Citation:
Kim, J.H.; Dong, J.; Le, B.H.;
Lonergan, Z.R.; Gu, W.; Girke, T.;
Zhang, W.; Newman, D.K.; Martins-
Green, M.
Pseudomonas aeruginosa
Activates Quorum Sensing,
Antioxidant Enzymes and Type VI
Secretion in Response to Oxidative
Stress to Initiate Biofilm Formation
and Wound Chronicity.
Antioxidants
2024
,
13
, 655. https://doi.org/
10.3390/antiox13060655
Academic Editor: Filippo Ren
ò
Received: 4 April 2024
Revised: 29 April 2024
Accepted: 13 May 2024
Published: 27 May 2024
Copyright:
©
2024 by the authors.
Licensee MDPI, Basel, Switzerland.
This article is an open access article
distributed under the terms and
conditions of the Creative Commons
Attribution (CC BY) license (https://
creativecommons.org/licenses/by/
4.0/).
antioxidants
Article
Pseudomonas aeruginosa
Activates Quorum Sensing, Antioxidant
Enzymes and Type VI Secretion in Response to Oxidative Stress
to Initiate Biofilm Formation and Wound Chronicity
Jane H. Kim
1
, Julianna Dong
1
, Brandon H. Le
2,3
, Zachery R. Lonergan
4
, Weifeng Gu
1
, Thomas Girke
2,3
,
Wei Zhang
2,3
, Dianne K. Newman
4,5
and Manuela Martins-Green
1,
*
1
Department of Molecular, Cell and Systems Biology, University of California, Riverside, CA 92521, USA
2
Institute for Integrative Genome Biology, University of California, Riverside, CA 92521, USA
3
Department of Botany and Plant Sciences, University of California, Riverside, CA 92521, USA
4
Division of Biology and Biological Engineering, California Institute of Technology, Pasadena, CA 91125, USA
5
Geological and Planetary Sciences, California Institute of Technology, Pasadena, CA 91125, USA
*
Correspondence: manuela.martins@ucr.edu
Abstract:
Pseudomonas aeruginosa
(
PA
) is an opportunistic pathogen frequently isolated from cutaneous
chronic wounds. How
PA
, in the presence of oxidative stress (OS), colonizes chronic wounds and
forms a biofilm is still unknown. The purpose of this study is to investigate the changes in gene
expression seen when PA is challenged with the high levels of OS present in chronic wounds. We
used a biofilm-forming
PA
strain isolated from the chronic wounds of our murine model (RPA)
and performed a qPCR to obtain gene expression patterns as RPA developed a biofilm
in vitro
in
the presence of high levels of OS, and then compared the findings
in vivo
, in our mouse model of
chronic wounds. We found that the planktonic bacteria under OS conditions overexpressed quorum
sensing genes that are important for the bacteria to communicate with each other, antioxidant
stress genes important to reduce OS in the microenvironment for survival, biofilm formation genes
and virulence genes. Additionally, we performed RNAseq
in vivo
and identified the activation of
novel genes/pathways of the Type VI Secretion System (T6SS) involved in RPA pathogenicity. In
conclusion, RPA appears to survive the high OS microenvironment in chronic wounds and colonizes
these wounds by turning on virulence, biofilm-forming and survival genes. These findings reveal
pathways that may be promising targets for new therapies aimed at disrupting
PA
-containing biofilms
immediately after debridement to facilitate the treatment of chronic human wounds.
Keywords:
anaerobic respiration; bacteria; antioxidant enzymes; RNAseq; chronic wounds
1. Introduction
Cutaneous wound healing involves not only the cells of the skin and immune cells but
also the microbes present in the microbiota of the skin [
1
–
4
], which are crucial for proper
wound healing; a sterile wound does not lead to a better healing outcome. Skin microbes
play an important role in chemoattracting circulating neutrophils and monocytes to infil-
trate the wounded tissue. Neutrophils release free oxygen species, cytotoxic enzymes, and
proteases into the wound microenvironment and kill pathogens [
5
,
6
]. They are followed by
monocytes that, when reaching the tissue, differentiate into pro-inflammatory macrophages
which control microbial invasion and remove dead neutrophils and other cells. Subsequent
to this phase of healing, pro-healing macrophages initiate the repair of the tissue by regu-
lating inflammation, stimulating fibroblast proliferation, angiogenesis and keratinocytes’
proliferation and migration, to close the wound [
7
,
8
]. When chronic wounds develop,
the tissue becomes ischemic, necrotic and hypoxic, providing a microenvironment that is
conducive for the colonization and proliferation of potential pathogenic microbes. This also
Antioxidants
2024
,
13
, 655. https://doi.org/10.3390/antiox13060655
https://www.mdpi.com/journal/antioxidants
Antioxidants
2024
,
13
, 655
2 of 16
leads to chronic inflammation. It is in this destructive microenvironment that opportunistic
pathogens take advantage of host nutrients and molecules to form a biofilm [9–13].
Bacterial biofilms can harbor a multitude of aerobic and anaerobic bacterial species
from across the phylogenetic tree. Each species in the biofilm differs significantly from its
planktonic counterpart in its morphology, mode of communication and metabolism [
14
–
16
].
The biofilm provides a unique environment for cell signaling through the production of
quorum sensing molecules, which promote collective behavior such as the optimization of
nutrients and acquisition and regulation of virulence, leading to sustained pathogenicity in
the wound [
12
,
17
]. Further research on microbial biofilms is needed because the bacteria
in these biofilms are becoming more resistant to conventional antibiotic therapies. They
are also difficult to disassemble and deconstruct due to the complicated structure of the
extracellular polymeric substances (EPSs) that comprise the biofilm matrix, which include
extracellular DNA, proteins/peptides, carbohydrates and lipids [18–20].
Of the many bacterial species found to colonize chronic wounds in humans,
Pseudomonas aeruginosa
(
PA
) is a very common species that is difficult to control once
it colonizes chronic wounds, because many strains of
PA
have already developed a resis-
tance to antibiotics [
21
–
24
].
PA
can form impenetrable biofilms that are difficult to remove
even with the wide range of therapeutics available to clinicians. In addition,
PA
poses
a significant health risk to patients because its genome contains a large arsenal of viru-
lence factors that enhance pathogenesis [
25
–
27
]. Quorum sensing (QS) is an integral part
of the pathogenesis of
PA
because it allows
PA
to adapt to diverse microenvironments.
Two gene tandems are important for QS:
lasR/lasI
and
rhlR/rhlI
[
28
]. The transcription
of these genes is highly expressed in their early stationary phase and their products can
activate the expression of many other genes, including those involved in virulence.
PA
can survive in both aerobic and anerobic conditions, but its most efficient growth
occurs during aerobic respiration [
29
,
30
]. Due to this fast metabolism, the levels of its
reactive oxygen species (ROS), e.g., superoxide (O
2
·−
) and H
2
O
2
, which are dangerous
by-products of aerobic respiration, must be controlled.
PA
has two superoxide dismutase
(SOD) enzymes with either manganese or iron co-factors [
29
]. When iron levels are high,
sodB
expression, which encodes for the enzyme Fe-SOD, is high; otherwise,
sodA
expression,
which encodes for the enzyme Mn-SOD, is activated. Both enzymes dismutate the highly
damaging radical O
2
·−
to H
2
O
2
, which is less damaging.
PA
also expresses catalase, which
is encoded by the two genes
katA
and
katB
, which breakdown H
2
O
2
into H
2
O and O
2
[
29
].
Chronic wounds in humans are a very serious condition; they have been considered by
many as a silent epidemic that affects a large fraction of the world population [
31
]. Indeed,
chronic wounds impact ~8.5 M people and cost ~USD 30 B/year in the US alone [
29
]. Many
people with chronic wounds have other comorbidities, one of which is diabetes. Over
30 M people live with diabetes in the United States and it is projected that there will be
60.6 M diabetic Americans by 2060 [
29
]. Diabetic foot ulcers (DFUs) account for 25–50%
of all diabetes-related hospital costs in the US, totaling billions of dollars per year [
32
].
PA
is a large contributor to these wounds staying chronic because
PA
biofilms readily
return after debridement. Therefore, it is important to understand how
PA
responds to the
high oxidative stress environments in these wounds to form biofilms. Using both
in vitro
and
in vivo
approaches, we found that, in response to high levels of oxidative stress,
PA
expresses global transcriptional activators, antioxidant enzymes and SOD to break down
ROS and expresses several genes that contribute to the formation of a biofilm. A bacterial
transcriptomic analysis of chronic wounds showed that the Type VI Secretion System (T6SS)
was significantly upregulated in chronic wounds. These results point to the potential of
developing new approaches to target the
PA
genes contributing to its survival
in vivo
as
a biofilm.
Antioxidants
2024
,
13
, 655
3 of 16
2. Materials and Methods
2.1. Pseudomonas aeruginosa’s Isolation, Growth and Treatment In Vitro
All experiments used Lysogeny Broth (LB; Difco, Franklin Lakes, NJ, USA) [
33
] as
the growth medium. Aerobic liquid cultures were incubated at 37
◦
C and shaken at
220 RPM. For our studies, we used a
Pseudomonas aeruginosa
(
PA
) strain RPA (Riverside
PA
) that was previously isolated from the chronic wounds of our
db
/
db
mouse model of
chronic wounds [
34
,
35
]. Briefly, the identity of RPA was confirmed after PCR amplification,
cloning and Sanger sequencing. The
in vitro
experiments were conducted in sterile 96-well
plates, with absorbance at OD
600
used to measure the concentration of the bacteria colony
forming units per ml (CFU/mL). Serial dilutions were made to achieve a concentration of
10
6
CFU/200
μ
L and pipetted into each well. The 96-well plates were incubated at 37
◦
C
without shaking. A bacterial treatment with hydrogen peroxide (H
2
O
2
) was performed
by delivering 20
μ
L of a higher concentration of H
2
O
2
to achieve a final concentration of
250
μ
M, 500
μ
M or 1000
μ
M of H
2
O
2
in wells containing 200
μ
L of bacteria in LB. For
the
in vitro
experiments, biofilm samples were not washed and contained both planktonic
and attached cells. For
in vivo
experiments, absorbance at OD
600
was used to measure
the concentration of the bacteria to achieve a concentration of 10
7
CFU/50
μ
L, which was
applied on top of the wound 24 h after wounding.
2.2. Chronic Wounds Caused by Infection with RPA
All experiments were completed in accordance and compliance with federal regu-
lations and the University of California’s policy and procedures. Animal experimental
protocol no. 11 was approved in July 2023 by the University of California Riverside (UCR)
Institutional Animal Care and Use Committee (UCR-IACUC). The description of how to
obtain chronic wounds in
db
/
db
mice has been published in detail by us
previously [35–37]
.
Briefly,
db
/
db
mice were bred in our conventional vivarium from B6.BKS(D)-
Lepr
db
/J het-
erozygotes obtained from the Jackson Laboratories (Stock no. 000697). Only
db
/
db
that are
at least 5 months old were used to create chronic wounds, because at this age they are fully
diabetic and obese. Twenty min prior to performing the surgical wounds, 3-amino-1,2,4-
triazole (ATZ), an inhibitor of catalase, was injected intraperitoneally at 0.75 g ATZ/kg
of mouse weight in sterile PBS. The skin of the mouse was wiped and disinfected with
iodine and 70% ethanol immediately prior to surgery to remove as much of the natural
skin microbiota as possible. As soon as a 7 mm full thickness skin excision wound was
made, it was covered with sterile Tegaderm, which provides a barrier to external contami-
nants and environmental bacteria from the cage and bedding. Using an insulin syringe,
mercaptosuccinic acid (MSA), at 150 mg MSA/kg of mouse weight in sterile PBS, was
deposited on top of the wound to inhibit glutathione peroxidase. All mice were treated
for pain intraperitoneally with 0.05 mg buprenorphine/kg of mouse weight in sterile PBS
before surgery and 6 h after surgery, and after that as needed.
2.3. RNA Collection from Mouse Wounds
RNA from the bacteria was collected from wounds at 24, 48, and 72 h after wounding,
from both healing and chronic wounds. Briefly, 100
μ
L of sterile nuclease-free water
was injected into the wound bed with an insulin syringe piercing through the Tegaderm.
Immediately, 100
μ
L of exudate mixed with the sterile nuclease-free water was extracted
with the same insulin syringe. Samples were stored in
−
80
◦
C before extraction with
RNeasy Mini Kit Ref. 74106 (Qiagen, Venlo, The Netherlands).
2.4. RNA Extraction, DNase1 Treatment and RNA Cleanup
The total RNA extracted from the RPA strains was used for PCR and RT-qPCR amplifi-
cation. RNA was extracted using the RNeasy Mini Kit Ref. 74106 (Qiagen) after the bacteria
were sonicated for 30 sec in RLT buffer. The extracted total RNA was incubated for 3 h in
DNaseI (Qiagen) and subsequently cleaned with vacuum grease and isopropyl alcohol.
Briefly, the DNA-free RNA was added to a microcentrifuge tube with a vacuum grease
Antioxidants
2024
,
13
, 655
4 of 16
phase lock. One volume of cold acid phenol was added, and the tube was hand-shaken to
mix the water and acid phenol to obtain an emulsion. Samples were centrifuged for 4 min
at 12,000 RPM. The aqueous phase was transferred to a new tube and combined with
1.5
μ
L
of 15 mg/mL glycogen, 0.1X volume of 3 M sodium acetate, pH 5.2 and
1.2X volume
of
room-temperature isopropyl alcohol. The tube was mixed by inversion, not vortexed. Sam-
ples were incubated for 30 min at
−
20
◦
C or for 10 min at
−
80
◦
C before being centrifuged
for 15 min at 12,000 RPM and 4
◦
C. The supernatant was removed while being careful
not to disturb the white RNA pellet at the bottom of the tube. The pellet was rinsed with
500
μ
L of cold 75% ethanol and allowed to dry before dissolving the RNA with ultrapure
nuclease-free water.
2.5. Reverse Transcription and Quantitative Polymerase Chain Reaction
The reverse transcriptase to obtain cDNA was conducted with the Superscript
™
IV
First-Strand Synthesis System (Invitrogen, Waltham, MA, USA). The primers used for the
reaction are listed in Table 1. For RT-qPCR amplification, the total reaction volume was
20
μ
L, including 10
μ
L of SYBR Green (Biorad, Biotech, Dalian, China), 1
μ
L each of the
forward and reverse primers (10
μ
M), 7
μ
L of sterile water, and 1
μ
L of the purified bacterial
cDNA as a template. A Biorad CFX Connect System (Roche, Switzerland) was used for
thermal cycling, as follows: an initial denaturation of DNA at 95
◦
C for 30 s, followed by
40 cycles of denaturation at 95
◦
C for 5 s and annealing at 55
◦
C for 60 s. The qPCR assay
was performed in duplicates with parallel analysis, in 96-well plates. Sterile water was used
in place of the DNA template as a negative control to ensure the absence of contaminants.
Log
2
FC was calculated using
proC
as the stable housekeeping gene [38].
Table 1.
List of Primers used for RT-qPCR.
Gene
Forward Primer
Reverse Primer
Gene Accession Number
sodA
CAACCACTCGCTGTTCTGGA
CTTGGTGAACGCATCCTTGAAC
PA4468
sodB
AACACCTACGTGGTGAACCTGA
TGACGATCTCTTCGAGGCTCTT
PA4366
katA
GAACAGCTTCAACCAGTGGCAG
CTCGTCGGTGAACAGATGGAAC
PA4236
katB
CGCTTCGATTTCTTCTCCCACG
CTTGTAGGCATGCACGCTGTTG
PA4613
lasI
GCCCCTACATGCTGAAGAACAC
CCTCCAGCGTACAGTCGGAAAA
PA1432
lasR
ATGGCCTTGGTTGACGGTTTTC
CCTAAGGACAGCCAGGACTACG
PA1430
rhlR
CCTCGGAAATGGTGGTCTGGAG
GGAAAGCACGCTGAGCAAATTG
PA3477
rhlI
CGACCAGGAATTCGACCAGTTC
GTTTCGCTGCACAGGTAGGC
PA3476
pelA
AGTACTACGCGCCGATCATCAA
AAGTGGTAGTACAGGTGCAGGC
PA3064
pelD
TGCCTGTATGCCTTCGAGTTGA
GAAGTCAGCGGCAACAACACC
PA3061
pslA
GCTACAACAACCGGCTGATCTG
GATGCTGGTCTTGCGGATGAAG
PA2231
pslB
CCTCAACACCAACGAATCCACC
CGTAGATGTCGTTGAAGCGGAC
PA2232
recA
TCACCGGCAATATCAAGAACGC
GACCGAGGCGTAGAACTTCAGT
PA3617
lexA
GCGAGGAGGTCACGGTGAAA
GCCTTCGATGATCAGTTCCTGC
PA3007
proC
CAGGCCGGGCAGTTGCTGTC
GGTCAGGCGCGAGGCTGTCT
PA0393
oxyR
CCGCTGTACATCGAGGAGAACT
ACATAGAAGGGCTCGTCGAACA
PA5344
phzA1&2
GACCGAGGATCCGAACCACTTC
CGTTTTATCCGGCCGTTCTCG
PA4210, PA1899
phzM
GTGGCCTTCGAGATCTTCCAGG
GGAACTCCTCGCCGTAGAACAG
PA4209
Antioxidants
2024
,
13
, 655
5 of 16
2.6. RNA Sequencing
The integrity of the DNase-treated RNA samples was assessed by a Bioanalyzer (Ag-
ilent, Hong Kong, China). All samples had RIN (RNA integrity number) values
≥
5.0.
Ribosomal transcripts were removed using the Illumina
®
Stranded Total RNA Prep, Liga-
tion with Ribo-Zero Plus (Illumina, cat no. 20040525, San Diego, CA, USA). Following the
manufacturer ’s instructions, 500 ng of total, DNase-treated RNA was used as the input for
the ribo-depletion procedure. RNA samples were then barcoded with IDT
®
for Illumina
®
RNA UD Indexes Set A Ligation (96 Indexes, 96 Samples) (Illumina, cat no. 20040553). The
quality of the barcoded libraries was assessed with a Bioanalyzer/Tapestation and pooled
equilmolar before submission. Libraries were sequenced with Illumina NextSeq 2000, using
P2 100-cycle kit (100 bp, Single End) for up to 4000 M reads with 20–25 M reads per sample.
2.7. Sequencing Genomic DNA and Annotations
Sequencing services were provided by SeqCoast Genomics, LLC (Portsmouth, NH,
USA). The genomic DNA of RPA was sequenced first using the Illumina Sequencing Meth-
ods. Briefly, RPA sample libraries were prepared using the Illumina DNA Prep kit and
IDT 10 bp UDI indices and sequenced on an Illumina NextSeq 2000, producing
2
×
151 bp
reads. Demultiplexing, quality control and adapter trimming were performed with Illu-
mina bcl-convert (v3.9.3). We then used the Oxford Nanopore Sequencing Method. Briefly,
sample libraries were prepared using Oxford Nanopore Technologies (ONT) Ligation Se-
quencing Kit (SQK-LSK109), with NEBNext
®
Companion Module (E7180L) in addition to
Native Barcode Kits (EXP-NBD104, EXP-NBD114), to the manufacturer’s specifications.
All samples were run on Nanopore R9.4.1 flow cells and a MinION Mk1B device. Post-
sequencing, Guppy (v5.0.16) was used for high-accuracy base calling (HAC) and demulti-
plexing. Genome hybrid assembly was performed using Unicycler (version 0.4.4) [
39
], with
error correction and contig assembly using SPAdes [
40
,
41
], mapping using Bowtie2 [
42
]
and variant identification and assembly polishing with Pilot [
43
], which produced a single
chromosomal contig of 6,839,924 bp and second plasmid contig of
7127 bp
. Gene prediction
and functional annotation were performed with BAKTA (version 1.5.1) [43–58].
2.8. Bioinformatic Analysis
To analyze the results, raw data were processed for quality control using Fastqc [
59
].
Low-quality bases and adapter sequences were removed using trim_galore (v0.6.7) [
60
].
High-quality trimmed reads were aligned to the RPA reference genome using the STAR
aligner (v.2.7.9a) [
61
] with parameters –outFilterMismatchNmax 0 and –outFilterMultimap-
Nmax 5 to select reads with a perfect match and that, at most, mapped to five locations.
QC data were summarized using MultiQC (v1.16) [
62
]. Approximately 21–30 million reads
were obtained per library, with the majority of the reads remaining after adapter trimming
with Cutadapt. Fastq-screen (v0.15.2) [
63
] was used to sample each library (
1 M reads
)
and quickly align them to known references including humans, mouse,
E. coli
,
Arabidopsis
,
vectors, adapters, etc. This analysis provided a quick assessment of the extent of the contam-
inants in the libraries. Sortmerna (4.3.6) [
64
] was used to determine the extent of the rRNA
contamination within each library. The reads were aligned against the rRNA references
from the SILVA database. Featurecount (v2.0.3) [
65
] quantified the mapped reads with a
genome annotation to generate a count matrix for each transcript/gene. The current fea-
turecount setting only accepts uniquely mapped reads and ignores multiple-mapped reads.
A differential expression analysis was performed using DESeq2 (1.38.0) [
66
]. Statistical
significance was determined using log
2
FC > 1 and an adjusted
p
-value (FDR) < 0.1.
3. Results
3.1.
Differential Gene Expression In Vitro between Biofilm-Forming PA and PA Already in
a Biofilm
Biofilm-forming RPA was used to understand the response of
PA
to OS as it develops
a biofilm in chronic wounds. RPA is a strain isolated from the biofilm of a fully chronic
mouse wound that is very effective at forming a biofilm
in vivo
[
34
], and initial experiments
Antioxidants
2024
,
13
, 655
6 of 16
showed that this strain also grows and produces a strong biofilm after 24 h under stressful
conditions
in vitro
(incubation at 37
◦
C without shaking). To understand how RPA develops
a biofilm, we first performed experiments
in vitro
in response to stress by investigating the
transcriptional response of RPA to the OS created by H
2
O
2
and compared the expression of
genes known to be important for the response to stress and biofilm formation. Specifically,
we compared the gene expression between planktonic bacteria as they develop into a
biofilm (0–24 h) and their gene expression once they are already in a biofilm (24–48 h).
RPA was grown in sterile 96-well plates and treated with 250
μ
M, 500
μ
M or
1000
μ
M
of
H
2
O
2
with a vehicle (sterile water) as control, to simulate bacterial exposure to low, medium,
and high levels of OS. One set of cultures was treated from time zero of culture (planktonic)
and the other allowed to form a biofilm for 24 h before treatments (Figure 1). Significant
differences in gene expression were found between planktonic RPA cells growing to form a
biofilm in response to OS and RPA already living in a biofilm culture when exposed to OS
(Figure 2). In RPA planktonic cells, all QS genes,
lasI
,
lasR
,
rhlI
and
rhlR
, were upregulated
(Figure 2A). The expression of
oxyR
, a gene that functions in OS defense and DNA repair,
was also significantly upregulated. In contrast, these genes were downregulated when
RPA was already in its biofilm form when exposed to H
2
O
2
(Figure 2A). In particular, we
observed that the expression of
lasI
,
lasR
,
rhlI
,
rhlR
and
oxyR,
when the bacteria were in a
biofilm, was significantly downregulated with the 500
μ
M H
2
O
2
treatment. The lack of the
same strong downregulation when the bacteria were treated with 1000
μ
M could be due to
these high levels of H
2
O
2
causing toxicity to the bacteria.
PA
has two SOD enzymes:
sodB,
which encodes for Fe-SOD, whose expression is high
when iron is high; otherwise,
sodA
, which encodes for Mn-SOD, is highly expressed. Both
enzymes dismutate the highly damaging radical O
2
·−
to H
2
O
2
, which is less damaging.
PA
also has two catalase genes that encode KatA and KatB, which breakdown H
2
O
2
into
H
2
O + O
2
[29]
. We found that, under planktonic conditions,
sodA
,
sodB
and
katA
are
all highly upregulated, whereas, when in a biofilm, RPA does not activate these genes
(Figure 2B).
When the genes responsible for Pel and Psl synthesis,
pelA
and
pelD
and
pslA
and
pslB
,
respectively, were studied, we observed that, again, under planktonic conditions, these
genes were upregulated, whereas, when the RPA was already in its biofilm, the same genes
were downregulated, except for
pslB
, which was slightly upregulated (Figure 2C). Similar
observations were made with the pathogenic genes,
phzA
,
phzB
,
lexA
and
rpsL
(Figure 2D).
Antioxidants
2024
,
13
, x FOR PEER REVIEW
6 of 16
3. Results
3.1. Differential Gene Expressi
on In Vitro between Biofilm-For
ming PA and PA Already in a
Biofilm
Biofilm-forming RPA was used to understand the response of
PA
to OS as it develops
a biofilm in chronic wounds. RPA is a strain isolated from the biofilm of a fully chronic
mouse wound that is very effective at forming a biofilm in vivo [34]
,
and initial experi-
ments showed that this strain also grows an
d produces a strong biofilm after 24 h under
stressful conditions in vitro (incubation at 37 °C without shaking). To understand how
RPA develops a biofilm, we first performed experiments in vitro in response to stress by
investigating the transcriptional response of RPA to the OS created by H
2
O
2
and compared
the expression of genes known to be importan
t for the response to stress and biofilm for-
mation. Specifically, we compared the gene
expression between planktonic bacteria as
they develop into a biofilm (0–24 h) and their
gene expression once they are already in a
biofilm (24–48 h).
RPA was grown in sterile 96-well plates and treated with 250 μM, 500 μM or 1000
μM of H
2
O
2
with a vehicle (sterile water) as control, to simulate bacterial exposure to low,
medium, and high levels of OS. One set of cultures was treated from time zero of culture
(planktonic) and the other allowed to form a biofilm for 24 h before treatments (Figure 1).
Significant differences in gene expression were found between planktonic RPA cells
growing to form a biofilm in response to OS and RPA already living in a biofilm culture
when exposed to OS
(Figure 2). In RPA planktonic cells, all QS genes,
lasI
,
lasR
,
rhlI
and
rhlR
, were upregulated (Figure 2A). The expression of
oxyR
, a gene that functions in OS
defense and DNA repair, was also significantly
upregulated. In contrast, these genes were
downregulated when RPA was already in its biofilm form when exposed to H
2
O
2
(Figure
2A). In particular, we observed that the expression of
lasI
,
lasR
,
rhlI
,
rhlR
and
oxyR,
when
the bacteria were in a biofilm, was significantly downregulated with the 500 μM H
2
O
2
treatment. The lack of the same strong down
regulation when the bacteria were treated
with 1000 μM could be due to these high levels of H
2
O
2
causing toxicity to the bacteria.
Figure 1.
Experimental design of in vitro H
2
O
2
treatment of RPA.
RPA was grown in 96-well plates
to test its transcriptomic response to H
2
O
2
in a dose-dependent manner. Planktonic cells were
treated for 24 h before they were
collected for analysis. RPA, in biofilm form, was undisturbed for
24 h to allow it to form a biofilm. At 24 h it was treated with H
2
O
2
. After another 24 h of treatment,
RPA cells in biofilm were collected for analysis.
Figure 1.
Experimental design of
in vitro
H
2
O
2
treatment of RPA. RPA was grown in 96-well plates
to test its transcriptomic response to H
2
O
2
in a dose-dependent manner. Planktonic cells were treated
for 24 h before they were collected for analysis. RPA, in biofilm form, was undisturbed for 24 h to
allow it to form a biofilm. At 24 h it was treated with H
2
O
2
. After another 24 h of treatment, RPA
cells in biofilm were collected for analysis.
Antioxidants
2024
,
13
, 655
7 of 16
Antioxidants
2024
,
13
, x FOR PEER REVIEW
7 of 16
PA
has two SOD enzymes:
sodB,
which encodes for Fe-SOD, whose expression is high
when iron is high; otherwise,
sodA
, which encodes for Mn-SOD, is highly expressed. Both
enzymes dismutate the highly damaging radical O
2
·
−
to H
2
O
2
, which is less damaging.
PA
also has two catalase genes that enco
de KatA and KatB, which breakdown H
2
O
2
into H
2
O
+ O
2
[29]. We found that, under planktonic conditions,
sodA
,
sodB
and
katA
are all highly
upregulated, whereas, when in a biofilm, RP
A does not activate these genes (Figure 2B).
When the genes responsible for Pel and Psl synthesis,
pelA
and
pelD
and
pslA
and
pslB
, respectively, were studied, we observed that, again, under planktonic conditions,
these genes were upregulated, whereas, when
the RPA was already in its biofilm, the same
genes were downregulated, except for
pslB
, which was slightly upregulated (Figure 2C).
Similar observations were made with the pathogenic genes,
phzA
,
phzB
,
lexA
and
rpsL
(Figure 2D).
Figure 2.
Transcriptional responses of planktonic RPA
and of RPA in an in vitro biofilm culture
when treated with H
2
O
2
. RPA was incubated with either H
2
O or 250 μM, 500 μM or 1000 μM of
H
2
O
2
. Samples were collected in triplicate within 1 mi
n for the 0 h timepoint and then at the specific
times indicated in the figure. (
A
) Log
2
FC trends of important quorum sensing genes,
lasI
and
lasR
,
and
rhlI
and
rhlR
and
oxyR,
that function in OS defense and DNA repair. (
B
) Log
2
FC trends of genes
coding enzymes that metabolize H
2
O
2
:
katA
and
katB
and
sodA
and
sodB
. (
C
) Log
2
FC trends of critical
biofilm genes for Pel formation,
pelA
and
pelD
, and Psl formation,
pslA
and
pslB
. (
D
)
Log
2
FC trends
of genes aiding in pathogenesis
through the biosynthesis of a crucial redox-sensitive phenazine mol-
ecule, pyocyanin,
phzA
and
phzM,
and metabolism,
lexA
and
rpsL
.
Bars represent average log
2
FC
and error bars represent standard deviation. Stat
istical significance was calculated by comparing
H
2
O
2
treatments with the vehicle control using Student’s
t-
test. *
p-
value
<
0.05, **
p-
value
<
0.01.
3.2. Differential Gene Expression
of Biofilm-Forming PA In Vivo
For the studies in vivo, we used our mouse model of chronic wounds, from which
RPA was isolated [37]. Five
-month-old, diabetic, obese
db
/
db
mice were used to perform
excision wounds (Figure 3). For these experi
ments, the mouse skin was depleted of its
microbiome, wounds were made and chronicity was induced. At 24 h after wounding,
RPA was introduced under the Tegaderm and biofilm formation began to proceed in the
Figure 2.
Transcriptional responses of planktonic RPA and of RPA in an
in vitro
biofilm culture when
treated with H
2
O
2
. RPA was incubated with either H
2
O or 250
μ
M, 500
μ
M or 1000
μ
M of H
2
O
2
.
Samples were collected in triplicate within 1 min for the 0 h timepoint and then at the specific times
indicated in the figure. (
A
) Log
2
FC trends of important quorum sensing genes,
lasI
and
lasR
, and
rhlI
and
rhlR
and
oxyR,
that function in OS defense and DNA repair. (
B
) Log
2
FC trends of genes
coding enzymes that metabolize H
2
O
2
:
katA
and
katB
and
sodA
and
sodB
. (
C
) Log
2
FC trends of critical
biofilm genes for Pel formation,
pelA
and
pelD
, and Psl formation,
pslA
and
pslB
. (
D
) Log
2
FC trends
of genes aiding in pathogenesis through the biosynthesis of a crucial redox-sensitive phenazine
molecule, pyocyanin,
phzA
and
phzM,
and metabolism,
lexA
and
rpsL
. Bars represent average log
2
FC
and error bars represent standard deviation. Statistical significance was calculated by comparing
H
2
O
2
treatments with the vehicle control using Student’s
t
-test. *
p
-value < 0.05, **
p
-value < 0.01.
3.2. Differential Gene Expression of Biofilm-Forming PA In Vivo
For the studies
in vivo
, we used our mouse model of chronic wounds, from which
RPA was isolated [
37
]. Five-month-old, diabetic, obese
db
/
db
mice were used to perform
excision wounds (Figure 3). For these experiments, the mouse skin was depleted of its
microbiome, wounds were made and chronicity was induced. At 24 h after wounding,
RPA was introduced under the Tegaderm and biofilm formation began to proceed in the
wound environment. We observed an activation of all of the QS genes we tested,
lasI
,
lasR
,
rhlI
,
rhlR
and
oxyR,
by 24 h after the application of RPA (Figure 4A). However, for
the antioxidant genes, we found that only
katA
and
sodB
were activated to decrease OS
in the wound, whereas
katB
and
sodA
were downregulated (Figure 4B). The Pel and Psl
genes that contribute to biofilm formation were all upregulated and so were all of the
pathogenic genes we tested,
phzA
,
phzM
,
recA
and
lexA
(Figure 4C,D).
phzA
showed the
highest upregulation, which increased with time as the wound progressed to chronicity.
3.3. Identification of Gene Expression in RPA-Induced Biofilm In Vivo by RNA-seq Detection
RPA-infected wounds were collected more than 24 h after inoculation for both chronic
and nonchronic wounds at three timepoints: 24, 48 and 72 h after infection. Approximately
21–30 million reads were obtained per sample library and a FASTQC analysis of the raw
reads from the sequenced libraries indicated that the reads are of high quality (e.g., few
adapter sequences trimmed, overall base quality (q) > 30). The reads were aligned to the
Antioxidants
2024
,
13
, 655
8 of 16
genome sequenced for RPA, with annotations predicted with BAKTA (see Materials and
Methods). Over 150 genes were found to be significantly differentially expressed in the
24 h wound tissues, FDR < 0.1. No genes were observed to be differentially expressed at
48 h and 72 h and, therefore, these timepoints were excluded from subsequent analyses.
Among the differentially expressed genes associated with metabolism and respiration at
24 h of infection, many genes of the Type VI Secretion System (T6SS) were significantly
overexpressed (Figure 5). Most significantly, genes that participated in the biosynthesis
of the secretion system’s HSI-II structure were upregulated in chronic wounds compared
to nonchronic wounds (Figure 5A). Genes that form the membrane complex,
hsiJ2
,
icmF2
and
dotU2
; baseplate,
vgrG4a
and
hsiF2
; and sheath complex,
hsiB2
and
hsiC2
, were sig-
nificantly upregulated. The ATPase that drives the secretion mechanism,
clpV2
, was also
upregulated. In the sheath complex of HSI-III cluster,
hsiB3
and
hsiC3
were significantly up-
regulated (Figure 5B). Several genes responsible for the effectors of T6SS were significantly
upregulated in chronic wounds at 24 h. Phospholipase effectors
tle5
,
tli5, tli5b1
and
tli5b2
were also significantly upregulated (Figure 5C). Several hemolysin coregulated proteins
(hcp) and other effectors such as
fha2
and
orfX
were also upregulated in chronic wounds.
RpoN/Sfa2-dependent activation may play a role in the regulation of the T6SS in the RPA
in chronic wounds (Figure 5D). The quorum sensing genes
lasI
and
lasR
were upregulated
in the RPA RNAseq analysis, confirming the results obtained through RT-qPCR. The prefer-
ential upregulation of
katA
and
sodB
seen in the RT-qPCR
in vivo
also correlated with the
differentially expressed gene analysis after sequencing the RPA transcriptomics (Figure 6).
Antioxidants
2024
,
13
, x FOR PEER REVIEW
8 of 16
wound environment. We observed an activa
tion of all of the QS genes we tested,
lasI
,
lasR
,
rhlI
,
rhlR
and
oxyR,
by 24 h after the application of RPA (Figure 4A).
However, for the
antioxidant genes, we found that only
katA
and
sodB
were activated to decrease OS in the
wound, whereas
katB
and
sodA
were downregulated (Figure 4B). The Pel and Psl genes
that contribute to biofilm formation were all
upregulated and so were all of the pathogenic
genes we tested,
phzA
,
phzM
,
recA
and
lexA
(Figure 4C,D).
phzA
showed the highest up-
regulation, which increased with time as the wound progressed to chronicity.
Figure 3.
Chronic wound progression with RPA infection. RPA strain was isolated and identified
from previous naturally infected chronic wounds. (A) For the in vivo studies, clean wounds were
infected with RPA 24 h after surgery and the stimulation of chronicity. Bacterial samples were col-
lected 24, 48 and 72 h after RPA was introduced in
to the wounds and analyzed for gene expression
using qPCR. (
B
) Picture of a wound 48 h after injury wh
ich then progressed to a fully chronic
wound.
Figure 3.
Chronic wound progression with RPA infection. RPA strain was isolated and identified from
previous naturally infected chronic wounds. (
A
) For the
in vivo
studies, clean wounds were infected
with RPA 24 h after surgery and the stimulation of chronicity. Bacterial samples were collected 24, 48
and 72 h after RPA was introduced into the wounds and analyzed for gene expression using qPCR.
(
B
) Picture of a wound 48 h after injury which then progressed to a fully chronic wound.
Antioxidants
2024
,
13
, 655
9 of 16
Antioxidants
2024
,
13
, x FOR PEER REVIEW
9 of 16
Figure 4.
RPA activates quorum sensing,
OS-combating
, biofilm and pathogenesis genes in vivo.
Non-chronic wounds (NCWs) and chronic wounds
(CWs) were infected with RPA and the exudate
was collected for RPA transcriptomics. Gene expres
sion in CWs was compared to that in NCWs at
24 h, 48 h and 72 h (
n
= 6 at each timepoint for each condition,
n
= 36 total). (
A
) The log
2
FC of quorum
sensing genes for
lasI
,
lasR
,
rhlI
and
rhlR
was significantly upregulated in CWs.
OxyR
was not sig-
nificantly upregulated. (
B
) RPA preferentially expressed
katA
and
sodB
genes in CWs, whereas
katB
and
sodA
were significantly downregulated in CWs. (
C
) RPA upregulated the biofilm genes
pelA
,
pelD
,
pslA
and
pslB
in CWs. (
D
)
phzA
in the phenazine biosynthesis pathway was significantly ex-
pressed by RPA in CWs, whereas their
RecA
and
lexA
expression decreased over time. Bars repre-
sent average log
2
FC and error bars represent standard devi
ation. Statistical significance was calcu-
lated by comparing CWs with NCWs using Student’s
t
-test. *
p
-value
<
0.05, **
p
-value
<
0.01, ***
p
-value < 0.001. + FDR < 0.1.
3.3. Identification of Gene Ex
pression in RPA-Induce
d Biofilm In Vivo by RNA-seq Detection
RPA-infected wounds were collected more than 24 h after inoculation for both
chronic and nonchronic wounds at three time
points: 24, 48 and 72 h after infection. Ap-
proximately 21–30 million reads were obtained per sample library and a FASTQC analysis
of the raw reads from the sequenced libraries indicated that the reads are of high quality
(e.g., few adapter sequences trimmed, overall base quality (q) > 30). The reads were
aligned to the genome sequenced for RPA,
with annotations predicted with BAKTA (see
Materials and Methods). Over 150 genes were found to be significantly differentially ex-
pressed in the 24 h wound tissues, FDR < 0.1. No genes were observed to be differentially
expressed at 48 h and 72 h and, therefor
e, these timepoints were excluded from
Figure 4.
RPA activates quorum sensing,
OS-combating
, biofilm and pathogenesis genes
in vivo
.
Non-chronic wounds (NCWs) and chronic wounds (CWs) were infected with RPA and the exudate
was collected for RPA transcriptomics. Gene expression in CWs was compared to that in NCWs
at 24 h,
48 h
and 72 h (
n
= 6 at each timepoint for each condition,
n
= 36 total). (
A
) The log
2
FC of
quorum sensing genes for
lasI
,
lasR
,
rhlI
and
rhlR
was significantly upregulated in CWs.
OxyR
was
not significantly upregulated. (
B
) RPA preferentially expressed
katA
and
sodB
genes in CWs, whereas
katB
and
sodA
were significantly downregulated in CWs. (
C
) RPA upregulated the biofilm genes
pelA
,
pelD
,
pslA
and
pslB
in CWs. (
D
)
phzA
in the phenazine biosynthesis pathway was significantly
expressed by RPA in CWs, whereas their
RecA
and
lexA
expression decreased over time. Bars
represent average log
2
FC and error bars represent standard deviation. Statistical significance was
calculated by comparing CWs with NCWs using Student’s
t
-test. *
p
-value < 0.05, **
p
-value < 0.01,
***
p
-value < 0.001. + FDR < 0.1.
Antioxidants
2024
,
13
, 655
10 of 16
Antioxidants
2024
,
13
, x FOR PEER REVIEW
10 of 16
subsequent analyses. Among th
e differentially expressed gene
s associated with metabo-
lism and respiration at 24 h of infection, many genes of the Type VI Secretion System
(T6SS) were significantly overexpressed (Figure 5). Most significantly, genes that partici-
pated in the biosynthesis of the secretion system’s HSI-II structure were upregulated in
chronic wounds compared to nonchronic woun
ds (Figure 5A). Genes that form the mem-
brane complex,
hsiJ2
,
icmF2
and
dotU2
; baseplate,
vgrG4a
and
hsiF2
; and sheath complex,
hsiB2
and
hsiC2
, were significantly upregulated. The ATPase that drives the secretion
mechanism,
clpV2
, was also upregulated. In the sh
eath complex of HSI-III cluster,
hsiB3
and
hsiC3
were significantly upregulated (Figure
5B). Several genes responsible for the
effectors of T6SS were significantly upregulated in chronic wounds at 24 h. Phospholipase
effectors
tle5
,
tli5,
tli5b1
and
tli5b2
were also significantly upre
gulated (Figure 5C). Several
hemolysin coregulated proteins (hcp) and other effectors such as
fha2
and
orfX
were also
upregulated in chronic wounds. RpoN/Sfa2-depe
ndent activation may play a role in the
regulation of the T6SS in the RPA in chronic wounds (Figure 5D). The quorum sensing
genes
lasI
and
lasR
were upregulated in the RPA RNAseq analysis, confirming the results
obtained through RT-qPCR. The
preferential upregulation of
katA
and
sodB
seen in the
RT-qPCR in vivo also correlated with the diff
erentially expressed gene analysis after se-
quencing the RPA transcriptomics (Figure 6).
Figure 5.
The Type VI Secretion System (T6SS) was si
gnificantly upregulated in chronic wounds.
The RPA in chronic and nonchron
ic wound exudate was collected at 24 h, and its RNA was se-
quenced (
n
= 3 for both chronic woun
ds and nonchronic wounds,
n
= 6 total).
(
A
) Many T6SS struc-
tural proteins of the HSI-II loci were found to be upregulated in chronic wounds, including those of
the membrane complex (
hsiJ2
,
icmF2
and
dotU2
), ATPase (
clpV2
), the baseplate (
vgrG4a
and
hsiF2
)
and the sheath complex (
hsiB2
and
hsiC2
). (
B
) Two components of the HIS-III sheath complex were
significantly upregulated. (
C
) Effectors that have phospholipase activity, the Tli- Tle system, were
also activated in CWs. (
D
) Hemolysin-coregulated proteins (
hcpA
,
hcpB
and
hcpC
) and other effec-
tors (
fha2
and
orfX
) of T6SS were also upregulated. RpoN/Sfa2 may regulate T6SS activity. Bars
Figure 5.
The Type VI Secretion System (T6SS) was significantly upregulated in chronic wounds. The
RPA in chronic and nonchronic wound exudate was collected at 24 h, and its RNA was sequenced
(
n
= 3
for both chronic wounds and nonchronic wounds,
n
= 6 total). (
A
) Many T6SS structural
proteins of the HSI-II loci were found to be upregulated in chronic wounds, including those of the
membrane complex (
hsiJ2
,
icmF2
and
dotU2
), ATPase (
clpV2
), the baseplate (
vgrG4a
and
hsiF2
) and
the sheath complex (
hsiB2
and
hsiC2
). (
B
) Two components of the HIS-III sheath complex were
significantly upregulated. (
C
) Effectors that have phospholipase activity, the Tli- Tle system, were
also activated in CWs. (
D
) Hemolysin-coregulated proteins (
hcpA
,
hcpB
and
hcpC
) and other effectors
(
fha2
and
orfX
) of T6SS were also upregulated. RpoN/Sfa2 may regulate T6SS activity. Bars represent
average log
2
FC and error bars represent lfcSE.
Statistical significance was determined using
log
2
FC > 1
and an adjusted
p
-value (FDR) < 0.1, comparing CW to NCW samples. + FDR < 0.1, * FDR < 0.05.
Antioxidants
2024
,
13
, x FOR PEER REVIEW
11 of 16
represent average log
2
FC and error bars represent lfcSE. Statistical significance was determined us-
ing log
2
FC > 1 and an adjusted
p
-value (FDR) < 0.1, comparing CW to NCW samples. + FDR < 0.1, *
FDR < 0.05.
Figure 6.
RNAseq confirmed gene expression:
the differentially expresse
d quorum sensing genes,
lasI
and
lasR
, and antioxidant genes,
katA
,
katB
,
sodA
and
sodB
, obtained through RNAseq validate
the genes expressed in vivo via the RT-qPCR results found in Figure 4.
Bars represent average
log
2
FC and error bars represent lfcSE. Statistical significance was determined using log
2
FC > 1 and
an adjusted
p
-value (FDR) < 0.1, comparing CW to NCW
samples. + FDR < 0.1, * FDR < 0.05, ** FDR
< 0.01.
4. Discussion
Pseudomonas aeruginosa
is a clinically significant, opportunistic pathogen in humans,
and the strength of its pathogenesis in human infections relies on many virulence factors
and effectors that are encoded in its large ge
nome [67]. We have sh
own that RPA activates
quorum sensing, virulence, biofilm formation and antioxidant enzyme genes in the high
OS microenvironment present in chronic wo
unds and colonizes these wounds success-
fully by forming a biofilm. Our finding that RP
A induces different genes when it exists as
a biofilm serves as reminder that treatments in the clinic must appropriately target the
metabolic state of biofilm-forming bacteria. Our transcriptomic analysis of the RPA in
chronic wounds showed that many structural
components and effectors of the recently
described T6SS are upregulated. RPA could be using T6SS to destroy or modulate com-
petitive bacteria, as well as to interact with and inactivate host immune cells and skin cells
important for wound healing.
The importance of the different metabolic
states of planktonic
vs. biofilm RPA was
reflected by their different responses to H
2
O
2
in vitro. Planktonic RPA strongly activated
its quorum sensing, OS response, virulence, biofilm formation and oxidative stress genes,
whereas, in a mature biofilm, it did not upregu
late any of these genes, suggesting that in
biofilm conditions
PA
is already in a physiologic state that is not affected by the inhospi-
table environment of chronic wounds. Therefore,
it is possible that the activation of these
genes could put the bacteria in a “planktonic” state and render them more susceptible to
standard antibiotic treatments.
The in vivo model of
PA
infection showed that RPA initially survives the harsh, high
OS microenvironment present in chronic woun
ds and colonizes these wounds as a biofilm
by turning on the quorum sensing genes (F
igure 5A) responsible for coordinating the
genes that contribute to biofilm development (Figure 5C).
pelD
, for example, is an essential
gene for Pel polysaccharide production. It regu
lates the production of the polysaccharide
through c-di-GMP, a second messenger used in
signal transduction that controls the cel-
lular processes that contribute to surface adaptation, biofilm formation, cell cycle
Figure 6.
RNAseq confirmed gene expression: the differentially expressed quorum sensing genes,
lasI
and
lasR
, and antioxidant genes,
katA
,
katB
,
sodA
and
sodB
, obtained through RNAseq validate the
genes expressed
in vivo
via the RT-qPCR results found in Figure 4. Bars represent average log
2
FC and
error bars represent lfcSE. Statistical significance was determined using log
2
FC > 1 and an adjusted
p
-value (FDR) < 0.1, comparing CW to NCW samples. + FDR < 0.1, * FDR < 0.05, ** FDR < 0.01.
Antioxidants
2024
,
13
, 655
11 of 16
4. Discussion
Pseudomonas aeruginosa
is a clinically significant, opportunistic pathogen in humans,
and the strength of its pathogenesis in human infections relies on many virulence factors
and effectors that are encoded in its large genome [
67
]. We have shown that RPA activates
quorum sensing, virulence, biofilm formation and antioxidant enzyme genes in the high OS
microenvironment present in chronic wounds and colonizes these wounds successfully by
forming a biofilm. Our finding that RPA induces different genes when it exists as a biofilm
serves as reminder that treatments in the clinic must appropriately target the metabolic state
of biofilm-forming bacteria. Our transcriptomic analysis of the RPA in chronic wounds
showed that many structural components and effectors of the recently described T6SS
are upregulated. RPA could be using T6SS to destroy or modulate competitive bacteria,
as well as to interact with and inactivate host immune cells and skin cells important for
wound healing.
The importance of the different metabolic states of planktonic vs. biofilm RPA was
reflected by their different responses to H
2
O
2
in vitro
. Planktonic RPA strongly activated
its quorum sensing, OS response, virulence, biofilm formation and oxidative stress genes,
whereas, in a mature biofilm, it did not upregulate any of these genes, suggesting that in
biofilm conditions
PA
is already in a physiologic state that is not affected by the inhospitable
environment of chronic wounds. Therefore, it is possible that the activation of these genes
could put the bacteria in a “planktonic” state and render them more susceptible to standard
antibiotic treatments.
The
in vivo
model of
PA
infection showed that RPA initially survives the harsh, high
OS microenvironment present in chronic wounds and colonizes these wounds as a biofilm
by turning on the quorum sensing genes (Figure 5A) responsible for coordinating the
genes that contribute to biofilm development (Figure 5C).
pelD
, for example, is an essential
gene for Pel polysaccharide production. It regulates the production of the polysaccharide
through c-di-GMP, a second messenger used in signal transduction that controls the cellular
processes that contribute to surface adaptation, biofilm formation, cell cycle progression
and virulence [
68
].
pelA
is an important gene necessary for the modification of the Pel
polymer after its assembly and secretion through a complex modified after polymerization
in the periplasm [
68
]. Note that both
pelA
and
pelD
are upregulated by an exposure to H
2
O
2
in vitro (Figure 2D).
Four antioxidant enzymes critical for the removal or breakdown of hydrogen peroxide
were found to be differentially expressed both
in vitro
and
in vivo
. The gene expression of
two SOD genes,
sodA
(manganese cofactor) and
sodB
(iron cofactor), showed that sodB may
be more important than sodA for aerobic growth [
68
].
SodB
expression was upregulated,
while
sodA
expression was downregulated. This could be explained by the fact that, in
the early stages of the response to injury, iron is very abundant in the wound [
69
,
70
] and
can be used by SodB for its activation [
71
–
73
]. In addition, the controlled regulation of
the lasR/lasI and rhlR/rhlI system has been found to be crucial for
PA
to form a biofilm
and respond to oxidative stress [
68
]. A mutation in either or both of these systems results
in
PA
becoming more sensitive to H
2
O
2
because the expression levels of the genes for
sodA
,
sodB
and
katA
are decreased [
74
]. For the catalase genes required for a resistance to
peroxides and osmotic stresses, we found that
katA
was significantly upregulated with
exposure to the higher concentration of hydrogen peroxide, whereas
katB
expression was
significantly downregulated. This is not surprising, because katA is the major catalase
of
PA
that detoxifies H
2
O
2
, a reactive oxygen species that is generated during aerobic
respiration [75].
In addition to QS, SOD and catalase, which maximize growth under excess OS con-
ditions,
PA
also produces enzymes such as alkaline protease, elastase and lasA protease,
and metabolites such hemolysin, rhamnolipids and phenazines (e.g., pyocyanin (PYO)), to
combat host immune cells during infection [
29
,
76
]. PYO is a blue secondary metabolite;
its synthesis is partly regulated by QS. As a zwitterion, it can easily penetrate and cross
biological membranes. It is a powerful metabolite because it is redox-active and known to
Antioxidants
2024
,
13
, 655
12 of 16
play a crucial role in the virulence and infection of
PA
in humans and in animal models.
PYO can inactivate catalase, but not SOD, in a human epithelial cell line [
77
]. It modulates
the normal glutathione (GSH) cycle by depleting cellular GSH and increasing oxidized
GSH (GSSG) in epithelial cells [
78
]. The synthesis of PYO is controlled by two phenazine
operons (
phzABCDEFG
) and by the
phzH
,
phzM
and
phzS
genes [
79
]. In addition, two other
systems are involved in
PA
’s virulence and biofilm formation: the Pel and Psl systems
of exopolysaccharides, which help facilitate the formation of biofilms while impairing
bacterial clearance [28].
Our bacterial transcriptomic analysis of RPA during the first 24 h of chronic wound
initiation showed that the T6SS was significantly upregulated. This system is a recently
described system of
PA
which adds another virulent weapon to its ever-growing arsenal to
compete and thrive in diverse environments. The T6SS has several structural components
including a puncturing structure that has been described in bacteriophages [
80
].
PA
benefits
from its T6SS by delivering toxins to its neighboring pathogens and translocating protein
effectors into the host cells. The T6SS also takes part in biofilm formation [
80
].
PA
has
been found to use the T6SS to compete in multi-bacterial species communities, providing
growing and fitness advantages over other species in a mixed-species biofilm [
81
].
PA
has
been described to have the ability to invade and infect eukaryotic cells using this secretion
system. Mutations in the system have shown a decreased or weakened virulence phenotype
in a rat model of chronic lung infection [
80
,
82
].
PA
also has the ability to target and infect
human lung epithelial cells through quorum sensing Las and Rhl systems and induces
the host Akt pathway, mediated through the phosphatidylinositol 3-kinase-dependent
pathway [83].
In conclusion
,
Pseudomonas aeruginosa
activates quorum sensing, OS-combating,
biofilm-forming and virulence genes in the high OS microenvironment present in chronic
wounds and colonizes these wounds successfully by forming a biofilm. The T6SS provides
an opportunity to target the ability of
PA
uses to destroy other bacteria in the biofilm as well
as host cells. Effective biofilm removal may be accomplished by disrupting these systems
and dismantling the biofilm, leaving the planktonic bacteria unprotected and susceptible to
antibiotic treatment.
Author Contributions:
Conceptualization, J.H.K. and M.M.-G.; Methodology, J.H.K., B.H.L., W.Z.,
W.G., and M.M.-G.; Software, B.H.L. and Z.R.L.; Validation, J.H.K. and B.H.L.; Formal analysis, J.H.K.;
Resources, W.Z.; Data curation, J.H.K. and J.D.; Writing—original draft, J.H.K.; Writing—review &
editing, J.H.K. and M.M.-G.; Supervision, M.M.-G.; Funding acquisition, T.G., D.K.N. and M.M.-G.
All authors have read and agreed to the published version of the manuscript.
Funding:
This research was funded by the NIH 1 R21 AI156688-01, NIH/NIAU19AG023122. ZRL
was supported by a fellowship from the Jane Coffin Childs Memorial Fund for Medical Research.
Institutional Review Board Statement:
Not applicable.
Informed Consent Statement:
Not applicable.
Data Availability Statement:
The bacterial genome sequences for RPA have been deposited in the
National Center for Biotechnology Information (NCBI)’s Sequence Read Archive (SRA) under the
BioProject Accession Number PRJNA1112363. The bacterial RNAseq sequences and raw count matrix
file have been deposited in NCBI’s Gene Expression Omnibus (GEO) under the GEO Accession
Number GSE267862.
Conflicts of Interest:
The authors declare no conflicts of interest.
References
1.
Kim, J.H.H.; Ahamed, A.; Chen, K.; Lebig, E.G.G.; Petros, B.; Saeed, S.; Martins-Green, M. Skin microbiota and its role in health
and disease with an emphasis on wound healing and chronic wound development. In
Microbiome, Immunity, Digestive Health and
Nutrition: Epidemiology, Pathophysiology, Prevention and Treatment
; Academic Press: Cambridge, MA, USA, 2022. [CrossRef]
2.
Loomis, K.H.; Wu, S.K.; Ernlund, A.; Zudock, K.; Reno, A.; Blount, K.; Karig, D.K. A mixed community of skin microbiome
representatives influences cutaneous processes more than individual members.
Microbiome
2021
,
9
, 22. [CrossRef] [PubMed]