Antibiotic Discovery with Synthetic Fermentation: Library Assembly, Phenotypic Screening, and Mechanism of Action of β-Peptides Targeting Penicillin-Binding Proteins

Article abstract of DOI:10.1021/acschembio.9b00227

In analogy to biosynthetic pathways leading to bioactive natural products, synthetic fermentation generates mixtures of molecules from simple building blocks under aqueous, biocompatible conditions, allowing the resulting cultures to be directly screened for biological activity. In this work, a novel β-peptide antibiotic was successfully identified using the synthetic fermentation platform. Phenotypic screening was carried out in an initially random fashion, allowing simple identification of active cultures. Subsequent deconvolution, focused screening, and structure-activity relationship studies led to the identification of a potent antimicrobial peptide, showing strong selectivity for our model system Bacillus subtilis over human HEK293 cells. To determine the antibacterial mechanism of action, a peptide probe bearing a photoaffinity tag was readily synthesized through the use of appropriate synthetic fermentation building blocks and utilized for target identification using a quantitative mass spectrometry-based strategy. The chemoproteomic approach led to the identification of a number of bacterial membrane proteins as prospective targets. These findings were validated through binding affinity studies with penicillin-binding protein 4 using microscale thermophoresis, with the bioactive peptide showing a dissociation constant (K_d) in the nanomolar range. Through these efforts, we provide a proof of concept for the synthetic fermentation approach presented here as a new strategy for the phenotypic discovery of novel bioactive compounds.

Full text of DOI:10.1021/acschembio.9b00227

Subscriber access provided by UNIV AUTONOMA DE COAHUILA UADEC
Article
Antibiotic discovery with synthetic fermentation: library
assembly, phenotypic screening, and mechanism of action
of beta-peptides targeting penicillin-binding proteins
Iain A. Stepek, Trung Cao, Anika Koetemann, Satomi Shimura, Bernd Wollscheid, and Jeffrey W. Bode
ACS Chem. Biol., Just Accepted Manuscript • DOI: 10.1021/acschembio.9b00227 • Publication Date (Web): 16 Apr 2019
Downloaded from http://pubs.acs.org on April 17, 2019
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 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 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.
is published by the American Chemical Society. 1155 Sixteenth Street N.W.,
Washington, DC 20036
Published by American Chemical Society. Copyright © American Chemical Society.
However, no copyright claim is made to original U.S. Government works, or works
produced by employees of any Commonwealth realm Crown government in the
course of their duties.

Page 1 of 23
ACS Chemical Biology
1
2
3
4
5
6
7
8
Antibiotic discovery with synthetic fermentation:
library assembly, phenotypic screening, and mechanism of action
of beta-peptides targeting penicillin-binding proteins
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
Iain A. Stepek1, Trung Cao1, Anika Koetemann2, Satomi Shimura1,
2
*
1, 3
*
Bernd Wollscheid and Jeffrey W. Bode
1Laboratorium für Organische Chemie, Department of Chemistry and Applied Biosciences,
ETH-Zürich, 8093 Zürich, Switzerland
2 Department of Health Sciences and Technology & Institute of Molecular Systems Biology
& BioMedical Proteomics Platform (BMPP), ETH Zurich, Zurich, Switzerland
*e-mail: wbernd@ethz.ch
3 Institute of Transformative Bio-Molecules (WPI-ITbM), Nagoya University, Chikusa,
Nagoya 464-8602, Japan
*e-mail: bode@org.chem.ethz.ch
Abstract
In analogy to biosynthetic pathways leading to bioactive natural products, synthetic
fermentation generates mixtures of molecules from simple building blocks under aqueous,
biocompatible conditions, allowing for the resulting cultures to be directly screened for
biological activity. In this work, a novel -peptide antibiotic was successfully identified using
the synthetic fermentation platform. Phenotypic screening was carried out in an initially
random fashion, allowing for simple identification of active cultures. Subsequent
deconvolution, focused screening and structure-activity relationship studies led to the
identification of a potent antimicrobial peptide, showing strong selectivity for our model
system Bacillus subtilis over human HEK293 cells. To determine the antibacterial
mechanism of action, a peptide probe bearing a photoaffinity tag was readily synthesized
through the use of appropriate synthetic fermentation building blocks and utilized for target
identification using a quantitative mass spectrometry-based strategy. The chemoproteomic
approach led to the identification of a number of bacterial membrane proteins as prospective
targets. These findings were validated through binding affinity studies with penicillin-binding
Page 1 of 22
ACS Paragon Plus Environment

ACS Chemical Biology
Page 2 of 23
1
2
3
4
5
6
7
8
protein 4 using microscale thermophoresis, with the bioactive peptide showing a dissociation
constant (Kd) in the nanomolar range. Through these efforts, we provide a proof of concept
for the synthetic fermentation approach presented here as a new strategy for the phenotypic
discovery of novel bioactive compounds.
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
Introduction
Synthetic fermentation allows for the rapid, operationally simple preparation of libraries of
bioactive molecules by covalent bond forming reactions that occur simply upon mixing the
constituent building blocks in aqueous buffer. This process is based on chemoselective
couplings that do not require reagents or produce toxic byproducts, allowing the resulting
mixtures to be screened directly in a biological assay.1, 2 The active components can be
identified by straightforward deconvolution strategies and – in analogy to natural products
3
isolation
– the most potent compounds purified by chromatographic methods and
characterized. In contrast to most other library-based approaches, synthetic fermentation hits
can be rapidly resynthesized from the constituent building blocks without the need for
traditional multi-step organic synthesis. By adjustment of building block selection, hit
optimization and structure activity relationship studies can be initiated even before the active
2
component is conclusively identified. The potential of synthetic fermentation using the -
ketoacid–hydroxylamine (KAHA) amide formation as a platform for the identification of
bioactive molecules was first demonstrated with the discovery of a -peptide HCV NS3/4A
protease inhibitor with an IC50 of 1.0 M (Figure 1). 1
Previous work: Protease inhibitor discovery with synthetic fermentation and target-based screening
O
O
OMe
HN
O
O
HN
RM
O
H
H
H
H
H
OH
RI
N
HO2
C
N
N
N
N
O
RT
NS3/4A HCV protease inhibitor
N
RT
O
O
H
O
O
O
O
O
O
IC50 = 1.0 M
Cl
I
M
T
CO2tBu
This work: Antibiotic discovery with synthetic fermentation and phenotypic screening
O
T1
M2
M1
M4
I
H
H
H
T3
M1
M3
I
N
N
N
N
T2
M5
M2
M1
M4
I
N
Antibacterial peptide
MIC (Bacillus subtilis) = 5.7 g/mL
Kd (PBPD): 364 (±19.1) nM
M1
M3
I
M1
I
M5
H
tBuOH/H2
45 ºC
O
tBuOH/H2
45 ºC
O
O
O
O
O
M3
I
I
I
M2
I
Cl
Cl
Bioactive synthetic fermentation products
Initiation
Initiation
EElolonnggaatitoionn
Termination
Figure 1: Identification of bioactive compounds through screening with the synthetic fermentation
1
platform.
Although synthetic fermentation produces mixtures of compounds that must eventually be
deconvoluted, it has the advantage of being ideally suited for phenotypic screening. The
Page 2 of 22
ACS Paragon Plus Environment

Page 3 of 23
ACS Chemical Biology
1
2
3
4
5
leading display technologies, including DNA-encoded libraries 4, 5 and mRNA-display 6, 7, can
produce far larger numbers of molecules but the minuscule amounts employed and large
6
7
8
tags present preclude screening on living systems. As alternatives medium-throughput
8,
approaches including diversity-oriented synthesis
9, multicomponent reactions forming
9
11,
sequence-controlled peptides 10, sequence-defined cation oligomers
12, synthetic
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
macrocycles 13, 14 , and even entirely random peptide cocktails 15 have been employed for the
discovery of new pharmaceutical lead structures to great success. With synthetic
fermentation, the quantities of compounds produced, the approximate numbers of distinct
molecules, and their general sequences can be controlled by the initial setup. Upon
completion of the assembly, the aqueous mixtures are diluted and used directly for screening
in cell-based or other phenotypic assays. With fifty initiators, monomers and terminators, a
prospective fermentation library of up to 300 million chemically distinct products is
accessible. 1
As a first demonstration of synthetic fermentation for the phenotypic screening of readily
prepared compound libraries, we sought to identify new antibiotics simply by combining a
small set of monomeric building blocks with basic liquid handling techniques. In this report,
we document the successful identification of a peptide targeting penicillin-binding proteins
using synthetic fermentation for hit identification and optimization, target identification, and
initial binding studies. The entire process can be conducted in a safe and reproducible
manner without any technical knowledge or practical experience in organic chemistry. Using
just 50 building blocks, we successfully identified a -peptide -ketoamide that exhibits
excellent antibacterial activity against an indicator strain, prepared a derivative for a chemical
proteomics approach to identify the molecular target, and confirmed nanomolar binding to a
penicillin binding protein. This work therefore not only establishes a practical approach to
preparing and screening thousands of molecules on a live organism, but also uncovers a
new antibiotic and its protein target.
RESULTS AND DISCUSSION
Library Design: At the outset of our studies, we sought to employ our robust conditions for
16
the formation of -peptide oligomers with KAHA ligation
toward the preparation of short
sequences that inhibit the growth of a microbial target. Based on our success in identifying
1
protease inhibitors in our first report
and previous reports from Gellman and others
17,18,19
showcasing -peptides as antimicrobal agents,
we elected to maintain the basic
approach of combining -ketoacid initiators, isoxazolidine monomers, and terminators. The
Page 3 of 22
ACS Paragon Plus Environment

ACS Chemical Biology
Page 4 of 23
1
2
3
4
5
6
greatest diversity and increase in compound numbers arises from the elongation monomers,
and for this project we sought to complement our established library with new monomer
building blocks containing a variety of side chain structures.
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 2. Elongation monomer building blocks for synthetic fermentation. (a) Synthetic route to
isoxazolidine elongation monomers. (b) Monomer side-chains synthesized for phenotypic screening
against Bacillus subtilis.
Using our previously developed unified route to isoxazolidine elongation monomers (Figure
1, 20, 21
2a),
we prepared the compounds shown in Figure 2b by nitrone cycloadditions of
aldehydes with gulose-derived hydroxylamine 1. Following crystallization to afford
diastereomerically pure material and auxiliary removal with HClO4, the isoxazolidine
monomers were converted into hydrochloride salts for long-term storage and ease of use.
The resulting salts are bench-stable solids, and are typically stored at -20 °C; the vast
majority of these compounds remain stable for years under these conditions, without any
significant precautions required. This two step protocol is suitable for both aliphatic and
aromatic aldehydes; the enantiomeric monomers can be produced simply by using the
readily available (L)-gulose derived auxiliary. 22
Page 4 of 22
ACS Paragon Plus Environment

Page 5 of 23
ACS Chemical Biology
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 3: Synthetic fermentation of -peptide oligomers by KAHA ligation. (a) The coupling of the
initiator and elongation monomer unveils another -ketoacid, which can react with further monomer
units or a hydroxylamine terminator. (b) Reaction conditions and sample products generated in a
typical three-monomer synthetic fermentation used in this study. 1 equivalent of initiator and each
monomer was used, while terminators were added in slight excess (1.5 equiv) to ensure complete
1, 20
capping.
We began by preparing a modest number of compounds by using one initiator, one
terminator, and three elongation monomers per well, in a 96-well plate format (see Figure 3
for a representative example). Initiators were selected from the large range of commercially
available -ketoacids. Monomers incorporating natural and non-natural alkyl and aryl amino
acid side-chains were included, as were monomers bearing side-chain esters, amines, and
aromatic and saturated heterocycles. Terminators capable of generating -ketoamides – well
known pharmacophores for enzyme inhibitors 23, 24 – as well as -ketoesters, aldehydes and
carboxylic acid functionalities in the final products were utilized. We hypothesized that the
modular nature of the synthetic fermentation process would allow for facile modification and
optimization of any prospective leads, aiding in the identification of pharmacophore regions
and enabling rapid and efficient structure-activity relationship studies.
Page 5 of 22
ACS Paragon Plus Environment

ACS Chemical Biology
Page 6 of 23
1
2
3
4
5
6
7
8
9
Lead Identification: The first culture prepared in this study (Culture 1) consisted of a
randomized, broad-ranging screen of three initiators, 24 monomers and four terminators
(Figure 4a). The monomers were divided into three general categories – aliphatic, aromatic,
and hydrophilic – based on the chemical properties of their side chains; one from each group
was added to each well (please see the Supporting Information for specific plate design).
Each well therefore contained one initiator, three monomers and one terminator. If only
products of type I-Mn-T (where n = 0-3) are considered, mixtures of this composition
generate around 40 compounds per well. A 96-well plate prepared in this manner from less
than 50 milligrams of total starting material allows for screening of around 4000 compounds
for biological activity – the total time required for the synthetic fermentation and anti-bacterial
assay is less than 48 hours.
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 fermentation reactions were conducted by the addition of 5 L each of initiator and
monomer stock solutions (tBuOH/H2O, 5:1, 0.1 M) to the desired wells of a 96-well plate. The
plate was sealed and heated to 45 °C for two hours in a PCR thermal cycler. Termination
was achieved by the addition of 7.5 L of a 0.1 M terminator stock solution to each well and
heating the plate to 45 °C for a further 16 h. The gram-positive bacterium Bacillus subtilis
was chosen as our model system at the outset of our study due to its ease of handling, fast
25
growth and recognized utility as an indicator strain for antibiotic screening. To determine
antibacterial activity, aliquots from each well were transferred to an assay plate containing a
sample of Bacillus subtilis in bacterial growth medium, and the assay plates were incubated
at 37 °C for 24 hours. Bacterial growth inhibition was determined by optical density
26
measurements at 600 nm using a plate reader.
In this initial screen, wells which were
visibly clear and showed an optical density of < 0.2 were considered to be active. Positive
controls were provided by 2.5 μL and 5 μL of a 10 g/mL tetracycline solution, while a
t
negative control was provided by 2.5 μL BuOH/H2O (5:1). A blank well containing no
additional material was also measured for reference (see Supporting Information for full
assay setup).
Page 6 of 22
ACS Paragon Plus Environment

Page 7 of 23
ACS Chemical Biology
1
2
3
4
Culture 1
Monomers
Terminators
T2
Initiators
5
6
7
8
(a)
O
O
O
OMe
OMe
M1
M8
M2
M9
M3
M4
M11
M17
M6
M7
HN
HN
O
I1
OH
T1
O
O
O
M10
M12
M18
M13
Me
Me
HO
O
9
I2
OH
O
OMe
HN
O
OMe
M14 M15
M16
M23
M19
M28
H
HN
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
O
NH2
T3
T4
N
O
O
HO
OH
I3
M21
M22
M24
M26
Me
Me
O
O
I1
I2
I3
(b)
M1 M6
M4 T4
M1 M21
M13 T4
M1 M18
M23 T4
M1 M17
M14 T2
M1 M10
M15 T2
M1 M9
M16 T2
M1 M8
M1 M7
M28 T3
M1 M6
M4 T3
M1 M21
M13 T1
M1 M18
M23 T1
M1 M17
M14 T1
M24 T3
M2 M7
M13 T4
M2 M6
M23 T4
M2 M21
M14 T4
M2 M18
M15 T2
M2 M17
M16 T2
M2 M10
M24 T2
M2 M9
M28 T3
M2 M8
M4 T3
M2 M7
M2 M6
M23 T1
M2 M21
M14 T1
M2 M18
M15 T1
M13 T3
M3 M8
M23 T1
M3 M7
M14 T1
M3 M6
M15 T1
M3 M21
M16 T3
M3 M18
M24 T3
M3 M17
M28 T3
M3 M10
M4 T4
M3 M9
M13 T4
M3 M8
M23 T4
M3 M7
M14 T2
M3 M6
M15 T2
M3 M21
M16 T2
M11 M9
M14 T1
M11 M8
M15 T1
M11 M7
M16 T1
M11 M6
M24 T3
M11 M21
M28 T3
M11 M18
M4 T3
M11 M17
M13 T4
M11 M10
M23 T4
M11 M9
M14 T4
M11 M8
M15 T2
M11 M7
M16 T2
M11 M6
M24 T2
M12 M10 M12 M9
M15 T2 M16 T2
M12 M8
M24 T2
M12 M7
M28 T4
M12 M6
M4 T4
M12 M21 M12 M18
M12 M17
M14 T1
M12 M10
M15 T1
M12 M9
M16 T3
M12 M8
M24 T3
M12 M7
M28 T3
M13 T4
M23 T1
M26 M17 M26 M10
M16 T2 M24 T2
M26 M9
M28 T2
M26 M8
M4 T4
M26 M7
M13 T4
M26 M6
M23 T4
M26 M21 M26 M18
M26 M17
M16 T1
M26 M10
M24 T3
M26 M9
M28 T3
M26 M8
M4 T3
M14 T1
M15 T1
M19 M18 M19 M17
M24 T3 M28 T3
M19 M10
M4 T3
M19 M9
M13 T1
M19 M8
M23 T1
M19 M7
M14 T1
M19 M6
M15 T2
M19 M21 M19 M18
M19 M17
M28 T4
M19 M10
M4 T4
M19 M9
M13 T4
M16 T2
M24 T2
M22 M21 M22 M18
M28 T3 M4 T3
M22 M17 M22 M10
M13 T3 M23 T1
M22 M9
M14 T1
M22 M8
M15 T1
M22 M7
M16 T2
M22 M6
M24 T2
M22 M21 M22 M18
M22 M17
M13 T4
M22 M10
M23 T4
M28 T2
M4 T4
(c)
Dilution
to required
concentration
Add initiators
Add monomers
Add terminators
(7.5 L / unit)
(5 L / unit)
Termination
45 °C, 16 h
Building block stock solutions
Assay directly
Elongation
45 °C, 2 h
Fermentation library
(tBuOH/H2O (5:1), 0.1 M)
no purification required
Page 7 of 22
ACS Paragon Plus Environment

ACS Chemical Biology
Page 8 of 23
1
2
3
4
Culture 2
I1
5
6
Terminators
(d)
(e)
Initiators
7
8
9
O
OMe
HN
H
O
N
M4
M4 M4
M8 T26
T4
I1
OH
N
O
T26
O
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
M4 M4
T26
M4 M32
T26
Monomers
M8
M4 M32
M32 T26
O
O
O
O
HN
HN
T26
O
O
M4
M8
O
HN
M26
O
Cl
O
O
O
M8 M8
T26
M4 M4
O
N
Cl
M32
T26
M32
M8 M32
T26
T26
(f)
O
M32 M32
T26
M8 M32
OH
M32
O
T26
O
H
H
i) tBuOH/H2O (5:1), 45 °C, 2 h
5 mg
N
N
N
N
H
I1
O
O
O
M4
M8
M8
M8
OMe
O
ii)
HN
H
N
T26
2
M32 T26
Cl
Cl
O
O
HN
N
O
15 mg
3 mg (16 % yield)
O
I1 M8 T4
Cl
T4
tBuOH/H2O (5:1), 45 °C, 16 h
M4 M8
M8 T26
M8
M4
O
11 mg
M32 T26
MIC (Bacillus subtilis): 23-35 µg/mL
IC50 (Hek293): 16-32 µg/mL
M8
Cl
Figure 4. Design of Culture 1 and Culture 2 and results of the coupled antibacterial assays. (a)
Synthetic fermentation building blocks utilized in Culture 1 (See Figure 2b for monomer side-chains).
Three initiators, 24 monomers and four terminators were used. (b) Well design of Culture 1. The
building blocks were distributed throughout the plate; one initiator, three monomers, and one
terminator were used per well. Two active wells (OD600 < 0.2) were identified (highlighted in yellow). (c)
Typical synthetic fermentation workflow. Building block stock solutions were transferred to a 96-well
plate using standard liquid handling procedures, and the reaction mixtures were incubated at 45 °C to
induce elongation. Terminators were added using the same procedure. The synthetic fermentation
cultures can be directly assayed after dilution to appropriate concentration. (d) Well design of Culture
2. Building blocks found in biologically active well F4 from Culture 1 (Figure 4b) were identified and
distributed in every logical permutation. The effects of varying stoichiometries were also tested. Each
monomer listed corresponds to one equivalent of building block. Wells showing growth inhibition
(OD600 < 0.35) are highlighted in yellow. (e) Synthetic fermentation building blocks utilized in Culture 2.
One initiator, three monomers and one terminator were used. (f) Milligram scale production and
biological properties of lead molecule 2.
From this initial screen of 96 wells, two cultures showing significant growth inhibition on
Bacillus subtilis were identified: one containing I1, M7, M12, M28 and T4 and one containing I1,
,
M4 M8, M26, and T4 (Figure 4 a, b, c). A second round of screening was carried out on the
,
constituents of these wells to determine the active components. These deconvolution
cultures were designed so that each new well would contain a fraction of the possible
products from the active mother well (Figure 4d, e). This was achieved by oligomerization
using each monomer individually, as well as all possible two-monomers-per-well
Page 8 of 22
ACS Paragon Plus Environment

Page 9 of 23
ACS Chemical Biology
1
2
3
4
5
6
permutations. Monomer stoichiometry was also varied to probe the effect of peptide length
on biological activity – for example, when three equivalents of elongation monomers are
used, products containing two to four monomer units will be most prominently observed.
7
8
9
Wells showing bacterial growth inhibition in this focused library were analyzed by LC/MS,
revealing that the two most active mixtures prominently featured peptide I1-M8-T4 (2). As
peptide 2 was also observed in the active mother well in Culture 1, we postulated that this
was the active compound. Synthesis of 2 on milligram scale was achieved simply by mixing
phenylpyruvic acid (I1) and M8 in tBuOH / H2O (5:1) and heating to 45 °C for 2 h, followed by
addition of T4 and allowing the reaction to stand at 45 °C for 16 h; 3 mg of the desired
peptide was successfully isolated after purification by HPLC (16 % yield, Figure 4f). The
ability to rapidly prepare adequate quantities of the hit compound simply by mixing the
components is a key feature of synthetic fermentation. Quantitative testing of 2 for bacterial
growth inhibition against Bacillus subtilis using optical density measurement revealed a
minimum inhibitory concentration (MIC) of 23–35 μg/mL. However, cytotoxicity tests carried
out against human HEK293 cells showed an IC50 of 16–32 μg/mL, indicating a lack of
selectivity for bacterial cells. Nonetheless, 2 was considered as a promising lead compound
for further investigation.
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
Using hit compound 2 as a starting point, we carried out a broader screening of monomer
and terminator building blocks. A terminator screen (Culture 3) was prepared in which each
well contained I1, M8, and one of 16 different terminators, including several that had not been
previously employed (Figure 5a, c). The assay results obtained from this screen showed that
while some of these new terminator types displayed some inhibitory activity, the well
containing T4 remained the most active.
Page 9 of 22
ACS Paragon Plus Environment

ACS Chemical Biology
Page 10 of 23
1
2
3
4
5
Cultures 3 and 4
I1
M8
I1
T4
(a)
(b)
(c)
6
7
8
9
Terminators
T7
T5
T6
T8
T9
M12
M22
T1
M1
O
HN
OMe
OMe
O
O
HN
No T
O
OMe
OMe
HN
HN
O
O
Cl
O
O
T2 T10
T3 T11
T4 T12
T5 T13
T6 T14
T7 T15
T8 T16
M2 M13 M23
M3 M14 M24
M6 M15 M25
M7 M17 M27
M8 M18 M29
M9 M20 M30
M10 M21 M31
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
Cl
T9
T10
T11
T12
O
OMe
NH2
O
OMe
N
HN
HN
O
HN
O
OMe
OEt
HN
H
O
Cl
N
O
O
Me
Me
O
N3
MeO
T13
T14
Me
T15
T16
O
OMe
O
HN
O
OMe
HN
H
HN
H
N
OEt
H
N
MeO2C
OBz
N
NH2
O
O
Me
Me
Me
Cl
(d)
(e)
O
H
H
O
N
N
N
3 (I1-M29-T4)
O
i) tBuOH/H2O (5:1), 45 °C, 2 h
O
HN
N
H
OH
O
O
O
O
O
OMe
O
ii)
HN
O
H
N
I1-M29-M29-T4
3
I1
M29
Cl
I1 M29 T4
N
O
5 mg
11 mg
15 mg
3 mg (14 % yield)
T4
Cl
tBuOH/H2O (5:1), 45 °C, 16 h
MIC (Bacillus subtilis): 53 µg/mL
IC50 (Hek293): 64 µg/mL
Figure 5. Design of Culture 3 and Culture 4 and results of the antibacterial assays. (a) Well design of
Culture 3. I1 and M8 were kept constant and tested alongside 16 different terminators. The most active
well is highlighted in yellow. (b) Well design of Culture 4. I1 and T4 were kept constant and tested
alongside 24 monomers. The most active well is highlighted in yellow. (c) New terminator structures
utilized in Culture 3. (d) Milligram scale production and biological properties of lead molecule 3. (e)
Preparative HPLC trace from the synthetic fermentation of 3.
A conceptually analogous monomer screen (Culture 4, Figure 5a, b) was also carried out, in
which 24 monomer types were tested in the presence of I1 and T4. The previously untested
monomer M29 stood out as a promising candidate, and I1-M29-T4 (3) was quickly synthesized
by the same method described for 2 (Figure 5d, e). The preparative HPLC trace shown in
Figure 5e highlights the clean nature of these scale-up processes. Desired I-M-T peptide 3
was easily isolated as the major peak; the only other significant product observed was the
analogous I-M-M-T peptide incorporating two monomer units. Encouragingly, peptide 3
displayed antibacterial properties (MIC = 53 μg/mL) as well as lowered cytotoxicity against
HEK293 cells (IC50 = 64 μg/mL). However, this peptide showed reduced inhibition of Bacillus
subtilis compared to lead compound 2 (Figure 5d).
Structure-Activity Relationship Studies: With these preliminary results in hand, we
embarked upon structure-activity relationship (SAR) studies. These efforts had two goals: to
Page 10 of 22
ACS Paragon Plus Environment

Page 11 of 23
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
identify a potent antibiotic peptide with improved selectivity for Bacillus subtilis over human
HEK293 cells, and to elucidate the relationship between the structure of lead compound 2
and its biological function. The compounds investigated were synthesized by scaled up
synthetic fermentation procedures and isolated in milligram quantities after purification by
HPLC.
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
Lead Compounds
Peptides from M8
(a)
O
O
O
O
H
N
H
N
H
N
H
N
H
N
H
N
H
N
H
N
H
N
H
N
H
N
N
N
N
N
N
H
N
H
N
H
N
H
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
2
3
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
MIC (Bacillus subtilis): > 28 µg/mL
MIC (Bacillus subtilis): > 35 µg/mL
MIC (Bacillus subtilis): 23-35 µg/mL
IC50 (Hek293): 16-32 µg/mL
MIC (Bacillus subtilis): 53 µg/mL
IC50 (Hek293): 64 µg/mL
Peptides from M29
Initiator Modifications
O
O
H
N
H
N
H
N
H
N
H
N
H
N
MeO
N
N
N
H
N
H
O
O
O
O
O
O
O
O
O
O
H
N
H
N
H
N
H
N
N
N
4
HO2C
N
H
N
H
O
O
O
O
O
O
Cl
Cl
Cl
Cl
HO
Cl
Cl
Cl
Cl
MIC (Bacillus subtilis): 5.7 µg/mL
IC50 (Hek293): > 100 µg/mL
MIC (Bacillus subtilis): 7.3 µg/mL
IC50 (Hek293): > 100 µg/mL
MIC (Bacillus subtilis): > 100 µg/mL
MIC (Bacillus subtilis): > 100 µg/mL
Me
Me
O
H
N
H
N
O
H
N
H
N
H
N
H
N
N
N
H
N
N
H
Terminator Modifications
O
O
O
O
O
O
O
O
O
H
N
H
N
O
H
N
H
N
N
Cl
N
H
N
Cl
Cl
Cl
N
H
O
O
O
O
O
O
MIC (Bacillus subtilis): 18.1 µg/mL
MIC (Bacillus subtilis): > 100 µg/mL
IC50 (Hek293): 75 µg/mL
Cl
Cl
Me
Cl
Cl
MIC (Bacillus subtilis): > 100 µg/mL
IC50 (Hek293): 100 µg/mL
MIC (Bacillus subtilis): 52 µg/mL
IC50 (Hek293): 75 µg/mL
Peptides from M8 and M36
O
O
H
H
N
H
N
H
N
H
N
H
N
O
O
H
N
H
N
H
N
N
N
H
N
H
N
N
H
N
H
N
N
N
N
H
N
H
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
O
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
Cl
MIC (Bacillus subtilis): 14.3 µg/mL
IC50 (Hek293): > 100 µg/mL
MIC (Bacillus subtilis): > 35 µg/mL
MIC (Bacillus subtilis): 27-53 µg/mL
MIC (Bacillus subtilis): > 100 µg/mL
(b)
O
H
N
H
N
H
N
O
4 (I1-M29-M29-T4
)
i) tBuOH/H2O (5:1), 45 °C, 2 h
N
N
O
O
HN
H
OH
I1-(M29
)
3-T4
O
O
O
O
O
OMe
O
ii)
O
HN
4
O
M29
I1-M29-T4
H
N
I1
Cl
M29 M29
Cl
T4
N
O
8 mg
36 mg
25 mg
5 mg (12 % yield)
T4
I1
Cl
tBuOH/H2O (5:1), 45 °C, 16 h
Figure 6. Chemical structure and biological properties of peptides synthesized in structure-activity
relationship studies. (a) A tight SAR profile was observed when modulating building blocks and
peptide length. The most active peptide identified (4) is highlighted in grey. (b) Synthesis and isolation
of active peptide 4. Preparative HPLC trace and mass spectrum of the purified compound shown.
We initiated a systematic study of lead compound 2 through the sequential modification of
each region of the peptide (Figure 6a). Replacement of the benzyl subunit of the initiator with
either its phenol analogue or an alkyl carboxylic acid was not found to improve inhibition of
Bacillus subtilis growth. A tight structure-activity relationship was also seen in the terminator
region of the molecule, with the biological activity proving sensitive to even minor modulation
of the tertiary amine. Substitution of the pyrrolidine tail with other saturated N-heterocycles
significantly impaired the biological activity, as did replacement of the benzyl side-chain with
phenyl or n-propyl groups. We therefore concluded that the use of terminator T4 was
Page 11 of 22
ACS Paragon Plus Environment

ACS Chemical Biology
Page 12 of 23
1
2
3
4
5
6
7
8
important for the activity of the lead peptide. Interestingly, these observations echo the
findings of Hergenrother, who proposed that amphiphilic small molecules bearing basic
amines are best suited to penetrate bacterial cell membranes and serve as broad-spectrum
antibiotics. 27
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
Further SAR studies were carried out investigating the effect of peptide elongation on
biological activity (Figure 6). For this study, peptides of type I-M-M-T and I-M-M-M-T were
synthesized. Our initial studies in this area focused on the impact of peptide elongation
through the incorporation of additional M8 subunits. While these compounds retained some
biological activity, no significant improvement in comparison to lead compound 2 was
observed. Exploration of longer peptides containing M29 units proved more effective. Peptides
I1-M29-M29-T4 and I1-M29-M29-M29-T4 exhibited a clear increase in inhibition compared to the
lead compound, as did mixed monomer compound I1-M8-M29-T4 Moreover, peptides of this
.
type showed significantly lowered cytotoxicity against the human HEK293 cell line. The most
potent of these, I1-M29- M29-T4 (4), showed an almost tenfold increase in bacterial growth
inhibition (MIC = 5.7 μg/mL) while displaying no measurable toxicity against HEK293 cells
(IC50 > 100 μg/mL, see Supporting Information for details) (Figure 6b).
These studies identified a tight SAR profile in many areas of the molecule, supporting our
confidence in synthetic fermentation as an effective medium-throughput screening technique.
Our goals for the investigation were further met by the discovery of 4, a potent inhibitor of
Bacillus subtilis growth bearing no apparent toxicity towards our human HEK293 cell model
system.
Target Identification: Although we initially anticipated that the antimicrobial activity might be
due to simple membrane disruption, the tight structure-activity relationships observed in 4
implied a discrete molecular target in bacteria. To investigate the mechanism of action, we
prepared a variant 5 suitable for photoaffinity labeling of receptor proteins and their
subsequent identification by a quantitative mass spectrometry (MS)-based strategy (Figure
28, 29, 30
7a).
For the synthetic fermentation of probe 5, an initiator (I4) was employed
containing both a diazirine as a photolabile group and an alkyne for the enrichment of target
proteins (Figure 7b), where the small size of these groups should ensure minimal
31
interference with target binding.
The ability to quickly prepare a molecule for target
identification is another advantage of this approach.
Page 12 of 22
ACS Paragon Plus Environment

Page 13 of 23
ACS Chemical Biology
1
2
3
(a)
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
(b)
N
N
O
BF3K
N
N
H
H
H
O
O
i) tBuOH/H2O (5:1), 45 °C, 2 h
N
N
N
N
N
O
O
H
+
I4
O
O
O
O
OMe
O
ii)
HN
H
N
O
5
HN
Cl
Cl
N
O
O
T4
O
I4 M29 M29 T4
10 mg (20 % yield)
M29
tBuOH/H2O (5:1), 45 °C, 16 h
Cl
(c)
(d)
Figure 7: Target identification for antibacterial oligopeptide 4 using a photoaffinity labeling approach.
(a) Workflow followed for target identification studies. Photoaffinity probe 5 was incubated with
bacterial suspensions, with and without the presence of a competition control. After light-induced
covalent bond formation between the probe and its target proteins, the bacterial cells were lysed and
the captured proteins were enriched by click chemistry to agarose beads and subjected to on-bead
tryptic digestion. The resulting purified peptides were identified by mass spectrometry. (b) Synthetic
fermentation of peptide 5 bearing a photoaffinity probe. (c) List of identified protein target candidates
and their cellular functions. (d) Enrichment of protein target candidates relative to a competition control
(n=3, FDR=0.01, log intensity difference ≥0.5).
We incubated probe conjugate 5 with suspensions of Bacillus subtilis in PBS for 20 minutes
to allow target binding, followed by irradiation at 350 nm for 20 minutes on ice. The reactive
carbene generated from the diazirine upon UV irradiation mediates covalent linkages with
31
adjacent biomolecules.
In a control sample, the bacterial suspensions underwent
simultaneous treatment with a five-fold excess of unconjugated peptide 4 to compete with
binding of diazirine probe 5 at specific binding partners. After cell lysis, proteins trapped by
the probe were enriched by click reaction of the alkyne with azide-bearing agarose beads.
Page 13 of 22
ACS Paragon Plus Environment

ACS Chemical Biology
Page 14 of 23
1
2
3
4
5
Following on-bead tryptic digestion and purification of resulting peptides, proteins were
identified and quantified relative to the competition control by MS-analysis.
6
7
8
9
Using stringent cut-offs for statistical significance and enrichment over the control, we
identified four high confidence protein candidates for binding partners, all of which are
located in the plasma membrane of the bacterial cell (Figure 7c, d). Among these candidates,
penicillin binding protein 4 (pbp4) seemed a promising molecular target to mediate the
antibiotic activity of 4, as it is involved in the essential biosynthesis of the bacterial cell wall,
and a known target of -lactam antibiotics such as penicillin. 32
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
-
SO3
N
N
N
N
-O3S
O
H
H
H
O
N
N
N
N
NH
N
H
O
O
O
O
6
N+
Cl
Cl
-O3S
Peptide-dye conjugate 6
Kd = 364 (±19.1) nM
Figure 8: Dose-response curve and dissociation constant obtained from binding studies between
peptide conjugate 6 and penicillin-binding protein 4 in three independent measurements using
microscale thermophoresis.
To validate these findings, we sought to determine the binding affinity of 4 with recombinant,
purified penicillin-binding protein 4. Initial binding studies using surface-based methods such
as surface plasmon resonance 33 and bio-layer interferometry 34 proved challenging due to a
high level of non-specific binding between the hydrophobic membrane protein and the
surface. However, solution-based microscale thermophoresis using peptide-dye conjugate 6
proved to be suitable. 35 A dose response curve for 6 was obtained, allowing determination of
a dissociation constant (Kd) of 364  19.1 nM, while a negative control peptide also bearing a
cyanine-3 dye unit showed no binding (Figure 8, see Supporting Information for details). This
value compares favorably with members of the 4-quinolone family that have been shown to
36
bind PBPs but show no antibacterial activity against Bacillus subtilis.
Notably, penicillin
follows a two-step serine acylation mechanism with penicillin-binding proteins and has a
surprisingly low dissociation constant prior to the formation of the covalent bond,
independently delineated as 0.9 mM by Lu 37 and 20 mM by Chesnel 38 for PBP2x.
Conclusions
Page 14 of 22
ACS Paragon Plus Environment

Page 15 of 23
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
The discovery of antibacterial peptide 4 provides a proof of concept of synthetic fermentation
as a platform for rapid and effective phenotypic screening. The robust synthesis of monomer
and terminator precursors has allowed the accumulation of a diverse building block library
capable of the efficient generation and screening of large numbers of compounds. Upon the
discovery of a hit, the modular nature of synthetic fermentation enables facile preparation
and modification of promising compounds, allowing for rapid investigation of pharmacophore
regions and structure-activity relationships. Probe materials – including fluorescent variants
and photoaffinity trapping reagents – for target identification studies are readily generated
through the use of bifunctional building blocks, faciliting the often tedious process of
preparing such derivatives. In this manner, we discovered that the peptide identified by rapid
molecular assembly and phenotypic screnning is a nanomolar binder of penicillin-binding
protein 4, a known target of beta-lactam antibiotics.
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 appeal of synthetic fermentation is the ability to construct relatively complex, non-natural
molecules simply upon mixing the building blocks in aqueous media. The resulting mixtures
can be directly screened for biological activity in both biochemical and phenotypic assays.
Although small mixtures are produced – requiring some deconvolution steps and HPLC
isolation of the active compounds – the ease of preparing new, focused libraries and
synthesizing derivatives suitable for target identification simplifies the overall approach to
drug discovery. Using this technique, we envision that simple kits of the constituent building
blocks will allow researchers with little experience with synthetic organic chemistry to prepare
and screen new molecules for specific biological activities.
METHODS
Synthetic Fermentation: The syntheses of cultures were carried out in 96-well plates. 0.1
t
M stock solutions of initiators and monomers in BuOH/H2O (5:1) were prepared, and the
required volumes were added to their respective wells using a micropipette. After addition of
initiators and monomers, the 96-well plate was capped, shaken and placed in a PCR thermal
cycler. The PCR machine (equipped with a heated lid to prevent solvent condensation on the
cap) was then heated to 45°C for 6 h. The plate was then removed and allowed to cool. 0.1
t
M stock solutions of terminators in BuOH/H2O (5:1) were then added to each well using a
micropipette. The plate was capped, stirred and heated at 45°C for 16 h.
Bacillus subtilis preparation: Bacillus subtilis bacteria were stored at –80°C in a Luria-
Broth (LB) medium containing 15% glycerol. A trace amount of bacteria was placed into a
Page 15 of 22
ACS Paragon Plus Environment

ACS Chemical Biology
Page 16 of 23
1
2
3
4
5
6
7
8
9
Falcon tube using a toothpick and diluted with 10 mL LB media solution. LB media solution
was prepared from 25 g LB media in 1 L MQ-H2O. The bacterial solution was incubated and
allowed to shake at 37 °C overnight, generating a bacterial preculture. 100 L of this
preculture was transferred to a Falcon tube, diluted with 10 mL LB media, and incubated and
shaken at 37 °C. Bacterial growth was monitored by spectrophotometric optical density
determination. Once optical density at 600 nm had reached 0.6, the bacterial solution was
stored at 0°C and used in the assay within 1 h. All steps were carried out under sterile
conditions using pre-sterilized equipment.
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
Growth inhibition assay: 66 μL of bacterial solution was transferred into a pre-sterilized
Falcon tube, and diluted with 10 mL LB media solution to generate a bacterial stock solution.
2.5 μL of crude mixture from each synthetic fermentation well was added to the respective
wells of an assay plate, and 150 μL of the bacterial stock solution was added to each well.
The assay plate was allowed to shake gently at 37 °C for 24 h. Positive controls were
provided by 2.5 μL and 5 μL of a 10 g/mL tetracycline solution. Negative control was
t
provided by 2.5 μL BuOH/H2O (5:1). Assay results were measured by inhibition of bacterial
growth using a plate reader, based on spectrophotometric optical density determination of
each well at 600 nm after 24 h incubation. This assay was always performed in triplicate, with
the final results given by the average value of the three measurements.
Minimum inhibitory concentration determination: Desired peptide was synthesized on a
milligram scale and purified by reverse phase HPLC. The peptide was dissolved in DMSO to
generate a 25 mM solution. 5 μL of this solution was added to 620 μL LB media, generating
a 200 μM stock solution. Different volumes of this solution were added to 50 μL Bacillus
subtilis solution in LB media, generating a gradient of final concentrations (e.g. 100, 75, 50,
37.5, 25, 10 and 5 μM) for measurement. The assay plate was incubated at 37 °C for 24
hours and analyzed as described above. Assay results were converted into g/mL units for
final publication.
39 40
Cytotoxicity determination:
Cytotoxicity tests were carried out on HEK293 cells using
the commercial product Cell Counting Kitmg8 (CCK-8), and experiments were performed
according to the manual provided. 100 μL of cell suspension (5000 cells/well) was dispensed
in a 96-well plate, which was incubated for 24 h at 37 °C. The media was then removed. 100
μL of seven different concentrations (200, 150, 100, 75, 50, 25, 10 μM) of I1-M36-M36-T26 in
DMSO were added to the plate, which was then incubated for a further 24 h at 37 °C. 10 μL
of CCK-8 solution was added to each well of the plate, which was incubated at 37 °C for a
Page 16 of 22
ACS Paragon Plus Environment

Page 17 of 23
ACS Chemical Biology
1
2
3
4
5
final 2 h. The absorbance was measured directly. Assay results were converted into g/mL
units for final publication.
6
7
8
9
Target identification using quantitative mass spectrometry: Cultures of Bacillus subtilis
(in log phase) were harvested by centrifugation (6000 xg for 10 min at room temperature)
and washed in phosphate buffered saline (PBS, Gibco). 200 mg bacterial pellet per sample
was resuspended in 10 mL PBS in thin-walled glass tubes and treated with 100 μg of the
peptide-diazirine-alkyne construct (0.07% DMSO end concentration) for 20 min at room
temperature. In a competition control, cells were simultaneously treated with 100 μg of the
peptide-diazirine-alkyne conjugate and 500 μg of the unconjugated peptide. Experiments
were conducted in triplicates per condition (n=3). Samples were cooled on ice and irradiated
for 20 min at 350 nm using a Rayonet RPR-200 photoreactor with Hitachi FL8BL-B light
tubes. Cells were transferred to 15 ml Falcon tubes, collected by centrifugation and
resuspended in 2 ml PBS supplemented with protease inhibitors (cOmplete ULTRA Tablets,
EDTA-free, Roche) on ice. Cells were lysed by three rounds in a French pressure cell press
(SLM Aminco). For protein solubilisation, samples were sonicated in a VialTweeter
(Hielscher, 30 s at 80% amplitude and 80% cycle time) with 0.1% RapiGest surfactant
(Waters) at 4 °C, and cell debris were removed by centrifugation (16.000 xg for 10 min at 4
°C).
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 enrichment of alkyne-labeled target proteins, lysates were incubated with 100 μL azide
beads (Jena Biosciences), 1 mM CuSO4, 5 mM THPTA, and 100 mM sodium ascorbate,
slowly rotating overnight at room temperature. Beads were washed in Biospin columns
(BioRad) connected to a Vac-Man Laboratory Vacuum Manifold (Promega) with 5 x 1 mL 1%
sodium dodecyl sulfate (SDS) in 100 mM Tris-base, pH 8, 5 x 1 mL 80% isopropanol, 5 x 1
mL 8M urea in 100 mM ammonium bicarbonate with 5 mM EDTA, 5 x 1 mL 5 M sodium
chloride and 5 x 1 mL 50 mM ammonium bicarbonate. Bead-bound proteins were reduced
for 30 min with 5 mM Tris(2-carboxyethyl)phosphine (TCEP, Pierce) in 50 mM ammonium
bicarbonate, 0.1% RapiGest, alkylated for 30 min with 5 mM iodoacetamide in 50 mM
ammonium bicarbonate in the dark, and digested overnight, rotating at 37 °C, with 1 μg
trypsin (Sequence grade modified trypsin, Promega) in 500 μL 50 mM ammonium
bicarbonate, 0.1% RapiGest. Supernatants were acidified to pH 2 by the addition of formic
acid, subjected to C18 purification using 3–30 μg UltraMicroSpin Columns (The Nest Group)
according to the manufacturer's instructions and dried in a speed-vac (Thermo Scientific).
Page 17 of 22
ACS Paragon Plus Environment

ACS Chemical Biology
Page 18 of 23
1
2
3
4
5
6
7
8
9
Peptides resuspended in 5% MeCN in H2O with 0.1% formic acid (FA) were separated by
reversed-phase chromatography on a high-pressure liquid chromatography (HPLC) column
(75-μm inner diameter, New Objective) that was packed in-house with a 15-cm stationary
phase (ReproSil-Pur C18-AQ, 1.9 μm) and connected to a nano-flow HPLC combined with
an autosampler (EASY-nLC II, Proxeon). The HPLC was coupled to an Orbitrap Fusion mass
spectrometer (Thermo Scientific) equipped with a nano-electrospray ion source (Thermo
Scientific). Peptides were loaded onto the column with 100% buffer A (99.9% H2O, 0.1% FA)
and eluted at a constant flow rate of 300 nl/min with a 50 min linear gradient from 5–35%
buffer B (99.9% MeCN, 0.1% FA), 5 min 35-50% and 5 min 50-90% buffer B, with a
subsequent washing step with 90% buffer B. In between batches of runs, the column was
cleaned with two steep consecutive gradients of MeCN (10-98%). Mass spectra were
acquired in a data-dependent manner (TopSpeed, charge state 2-6, dynamic exclusion
duration 30 s). High-resolution MS1-spectra were acquired at 120,000 resolution (automatic
gain control target value 2.0*10e5) to monitor peptide ions in the mass range of 395–1,500
m/z, followed by HCD MS/MS-scans (Quadrupole isolation, HCD 30%, Detector IonTrap,
Scan rate turbo, automatic gain control target value 1.0*10e2, Maximum injection time 250
ms, centroid).
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
RAW data were analysed using the MaxQuant software (version 1.6.0.16). Spectra were
searched against a Uniprot-based database of Bacillus subtilis proteins, and label-free
quantitation was performed using standard settings. Statistical analysis and data
visualization was done using the Perseus software (version 1.6.0.7). 42 Data were filtered for
contaminants, decoys (reverse sequences) and proteins that were only identified by a
modification site, and statistical significance was determined by a two-sample t-test (with at
least two valid values in at least one group). Proteins were considered candidates if they
showed statistical significance with a stringent false discovery rate cut-off of 0.01, and a
difference between log intensity values of at least 0.5.
Determination of protein binding affinity with microscale thermophoresis: A trace
amount of peptide-dye conjugate 6 or control conjugate 7 was dissolved in NanoTemper
MST buffer (200 L) and diluted fivefold until no visible colour was observed. A 11.5 M
solution of recombinant Bacillus subtilus penicillin binding protein 4 (Biomatik) in
NanoTemper MST buffer (20 L) was prepared, and used as the source for a 1:1 dilution
series of sixteen samples. To each resulting 10 L sample was added 10 L of peptide-dye
conjugate solution. Capillaries were used to take up a small amount of each solution, and
these were ordered sequentially by concentration and placed inside a NanoTemper Monolith
Page 18 of 22
ACS Paragon Plus Environment

Page 19 of 23
ACS Chemical Biology
1
2
3
4
5
6
7
8
9
NT.115. A cap scan was performed to determine optimal LED power, and the samples were
subjected to binding affinity testing by thermophoresis + T-Jump experiments. Experiments
were performed in triplicate, and Kd calculations were performed by NanoTemper NTAnalysis
software based on the resulting calibration curves. Error bars = standard deviation from three
independent measurements.
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
Experimental procedures and spectroscopic data for all new compounds. This material is
available free of charge via the Internet at http://pubs.acs.org.
Author Information
Corresponding Authors
*E-mail: bode@org.chem.ethz.ch
*E-mail: wbernd@ethz.ch
Notes
The authors declare no competing financial interest.
Acknowledgements. This work was supported by the ETH Research Commission (ETH 06
15-2) and Swiss National Science Foundation Agora Funding (CRAGP2_171667) to J.B. and
from the Swiss National Science Foundation SNSF #31003A_160259 to B.W.. We would like
to thank J. Hubert and Y. Stöferle for contributions to monomer synthesis, J. Sobek and S.
Oishi for assistance with the target validation studies, V. Pattabiraman for sourcing starting
materials, and R. Hofmann for helpful discussions.
Page 19 of 22
ACS Paragon Plus Environment

ACS Chemical Biology
Page 20 of 23
1
2
3
4
References
5
6
7
8
(1)
(2)
(3)
(4)
(5)
(6)
(7)
(8)
Huang, Y.-L., and Bode, J. W. (2014) Synthetic fermentation of bioactive non-ribosomal
peptides without organisms, enzymes or reagents Nat Chem. 6, 877–884.
Stepek, I. A., and Bode, J. W. (2018) Synthetic fermentation of bioactive molecules Curr. Opin.
Chem. Biol. 46, 18–24.
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
Sattely, E. S., Fischbach, M. A., and Walsh, C. T. (2008) Total biosynthesis: in vitro
reconstitution of polyketide and nonribosomal peptide pathways. Nat. Prod. Rep. 25, 757–793.
Zimmermann, G., and Neri, D. (2016) DNA-encoded chemical libraries: foundations and
applications in lead discovery Drug Discov. Today 21, 1828–1834.
Goodnow Jr., R. A., Dumelin, C. E., and Keefe, A. D. (2017) DNA-encoded chemistry: enabling
the deeper sampling of chemical space Nat. Rev. Drug Discov. 16, 131–147.
Obexer, R., Walport, L. J., and Suga, H. (2017) Exploring sequence space: harnessing
chemical and biological diversity towards new peptide leads Curr. Opin. Chem. Biol. 38, 52–61.
Roberts, R. W., and Szostak, J. W. (1997) RNA–peptide fusions for the in vitro selection of
peptides and proteins. Proc. Natl Acad. Sci. USA 94, 12297–12302.
Cordier, C., Morton, D., Murrison, S., Nelson, A., and O’Leary-Steele, C. (2008) Natural
products as an inspiration in the diversity-oriented synthesis of bioactive compound libraries
Nat. Prod. Rep. 25, 719–737.
(9)
Chow, S., Liver, S., and Nelson, A. (2018) Streamlining bioactive molecular discovery through
integration and automation Nature Rev. Chem. 2, 174–183.
(10) Wang, S., Tao, Y., Wang, J., Tao, Y., and Wang, X. (2019) A versatile strategy for the synthesis
of sequence- defined peptoids with side-chain and backbone diversity via amino acid building
blocks Chem. Sci. 10, 1531–1538.
(11) Porel, M., Thornlow, D. N., Artim, C. M., and Alabi, C. A. (2017) Sequence-Defined Backbone
Modifications Regulate Antibacterial Activity of OligoTEAs ACS Chem. Biol. 12, 715–723.
(12) Artim, C. M., Phan, N. N., and Alabi, C. A. (2018) Effect of Composition on Antibacterial Activity
of Sequence-Defined Cationic Oligothioetheramides ACS Infect. Dis. 4, 1257–1263.
(13) Porel, M., Thornlow, D. N., Phan, N. N., and Alabi, C. A. (2016) Sequence-defined bioactive
macrocycles via an acid-catalysed cascade reaction Nature Chem. 8, 590–596.
(14) Dow, M., Marchetti, F., Abrahams, K. A., Vaz, L., Besra, G. S., Warriner, S., and Nelson, A.
(2017). Modular Synthesis of Diverse Natural Product-Like Macrocycles: Discovery of Hits with
Antimycobacterial Activity. Chemistry 23, 7207–7211.
(15) Amsoa, Z., and Hayouka, Z. (2019) Antimicrobial random peptide cocktails: a new approach to fight
pathogenic bacteria Chem. Commun., Advance Article (DOI: 10.1039/C8CC09961H)
(16) Bode, J. W. (2017) Chemical Protein Synthesis with the α-Ketoacid–Hydroxylamine
Ligation Acc. Chem. Res. 50, 2104–2115
(17) Porter, E. A., Weisblum, B., Gellman, S. H. (2002) Mimicry of Host-Defense Peptides by
Unnatural Oligomers:ꢀ Antimicrobial β-Peptides J. Am. Chem. Soc. 124, 7324–7330
(18) Godballe, T.,Nilsson, L. L., Petersen, P. D., Jenssen, H. (2011) Antimicrobial β‐Peptides and
α‐Peptoids Chem. Biol. Drug Des. 77, 107–116
Page 20 of 22
ACS Paragon Plus Environment

Page 21 of 23
ACS Chemical Biology
1
2
3
4
5
6
(19) Epand, R. F., Raguse, T. L., Gellman, S. H., Epand, R. M. (2004) Antimicrobial 14-Helical β-
Peptides: Potent Bilayer Disrupting Agents Biochemistry 43, 9527–9535
(20) Chiang, Y.-L., Russak, J. A., Carrillo, N, and Bode, J. W. (2012) Synthesis of Enantiopure
Isoxazolidine Monomers for the Preparation of β3-Oligopeptides by Iterative α-Ketoacid–
Hydroxylamine (KAHA) Ligations Helv. Chim. Acta 95, 2481–2501.
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
(21) Gerfaud, T., Chiang, Y.-L., Kreituss, I., Russak, J., and Bode, J. W. (2016) Enantioselective,
chromatography-free synthesis of β3-amino acids with natural and unnatural side chains Org.
Process Res. Dev. 16, 687–696.
(22) Yu, S.; Ishida, H.; Juarez-Garcia, M. E.; Bode, J. W. (2010) Unified synthesis of enantiopure
β2h, β3h and β2,3-amino acids Chem. Sci. 1, 637–641.
(23) Han, W., Hu, Z., Jiang, X., and Decicco, C. P. (2000) α-Ketoamides, α-ketoesters and α-
diketones as HCV NS3 protease inhibitors Bioorg. Med. Chem. Lett. 10, 711–713.
(24) Zhou, J., Mock, E. D., Martella, A., Kantae, V., Di, X., Burggraaff, L., Baggelaar, M. P., Al-
Ayed, K., Bakker, A., Florea, B. I., et al (2019) Activity-Based Protein Profiling Identifies α-
Ketoamides as Inhibitors for Phospholipase A2 Group XVI ACS Chem. Biol. Article ASAP (DOI:
10.1021/acschembio.8b00969)
(25) Zeigler, D. R., Prágai, Z., Rodriguez, S., Chevreux, B., Muffler, A., Albert, T., Bai, R., Wyss, M.
and Perkins, J. B. (2008) The origins of 168, W23, and other Bacillus subtilis legacy strains. J.
Bacteriol. 190, 6983–6995.
(26) Campbell, J. (2012) High‐Throughput Assessment of Bacterial Growth Inhibition by Optical
Density Measurements Curr. Protoc. Chem. Biol. 2, 195–208.
(27) Richter, M. F., Drown, B. S., Riley, A. P., Garcia, A., Shirai, T., Svec, R. L., and Hergenrother,
P. J. (2017) Predictive Compound Accumulation Rules Yield a Broad-Spectrum Antibiotic
Nature 545, 299–304.
(28) Koh, M., Park. J., Koo, J. Y., Lim, D., Cha, M. Y., Jo, A., Choi, J. H., and Park, S. B. (2014)
Phenotypic Screening to Identify Small‐Molecule Enhancers for Glucose Uptake: Target
Identification and Rational Optimization of Their Efficacy Angew. Chem. Int. Ed. 53, 5102–5106.
(29) Burdine, L., and Kodadek, T. (2004) Target identification in chemical genetics: the (often)
missing link Chem. Biol. 11, 593–597.
(30) Parker, C. G., Galmozzi, A., Wang, Y., Correia, B. E., Sasaki, K., Joslyn, C. M., Kim, A. S.,
Cavallaro, C. L., Lawrence, R. M., Johnson, S. R. et al. (2017) Ligand and Target Discovery by
Fragment-Based Screening in Human Cells Cell 168, 527–541.
(31) Li, Z., Hao, P., Li, L., Tan, C. Y. J., Cheng, X. Chen, G. Y. J., Sze, S. K., Shen, H.-M., and Yao,
S. Q. (2013) Design and Synthesis of Minimalist Terminal Alkyne-Containing Diazirine Photo-
Crosslinkers and Their Incorporation into Kinase Inhibitors for Cell- and Tissue-Based Proteome
Profiling Angew. Chem. Int. Ed. 52, 8551–8556.
(32) Fisher, J. F., Meroueh, S. O., and Mobashery, S. (2005) Bacterial Resistance to β-Lactam
Antibiotics: Compelling Opportunism, Compelling Opportunity! Chem. Rev. 105, 395–424.
(33) Schuck, P. (1997) Use of surface plasmon resonance to probe the equilibrium and dynamic
aspects of interactions between biological macromolecules. Annu. Rev. Biophys. Biomol. Struct.
26, 541–566.
Page 21 of 22
ACS Paragon Plus Environment

ACS Chemical Biology
Page 22 of 23
1
2
3
4
5
6
7
(34) Abdiche, Y., Malashock, D., Pinkerton, A., and Pons, J. (2008) Determining kinetics and
affinities of protein interactions using a parallel real-time label-free biosensor, the Octet Anal.
Biochem. 377, 209–217.
8
9
(35) Seidel, S. A. I., Wienken, C. J., Geissler, S., Jerabek-Willemsen, M., Duhr, S., Reiter, A.,
Trauner, D., Braun, D., and Baaske, P. (2012) Label-Free Microscale Thermophoresis
Discriminates Sites and Affinity of Protein-Ligand Binding Angew. Chem. Int. Ed. 51, 10656–
10659.
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
(36) Shilabin, A. G., Dzhekieva, L., Misra, P., Jayaram, B., and Pratt, R. F. (2012) 4-Quinolones as
Noncovalent Inhibitors of High Molecular Mass Penicillin-Binding Proteins ACS Med. Chem.
Lett. 3, 592−595.
(37) Lu, W. P., Kincaid, E., Sun, Y., and Bauer M. D. (2001) Kinetics of beta-lactam interactions with
penicillin-susceptible and -resistant penicillin-binding protein 2x proteins from Streptococcus
pneumoniae. Involvement of acylation and deacylation in beta-lactam resistance. J. Biol. Chem.
276, 31494–31501.
(38) Chesnel, L., Pernot, L., Lemaire, D., Champelovier, D., Croize, J., Dideberg, O., Vernet, T., and
Zapun, A. (2003) The structural modifications induced by the M339F substitution in PBP2x from
Streptococcus pneumoniae further decreases the susceptibility to beta-lactams of resistant
strains. J. Biol. Chem. 278, 44448–44456.
(39) Ishiyama, M., Miyazono, Y., Sasamoto, K., Ohkura, Y., and Ueno, K. (1997) A highly water-
soluble disulfonated tetrazolium salt as a chromogenic indicator for NADH as well as cell
viability. Talanta 44, 1299–1305.
(40) Tominaga, H., Ishiyama, M., Ohseto, F., Sasamoto, K., Hamamoto, T., Suzuki, K., and
Watanabe, M. (1999) A water-soluble tetrazolium salt useful for colorimetric cell viability assay.
Anal. Commun. 36, 47–50.
(41) Cox, J. and Mann, M. (2008) MaxQuant enables high peptide identification rates, individualized
p.p.b. range mass accuracies and proteome-wide protein quantification, Nat. Biotechnol. 26,
1367–1372.
(42) Tyanova, S.; Temu, T.; Sinitcyn, P.; Carlson, A.; Hein, M. Y.; Geiger, T.; Mann, M. and Cox, J.
(2016) The Perseus computational platform for comprehensive analysis of (prote)omics data,
Nat. Methods 13, 731–740.
Page 22 of 22
ACS Paragon Plus Environment

Page 23 of 23
ACS Chemical Biology
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
282x63mm (300 x 300 DPI)
ACS Paragon Plus Environment