Chemical Probes Unravel an Antimicrobial Defense Response Triggered by Binding of the Human Opioid Dynorphin to a Bacterial Sensor Kinase

Article abstract of DOI:10.1021/jacs.7b01072

Host-microbe communication via small molecule signals is important for both symbiotic and pathogenic relationships, but is often poorly understood at the molecular level. Under conditions of host stress, levels of the human opioid peptide dynorphin are elevated, triggering virulence in the opportunistic pathogenic bacterium Pseudomonas aeruginosa via an unknown pathway. Here we apply a multilayered chemical biology strategy to unravel the mode of action of this putative interkingdom signal. We designed and applied dynorphin-inspired photoaffinity probes to reveal the protein targets of the peptide in live bacteria via chemical proteomics. ParS, a largely uncharacterized membrane sensor of a two-component system, was identified as the most promising hit. Subsequent full proteome studies revealed that dynorphin(1-13) induces an antimicrobial peptide-like response in Pseudomonas, with specific upregulation of membrane defense mechanisms. No such response was observed in a parS mutant, which was more susceptible to dynorphin-induced toxicity. Thus, P. aeruginosa exploits the ParS sensing machinery to defend itself against the host in response to dynorphin as a signal. This study highlights interkingdom communication as a potential essential strategy not only for induction of P. aeruginosa virulence but also for maintaining viability in the hostile environment of the host.

Full text of DOI:10.1021/jacs.7b01072

Subscriber access provided by HACETTEPE
UNIVERSITESI KUTUPHANESI
Article
Chemical probes
unravel an antimicrobial
defense response
triggered by binding
of the human opioid
dynorphin to a
bacterial sensor kinase
Megan Holly Wright, Christian
Fetzer, and Stephan A. Sieber
Journal of the American
Chemical Society is
published by the American
Chemical Society. 1155
Sixteenth Street N.W.,
Washington, DC 20036
Published by American
Chemical Society.
Copyright © American
Chemical Society.

Subscriber access provided by HACETTEPE
UNIVERSITESI KUTUPHANESI
J. Am. Chem. Soc., Just Accepted
Manuscript • DOI: 10.1021/jacs.7b01072
•
Publication Date (Web): 28 Mar 2017
Downloaded from http://
pubs.acs.org on March 28, 2017
Just Accepted
“Just Accepted” manuscripts have been pee
online prior to technical editing, formatting fo
Society provides “Just Accepted” as a fre
dissemination of scientific material as soon a
appear in full in PDF format accompanied by
Journal of the American
Chemical Society is
published by the American
Chemical Society. 1155
Sixteenth Street N.W.,
Washington, DC 20036
Published by American
Chemical Society.
Copyright © American
Chemical Society.

Subscriber access provided by HACETTEPE
UNIVERSITESI KUTUPHANESI
fully peer reviewed, but should not be conside
readers and citable by the Digital Object Iden
to authors. Therefore, the “Just Accepted” W
in the journal. After a manuscript is technica
Accepted” Web site and published as an ASA
changes to the manuscript text and/or grap
and ethical guidelines that apply to the jou
or consequences arising from the use of info
Journal of the American
Chemical Society is
published by the American
Chemical Society. 1155
Sixteenth Street N.W.,
Washington, DC 20036
Published by American
Chemical Society.
Copyright © American
Chemical Society.

P aa Jg. oe u1r no af l1 o7 f the American Chemical Society
signal
receptor?
1
2
3
4
5
host cell
bacterium
response
b.
6
7
8
9
DYN Y G G F L R R
SCR L F K Y R R
CC L S A T R H Q R V R V K - NH2
I
R P K L K - NH2
I
G P R G K L - NH2
I
c.
10
11
12
13
14
15
16
17
18
19
1
.2
n=3
24
OD PYO
20
0.8
0.4
0
16
12
8
4
0
0
4
11 33 100 300
DYN (µM)
2
d0 .
1
0
0
.2
.8
.4
n=3
10
OD
PYO
21
22
23
24
25
26
27
28
29
30
8
6
4
2
0
0.0
ACS Paragon Plus Environment
DMSO
CC
SCR
DYN
Peptide (50 µM)

a.
photo-group
Journal of the American Chemical Societ cy .
Page 2 of 17
1. x-link
. cell lysis
in-gel
2
fluorescence
analysis
1
2
3
tag
CuAAC
Bpa
m/z
MS-based
proteomics
Pra
bacterial culture
4
b.
5
Residue
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
6
DYN
Tyr Gly Gly Phe Leu Arg Arg Ile Arg Pro Lys Leu Lys
Tyr Gly Gly Bpa Leu Arg Arg Ile Arg Pro Lys Leu Lys Pra
7
DYN1
8
DYN3
Tyr Gly Gly Phe Leu Arg Arg Ile Arg Pro Lys Leu Lys Bpa Pra
Pra Bpa Lys Tyr Arg Arg Ile Gly Pro Arg Gly Lys Leu
Tyr Gly Gly ArN3 Leu Arg Arg Ile Arg Pro Lys Pra Lys
Tyr Gly Gly Phe Leu Arg Arg Ile Arg P(Di) Lys Pra Lys
Tyr Gly Gly Phe Leu Arg Arg Ile K(Di) Pro Lys Pra Lys
9
1
SCR1
D0YN4
DYN5
ACS Paragon Plus Environment
1
1
ArN3
P(Di)
K(Di)
DYN9
12
3
1

a .P a Jg oe u3r no af l1 o7 f the American Chemical Society
1.2
OD
PYO
n=3
1
8
4
0
2
1 0.8
2
3
4
5
6
7
8
0.4
0.0
DMSO DYN DYN5 DYN4 DYN1 SCR1
Peptide (50 µM)
9
b.
Diazirine
ArN3
Bpa
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
0
1
Probe
DMSO DYN5 DYN9 DYN4 DYN3
2Irradiation
3
-
+
-
+
-
+
-
+
-
+
9
6
8 -
3 -
4
5
6
7
8
9
0
1
2
4
0 -
2 -
3
21 -
11 -
c.
d.
Probe DYN1
DYN5
3
4
5
6
7
8
DYN1
LPS
Prot.K
-
+
-
+
1
9
55
8
- 155
- 98
63
-
63
940
032
1
-
40
ACS Paragon Plus Environment
-
-
32
21
21
2
3

Journal of the American Chemical Society
Page 4 of 17
a.
c.
7
6
5
4
3
2
1
0
2
8
Other
PagL
Significant
2
6
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
Competition at 5x DYN
ParS
Rnk
PA3226
24
22
20
PA3435
PA1243, FruR, PA3341
PA3440 & PdxY
**
FDR 0.01, s0 2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
PA3226
PagL
PA3341
ParS
RnK
PA3435
PA1243
-4
-2
0
2
4
6
8
10
Enrichment log (DYN5/DMSO)
2
b.
5
4
3
2
1
0
d.
Uniprot Gene
Other
Significant
Enriched
ParS
Protein
Q9HYQ4 PA3341 transcriptional
regulator
Q9HZ14 PA3226 hydrolase
RnK
PA3435
PA3440
ThiD
Q9HTS6 PA5274 Nucleoside
diphosphate kinase
regulator (Rnk)
FDR 0.05, s0 0.5
Q9HYH1 PA3435 Uncharacterized
(oxidoreductase)
PA3341
PA1243
Q9HVD1 PA4661 Lipid A deacylase
PagL
PA3226
Q9I495
PA1243 sensor/response
regulator hybrid
-3
-1
1
3
5
Q9I2U4
PA1798 Two-component
sensor, ParS
ACS Paragon Plus Environment
Enrichment log (DYN5/(DYN5:DYN 1:2 competition))
2

Pa a. ge 5 of 17
Journal of the American Chemical Society c.
1
0
8
6
4
2
0
OprD
ArnT
PmrA
PmrB
MexI
LecA
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
MexH
SpeE2
MexX(AmrA)
ArnB
MexY
AmrB)
(
OpmD ArnA
Response to
antibiotic
MexG
0
1
2
3
4
5
6
7
8
9
FDR 0.01, s0 1
Log (DYN/DMSO)
2
-6
-4
-2
0
2
4
6
8
10
1.38
9.85
Fold-change Stat.10 Log (DYN 10 µM / DMSO)
2
siderophore
sulfur cpd metab.
response antibiotic/efflux
arn operon
PA4774-7 (PmrA,B & others)
b.
6
2
0
1
2
3
4
5
6
7
8
9
0
1
2
3
PmrA
PmrB
2
2
2
2
2
2
2
2
2
3
3
3
3
ArnA-D
PA3559
4
2
0
ArnT
SpeE2
FDR 0.05, s0 1
arn operon
- Lipid A modification
ACS Paragon Plus Environment
-6
-4
-2
0
2
4
6
8
10
Fold-change Exp.5 Log (DYN 5µM / DMSO)
2

1
.2
a.
Journal of the American Chemical So bc .iety
Page 6 of 17
8
6
4
8
0.8
1
2
3
4
5
6
7
8
9
6
4
2
0
PmrA
PmrB
0.4
0.0
wt
ParS
ParR
Strain
0 µM 3 µM 10 µM 30 µM 50 µM
1
1
1
1
1
1
1
1
1
1
2
2
2
2
0 2
1
2
3
c.
DYN
membrane
defence
LPS
ParS
4 0
-6
-2
2
6
10
-6
-2
2
6
10
5
6
7
8
9
0
1
2
Fold-change wild-type
Fold-change parS mutant
Log (DYN1µM / DMSO)
PmrA
PmrB
Arn
proteins
Log (DYN1µM / DMSO)
2
2
ParR
Proteins in arn operon
other hit
Efflux pumps e.g.MexXY
?
PA4774-7 (PmrA,B & others)
hit but protein differs significantly
ACS Paragon Plus Environment
virulence
expression
between wt and parS mutant
antibiotic
defence
3

Page 7 of 17
Journal of the American Chemical Society
1
2
3
4
5
6
Chemical probes unravel an antimicrobial defense response trig-
gered by binding of the human opioid dynorphin to a bacterial
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
sensor kinase.
†
*
*
Megan H. Wright, C. Fetzer, S. A. Sieber .
Department of Chemistry, Technical University of Munich, Lichtenbergstr. 4, D-85748, Garching, Germany.
ABSTRACT: Host-microbe communication via small molecule signals is important for both symbiotic and pathogenic
relationships, but is often poorly understood at the molecular level. Under conditions of host stress, levels of the human
opioid peptide dynorphin are elevated, triggering virulence in the opportunistic pathogenic bacterium Pseudomonas ae-
ruginosa via an unknown pathway. Here we apply a multilayered chemical biology strategy to unravel the mode of action
of this putative interkingdom signal. We designed and applied dynorphin-inspired photoaffinity probes to reveal the pro-
tein targets of the peptide in live bacteria via chemical proteomics. ParS, a largely uncharacterized membrane sensor of a
two-component system, was identified as the most promising hit. Subsequent full proteome studies revealed that dy-
norphin(1-13) induces an antimicrobial peptide-like response in Pseudomonas, with specific upregulation of membrane
defence mechanisms. No such response was observed in a parS mutant, which was more susceptible to dynorphin-
induced toxicity. Thus, P. aeruginosa exploits the ParS sensing machinery to defend itself against the host in response to
dynorphin as a signal. This study highlights interkingdom communication as a potential essential strategy not only for
induction of P. aeruginosa virulence but also for maintaining viability in the hostile environment of the host.
in the environment, integrating multiple inputs in the
decision to switch to the costly virulence phenotype. Bac-
Introduction
7
The dynorphins are important endogenous human pep-
tide hormones that act primarily through binding to the
κ-opioid receptor (OPRK) and have roles in pain re-
teria have receptors that recognize other host molecules
8
9
(
e.g. interferon-γ and adrenergic hormones ). However,
the putative dynorphin receptor and the majority of the
response circuitries mediating bacterial response to opi-
oids are completely unknown (Figure 1a). Targeting viru-
lence or manipulating host-microbe or microbe-microbe
communications have been put forward as possible non-
traditional approaches to tackle the rise of multidrug re-
sistant bacteria. However, the pathways and environmen-
tal conditions, such as host stress and interkingdom sig-
naling, that lead to microbial virulence are poorly ex-
plored.
1
sponse, stress and addiction, amongst others. Opioids
and their receptors are important components of the pe-
ripheral nervous system, but they also accumulate during
2
host stress at sites of inflammation. Coupled with evi-
dence that bacteria activate virulence in response to host
3
stress, these observations suggest that opioids may be
recognized by bacteria as signaling molecules. A small
molecule OPRK agonist and the peptide hormone dy-
norphin-A (Dyn-A) were shown to activate quorum sens-
ing and result in enhanced virulence of the opportunistic
Here, we use a chemical biology approach to unravel the
mechanism of action of the human opioid peptide Dyn-A
in PA. Phenotypic studies first confirm that the virulence-
inducing effect of a truncated dynorphin analogue (DYN)
is peptide sequence-dependent. We then employ a chem-
ical proteomic approach using DYN-inspired photoaffini-
ty probes to identify the membrane sensor ParS as a puta-
tive binding partner of DYN in live bacteria. In-depth
global proteomic analyses further define the specific de-
fense response evoked in PA by DYN. Finally, we show
that this defense response is abrogated in a mutant lack-
ing ParS, functionally linking peptide binding to ParS
with the phenotypic response.
4
pathogen Pseudomonas aeruginosa (PA). PA is one of the
most commonly isolated microbes in patients with blood-
stream and lung infections in acute care settings in Eu-
rope, and the leading cause of death in patients with cyst-
5
ic fibrosis. The pathogen is also very challenging to treat
due to its high intrinsic resistance to antibiotics and its
ability to modulate its behavior in widely different envi-
ronments and in response to a host. This ability may stem
in part from the organism’s large number of two-
component sensors, its large genome, and its complex
6
quorum sensing networks. In a rodent stress model,
Dyn-A was shown to be released in the intestine where it
4
bound to and was taken up by bacteria. A subsequent
study showed that PA regulates its response to a small
molecule opioid depending on the abundance of nutrients
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 8 of 17
Results
properties, it is difficult to predict which photogroup will
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
perform best in any one system; we therefore investigated
all three. Rather than modify the peptide with a large la-
bel for detection, we incorporated a terminal alkyne into
the structure of probes; this pre-tag allows subsequent
attachment of a fluorophore or biotin affinity label via
copper-catalysed azide-alkyne cycloaddition (CuAAC) for
enrichment and visualization (Figure 2a). Probes were
synthesized using Fmoc-SPPS chemistry. The bulkiest
photogroup, Bpa, was incorporated at two different posi-
tions: in place of phenylalanine (probe DYN1) and at the
C-terminus of the sequence (DYN3; Figure 2b).
Dynorphin and derivatives induce Pseudomonas
virulence
Since PA virulence is known to be highly environment-
dependent, we first sought to define the conditions under
which dynorphin induced a response. Previous work has
linked the peptide Dyn-A (dynorphin(1-17)-OH) and the
small molecule opioid U50,488 to induction of virulence
in PA strain PAO1. We confirmed that in our hands the-
se compounds also induced the production of the blue-
green toxin pyocyanin under similar conditions of nutri-
ent limitation, as did a truncated Dyn-A analogue with a
4
,7
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
C-terminal amide (DYN, dynorphin(1-13)-NH ; Figure S1).
The N-terminus of dynorphin is critical for signaling via
human opioid receptors and this truncated peptide is
2
1
0
equipotent in the human system. However, although
they induced a similar response to each other, the two
peptides behaved quite differently compared to the small
molecule: there was significant growth reduction at high-
er concentrations of peptide and a corresponding drop-off
of pyocyanin production, whereas U50,488 was less toxic
to bacteria and only induced pyocyanin at higher concen-
trations (Figure S1). It is therefore not clear that dy-
norphin-based peptides and U50,488 have the same phe-
notype and thus mode of action. We decided to focus on
the endogenous peptide signals, and, given the similar
responses of the two peptides, chose the shorter dy-
norphin analogue, DYN, for synthetic tractability. We
used a chemically defined MOPS-buffered succinate me-
dium to optimize the iron and phosphate contents, since
these components are important for the virulence re-
7
sponses of PA (Figure S1). Our optimized conditions gen-
erated more reproducible data and we observed that DYN
Figure 1b) induced mild toxicity and production of pyo-
cyanin in a dose-dependent manner (Figure 1c).
(
To establish whether the sequence of DYN is important
for the observed phenotype, we tested a sequence-
scrambled peptide (SCR) and a charge-control peptide
(
CC), which has a different composition but is still highly
basic (Figure 1b). The charge control showed no signifi-
cant effect on either growth or virulence, and the scram-
bled peptide a greatly reduced effect (Figure 1d), suggest-
ing a highly specific response to DYN.
Photoaffinity profiling of dynorphin targets
Intrigued by this phenotype, we next sought to identify
protein interacting partners of DYN in live bacteria using
photoprobes. Photoaffinity labelling is a technique used
Figure 1. DYN induces virulence in Pseudomonas aeru-
ginosa in a sequence-specific manner. a. Dynorphin is a
human peptide hormone released by host cells during stress.
It triggers a virulence response in the bacterium PA via an
unknown pathway and putative receptor. b. Sequences of
DYN and control peptides. c. The effect of DYN on growth,
as measured by optical density (OD, grey bars), and on the
production of the virulence factor pyocyanin (PYO, blue
bars) in PAO1. Error bars show standard deviation from 3
independent biological experiments. OD and PYO were
normalized to DMSO control in each case. d. Analyzing the
sequence specificity of the DYN response using the peptides
shown in b.
11,12
to explore small molecule-protein binding. A chemical
probe mimicking the molecule of interest is equipped
with a photoreactive group and a label or tag for detec-
tion. The probe is incubated with lysates or live cells, and,
upon UV irradiation, a highly reactive intermediate is
generated from the photoreactive group which crosslinks
the probe to its biomolecular binding partners (Figure
2
a). Common photoreactive groups include benzophe-
none (Bpa), aryl azides (ArN ) and diazirines (Di) (Figure
3
2
b), which possess different photo-properties and steric
1
3
footprints. Given their distinct and complementary
ACS Paragon Plus Environment

Page 9 of 17
Journal of the American Chemical Society
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 2. Design of photoreactive probes mimicking DYN. a. Workflow for identifying putative DYN protein interactors via
chemical proteomics. Probes are incubated with bacterial cultures and the cultures are irradiated with UV light, thereby photo-
crosslinking the probe to its protein binding partners. Probes contain a clickable alkyne for ligation to azide-functionalised
fluorophores (for visualization by in-gel fluorescence) or biotin (for affinity enrichment and gel-free LC-MS/MS analysis of tryp-
tic peptides). b. Sequences of DYN and probes bearing different photocrosslinkers (all C-terminal amides). Bpa = benzophenone;
ArN = aryl azide; P(Di) = photoproline (diazirine-bearing proline); K(Di) = diazirine-modified lysine; Pra: propargyl-glycine. c.
3
Structures of the unnatural amino acid residues.
Similarly, an aryl azide moiety was incorporated in place
of phenylalanine in DYN4, and diazirine-containing ‘pho-
hypothesis, a scrambled Bpa-containing probe (SCR1, Fig-
ure 2b), designed based on the sequence of the scrambled
control peptide SCR, gave a similar LPS-dominated label-
ling pattern (Figure 3d).
1
4
toproline’ in place of proline in DYN5. A synthetically
facile orthogonal protection strategy also yielded probe
DYN9, where a short diazirine building block was coupled
to a lysine side chain. Probes DYN5, DYN4, DYN3 and
DYN1 all triggered a clear virulence response in PA, whilst
the scrambled Bpa probe SCR1 showed very little effect, as
anticipated (Figure 3a, S2). Unexpectedly, DYN9 had little
effect on pyocyanin production (Figure S2), possibly indi-
cating high sequence sensitivity of the virulence response.
We initially focused on the Bpa probe DYN1, since probes
with this photogroup gave the strongest labelling and
high pyocyanin induction in PA. Proteins were subject to
CuAAC with biotin-azide, enriched on avidin beads and
digested. Peptides from control (DMSO vehicle treated)
1
6
or DYN1 samples were isotopically dimethyl labelled for
quantification, mixed and analyzed by gel-free LC-
MS/MS. Consistent with the high level of labelling ob-
served on-gel, large numbers of proteins were apparently
enriched in DYN1 over control samples, suggestive of high
To perform labelling experiments in live cells, PA was
grown to early stationary phase, resuspended in PBS, in-
cubated with probes at 4 µM (the onset of virulence in-
duction) and samples were irradiated. Cells were washed,
lysed and lysates subject to CuAAC to attach a rhodamine
fluorophore. Separation of proteins by SDS-PAGE and
fluorescence scanning revealed some striking differences
between probes bearing different photoreactive groups
(Figure 3b). All probes exhibited irradiation-dependent
labelling, but whilst aryl azide probe DYN4 labelled dis-
tinct bands, both diazirine and Bpa probes additionally
gave rise to diffuse bands and a ‘laddering’ pattern, remi-
niscent of heterogeneous glycolipids. Lipopolysaccharide
1
7-19
non-specific binding or photoreactivity (Figure S3).
We next performed chemical proteomic analyses with
1
DYN4 and using a similar workflow but label-free MS
20
intensity-based quantification (MaxLFQ), which enables
comparison across more samples. We included a competi-
tion experiment where bacteria were first incubated with
5
-fold excess DYN and then with DYN4, with the aim of
differentiating between genuine DYN binders over high
abundance or false positive hits. DYN4 was more specific
than DYN1 but still gave a large number of potential hits
(Figure S3, Table S1). Outer membrane proteins (OMPs)
figured prominently and biochemical separation of inner
and outer membranes supported this result (Figure S4).
We also observed a surprising enrichment of ribosomal
proteins (Table S1). To check for effects of DYN on pro-
tein synthesis, we treated early log-phase bacteria with
DYN at 10 µM for 2 hours and then incubated aliquots for
30 min with the methionine analogue azidohomoalanine
(AHA), which is incorporated into newly synthesized pro-
(
LPS) forms a large part of the surface coat of PA and oth-
er gram-negative bacteria and consists of lipid A, a core
oligosaccharide and a variable O-antigen polysaccharide
1
5
side chain. Treatment of samples after CuAAC with a
protease, proteinase K, indeed showed that the diffuse
labeling and laddered bands were not protein (Figure 3c).
Furthermore, isolated PA LPS labelled with probes gener-
ated a similar pattern (Figure 3d; Figure S2). The DYN
probes are positively charged at physiological pH and
charge-charge interactions may mediate binding to the
negatively charged phosphates of LPS. Supportive of this
21,22
teins.
Although total protein yield was lower for the
ACS Paragon Plus Environment

Journal of the American Chemical Society
DYN-treated bacteria due to reduction in growth, the rate
Page 10 of 17
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
of protein synthesis as measured by pulsed incorporation
of AHA was the same (Figure S4). Furthermore, a simple
1
9
aryl azide ‘minimal’ probe labelled proteins in the solu-
ble fraction in a similar pattern to DYN4 (Figure S4). To-
gether these data suggest that aryl azide probes exhibit a
high degree of non-specific binding in this system. Lack of
LPS labelling by DYN4 (Figure S4) also suggested that this
probe is biased in its reactivity. This may be related to the
propensity of aryl azides to undergo ring expansion upon
1
3
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
irradiation.
Finally, to investigate the protein targets of diazirine
probe DYN5, bacteria were incubated with DYN at vary-
ing concentrations followed by DYN5 (ratios of 1:1, 1:2 and
1
:5 of DYN5:DYN) to generate competition data. Enrich-
ment and label-free proteomics were carried out as be-
fore. Analysis of the data revealed more selective labelling
compared to DYN1 and DYN4 (Figure 4a), as expected
given the lower intensity of in-gel labelling. A subset of
enriched proteins was significantly outcompeted at 5-fold
excess DYN (Figure S5) and three proteins were also out-
competed at 2-fold excess DYN (Figure 4b; Table S2). Of
these, the protein ParS was alone in being completely
absent from DMSO controls. ParS was also detected in
DYN4 labelling, although only quantified in two out of
three replicates. We analyzed the competition response
profiles (LFQ intensity, raw intensity and sequence cover-
age) of all possible hits (Figure S5; Figure 4c); ParS
showed the most pronounced dose-dependent response.
ParS is a putative membrane sensor kinase and is predict-
ed to be co-expressed with response regulator ParR in a
23
two-component system. The ParRS system has been
linked to the response of PA to cationic antimicrobial
24
peptides (AMPs). AMPs include endogenous host pep-
tides involved in the immune defense against infection
and synthetic compounds in clinical use. Although in
many cases their mode of action is unclear, AMPs often
bind to the bacterial membrane and rapidly kill cells
Figure 3. Photoaffinity labelling in live bacteria. a. De-
termination of growth (OD, grey) and PYO (blue) induction
by probes.. Error bars show standard deviation from 3 biolog-
ical replicates. b. Gel-based analysis after photolabeling with
probes (DYN5, 9, 4 and 3; see Figure 2) in live PA. Probes
were incubated at 5 µM. Irradiation with UV was performed
as indicated; after lysis and separation into soluble (PBS) and
insoluble fractions, proteins were subject to CuAAC with
rhodamine-azide (RhN3), precipitated with acetone, separat-
ed by gel and imaged for fluorescence. Insoluble fraction
shown here. c. Analysis of probe labeling of bacterial lipopol-
ysaccharide (LPS). Bacteria were labeled with DYN1 or DYN5
and lysed in 0.2% TX100 in PBS. Following CuAAC with
RhN3 and precipitation, samples were resuspended, split into
two portions and treated with proteinase K (Prot.K) or buffer
25
through formation of pores or membrane disruption.
Select peptides appear to act via ParRS to induce adaptive
resistance to other AMPs – a response where the bacterial
cell activates specific mechanisms that render it tempo-
24
rarily more resistant to attack. Although DYN only
slightly reduced growth (Figure 1c), it is cationic. Thus we
selected ParS out of the list of possible hits (Figure 4d) as
the most promising protein for further analysis.
Total proteome analysis of DYN-treated Pseudo-
monas reveals an antimicrobial peptide-like de-
fense response
If DYN binds ParS to induce an AMP-like response in PA,
changes in the expression of specific proteins and path-
ways should be detectable. We therefore undertook a
global proteome analysis of bacteria treated with the pep-
tide. Bacteria were grown for 16 hours in the presence of
10 µM DYN (a concentration resulting in a small drop in
growth and a measurable increase in pyocyanin; Figure
S6).
(
control) to digest proteinaceous material. Samples were
separated on SDS PAGE and gels fluorescent scanned. d.
Left: probe DYN1 labelling of purified LPS. Right: Scrambled
Bpa probe SCR1 labelling of bacteria (as for c). See full gels
and Coomassie staining in Figure S2.
ACS Paragon Plus Environment

Page 11 of 17
Journal of the American Chemical Society
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
Figure 4. Chemical proteomics with diazirine probe DYN5 reveals ParS as a potential target. Bacteria were incubated
with 3 μM DYN5, or DYN5 plus increasing concentrations of DYN (1, 2 or 5-fold), or DMSO control, irradiated, lysed, proteins
ligated by CuAAC to biotin-N , enriched on avidin beads, digested by trypsin and peptides analyzed by LC-MS/MS. Three inde-
3
pendent biological replicate labelling experiments were performed on different days. Data were analyzed by MaxLFQ, filtered to
retain only proteins identified by multiple peptides and quantified in 2/3 biological replicates, and missing values were imputed
to mimic noise. Two-sample permutation-corrected t-tests with FDR (false discovery rate) control were performed to compare
samples. a. Volcano plot depicting enrichment and significance of enrichment in chemical proteomics experiments with 3 μM
DYN5. In red: proteins shown to be outcompeted by 5x excess DYN (see Figure S5). n=3 biological replicates. b. Volcano plot
showing results of competition between DYN5 and 2x excess DYN. In blue: proteins enriched over DMSO controls (see a). c.
1
Profile plots of potential hits (raw MS intensity). * = value of zero in DMSO control for these proteins. Proteins selected showed
a dose-response (see Figure S5), significant enrichment over DMSO controls and significant out-competition by DYN. FruR,
PA33440 and PdxY did not show a consistent dose-response. d. The most promising hits based on competition and enrichment,
and their putative roles. See Table S2.
Cells were subsequently lysed, lysates normalized based
on total protein amount and samples prepared for analy-
sis using the FASP protocol (filter-aided sample prepara-
tion), which enables lysis of samples in buffers containing
high detergent concentrations to maximize solubilization
processes were overrepresented amongst downregulated
proteins, indicating moderate changes in iron homeosta-
sis and uptake (Figure S7). Proteins upregulated by DYN
treatment included those involved in lipid A biosynthesis
and belong to the arn operon (ArnA, B, C, D and T; Figure
5 and Figure S8), which is upregulated in response to di-
verse AMPs. ArnT adds 4-aminoarabinose to lipid A
phosphate groups, reducing the negative charge of the
26
of membrane proteins. Peptides were analyzed by LC-
20
MS/MS and quantified by MaxLFQ. Over 3000 proteins
were quantified in at least three biological replicates, with
high reproducibility (Pearson correlations of 0.96-0.99,
Figure S6). We observed a specific response, with only
1
5
membrane and promoting resistance to AMPs. The arn
operon is under the control of various two-component
2
4,27,28
5
.5% of proteins significantly affected by DYN treatment
systems in PA, including ParRS, PmrAB and CprRS.
27
(t-test; Figure 5a and Table S3). Gene ontology analysis
pmrA & pmrB are also upregulated by AMPs, and expres-
sion levels of these two proteins increased strongly in
response to DYN in our analysis.
suggested that siderophore biosynthesis and transport,
sulfur compound metabolism, localization and transport
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 12 of 17
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
Figure 5. Global proteomics reveals specific changes in response to DYN. a. Volcano plot visualization of global prote-
omics experiment “Stat.10”. PA was grown in the presence of DYN at 10 µM for 16 h until stationary phase. Bacteria were harvest-
ed, lysed and samples prepared by FASP for LC-MS/MS. Quantification was carried out with MaxLFQ. n=5 (biological repli-
cates). After filtering to discard proteins not identified in at least four replicates, missing values were imputed and a two-sample
t-test (permutation-based, FDR 0.01, s0 1) carried out to compare DMSO and DYN treated samples. Proteins of particular inter-
est are labelled and color-coded by putative function/pathway. See also Table S3. b. Global proteomics – exponential phase, da-
taset “Exp.5”. PA was grown to OD 0.2-0.3 then incubated with DYN at 5 µM for 2 h during exponential growth. Samples were
analyzed as above. n=3 (replicates grown from independent starter cultures). Only proteins identified in at least duplicate were
used in analysis. Missing values were imputed (downshift 1.8, width 0.3) and a two-sample t-test (permutation-based, FDR 0.05,
s0 1) carried out to compare DMSO and DYN treated samples. See also Supplementary Table S4. c. Cytoscape analysis of upregu-
lated proteins (from a.; see also Figure S8). Edges between nodes represent predicted functional or physical interactions. Analy-
sis of downregulated proteins is shown in Figure S7.
Another highly upregulated protein with a known role in
All proteins from the mexGHI-opmD operon (PA4205-8)
were significantly upregulated, as were MexX and MexY
(PA2018-19; Table. S3). In contrast, the porin OprD, anno-
tated as a basic amino acid, basic peptide and imipenem
outer membrane transporter, was downregulated. Inter-
estingly, the ParRS system has been previously linked to
29
enhanced AMP resistance was SpeE2 (PA4774), which,
together with another upregulated product (PA4775), is
predicted to be in the same operon as pmrAB. Another set
of proteins that responded to DYN belong to the re-
sistance-nodulation-division (RND) family and have
known or putative roles in drug efflux. These tripartite
complexes consist of an RND transporter, a periplasmic
membrane fusion protein and an outer membrane factor,
3
2
these changes.
3
0,31
and act to pump antibiotics across the cell membrane.
ACS Paragon Plus Environment

Page 13 of 17
Journal of the American Chemical Society
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 6. ParS mediates the AMP-response induced by DYN treatment. a. Volcano plot visualization of Stat.1 dataset (wt or
parS mutant bacteria grown for 16 h in the presence of 1 μM DYN or DMSO control). n=3 replicates (grown from separate starter
cultures on the same day). Bacteria were harvested, lysed and samples prepared by FASP for LC-MS/MS. Quantification was car-
ried out with MaxLFQ. After filtering to discard proteins not identified in at least two replicates, missing values were imputed
(
downshift 1.8, width 0.3) and a two-sample t-test (permutation-based, FDR 0.05, s0 1) carried out to compare DMSO and DYN
treated samples for each strain. No significant hits were obtained in the parS mutant. See also Table S5. The wt and mutant pro-
teomes are compared in Figure S10. b. parS and parR transposon mutants are more susceptible to treatment with DYN. Error
bars show standard deviation from technical duplicates. Response to pyocyanin for these mutants and others is shown in Figure
S9. c. Proposed model for DYN signaling via ParS to promote membrane defense mechanisms and induce virulence. DYN inter-
acts with ParS sensor kinase, presumably triggering phosphorylation of the response regulator ParR and transcription of proteins
in the pmrAB and arn operons. Proteins from these operons are involved in the antimicrobial peptide-liked response, specifically
modifying the structure of LPS lipid A in the case of ArnT. Efflux pumps are also upregulated. Thus ParS signaling triggers bacte-
rial defense against antimicrobial peptides and antibiotics. The link to virulence expression (e.g. pyocyanin induction and up-
regulation of biofilm-related protein LecA) is yet to be elucidated.
Another virulence-related protein, PA-I galactophilic lec-
tin (LecA) was strongly upregulated in our data. LecA is a
cytotoxic lectin and adhesion molecule implicated in bio-
In all analyses ParR and ParS proteins themselves were
unaffected by DYN treatment; however, this is not incon-
sistent with their role as mediators since a previous study
by Fernández et al. reported that the AMP indolicidin had
no effect on parRS expression, although this peptide trig-
gered responses via parRS. We compared our dataset to
3
3
film formation, and induced in response to another host
8
signal, interferon-γ. The data described above were col-
lected incubation of DYN throughout growth until sta-
tionary phase. To explore the effects of transient DYN
exposure, bacteria were grown to early exponential phase
and then incubated with a slightly lower concentration of
5 μM DYN (to limit the mild inhibition of growth that
DYN induces) and the global proteome analyzed (Table
S4). Only 0.75% of proteins showed a significant response
under these conditions: again, proteins in the arn and
pmrAB operons were the most prominent upregulated
hits (Figure 5b). Thus, arn and pmrAB upregulation is a
rapid, direct response to DYN under different growth
stages.
2
4
the microarray transcriptome data of Fernández et al.
Proteins in the arn and pmr operons, significantly upregu-
lated by DYN treatment in our analysis, were indeed in-
duced by indolicidin treatment in wild-type relative to a
parR mutant (Table S6).
ParS mediates the DYN-induced phenotype
We hypothesized that of the possible DYN targets (Figure
4
d), ParS was most likely responsible for the AMP-defense
phenotype. To test this hypothesis more directly, a PAO1
parS transposon mutant was obtained (PA Two-Allele
Library, University of Washington), validated by PCR
(
Figure S9) and assayed for response to DYN. Both this
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 14 of 17
and a parR mutant were dramatically more susceptible to system CprRS mediate the response to diverse AMPs in
28
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
DYN-induced toxicity than the original wild-type (wt)
strain and we did not observe any pyocyanin production
in response to the peptide at non-toxic concentrations
PA in a peptide-dependent manner. CprRS was not de-
tected in our analyses, suggesting that it is not involved in
DYN-dependent signaling. Interestingly, DYN is only
weakly toxic but induces strong upregulation of defense
pathways, consistent with previous data indicating that
the signaling and growth inhibition functions of AMPs are
(
Figure S9). Using a low DYN concentration of 1 μM to
minimize effects on growth, we repeated the global pro-
teome analysis of the parS mutant or wt PA treated with
DYN for 16 hours (Figure S6; Table S5). Even at this low
concentration, there was strong upregulation of Arn pro-
teins and PmrA & B in wt bacteria, whilst cells lacking
ParS showed no significant response at all (Figure 6b).
Although there are differences between the parS mutant
and wt strain proteomes, most proteins altered by DYN
treatment in the wt were expressed at comparable levels
24,27,28
not directly coupled.
More work is needed to under-
stand the structure-activity relationships of AMP signal-
ing and the exact binding site of the peptides, but it is
tempting to speculate that DYN binds at the predicted
negatively charged periplasmic loops of ParS, in a similar
manner to peptide binding to sensor PhoQ in
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
3
5
Salmonella. The mechanism of mild bacterial growth
inhibition by DYN also remains to be explored, but may
be due to stress induced by membrane binding.
(
Figure S10). Furthermore, although parR and parS mu-
tants were more susceptible to DYN treatment (Figure
6b), other strains from the library, such as a strain mutant
in the unrelated response regulator phoP, showed a simi-
lar growth and phenotypic response to the wt (Figure S9).
Together these data strongly link ParS to the AMP-
defense response of Pseudomonas to DYN.
Treatment of PA with Dyn-A has been previously linked
to the PQS (Pseudomonas Quinolone Signal) quorum
4
sensing (QS) system. AMPs, such as colistin, also induce
36
expression of QS genes, and ParRS has been linked to
37
modest changes in quorum sensing pathways in PA.
However, we observed no upregulation of any QS pro-
teins in our proteomic analyses. This could be due to the
timing of analysis, or differences in the culture medium,
which is a key determinate in the phenotypic response.
Our proteomic data also do not readily explain the in-
crease in pyocyanin production observed in response to
DYN; of all the proteins involved in the synthesis of this
toxin, only PhzB2 was upregulated. Again, this could re-
flect timing of analysis. How ParRS signaling may be
linked to the circuitry for enhanced pyocyanin production
is unclear, and it is possible that other sensors and path-
ways are involved in this aspect of the response. The other
chemical proteomics hits identified here (Figure 4d; e.g.
putative sensor/response regulator PA1243) are interest-
ing starting points for further study.
Discussion
Here we use chemical probes to identify a membrane sen-
sor kinase, ParS, as a binding partner of the human pep-
tide dynorphin and the mediator of a highly specific bac-
terial defense response.
Dynorphin(1-17) and the shorter (1-13) analogue (DYN)
both induced production of the virulence factor pyocya-
nin. Interestingly this phenotype was highly medium-
dependent, indicating that PA integrates multiple envi-
ronmental signals in its response to these compounds.
Pyocyanin induction was also peptide sequence-
dependent; however, binding of DYN-based photoaffinity
probes to the LPS was subsequently shown to be se-
quence-independent, suggesting that LPS interaction is
not responsible for induction of virulence but may be im-
portant for initial charge-mediated coordination with the
cell membrane. Photoaffinity probes with different photo-
reactive groups exhibited very distinct profiles, and the
We also observed DYN-induced upregulation of drug ef-
flux pumps. A previous study showed that, unlike wt bac-
teria, a parRS mutant did not upregulate MexXY and
downregulate the porin OprD in response to AMPs, and
that spontaneous mutants in parR & S had heightened
lack of specificity exhibited by both benzophenones and
aryl azides, as noted recently by us and others,
to focus on diazirine-containing probes. We identified the
membrane protein ParS, the sensing partner of a two-
component response system, as a promising hit.
17,19
3
2
led us
intrinsic resistance to diverse antibiotics. This may be
clinically relevant since the AMP colistin and other anti-
32,38
biotics are given sequentially to cystic fibrosis patients.
The present study thus adds to the increasing body of
evidence linking virulence and antibiotic resistance in
Pseudomonas, and suggests that host factors can trigger a
multi-resistant phenotype in this pathogen.
Subsequent in-depth global proteome analyses of bacteria
treated with DYN at different growth stages revealed a
highly specific, parS-dependent, response: a dramatic
increase in the levels of proteins that boost membrane
defense mechanisms against cationic peptides. Upregu-
lated proteins included those in the arn operon that mod-
ify lipid A, reducing the negative charge of the membrane
and thereby increasing AMP resistance. Such modifica-
tions also alter outer membrane permeability and host
immune recognition of LPS, and are common in clinical
Conclusion
In summary, we reveal that a human opioid peptide in-
duces a specific defense mechanism in P. aeruginosa via
direct binding of the sensor ParS. We provide strong evi-
dence that the antimicrobial peptide (AMP) response is
mediated by ParS. Concentrations of up to ~0.4 µM Dyn-
4
A have been reported in the PA-infected, injured gut – in
3
4
isolates of PA from cystic fibrosis patients. These data
are consistent with previous reports of parRS-dependent
arn upregulation and adaptive AMP resistance. Previous
the same range as DYN concentrations that induced a
AMP response – indicating that this effect could be bio-
logically relevant in vivo. Our data thus supports the in-
triguing possibility that Pseudomonas can specifically
2
4
work showed that ParRS and a second two-component
ACS Paragon Plus Environment

Page 15 of 17
Journal of the American Chemical Society
recognize human stress hormones and initiate mecha-
(8)
Wu, L.; Estrada, O.; Zaborina, O.; Bains, M.;
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Shen, L.; Kohler, J. E.; Patel, N.; Musch, M. W.; Chang, E. B.; Fu,
Y. X.; Jacobs, M. A.; Nishimura, M. I.; Hancock, R. E.; Turner, J.
R.; Alverdy, J. C. Science 2005, 309, 774.
nisms to defend itself against the hostile environment of
the host. This work highlights ParS as a promising drug
target for blocking virulence induction and for sensitizing
bacteria to antimicrobial therapies.
(
9)
Boedeker, E. C.; Sperandio, V. Proc Natl Acad Sci U S A 2006, 103,
0420.
Clarke, M. B.; Hughes, D. T.; Zhu, C.;
1
ASSOCIATED CONTENT
(
10)
Chavkin, C.; Goldstein, A. Proc Natl Acad Sci
Wright, M. H.; Sieber, S. A. Nat Prod Rep
Lapinsky, D. J. Bioorg Med Chem 2012, 20,
Geurink, P. P.; Prely, L. M.; van der Marel, G.
Supporting Information. Detailed synthetic and experi-
mental procedures, analytical (NMR, LCMS) characterization
of compounds and peptides. Supporting data: Figures S1-S10,
Tables S1-S5 (separate Excel files) and S6. This material is
available free of charge via the Internet at
http://pubs.acs.org.
U S A 1981, 78, 6543.
(11)
2016, 33, 681.
(12)
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
6237.
(13)
A.; Bischoff, R.; Overkleeft, H. S. Top Curr Chem 2012, 324, 85.
14) Srinivas, N.; Jetter, P.; Ueberbacher, B. J.;
(
Proteomics datasets. The mass spectrometry proteomics
Werneburg, M.; Zerbe, K.; Steinmann, J.; Van der Meijden, B.;
Bernardini, F.; Lederer, A.; Dias, R. L.; Misson, P. E.; Henze, H.;
Zumbrunn, J.; Gombert, F. O.; Obrecht, D.; Hunziker, P.;
Schauer, S.; Ziegler, U.; Kach, A.; Eberl, L.; Riedel, K.; DeMarco,
S. J.; Robinson, J. A. Science 2010, 327, 1010.
data have been deposited to the ProteomeXchange Consorti-
3
9
um via the PRIDE partner repository with the dataset iden-
tifier PXD006172.
AUTHOR INFORMATION
(
15)
Microbiol 2013, 11, 467.
16) Boersema, P. J.; Raijmakers, R.; Lemeer, S.;
Mohammed, S.; Heck, A. J. Nat Protoc 2009, 4, 484.
17) Park, J.; Koh, M.; Koo, J. Y.; Lee, S.; Park, S. B.
ACS Chem Biol 2016, 11, 44.
18) Park, H.; Koo, J. Y.; Srikanth, Y. V.; Oh, S.;
Lee, J.; Park, J.; Park, S. B. Chem Commun (Camb) 2016, 52, 5828.
(19) Kleiner, P.; Heydenreuter, W.; Stahl, M.;
Korotkov, V. S.; Sieber, S. A. Angew Chem Int Ed Engl 2017, 56,
396.
Needham, B. D.; Trent, M. S. Nat Rev
Corresponding Author
(
* stephan.sieber@mytum.de; M.H.Wright@leeds.ac.uk.
(
Present Addresses
†
School of Chemistry, University of Leeds, Leeds, LS2 9JT,
(
UK.
Author Contributions
All authors have given approval to the final version of the
manuscript.
1
(
20)
Nagaraj, N.; Mann, M. Mol Cell Proteomics 2014, 13, 2513.
21) Hatzenpichler, R.; Scheller, S.; Tavormina, P.
Cox, J.; Hein, M. Y.; Luber, C. A.; Paron, I.;
ACKNOWLEDGMENT
(
We gratefully acknowledge funding from the European Un-
ion’s Seventh Framework Programme under grant agreement
PIEF-GA-2013-625528, and the Center for Integrated Protein
Science Munich (CIPSM). We thank the Manoil lab (Univer-
sity of Washington) for transposon mutants from the Pseu-
domonas aeruginosa PAO1 transposon mutant library (grant
# NIH P30 DK089507). We also thank Katja Bäuml and Nina
Bach for mass spectrometry assistance, Mona Wolff for tech-
nical assistance, and Angela Weigert Muños, Mert Malay,
and Thomas Löhr for synthetic assistance (internships). We
thank Annabelle Hoegl and Matthias Stahl in the Sieber
group for critical reading of this manuscript.
L.; Babin, B. M.; Tirrell, D. A.; Orphan, V. J. Environ Microbiol
2014, 16, 2568.
(22)
Dieterich, D. C. Methods Mol Biol 2015, 1266, 199.
23) Winsor, G. L.; Griffiths, E. J.; Lo, R.; Dhillon,
Landgraf, P.; Antileo, E. R.; Schuman, E. M.;
(
B. K.; Shay, J. A.; Brinkman, F. S. Nucleic Acids Res 2016, 44,
D646.
(24)
Fernandez, L.; Gooderham, W. J.; Bains, M.;
McPhee, J. B.; Wiegand, I.; Hancock, R. E. Antimicrob Agents
Chemother 2010, 54, 3372.
(25)
Brogden, K. A. Nat Rev Microbiol 2005, 3,
238.
(26)
Wisniewski, J. R.; Zougman, A.; Nagaraj, N.;
Mann, M. Nat Methods 2009, 6, 359.
27) McPhee, J. B.; Lewenza, S.; Hancock, R. E.
Mol Microbiol 2003, 50, 205.
REFERENCES
(
(1)
Chavkin, C. Mol Pharmacol 2013, 83, 729.
(2)
Sobczak, M.; Salaga, M.; Storr, M. A.; Fichna,
(28)
Fernandez, L.; Jenssen, H.; Bains, M.;
J. J Gastroenterol 2014, 49, 24.
3) Wu, L. R.; Zaborina, O.; Zaborin, A.; Chang,
Wiegand, I.; Gooderham, W. J.; Hancock, R. E. Antimicrob
Agents Chemother 2012, 56, 6212.
(
E. B.; Musch, M.; Holbrook, C.; Turner, J. R.; Alverdy, J. C. Surg
Infect (Larchmt) 2005, 6, 185.
(29)
Johnson, L.; Mulcahy, H.; Kanevets, U.; Shi,
Y.; Lewenza, S. J Bacteriol 2012, 194, 813.
(4)
Zaborina, O.; Lepine, F.; Xiao, G.; Valuckaite,
(30)
Vargiu, A. V.; Pos, K. M.; Poole, K.; Nikaido,
V.; Chen, Y.; Li, T.; Ciancio, M.; Zaborin, A.; Petrof, E. O.; Turner,
J. R.; Rahme, L. G.; Chang, E.; Alverdy, J. C. PLoS Pathog 2007, 3,
e35.
H. Front Microbiol 2016, 7, 833.
(31)
Sun, J.; Deng, Z.; Yan, A. Biochem Biophys
Res Commun 2014, 453, 254.
(5)
Saiman, L. Paediatr Respir Rev 2004, 5 Suppl
(32)
Muller, C.; Plesiat, P.; Jeannot, K. Antimicrob
A, S367.
Agents Chemother 2011, 55, 1211.
(6)
Balasubramanian, D.; Schneper, L.; Kumari,
(33)
Diggle, S. P.; Stacey, R. E.; Dodd, C.; Camara,
H.; Mathee, K. Nucleic Acids Res 2013, 41, 1.
M.; Williams, P.; Winzer, K. Environ Microbiol 2006, 8, 1095.
(7)
Zaborin, A.; Gerdes, S.; Holbrook, C.; Liu, D.
(34)
Ernst, R. K.; Yi, E. C.; Guo, L.; Lim, K. B.;
C.; Zaborina, O. Y.; Alverdy, J. C. PLoS One 2012, 7, e34883.
Burns, J. L.; Hackett, M.; Miller, S. I. Science 1999, 286, 1561.
ACS Paragon Plus Environment

Journal of the American Chemical Society
Page 16 of 17
(35)
Bader, M. W.; Sanowar, S.; Daley, M. E.;
1
2
3
4
5
6
7
8
9
1
1
1
1
1
1
1
1
1
1
2
2
2
2
2
2
2
2
2
2
3
3
3
3
3
3
3
3
3
3
4
4
4
4
4
4
4
4
4
4
5
5
5
5
5
5
5
5
5
5
6
Schneider, A. R.; Cho, U.; Xu, W.; Klevit, R. E.; Le Moual, H.;
Miller, S. I. Cell 2005, 122, 461.
(
36)
J.; O'Gara, F. Microbiology 2009, 155, 2826.
37) Wang, D.; Seeve, C.; Pierson, L. S., 3rd;
Pierson, E. A. BMC Genomics 2013, 14, 618.
38) Flume, P. A.; O'Sullivan, B. P.; Robinson, K.
Cummins, J.; Reen, F. J.; Baysse, C.; Mooij, M.
(
(
A.; Goss, C. H.; Mogayzel, P. J., Jr.; Willey-Courand, D. B.; Bujan,
J.; Finder, J.; Lester, M.; Quittell, L.; Rosenblatt, R.; Vender, R. L.;
Hazle, L.; Sabadosa, K.; Marshall, B.; Cystic Fibrosis Foundation,
P. T. C. Am J Respir Crit Care Med 2007, 176, 957.
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
1
2
3
4
5
6
7
8
9
0
(39)
Vizcaino, J. A.; Csordas, A.; Del-Toro, N.;
Dianes, J. A.; Griss, J.; Lavidas, I.; Mayer, G.; Perez-Riverol, Y.;
Reisinger, F.; Ternent, T.; Xu, Q. W.; Wang, R.; Hermjakob, H.
Nucleic Acids Res 2016, 44, 11033.
ACS Paragon Plus Environment

Page 17 of 17
Journal of the American Chemical Society
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
11
ACS Paragon Plus Environment