor binding
To examine the formation of covalent enzyme adducts, selected active compounds
(inhibition >25%) were incubated with their respective target enzymes for one hour at elevated
concentrations (25 µM enzyme, 300 µM inhibitor) and analyzed by ESIꢀMS. To ensure
functionality of this approach, fosfomycin was reacted with MurA and a mass difference of
138 m/z in comparison to an unmodified sample was observed. This difference corresponds to an
addition of one molecule of fosfomycin per MurA enzyme.
8
ACS Paragon Plus Environment

Page 9 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
No covalent adducts of dimethylsulfonium salts 3a, 3i and 3k, as well as of
bromodihydroisoxazoles 5d and 5k with E. coli MurE could be detected. E. coli MetAP
remained equally unmodified on exposure to 5d. Coꢀincubation of MurB and imidazoleꢀ1ꢀ
carboxamide 7e furnished no adduct.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
In addition to the abovementioned experiments we also assessed adduct formation between 2a,
3a, 4a, 5a, 6a and 7a with E. coli MetAP, MurE and MurA, in order to study the possibility of
binding events without detectable influence on the enzyme activity, but no adduct formation was
observed for any of the combinations.
Cell viability assay
A selection of compounds was submitted to cell viability assays. HeLa cells were incubated for
24 h with the compounds of interest at varying concentrations and viability determined with a
reaszurin based assay. The EC50 values are represented in Table 5. Only compound 4a had an
EC50 value below 50 µM.
DISCUSSION
In this work we studied the inhibitory potential of six electrophilic moieties, covering a broad
range of reactivity, against several enzymes. Some of the enzymes contain catalytically active
serine and cysteine residues that we expected to have an increased nucleophilicity and therefore
to be reactive towards the tested electrophiles. The assayed enzymes represent a broad spectrum
of species (viral, bacterial, human targets), require different types of cofactors (small proteins,
metals) as well as diverse substrates (peptides, proteins, carbohydrates, modified carbohydrates).
The electrophiles were selected with respect to their reactivity and suitability for pharmacological
9
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 10 of 38
1
2
3
4
5
6
7
8
9
applications. Therefore, highly reactive electrophiles, like aldehydes or trifluoromethylketones
that are sometimes used during the target validation and early drug discovery stages, but without
conceivable application in approved drugs, were not analyzed in this study. For all six evaluated
reactive groups, exemplary applications in Medicinal Chemistry can be found in literature (cf.
Table 1). Allylamines/ꢀanilines (1) were used as nonꢀreactive references to estimate the part of
covalent interaction of the series of electrophiles (2ꢀ7). Depending on the enzyme, the
compounds were tested at concentrations of 25 or 50 µM, which is at or above the upper
concentration range that is typically used in compound screening. The concentrations were
chosen as a "worst case scenario" and to permit the detection of "weak" binders.
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 nonꢀreactive controls (1a-k, 11 and 12) show little or no inhibition of any of the enzymes
and did not react with GSH on a timescale relevant to the assays. Therefore, all significant
inhibitory activities can be attributed to the electrophilic moieties in the test compounds.
The acrylanilides and ꢀamides (2a-k) are inactive in all enzymatic assays. They were not
reactive towards GSH on a timescale relevant for the assays (<5% conversion over 3 h), but
formation of GSH adducts was detected by LCꢀMS after extended incubation times. Therefore,
these electrophiles appear to be suitable for the design of targeted covalent inhibitors, which are
first anchored to the target protein by nonꢀcovalent interactions, followed by the (slow) formation
of a covalent bond. The low promiscuity and offꢀtarget reactivity of covalent drugs with
acrylanilide and acrylamide electrophiles is clearly demonstrated by their inertness towards all
tested enzymes. It is highly remarkable that also the unsubstituted acrylamide, under the given
experimental conditions, showed only negligible binding to all enzymes and GSH. An earlier
report already indicated that acrylamide has a halfꢀlive of ~3 h at cellular levels of glutathione
(~3 mM).32
10
ACS Paragon Plus Environment

Page 11 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
Dimethylsulfonium salts (3a-k) have low unspecific reactivity. Interestingly, 3b-3d
decomposed immediately upon dilution in aqueous media, in reversal of the synthesis. Therefore,
these compounds were excluded from the screening. The other dimethylsulfonium salts (3a, 3e-
k) are stable under aqueous conditions and were not reactive towards GSH.
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
Chloroacetylanilides and ꢀamides (4a-i) show the most "pronounced" reactivity towards GSH
of all tested compounds (kGSH ~ 0.1 lꢁmolꢀ1ꢁsꢀ1), with aqueous halfꢀlives still above 200 h, but
virtually no inhibition of any enzyme. The group is therefore a slightly more reactive alternative
to the acrylamides (2a-k), but retains a surprisingly low random reactivity.
2ꢀBromodihydroisoxazoles (5a-k) show no GSH reactivity and a halfꢀlive above 200 h, making
them a rather stable electrophilic group. They exhibit limited inhibitory potential in the
enzymatic assays and also some promiscuous binding, although not very potent and with a focus
on the Murꢀenzymes. Substances 5c-e and 5g-k are relatively selective inhibitors of MurE. The
bromodihydroisoxazoles have been described as cysteineꢀtargeting electrophiles, but ecMurE
does not contain a solventꢀexposed cysteine residue. To pinpoint the reactive residue of ecMurE
that interacts with this warhead, four of the most active compounds (5d, 5g, 5i, 5k) were
submitted to docking experiments with MurE (PDBꢀID: 1E8C33). It was found that in all cases
the bromodihydroisoxazole moiety is located in a pocket close to the mDAPꢀbinding domain.
This pocket contains a lysine (Lys119) that is in a potentially reactive orientation and distance
towards the reactive moiety of these compounds (Figure 6). However, a covalent interaction with
the MurE protein could not be confirmed by mass spectrometry. Therefore, these compounds are
reversible or reversibleꢀcovalent inhibitors of MurE.
11
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 12 of 38
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: Docked structure of 5k in the active site of MurE. Lys119, a potential binding partner
for the ligand, is located in the center of the graphic. The carbon atoms of the ligand are shown in
yellow. PDB code of the target structure: 1E8C.33 This figure was generated using Chimera.34
Cyanoacetanilides and amides (6a-k) are not reactive towards GSH, show no substantial
inhibition or multiꢀenzyme reactivity and are stable, with halfꢀlives of over 200 h. Imidazoleꢀ1ꢀ
carboxamides (7a-k) show no GSH reactivity and halfꢀlives in the range of days. They show no
promiscuous binding to the screened enzymes.
Substances 6a-k and 7a-k also show no inhibition of DEN up to 500 µM (data not shown).
Active site serines are largely resistant towards inhibition by all electrophiles studied here.
DEN, WNV, THR and MurB, all containing activeꢀsite serines with increased nucleophilicity,
are neither inhibited by the 2ꢀcyanoacetamides 6a-k nor the more reactive aminomethylnitriles 9
and 10 and the imidazoleꢀ1ꢀcarboxamides (7a-m).
All compounds that exhibit inhibition of an enzyme were also tested for covalent adduct
formation with their corresponding targets, but no adducts were found. This was also extended to
all phenyl substituted substances (2aꢀ7a), because they ranged among the most reactive of their
group (cf: Chart 1, 4a), with E. coli MetAP, E. coli MurA and E. coli MurD. This was done to
12
ACS Paragon Plus Environment

Page 13 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
search for "unspecific" adducts with residues outside the active site. No adducts could be
identified. These findings further emphasize the limited unspecificity of these reactive groups.
Cell toxicity was assessed for all compounds exhibiting significant inhibition of any enzyme
and for all phenylꢀ (2aꢀ7a) and butylꢀ (2iꢀ6i) substituted substances as well as nitriles 9 and 10.
The only substance showing cell toxicity with an EC50 value below 50 µM is the
chloroacetylanilide 4a. When compared to 4i, with equal GSH reactivity and no cytotoxicity
below 100 µM, this can not solely be attributed to unspecific reactivity of 4a, but is probably due
to specific interaction with a vital target.
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
CONCLUSION
In this study, in which biochemical properties of 72 substances were determined in over 900
measurements, we assessed the selectivity of diverse electrophilic groups towards targets with
pharmacological relevance.
We found that acrylꢀ and chloroacetylamides/anilides as well as 2ꢀcyanoacetamides and
imidazoleꢀ1ꢀcarboxamides exhibit a surprisingly low offꢀtarget reactivity and can therefore be
considered as suitable warheads in targeted covalent inhibitors. These inhibitors must recognize
their targets by specific, nonꢀcovalent interactions that increase the residence time sufficiently to
raise the probability of the second, covalent binding event. We will apply these findings in
targeted covalent inhibitors that are currently under development in our group.
The 3ꢀbromoꢀ4,5ꢀdihydroisoxazole moiety selectively addresses the antibiotic target enzyme
MurE. Therefore, in future drug discovery efforts aimed at the development of antibiotics that
interfere with cytoplasmatic peptidoglycan precursor synthesis, this scaffold may be a valuable
13
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 14 of 38
1
2
3
4
5
6
fragment for inhibitor design. Apart from that, this reactive moiety did not show any random
reactivity with proteins or significant cytotoxicity.
7
8
9
An unexpected but significant consequence of the present study is the relatively low inhibitory
potential of the reactive compounds against the analyzed enzymes. Even in cytotoxicity assays
and when we looked for inhibitor enzyme adduct formation we did not find any elevated
cytotoxicity or unspecific modification of proteins. Particularly in the case of
chloroacetylamides/ꢀanilides (4) and dimethylsulfonium salts (3), which we consider to be among
the most reactive in this series, this is a promising results. From these results the following
consequences for moderately reactive groups in Medicinal Chemistry can be drawn. Promiscuous
reactivity and offꢀtarget effects of electrophiles with moderate reactivity may often be
overestimated. It also does not appear justified to generally exclude "reactive" moieties from
compound libraries for screening purposes, since the nonꢀspecific reactivity may turn out to be
much inferior than anticipated. A sufficiently close contact of the electrophilic group and the
nucleophile of the target and a sufficiently long residence time of the electrophile is required to
establish a covalent interaction. In the absence of specific, nonꢀcovalent interactions between
ligand and target, the covalent binding potential of electrophiles with intermediate reactivity is
low.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
EXPERIMENTAL SECTION
General. The stock solutions for all assays were prepared in DMSO (10 mM) and stored
at ꢀ20° C in brown glass vials with a teflon coated lid and further diluted with assay buffer as
required. Determination of biochemical data was performed in triplicate with standard deviations
typically below 15%.
14
ACS Paragon Plus Environment

Page 15 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
DEN and WNV overexpression and purification. The dengue virus (serotype 2) and Westꢀ
Nile Virus NS2BꢀNS3 proteases were expressed and purified according to the protocol described
by Steuer et al.22, 35 For both proteases the cofactor is covalently connected to the protease
domain by a glycineꢀserine linker (GGGGSGGGG).36
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
E. coli MetAP overexpression and purification. E. coli MetAP was obtained by
overexpression in E. coli BL21(DE3) using an Argꢀ175ꢀGln mutant kindly provided by Prof. W.
T. Lowther and Prof. B. Matthews following their protocol.24
HsMetAP-1 overexpression and purification. The gene for HsMetAPꢀ1 was synthesized
with an optimized nucleotide sequence for E. coli expression by a commercial supplier and
cloned into the pETꢀ28a(+) vector (Novagen) using the BamH1 and Sac1 cloning sites. The
enzyme with Nꢀterminal Hisꢀtag was overexpressed in E. coli BL21(DE3). The enzyme was
isolated by nickel affinity chromatography (Chelating Sepharose Fast Flow, Amersham
Biosciences; buffer: HEPES 50 mM pH 7.9, 0.5 KCl; imidazole gradient 10ꢀ250 mM).
MurA overexpression and purification. Was performed as previously reported.37 MurB,
MurC, MurD, MurE and MurF were obtained analogously.
Thrombin assay. Bovine thrombin was purchased from SigmaꢀAldrich. The thrombin assay
was performed as a continuous fluorimetric assay using a BMG Labtech Fluostar OPTIMA
microtiter fluorescence plate reader and black 96 well Vꢀbottom plates from Greiner BioꢀOne
(Germany). The excitation wavelength was 355 nm and the emission wavelength was 460 nm.
The protease was assayed using the substrate BocꢀValꢀProꢀArgꢀAMC, purchased from Bachem
(Germany). The final concentrations of the enzyme and substrate were 10 nM and 50 µM,
respectively. The inhibitors were preincubated with the enzyme for 15 min at a concentration of
25 µM. The cleavage reaction was initiated by addition of the substrate. The assay buffer
15
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 16 of 38
1
2
consisted of 50 mM TrisꢀHCl, pH 7.5, 150 mM NaCl and 0.05% Tween 20.38 The activity of the
enzyme was determined as the slope per second (RFU/s) and monitored for 10 min.
DEN and WNV assays. Were performed as previously reported.20, 39
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
MetAP assay by HPLC. The MetAP HPLC assay was perfomed as an endpoint assay as
follows: Into a 96ꢀwellꢀplate (clear PS, UꢀBottom from Greiner Bio One) 10 µl inhibitor solution
(250 µM), 60 µl masterꢀmix (0.125% w/v BSA, 0.167 M NaCl, 0.167 mM CoCl2 in 50 mM
TRISꢀbuffer at pH 7.5) and 10 µl enzyme (500 nM) were added per well. The content of each
well was gently stirred with the pipet tips followed by incubation at 37 °C for 20 min. Then
MGMM (20 µl, 2 mM) again followed by gentle stirring and incubation for 15 min (HsMetAPꢀ1)
or 20 min (EcMetAP).
This results in a total volume of 100 µl and the following final concentrations: CoCl2: 100 µM;
NaCl: 100 mM; BSA: 0.075% (w/v); enzyme: 50 nM; MGMM: 400 µM; inhibitor: 25 µM. The
enzyme reactions were stopped by adding 10 µl 4% TFA per well.
The HPLC elution was as follows: The product GMM was quantified by monitoring its UV
absorbance at 214 nm. Flow rate 1.0 ml/min at 30°C. Initial conditions: 99% water (0.1% TFA)
(solvent A) and 1% acetonitrile (0.1% TFA) (solvent B). These conditions were kept for
0.25 min. From 0.25 min to 2.5 min the gradient was ramped linearly to 60% B. The elution
proceeded with 100% B until 3.25 min. The column was then reꢀequilibrated at initial conditions
for 3 min. A ReproSilꢀPur ODS 3.5 µm, 50 x 2 mm column from Maisch was used.
MurA Assay. Was performed as previously reported.37
MurB, MurC, MurD, MurE, MurF Assays. MurB to MurF reactions were determined in
single assays by a LCꢀMS method, measuring product formation.
The respective NAGꢀsubstrates (Nꢀacetyl glucosamine) for each Mur enzyme assay were
generated by enzymatic conversion by incubating the preceding enzymes at 400 nM with their
16
ACS Paragon Plus Environment

Page 17 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
required substrates (UNAG 1mM, PEP 1mM, NADPH and amino acids 2 mM, ATP in
increasing amounts of 2 mM (MurD substrate), 3 mM (MurE substrate), 4 mM (MurF substrate)
for 5 h at 37 °C. Reactions were then terminated by heat denaturation (85 °C for 5 min). NAGꢀ
substrate preparations were stored at ꢀ20 °C until their use.
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
For inhibition studies, the respective E. coli wildꢀtype enzyme was preꢀincubated with
inhibitors at 37 °C for 10 min, after which in case of the Mur ligases the NAGꢀsubstrate mixtures
including the cosubstrate ATP, and in case of E. coli MurB NADPH, were added and incubated
for another 10 min. Reactions were initiated by the addition of the respective amino acid
cosubstrates: LꢀAla (MurC), DꢀAla (MurD), mDAP (MurE) or dipeptide DꢀAlaꢀDꢀAla (MurF).
MurB assays reactions were initiated with the EPꢀUNAGꢀsubstrate preparation.
For the Mur ligases, final concentrations in the assay were: enzyme 4 nM, ATP and respective
NAGꢀsubstrates approx. 200 µM, NH4HCO3 50 mM, MgCl2 2 mM (MurD, MurE) / MgCl2
10 mM (MurC), MgCl2 5 mM (MurF), BSA 0.07% (w/v), DMSO 1% (v/v).
For MurB, the final concentrations in the assay were: MurB 4 nM, NADPH 200 µM and EPꢀ
UNAG substrate approx. 200 µM, NH4HCO3 50 mM, KCl 5 mM, BSA 0.07% (w/v), DMSO 1 %
(v/v).
Reactions were quenched after 15 min at 37 °C by adding 5 µL of formic acid (10% (v/v)) per
100 µL total reaction volume. A volume of 5 µL was injected onto two coupled RP18ꢀAQ preꢀ
columns (ReproSilꢀPur C18ꢀAQ, 3 µm, 3x20 mm, Dr. Maisch, Ammerbuch, Germany) and
eluted with a rapid gradient. Solvent A: 50 mM ammonium formate + 0.2% HCOOH, solvent B:
500 mM ammonium formate + 1.6% HCOOH – methanol 10/90 (v/v). HPLC conditions were as
follows: 100% A for 0.5 min, gradient to 95% B in 1.0 min. Solvent B was kept at 95% for
0.5 min followed by a gradient to initial conditions in 0.1 min. Initial conditions were maintained
until a total analysis time of 7 min. Flow rate was 0.4 mL/min. The HPLC system consisted of an
17
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 18 of 38
1
2
3
4
5
6
7
8
9
Agilent 1200 series device (consisting of G1322A degasser, G1312B binary pump, G1367C
autosampler, G1316B column oven) coupled to a ESI massspectrometer micrOTOFꢀQ II (Bruker
Daltonik, Bremen, Germany) operating in electrospray negative ionization mode. Product
formation of each Mur reaction was quantified as peak area of the monitored EIC trace.
Reactivity towards glutathione.26 25 µL of a freshly prepared 10 mM solution of glutathione
in water was added to 950 µl of a 50 mM TRISꢀmaleateꢀbuffer at pH 7.5 and thoroughly mixed.
The reaction was started by adddition of 25 µL of the 10 mM compound stock in DMSO (final
concentration: 250 µM). 100 µl were transferred into HPLCꢀvials containing 10 µl of 10%
phosphoric acid at 0, 60, 180 min. These aliquots were analysed by RPꢀHPLC and the decrease
of the test compound monitored at a suitable wavelength.
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
Flow rate 1.0 ml/min at 30 °C. Initial conditions: 99% water (0.1% TFA) (solvent A) and 1%
acetonitrile (0.1% TFA) (solvent B). These conditions were kept for 0.25 min. From 0.25 min to
5 min the gradient was ramped linearly to 100% B. The elution proceeded with 100% B until
3.25 min. The column was then reꢀequilibrated at initial conditions for 3 min. A ReproSilꢀPur
ODS 3.5 µm, 50 x 2 mm column from Maisch was used.
ESI-MS analysis of covalent inhibitor binding: All experiments were performed in
NH4HCO3 buffer to ensure ESIꢀMS compatibility.
Enzymes (25 µM) were incubated for 60 min at 37 °C in NH4HCO3 50 mM, pH 7.9 with
300 µM test compound and individual ionic additives (E. coli MetAP: CoCl2 200 µM, E. coli
MurC: MgCl2 10 mM, E. coli MurD, MurE: MgCl2 2 mM, E. coli MurF: MgCl2 5 mM).
Incubation conditions were analogous to the enzymatic assay procedure: E. coli MetAP:
incubation of enzyme with inhibitor; E. coli MurA: preꢀincubation of enzyme with the substrate
UNAG 300 µM for 10 min at 37 °C prior to addition of inhibitor; E. coli MurB to MurF: preꢀ
incubation of enzyme with inhibitor for 10 min, 37 °C prior to addition of substrates ATP
18
ACS Paragon Plus Environment

Page 19 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
7
8
9
200 µM, sugar nucleotide substrate approx. 200 µM. To investigate the effects of substrate
exposure on the enzyme adduct profile, E. coli MurD was incubated without substrates and under
simultaneous addition of substrates and inhibitor, respectively. Total incubation volume was
100 µL. Untreated enzymes served as controls. Desalting and removal of excess substrates, nonꢀ
bound inhibitor and DMSO was performed by spinꢀdesalting using preꢀequilibrated (NH4HCO3
50 mM) Sephadex Gꢀ25ꢀcolumns, which were prepared according to Weingart.40 A volume of
20 µL was analyzed by flow injection analysis using an Agilent 1200 series HPLC device
coupled to a ESIꢀMS instrument (mircOTOFꢀQII, Bruker Daltonik, Bremen) operating in
positive ionization mode. Removal of DMSO was required for sufficient MS signal intensity
during flow injection analysis. Injection conditions were as follows: no column, isocratic flow of
H2O + 0.1% HCOOH / CH3CN + 0.1% HCOOH 95/5 (v/v) at 0.3 mL/min. Instrument
calibration and external mass calibration at the end of each analysis run were performed with
ESIꢀTunemix (Fluka) calibration standard. Mass spectra of analytes with multiple, variable
charges were deconvoluted using the Maximum Entropy Deconvolution algorithm (Compass
DataAnalysis Version 4.0 SP4, Bruker Daltonik).
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
41
Docking. The docking experiments were performed using GOLD 5.1 with Hermes on an
Intel Core2 Quad CPU Q9450 @ 2.66 GHz running open SuSE 11.0. The Xꢀray structures were
downloaded from the PDB.42 Ligands and all waters were extracted. Hydrogens were added
exhaustively using Hermes. The binding site was set according to the previously extracted ligand
plus 6 Å. GoldScore was used as scoringꢀfunction. All settings were used in the standard
configuration except for setting the search option to the "high flexibility" preset. Graphics were
created using Chimera.34
Cell viability assay. HeLa cells were seeded in triplicate into 96ꢀwell plates at 5 × 103 cells in
a final volume of 50 µl per well and incubated at 37 °C/5% CO2 for 24 h. To each well 50 µl of
19
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 20 of 38
1
2
3
4
5
6
7
8
9
the compounds as 2x concentrated stock solution in phenolꢀred free DMEM supplemented with
FBS, 100 IU/ml penicillin, 100 ꢂg/ml streptomycin, 1 mm sodium pyruvate, 4 mm lꢀglutamine
were added. After incubation at 37 °C/5% CO2 for 24 h, the media was replaced with fresh
phenol redꢀfree medium containing CellTiterꢀBlue reagent (100 ꢂl of medium + 20 ꢂl of reagent)
per well. Cells were incubated for a further 3 h at 37 °C/5% CO2, and then the absorbance at
570 nm and 610 nm was measured. After background subtraction, absorbance values were used
to calculate the percentage viability, expressed as a percentage of untreated controls which were
set as 100% viable.
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
Chemistry. NMRꢀSpectra were measured at room temperature (rt) on a Varian Mercury Plus
(300 MHz) and a Varian NMR System (500 MHz). The chemical shifts are given relative to
TMS (δ = 0 ppm) using the residual solvent peak as internal standard. High resolution mass
spectra were measured on a Bruker micrOTOFꢀQ II with sodium formate as calibrant. All
reactions were carried out in standard glassware; temperatures refer to the heating or cooling bath
temperatures. All reagents and solvents were unless otherwise stated purchased from Sigmaꢀ
Aldrich and used as received. MPLC was performed on a Biotage Isolera One using selfꢀpacked
columns with E60 flash silica gel from SiliCycle. Cyclohexane (technical grade) was purchased
from Grüssing and distilled prior to use. Ethyl acetate used for column chromatography was
distilled prior to use. Dibromoformaldoxim was prepared following the procedure provided by
Wade et al.54 The purity of all compounds was determined to be above 95% unless otherwise
stated by either HPLC (on an Agilent 1200 HPLC with DAD on a Reprosil Pur ODSꢀ3 50x2 mm
column from Maisch; Gradient: 0ꢀ0.5 min 1% B, linear gradient to 100% B at 8 min, hold 100%
B for 1 min and reꢀequilibrate at 1% B for 3 min; A: MeCN + 0.1% TFA. B: Water + 0.1% TFA;
detection wavelength was, unless otherwise stated 214 nm) or combustion analysis on a Vario EL
1
from Foss Heraeus. For all compounds known prior to this work HꢀNMRꢀspectra were used to
20
ACS Paragon Plus Environment

Page 21 of 38
Journal of Medicinal Chemistry
1
2
verify identity. For all unknown compounds 1Hꢀ, 13Cꢀ (where needed: 2D NMRꢀspectra) and high
resolution mass spectra were used to verify the identity. Detailed descriptions of the syntheses of
compounds known prior to this work can be found in the Supporting Information.
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
Procedure for the synthesis of dimethylsulfoniumbromide 3a (3b-k were synthesized
analogously, analytical data is provided in the Supporting Information).30 200 mg of 2ꢀ
bromoꢀNꢀphenylacetamide (2 mmol) were dissolved in 600 µL of dimethylsulfide and stirred in a
closed snapꢀcap vial over night. To the resulting solution 10 ml ether were added and the formed
precipitate collected by filtration. The solid was washed repeatedly with ether and dried in vacuo
1
to obtain 3a in 62% (161 mg) yield. H-NMR (500 MHz, DMSO): δ = 7.57 (d, J = 7.7 Hz, 2H,
2x CHArꢀo), 7.36 (t, J = 7.7 Hz, 2H, 2x CHArꢀm), 7.13 (t, J = 7.7 Hz, 1H, CHArꢀp), 4.64 (s, 2H,
CH2), 2.97 (s, 6H, S(CH3)2) ppm. 13C-NMR (125 MHz, DMSO): δ = 162.0 (CO), 137.8 (CqꢀAr),
129.0 (2x CHArꢀm), 124.4 (CHArꢀp), 119.5 (2x CHArꢀo), 48.2 (CH2), 24.8 (S(CH3)2) ppm. HRMS
(ESI) m/z: [M]+ calcd for [C10H14NOS]+: 196.0791; found: 196.0791.
Procedure for the synthesis of 3-bromo-4,5-dihydroisoxazole 5a (5b-k & 8 were
synthesized analogously, analytical data is provided in the Supporting Information).31
Dibromoformaldoxime (100 mg, 0.493 mmol) and Nꢀphenylacrylamide (2a, 87 mg, 0.592 mmol)
were dissolved in DMF (1 mL) at 0 °C. An aqueous solution (1.2 mL) of KHCO3 (123 mg,
1.232 mmol) was added to this solution dropꢀwise during 1 h. The resulting mixture was warmed
to rt and stirred for 1h. After addition of ethyl acetate and water, the aqueous layer was separated
and extracted three times with ethyl acetate. The combined organic layers were dried and
1
concentrated in vacuo. The residue was purified by MPLC to give 5a in 72% (95 mg) yield. H
NMR (300 MHz, CDCl3): δ = 8.40 (s, 1H, NH), 7.58 (d, 2H, J = 7.5 Hz, HAr), 7.36 (t, 2H, J =
7.5 Hz, HAr), 7.17 (t, 1H, J = 7.5 Hz, HAr), 5.17 (dd, 1H, J = 7.5, 9.8 Hz, CH), 3.68 (dd, 1H, J =
9.8, 17.9 Hz, CH2), 3.61 (dd, 1H, J = 7.5, 17.9 Hz, CH2) ppm. 13C NMR (75 MHz, CDCl3): δ =
21
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 22 of 38
1
2
3
4
5
6
7
8
9
167.56 (CO), 139.17 (CꢀBr), 136.39(Cq Ar), 129.12 (CHAr), 125.16 (CHAr), 119.93 (CHAr), 79.14
(CH), 45.82 (CH2) ppm. HRMS (ESI) (m/z) [M + Na+] calcd for [C10H9O2N2BrNa]+: 290.9740,
found: 297.9709.
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
ASSOCIATED CONTENT
Supporting Information. Complete analytical data and synthetic procedures; tabular
representation of halfꢀlives. This material is available free of charge via the Internet at
http://pubs.acs.org.
AUTHOR INFORMATION
Corresponding Author
*Phone: ++49ꢀ6221ꢀ554875. Eꢀmail: c.klein@uniꢀheidelberg.de
Funding Sources
Christoph Nitsche acknowledges funding by the Studienstiftung des deutschen Volkes. This
work was partially supported by grants from the Deutsche Krebshilfe and Deutsche
Forschungsgemeinschaft, KLꢀ1356/3ꢀ1.
ACKNOWLEDGMENT
We thank Michael Wacker, Heiko Rudy and Tobias Timmermann for technical assistance and
Dominik Graf for performing the cell viability assay. Mira Behnam kindly determined some of
the inhibition data. Christoph Nitsche acknowledges funding by the Studienstiftung des
deutschen Volkes. This work was partially supported by grants from the Deutsche Krebshilfe and
Deutsche Forschungsgemeinschaft, KLꢀ1356/3ꢀ1.
22
ACS Paragon Plus Environment

Page 23 of 38
Journal of Medicinal Chemistry
1
2
3
4
ABBREVIATIONS
5
6
7
8
EcMetAP, methionine aminopeptidase from E. coli; HsMetAPꢀ1, methionineꢀaminopeptidase
type 1 from Homo sapiens; MurA, UDPꢀNꢀacetylglucosamineꢀ1ꢀcarboxyvinyltransferase from E.
coli; MurB, uridine diphosphoꢀNꢀacetylenolpyruvylglucosamine reductase from E. coli; MurC,
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
UDPꢀNꢀacetylmuramateꢀalanine ligase from E. coli; MurD, UDPꢀNꢀacetylmuramoylꢀ
ꢀglutamate ligase from E. coli; MurE, UDPꢀNꢀacetylmuramyl tripeptide synthetase from E. coli;
MurF, UDPꢀmurNacꢀtripeptide ꢀalanylꢀ ꢀalanine ligase from E. coli; GSH, glutathione; DEN,
Lꢀalanineꢀ
D
D
D
protease from dengue virus; WNV, protease from West Nile virus; THR, bovine thrombin;
mDAP, meso diaminopimelic acid.
23
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 24 of 38
1
2
3
4
Table 1: Electrophiles for covalent protein modification.
5
6
7
8
Targeted
nucleophiles
# Electrophile
Example
ꢀ
Binding mode
ꢀ
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
Cysteine,
Afatinib43
2
Threonine
Syringolin A14
Transglutaminase
inhibitors
Cysteine
3
(M. Griffin et
al.)12, 44
Cysteine
Cysteine
Metazachlor11
4
5
Bromoacivicin16,
45, 46
NEPꢀinhibitors
(S. S. Ghosh et
al.)47
Serine
6
(proposed)
(proposed)
Cysteine
Balicatib48
O
Lꢀnorvaline
methylester
azolide18
Serine
7
N
N
24
ACS Paragon Plus Environment

Page 25 of 38
Journal of Medicinal Chemistry
1
2
3
4
Table 2. Nonꢀreactive moieties that were combined with the electrophiles.
5
6
7
8
9
a
g
b
h
c
i
d
j
e
f
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
k
25
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 26 of 38
1
2
3
4
Table 3: Target enzymes.
5
6
7
Name
Abbreviation Type
Native
Catalyzed reaction
Function
Nucleophilic
organism
residues in or near
the active site
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
Dengue virus
protease
DEN
WNV
THR
Serine endoꢀ
protease
Dengue
Virus
Cleavage of viral
polyprotein49
Production of
viral enzymes
SER13550
SER13551
SER19553
West Nile virus
protease
Serine endoꢀ
protease
WestꢀNile Cleavage of viral
Virus
Production of
viral enzymes
polyprotein
Thrombin
Serine endoꢀ
protease
Bos
taurus
Conversion of fibrinogen
(coagulation factor I) to
fibrin (factor Ia)52
Coagulation
E. coli Methionine
aminopeptidase
ecMetAP
Metallo exoꢀ
protease
E. coli
Cleavage of Nꢀterminal
methionine54
Protein
maturation
HIS79
H. sapiens
Methionine
aminopeptidase type
1
hsMetAPꢀ1 Metallo exoꢀ
Homo
sapiens
Cleavage of Nꢀterminal
methionine
Protein
maturation
HIS23155
protease
E. coli MurA
E. coli MurB
E. coli MurC
MurA
MurB
MurC
Transferase
E. coli
Transfer of enol pyruvate
from PEP to UDPꢀNꢀ
acetylglucosamine
Bacterial murein CYS11554
biosynthesis56
Oxidoreductase E. coli
Reduction of EPꢀUDPꢀNꢀ Bacterial murein SER22957
acetylglucosamine to UDPꢀ biosynthesis
MurꢀNAc
ATPꢀdependent E. coli
ligase
Ligation of LꢀAla and
UDPꢀMurꢀNAc
Bacterial murein ꢀ
biosynthesis
26
ACS Paragon Plus Environment

Page 27 of 38
Journal of Medicinal Chemistry
1
2
3
4
5
6
E. coli MurD
E. coli MurE
MurD
MurE
ATPꢀdependent E. coli
ligase
Ligation of DꢀGlu and
UDPꢀMurꢀNAcꢀLꢀAla
Bacterial murein CYS413
biosynthesis
ATPꢀdependent E. coli
Ligation of mDAP and
Bacterial murein (LYS391)
7
ligase
UDPꢀMurꢀNAcꢀLꢀAlaꢀDꢀ biosynthesis
Glu
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
E. coli MurF
MurF
ATPꢀdependent E. coli
ligase
Ligation of DꢀAlaꢀDꢀAla
and UDPꢀMurꢀNAcꢀLꢀAlaꢀ biosynthesis
DꢀGluꢀmDAP
Bacterial murein CYS371
27
ACS Paragon Plus Environment

Journal of Medicinal Chemistry
Page 28 of 38
1
2
3
4
5
6
Table 4: Enzyme inhibition data for compounds 8-12. n.r. – no reaction, k < 0.1 lꢁmolꢀ1ꢁsꢀ1; n.i.
– no inhibition (<15%); color code: 15ꢀ29%. Concentration of test compounds: 25 µM for
ecMetAP, hsMetAP1, THR & MurAꢀMurF, 50 µM for DEN and WNV.
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
n.r. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. 17 n.i. n.i.
n.r. n.i. 23 n.i. n.i. n.i. n.i. 17 n.i. n.i. n.i. n.i.
n.r. 17 18 n.i. n.i. n.i. n.i. n.i. n.i. 20 n.i. n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. 22 n.i. 18 17 n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i.
8
9
10
11
12
28
ACS Paragon Plus Environment

Page 29 of 38
Journal of Medicinal Chemistry
1
2
Chart 1: Enzyme inhibition data. n.r. – no reaction, k < 0.1 lꢁmolꢀ1ꢁsꢀ1; n.i. – no inhibition (<15%); color codes: 15ꢀ29%; 30ꢀ50%;
could not be determined (see text). Concentration of test compounds: 25 µM for ecMetAP, hsMetAP1, THR & MurAꢀMurF, 50 µM
for DEN and WNV.
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
n.r. n.i. n.i. n.i. n.i. 22 16 n.i. n.i. n.i. n.i. n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. 15
n.r. n.i. n.i. n.i. n.i. n.i. n.i. 20 n.i. n.i. n.i. n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. 20 n.i. n.i. n.i. n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. 17 n.i. n.i. n.i. n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. 17 n.i. n.i. n.i. n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. 19 n.i. n.i. n.i. n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i.
n.r. n.i. n.i. n.i. n.i. 17 n.i. n.i. n.i. n.i. n.i. n.i.
0.10 n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i.
0.13 n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i.
0.13 n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i.. n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i.. n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i.. n.i.
0.11 n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. 15 n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i.
n.r. n.i. 22 n.i. 19 n.i. n.i. n.i. n.i. n.i. n.i. n.i.
n.r. n.i. 21 n.i. n.i. n.i. n.i. n.i. n.i. n.i. 18 n.i.
n.r. n.i. n.i. n.i. 28 n.i. n.i. n.i. n.i. n.i. 44 n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. 23 21
n.r. n.i. n.i. n.i. n.i. n.i. n.i. 18 n.i. 16 n.i. 15
n.r. n.i. n.i. n.i. n.i. n.i. n.i. 16 n.i. 18 23 21
n.r. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. 15 16 n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. 27 n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. 39 n.i.
n.r. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. n.i. 23 n.i.
a
b
c
d
e
f
a
b
c
d
e
f
4
1
g
h
i
h
i
a
b
c
d
e
f
j
k
a
b
c
d
e
f
g
h
i
5
2
g
h
k
a
29
ACS Paragon Plus Environment