Prevalence and Correlates of Phenazine Resistance in Culturable
Bacteria from a Dryland Wheat Field
Elena K. Perry
,
a
Dianne K. Newman
a
,
b
a
Division of Biology and Biological Engineering, California Institute of Technology, Pasadena, California, USA
b
Division of Geological and Planetary Sciences, California Institute of Technology, Pasadena, California, USA
ABSTRACT
Phenazines are a class of bacterially produced redox-active natural anti-
biotics that have demonstrated potential as a sustainable alternative to traditional
pesticides for the biocontrol of fungal crop diseases. However, the prevalence of
bacterial resistance to agriculturally relevant phenazines is poorly understood, limit-
ing both the understanding of how these molecules might shape rhizosphere bacte-
rial communities and the ability to perform a risk assessment for off-target effects.
Here, we describe pro
fi
les of susceptibility to the antifungal agent phenazine-1-car-
boxylic acid (PCA) across more than 100 bacterial strains isolated from a wheat
fi
eld
where PCA producers are indigenous and abundant. We found that Gram-positive
bacteria are typically more sensitive to PCA than Gram-negative bacteria, and there
was signi
fi
cant variability in susceptibility both within and across phyla. Phenazine-
resistant strains were more likely to be isolated from the wheat rhizosphere, where
PCA producers were also more abundant, compared to bulk soil. Furthermore, PCA
toxicity was pH-dependent for most susceptible strains and broadly correlated with
PCA reduction rates, suggesting that uptake and redox-cycling were important deter-
minants of phenazine toxicity. Our results shed light on which classes of bacteria are
most likely to be susceptible to phenazine toxicity in acidic or neutral soils. In addi-
tion, the taxonomic and phenotypic diversity of our strain collection represents a val-
uable resource for future studies on the role of natural antibiotics in shaping wheat
rhizosphere communities.
IMPORTANCE
Microbial communities contribute to crop health in important ways. For
example, phenazine metabolites are a class of redox-active molecules made by diverse
soil bacteria that underpin the biocontrol of diseases of wheat and other crops. Their
physiological functions are nuanced. In some contexts, they are toxic. In others, they are
bene
fi
cial. While much is known about phenazine production and the effect of phena-
zines on producing strains, our ability to predict how phenazines might shape the com-
position of environmental microbial communities is poorly constrained. In addition,
phenazine prevalence in the rhizosphere has been predicted to increase in arid soils as
the climate changes, providing an impetus for further study. As a step toward gaining a
predictive understanding of phenazine-linked microbial ecology, we document the
effects of phenazines on diverse bacteria that were coisolated from a wheat rhizosphere
and identify conditions and phenotypes that correlate with how a strain will respond to
phenazines.
KEYWORDS
correlates, dryland wheat, ef
fl
ux pumps, phenazines, resistance,
rhizosphere, toxicity
D
iverse microorganisms produce natural antibiotics that can kill or inhibit the
growth of other microbes (1, 2). Several such compounds have been commercial-
ized as antimicrobial drugs for the treatment of infections, beginning with penicillin in
the 1940s (3, 4). Unfortunately, the selective pressures exerted by the widespread use
Editor
Arpita Bose, Washington University in
St. Louis
Copyright
© 2022 American Society for
Microbiology.
All Rights Reserved
.
Address correspondence to Dianne K.
Newman, dkn@caltech.edu.
The authors declare no con
fl
ict of interest.
Received
23 November 2021
Accepted
1 February 2022
Accepted manuscript posted online
9 February 2022
Published
March 2022 Volume 88 Issue 6 e02320-21
Applied and Environmental Microbiology
aem.asm.org
1
ENVIRONMENTAL MICROBIOLOGY
22
March
2022
Downloaded from https://journals.asm.org/journal/aem on 05 April 2022 by 131.215.249.7.
of antibiotics in medicine and agriculture have led to worrisome increases in the preva-
lence of antimicrobial resistance among human and animal pathogens over the past
several decades (5, 6). Yet, while this repercussion of human antibiotic use has been
well documented, comparatively little is known about the ecological effects of micro-
bially produced antibiotics in natural environments (7, 8). Recent studies have sug-
gested that natural antibiotics may serve a variety of functions for their producers
beyond the suppression of competing microbes (9, 10), including nutrient acquisition
(11, 12), conservation of energy in the absence of oxygen (13, 14), and cell-cell signal-
ing (15, 16). At the same time, toxicity to one or more microorganisms is a feature of
these molecules, but the extent to which this trait shapes their in
fl
uence on microbial
communities is unclear (2, 17, 18). In addition, for many if not most natural antibiotics,
the determinants and prevalence of susceptibility or resistance to their toxicity remain
unknown or poorly characterized (19). These gaps in knowledge hinder our ability to
understand and predict the impacts of these metabolites on microbial communities of
interest.
One environmental context in which natural antibiotics are thought to be particu-
larly abundant and ecologically relevant is the rhizosphere
—
the narrow plane of soil
immediately adjacent to plant roots (20
–
22). Natural antibiotics such as phenazines
and 2,4-diacetylphloroglucinol have been directly detected in the rhizospheres of mul-
tiple crops, including wheat, potato, and sugar beet (22
–
24), and phenazines have
been shown to increase the
fi
tness of their producers when competing with other
microbes in the rhizosphere (25, 26). Production of these molecules has also been
demonstrated to underpin the ability of certain bacteria to control fungal crop diseases
(23, 26
–
29), further indicating that natural antibiotics can act as agents of microbe-
microbe antagonism in the rhizosphere. As a result of this activity, phenazine-produc-
ing
Pseudomonas
strains have received attention as potential biocontrol agents that
could serve as a more sustainable alternative to traditional synthetic pesticides in agri-
culture (19, 28). However, several challenges remain concerning the practical applica-
tion of these strains, including inconsistent ef
fi
cacy under
fi
eld conditions (20, 28), lim-
ited understanding of the mechanisms and evolutionary dynamics of resistance to
phenazines (19, 30), and the possibility of off-target effects (28).
Importantly, with regard to the latter concern, phenazines are toxic to fungi and
some bacteria (31
–
33). Yet, while the utility of phenazine-producing pseudomonads
for biocontrol of fungal crop diseases has been extensively investigated, the potential
impact of these strains on nontarget bacterial residents of the rhizosphere is less well
understood. One study found that inoculation of
Pseudomonas
biocontrol strains
shifted the rhizosphere community of maize seedlings, pushing the ratio of Gram-posi-
tive to Gram-negative bacteria in favor of the latter. However, this analysis was based
on pro
fi
ling colony growth rates and whole-cell fatty acids from pooled cultured iso-
lates, greatly limiting the taxonomic resolution and making it dif
fi
cult to rule out
whether the Gram-negative biocontrol strains might themselves have directly contrib-
uted to the shift (34). On the other hand, at least three studies found no notable or
consistent effects of introduced
Pseudomonas
species on the rhizosphere bacterial
communities of wheat or potato (35
–
37), albeit the studies in wheat employed meth-
ods with limited discriminatory power (namely, carbon source utilization pro
fi
ling and
terminal restriction fragment length polymorphism). Given these mixed results and the
lack of
fi
ne-grained spatial or taxonomic resolution in most studies on this topic,
whether phenazine-producing bacteria actively shape the surrounding rhizosphere
bacterial community through antibiosis, potentially at the micron or millimeter scale,
remains an open question.
In addition to the lack of clarity regarding the effects of phenazines on rhizosphere
bacterial communities, the taxonomic and physiological correlates of phenazine resist-
ance in bacteria remain incompletely understood. The toxicity of phenazines is gener-
ally ascribed to the generation of reactive oxygen species (ROS) and interference with
respiration (33, 38
–
40). Previous studies have suggested that ef
fl
ux pump expression,
Perry and Newman
Applied and Environmental Microbiology
March 2022 Volume 88 Issue 6 e02320-21
aem.asm.org
2
Downloaded from https://journals.asm.org/journal/aem on 05 April 2022 by 131.215.249.7.
cell permeability, oxidative stress responses, and the composition of the respiratory
electron transport chain can affect bacterial susceptibility to phenazines (39
–
44). In
addition, a comparison of 14 bacterial strains indicated that Gram-negative bacteria as
a group may be more resistant to phenazine toxicity than Gram-positive bacteria (33).
However, all these studies focused on a speci
fi
c phenazine, pyocyanin, that is particu-
larly toxic (45) and best known for its role as a virulence factor during infections of
humans and animals (46, 47). Whether the same observations hold for more agricultur-
ally relevant phenazines, such as phenazine-1-carboxylic acid (PCA) (22, 23, 27, 48), is
unknown.
In this study, we set out to lay a foundation for addressing unresolved questions
about the ecological impact of phenazine toxicity in the rhizosphere by determining
the prevalence of phenazine resistance among bacteria isolated from a wheat
fi
eld in
the Inland Paci
fi
c Northwest, a region where phenazine production and the biocontrol
potential of indigenous
Pseudomonas
species have been studied for decades (22, 23,
25, 27). We designed a culture-based assay to measure sensitivity to PCA, which is the
best-studied and one of the most abundant phenazines in this environment (22, 23,
48). We also performed full-length 16S rRNA gene sequencing of our isolates to assess
the relationship between taxonomy and PCA resistance. Finally, to assess potential
physiological correlates of PCA resistance, we measured PCA reduction rates for a sub-
set of phenotypically diverse strains and investigated the effects of broad-spectrum
ef
fl
ux pump inhibitors on growth in the presence of PCA.
RESULTS
Taxonomic diversity of culturable bacteria from dryland wheat rhizospheres
and bulk soil.
A total of 166 strains of bacteria were isolated from 12 soil samples col-
lected from a wheat
fi
eld at Washington State University
’
s Lind Dryland Research
Station in early August 2019, shortly after the wheat harvest. The samples comprised 4
replicates each of wheat rhizosphere (
“
Wheat
”
), bulk soil collected between planted
rows (
“
Between
”
), and bulk soil from a virgin patch of uncultivated soil adjacent to the
fi
eld (
“
Virgin
”
). Full-length 16S rRNA gene sequencing revealed that the isolates repre-
sented 24 genera across 4 phyla: Actinobacteria, Bacteroi
detes, Firmicutes, and
Proteobacteria. Most isolates from the bulk soil samples (Between and Virgin) were
Actinobacteria or Firmicutes. In Wheat samples, on the other hand, the combined propor-
tions of these two phyla were lower, accounting for 18 to 50% and 10 to 25% of isolates,
respectively, while Proteobacteria accounted for approximately 25 to 60% of isolates
depending on the replicate, and 1 to 3 strains of Bacteroidetes were also detected in each
replicate (5 to 12% of isolates) (Fig. 1).
FIG 1
Taxonomic distribution of bacterial isolates from wheat rhizosphere and bulk soil samples. This
plot depicts the proportion of isolates from each sample that belonged to the 4 represented phyla.
Each column represents one soil sample, and each box within the columns represents an individual
isolate colored by the phylum to which it belongs (e.g., 6 boxes comprising one column indicate that
6 strains were isolated from that sample). B = Between (bulk soil), V = Virgin (bulk soil), W = Wheat
(rhizosphere).
Phenazine Resistance in Wheat Rhizosphere Community
Applied and Environmental Microbiology
March 2022 Volume 88 Issue 6 e02320-21
aem.asm.org
3
Downloaded from https://journals.asm.org/journal/aem on 05 April 2022 by 131.215.249.7.
Taxonomic and spatial distribution of PCA resistance phenotypes.
We screened
our isolates for resistance to PCA at both circumneutral and acidic pH (7.3 and 5.1,
respectively), because the toxicity of PCA to diverse organisms is known to vary
depending on pH (49, 50) and because the bulk soil and rhizosphere pH of wheat
fi
elds
in the Inland Paci
fi
c Northwest can vary by
.
3 units (pH 4.3 to 8.0) depending on geo-
graphic location and fertilizer treatment status (51). The pH-dependency of PCA toxic-
ity has been attributed to the fact that the deprotonated form of PCA is negatively
charged (Fig. S1). The negative charge on bacterial cell walls and the outer membrane
of Gram-negative bacteria (or the negative membrane potential of eukaryotic cells)
likely hinder the uptake of this species. The protonated form of PCA, on the other
hand, is neutral and presumably can passively diffuse across cell membranes given the
small size and hydrophobic nature of the molecule (50, 52). The pKa of PCA is 4.24 at
25°C. Thus, at pH 7, only 0.2% of PCA in solution is protonated compared to 14.8% at
pH 5, typically leading to greater toxicity at the lower pH (49). In designing the screen,
we additionally decided to focus on comparing growth at a single concentration of
PCA versus a PCA-free control condition, rather than performing an MIC assay based
on a dilution series, to enable resource-ef
fi
cient comparison of a large number of
strains. We chose 100
m
M as the working concentration of PCA as this is likely to be in
a physiologically relevant range based on concentrations measured both in pure cul-
tures and in the
fi
eld. In broth cultures of biocontrol strains of
Pseudomonas
, PCA accu-
mulates to concentrations ranging from dozens to hundreds of micromolar (53, 54). In
natural wheat rhizospheres, PCA has been detected at nanomolar concentrations (22),
but these bulk measurements almost certainly underestimate local concentrations at
the micron scale given that bacteria colonize the rhizosphere in a patchy manner (23).
Notably, PCA can accumulate in bio
fi
lms to concentrations 360-fold greater on a per-
volume basis compared to broth cultures (53), suggesting that local concentrations of
PCA in the rhizosphere, where biocontrol strains form robust bio
fi
lms (55), may be
orders of magnitude higher than the reported bulk values.
The screen was performed using 24-well plates and 0.1
tryptic soy agar (TSA) that
was either left unadjusted (pH 7.3 to 7.5) or adjusted to pH 5.1. Each isolate was spot-
ted onto agar in separate wells to prevent cross talk and antagonism between the
strains, and image analysis was used to derive quantitative information about the
growth of each strain over 7 days (with one time point per day). Because several strains
among our 166 isolates appeared to be duplicates of each other based on 16S
sequence and colony morphology, we restricted the screen to 108 strains that we
judged likely to be unique (Table 1). Where multiple strains appeared identical,
we chose a representative strain. We also included 30 strains obtained from public cul-
ture collections that represented species found among our isolates (Table 1), to investi-
gate the extent to which PCA resistance phenotypes are consistent within species
across different strains isolated from different geographical locations. Accurate quanti-
fi
cation of growth was not possible in this screen for certain strains that formed trans-
parent colonies, produced dark pigments, or tended to turn mucoid and spread
(Fig. S2). Nevertheless, this assay enabled us to derive detailed pro
fi
les of PCA sensitiv-
ity and resistance for most of our strains. Growth assays based on planktonic cultures
were precluded due to the tendency of
Streptomyces
species to form clumps as well as
poor growth of
Agromyces
,
Neobacillus
, and
Paenibacillus
strains in liquid 0.1
tryptic
soy broth.
To compare PCA susceptibility across strains, we
fi
rst focused on a single time point
snapshot of each strain
’
s phenotype taken at the equivalent of early stationary phase
(i.e., around the time that the spots on non-PCA control plates reached their maximal
density). In accordance with previous reports of the pH dependency of PCA toxicity,
we found that more of our strains were sensitive to PCA at pH 5.1 than at pH 7.3, and
strains that were mildly inhibited by PCA at pH 7.3 were typically inhibited more
strongly at pH 5.1 (Fig. 2A and B). At pH 5.1, strains of Actinobacteria and Firmicutes
were strongly inhibited by PCA, while most Proteobacteria were relatively resistant.
Perry and Newman
Applied and Environmental Microbiology
March 2022 Volume 88 Issue 6 e02320-21
aem.asm.org
4
Downloaded from https://journals.asm.org/journal/aem on 05 April 2022 by 131.215.249.7.
TABLE 1
Organisms used in this study
Species
Strain ID
Source
a
Agromyces albus
V4I21
This study
Agromyces ramosus
V3I3
This study
Arthrobacter agilis
W3I6
This study
Arthrobacter globiformus
V3I9
This study
Arthrobacter globiformus
V4I26
This study
Arthrobacter globiformus
W2I3
This study
Arthrobacter humicola
B3I4
This study
Arthrobacter pascens
V4I24
This study
Arthrobacter pascens
W1I14
This study
Arthrobacter pascens
W4I8
This study
Arthrobacter pokkalii
B2I5
This study
Arthrobacter woluwensis
W4I2
This study
Clavibacter michiganensis
B3I6
This study
Microbacterium murale
W2I7
This study
Microbacterium natoriense
W4I9-1
This study
Microbacterium phyllosphaerae
W4I20
This study
Microbacterium
sp.
W4I4
This study
Nocardia salmonicida
V2I2
This study
Pseudarthrobacter chlorophenolicus
B2I9
This study
Pseudarthrobacter de
fl
uvii
B1I2
This study
Pseudarthrobacter de
fl
uvii
W1I2
This study
Pseudarthrobacter equi
W3I10
This study
Pseudarthrobacter phenanthrenivorans
W4I12
This study
Pseudarthrobacter siccitolerans
B3I9
This study
Pseudarthrobacter siccitolerans
V1I9
This study
Pseudarthrobacter siccitolerans
V3I10
This study
Pseudarthrobacter siccitolerans
W1I19
This study
Pseudarthrobacter siccitolerans
W1I3
This study
Pseudarthrobacter siccitolerans
W4I7
This study
Pseudarthrobacter
sp.
V4I6
This study
Streptomyces afghaniensis
W1I20
This study
Streptomyces africanus
B3I12
This study
Streptomyces bobili
V4I16
This study
Streptomyces bungoensis
B4I12
This study
Streptomyces canus
B4I3
This study
Streptomyces canus
V4I20
This study
Streptomyces canus
V4I22
This study
Streptomyces cyaneofuscatus
W4I9-2
This study
Streptomyces ederensis
V2I4
This study
Streptomyces
fi
ldesensis
B3I13
This study
Streptomyces galbus
W1I17
This study
Streptomyces luteogriseus
W4I19-1
This study
Streptomyces novaecaesareae
B3I7
This study
Streptomyces novaecaesareae
B3I8
This study
Streptomyces phaeochromogenes
V1I8
This study
Streptomyces plumbiresistens
V4I2
This study
Streptomyces rishiriensis
B2I10
This study
Streptomyces rishiriensis
B2I6
This study
Streptomyces rishiriensis
B4I13
This study
Streptomyces rishiriensis
W4I19-2
This study
Streptomyces
sp.
B1I3
This study
Streptomyces
sp.
B2I16
This study
Streptomyces
sp.
V1I6
This study
Streptomyces
sp.
V3I7
This study
Streptomyces
sp.
V4I23
This study
Streptomyces
sp.
V4I8
This study
Streptomyces
sp.
V1I1
This study
Streptomyces
sp.
V2I9
This study
Streptomyces tauricus
V3I8
This study
Streptomyces turgidiscabies
V4I19
This study
Streptomyces turgidiscabies
W2I16
This study
Chitinophaga ginsengisegetis
W2I13
This study
(Continued on next page)
Phenazine Resistance in Wheat Rhizosphere Community
Applied and Environmental Microbiology
March 2022 Volume 88 Issue 6 e02320-21
aem.asm.org
5
Downloaded from https://journals.asm.org/journal/aem on 05 April 2022 by 131.215.249.7.
TABLE 1
(Continued)
Species
Strain ID
Source
a
Chitinophaga ginsengisegetis
W3I9
This study
Chryseobacterium ginsenosidimutans
W1I9
This study
Chryseobacterium shigense
W4I1
This study
Pedobacter kyongii
W3I1
This study
Pedobacter kyongii
W4I14
This study
Brevibacterium frigoritolerans
B2I3
This study
Bacillus atrophaeus
V3I1
This study
Bacillus idriensis
V2I10
This study
Bacillus pumilus
B3I3
This study
Neobacillus drentensis
B4I4
This study
Neobacillus drentensis
B4I6
This study
Neobacillus drentensis
V3I5
This study
Neobacillus niacini
V4I10
This study
Neobacillus niacini
V4I12
This study
Neobacillus niacini
V4I25
This study
Paenibacillus alginolyticus
V4I5
This study
Paenibacillus alginolyticus
V4I7
This study
Paenibacillus marchantiophytorum
V4I3
This study
Paenibacillus mobilis
W2I17
This study
Paenibacillus mobilis
W4I10
This study
Paenibacillus nebraskensis
V4I9
This study
Paenibacillus rhizoryzae
B1I5
This study
Paenibacillus rhizoryzae
B2I8
This study
Peribacillus simplex
B2I15
This study
Peribacillus simplex
B2I2
This study
Peribacillus simplex
V2I11
This study
Priestia megaterium
B1I1
This study
Priestia megaterium
B1I6
This study
Priestia megaterium
W1I8
This study
Inquilinus ginsengisoli
V4I1
This study
Luteibacter jiangsuensis
W1I16
This study
Massilia niastensis
W1I21
This study
Massilia plicata
W1I10
This study
Pantoea agglomerans
W2I1
This study
Paraburkholderia graminis
W1I13
This study
Phyllobacterium ifrigiyense
W4I11
This study
Pseudomonas brenneri
W2I6
This study
Pseudomonas cedrina
W1I11
This study
Pseudomonas
fl
uorescens
W3I7
This study
Pseudomonas orientalis
W4I3
This study
Pseudomonas synxantha
W2I4
This study
Sphingomonas faeni
W4I17
This study
Variovorax boronicumulans
W1I1
This study
Variovorax boronicumulans
W2I14
This study
Variovorax paradoxus
W1I18
This study
Variovorax paradoxus
W2I8
This study
Agromyces agilis
DSM 20550
DSMZ
Massilia niastensis
DSM 21313
DSMZ
Massilia plicata
DSM 17505
DSMZ
Pseudarthrobacter equi
DSM 23395
DSMZ
Pseudarthrobacter phenanthrenivorans
DSM 18606
DSMZ
Pseudomonas brenneri
DSM 15294
DSMZ
Sphingomonas faeni
DSM 14747
DSMZ
Streptomyces luteogriseus
DSM 40483
DSMZ
Agromyces albus
NBRC 103057
NBRC
Agromyces ramosus
NBRC 13899
NBRC
Arthrobacter globiformus
NBRC 12137
NBRC
Chitinophaga ginsengisegetis
NBRC 109750
NBRC
Neobacillus drentensis
NBRC 102427
NBRC
Peribacillus simplex
NBRC 15720
NBRC
Priestia megaterium
NBRC 15308
NBRC
Streptomyces rishiriensis
NBRC 13407
NBRC
Variovorax boronicumulans
NBRC 103145
NBRC
(Continued on next page)
Perry and Newman
Applied and Environmental Microbiology
March 2022 Volume 88 Issue 6 e02320-21
aem.asm.org
6
Downloaded from https://journals.asm.org/journal/aem on 05 April 2022 by 131.215.249.7.
Members of Bacteroidetes exhibited variable responses, ranging from mild to severe
growth inhibition (Fig. 2A and B). At pH 7.3, there was considerably more phenotypic
variation among the Actinobacteria (Fig. 2A and B), particularly within the
Streptomyces
ge-
nus. In some cases, there was noticeable variation within the same putative species of
Streptomyces
(Fig. S3). In addition, two strains of
Microbacterium
(W2I7 and W4I20) and
one strain of
Arthrobacter
(W3I6) grew slightly better in the presence of PCA at pH 7.3 com-
pared to the control condition, perhaps suggesting that they can use PCA as a carbon
source or otherwise bene
fi
t from its presence at a pH where toxicity is limited (Fig. 2B). On
the other hand, in contrast to the phenotypic variation seen among Actinobacteria,
Firmicutes remained almost universally sensitive to PCA at pH 7.3 (albeit somewhat less so
than at pH 5.1), while all members of Bacteroidetes were resistant to PCA at pH 7.3, which
contrasts with their sensitivity to PCA at pH 5.1 (Fig. 2A and B). Proteobacteria remained
generally resistant (Fig. 2A and B) with the most notable exception being a strain of
Sphingomonas faeni
, isolate W4I17, that was inhibited by PCA at both pH 7.3 and pH
5.1 (Fig. 2A and B).
Interestingly, while most strains yielded qualitatively consistent results across inde-
pendent biological replicates that were separated by months (Fig. S4 to S7), a few
strains displayed variable growth and lag times at pH 5.1 either in the absence of PCA
(mostly among Actinobacteria and Firmicutes) or in the presence of PCA (a few strains
among the Bacteroidetes). Growth at pH 7.3 was generally more consistent with the
exception being two strains of
Neobacillus niacini
(V4I10 and V4I25) that initially failed
to grow in the presence of PCA but grew with minimal to no inhibition in subsequent
replicates. The reasons for these discrepancies are unclear but may be related to varia-
tions in how long the strains had been in stationary phase before inoculating the ex-
perimental cultures, which was dif
fi
cult to control precisely due to different growth
rates across a large number of strains. The trace nutrient content may also have varied
between different batches of media because the
fi
rst replicate was performed using a
different lot of tryptic soy broth compared to subsequent replicates. Thus, the growth
and PCA sensitivity of some strains may be in
fl
uenced by environmental factors
beyond pH that remain to be elucidated.
We next examined whether there was any evidence of a correlation between PCA
resistance phenotypes and the type of soil each strain was isolated from (Between,
Virgin, or Wheat). A previous study based on samples taken from the same wheat
fi
eld
found that the relative abundance of phenazine producers was higher in the wheat rhi-
zosphere compared to adjacent bulk soil (48), a
fi
nding that we also con
fi
rmed using
shotgun metagenomic sequencing of our soil samples (Fig. 2C). We therefore hypothe-
sized that if PCA-mediated antibiosis had shaped the bacterial community composition
of this
fi
eld, the prevalence of PCA resistance would be higher among isolates from the
rhizosphere samples. Indeed, the Wheat samples had the highest proportion of PCA-
TABLE 1
(Continued)
Species
Strain ID
Source
a
Variovorax paradoxus
NBRC 15149
NBRC
Bacillus atrophaeus
NRS-213
NRRL
Bacillus pumilus
NRS-272
NRRL
Streptomyces africanus
B-24243
NRRL
Streptomyces aurantiacus
ISP-5412
NRRL
Streptomyces bobili
B-1338
NRRL
Streptomyces bungoensis
B-24305
NRRL
Streptomyces canus
B-3980
NRRL
Streptomyces ciscaucasicus
ISP-5275
NRRL
Streptomyces cyaneofuscatus
B-2570
NRRL
Streptomyces novaecaesareae
B-1267
NRRL
Streptomyces peucetius
B-3826
NRRL
Streptomyces phaeochromogenes
B-1248
NRRL
a
DSMZ, Deutsche Sammlung von Mikroorganismen und Zellkulturen; NBRC, Biological Resource Center, NITE
(NBRC); NRRL, ARS Culture Collection (NRRL).
Phenazine Resistance in Wheat Rhizosphere Community
Applied and Environmental Microbiology
March 2022 Volume 88 Issue 6 e02320-21
aem.asm.org
7
Downloaded from https://journals.asm.org/journal/aem on 05 April 2022 by 131.215.249.7.
resistant isolates, regardless of whether resistance was assessed at pH 5.1 (Fig. 2D) or
pH 7.3 (Fig. 2E). In contrast, all strains from Between or Virgin samples were highly sen-
sitive to PCA at pH 5.1, although at pH 7.3, the resistance phenotypes of isolates from
either type of bulk soil spanned the full range from highly sensitive to completely
FIG 2
Distribution of PCA resistance phenotypes across phyla and soil sample type. (A and B) Heat maps depicting the growth of the strains at pH 5.1 (A)
and pH 7.3 (B). Each row represents a strain. The left columns are colored according to the ratio of growth on PCA-containing agar versus PCA-free agar.
Magenta indicates sensitivity to PCA while green indicates resistance to PCA. The right two columns in each heat map are colored according to the
separate values for growth on PCA-containing agar (
1
PCA) or solvent control agar (
2
PCA) with darker green indicating more growth. Values are the mean
of two to four biological replicates that each comprised three technical replicates. Asterisks indicate strains that displayed markedly variable s
usceptibility
to PCA across biological replicates. Strains for which fewer than two biological replicates grew to stationary phase in the PCA-free control at the gi
ven pH,
or for which color interfered with growth quanti
fi
cation, were omitted from this analysis. See Materials and Methods for a description of how growth was
quanti
fi
ed. (C) Relative abundance of putative phenazine producers (
phz
1
bacteria) across the three types of soil samples. Error bars represent the
standard deviation. (D and E) Density plots (i.e., smoothed histograms) representing the distribution of PCA resistance phenotypes at pH 5.1 (D) and
pH 7.3
(E) among strains isolated from Between (bulk), Virgin (bulk), or Wheat (rhizosphere) soil. Higher values along the
x
-axis indicate greater resistance to PCA.
If multiple identical strains were isolated from the same soil type, only one representative was counted. Thus, 26 strains were isolated from Between
,39
from Virgin, and 45 from Wheat. Colored tick marks above the
x
-axis represent where individual isolates fall in the range of PCA resistance phenotypes.
Perry and Newman
Applied and Environmental Microbiology
March 2022 Volume 88 Issue 6 e02320-21
aem.asm.org
8
Downloaded from https://journals.asm.org/journal/aem on 05 April 2022 by 131.215.249.7.
resistant. However, these
fi
ndings are con
fl
ated with the fact that all our Bacteroidetes
strains and all but one strain of Proteobacteria were exclusively isolated from the
Wheat samples and that the members of these two phyla were generally relatively re-
sistant to PCA, especially at pH 7.3. Further experiments will be necessary to distin-
guish whether the enrichment of PCA-resistant phenotypes in the rhizosphere samples
is a consequence of PCA-mediated antibiosis or merely an indirect re
fl
ection of other
factors that favor the growth of Proteobacteria and Bacteroidetes in the rhizosphere.
Finally, visualizing the growth of each strain over the full 7-day time course revealed
more nuanced variations among the PCA resistance phenotypes at both pH 5.1 and pH
7.3 (Fig. 3 and Fig. S4 to S7). For example, while some strains of Bacteroidetes dis-
played increased lag and/or lower growth rates in the presence of PCA at pH 5.1, the
growth on PCA often eventually caught up to the PCA-free controls. At pH 7.3, the
effect of PCA on
Peribacillus
species was a combination of increased lag and lower
fi
nal
cell density, but not decreased maximal growth rate, compared to no growth at all on
FIG 3
Growth of representative strains over time with and without PCA. Growth was quanti
fi
ed as described in the Methods. Solid lines represent the
growth of spotted cultures on PCA-free agar, and dashed lines represent the growth of spotted cultures on agar containing 100
m
M PCA. Blue represents
growth at pH 7.3 and pink represents growth at pH 5.1. Data points are the mean of three technical replicates from a representative biological replicat
e
for each strain, and the shaded ribbon represents the standard deviation. The selected strains depicted here cover the range of PCA susceptibility
phenotypes observed in each genus. The full set of biological replicates for all tested isolates is in Fig. S4 to S7.
Phenazine Resistance in Wheat Rhizosphere Community
Applied and Environmental Microbiology
March 2022 Volume 88 Issue 6 e02320-21
aem.asm.org
9
Downloaded from https://journals.asm.org/journal/aem on 05 April 2022 by 131.215.249.7.
PCA at pH 5.1. This was distinct from the lower growth rates seen for certain strains of
Bacillus
,
Neobacillus
, and
Paenibacillus
at pH 7.3. A few strains of the latter three genera
appeared to be still unable to grow on PCA at pH 7.3 based on the values derived from
image analysis (Fig. S5). However, examination of the plate images by eye revealed
that this sometimes re
fl
ected a limitation in the sensitivity of our imaging and quanti
fi
-
cation methods to low levels of growth, rather than a true lack of growth (Fig. S2).
PCA reduction rates in relation to PCA susceptibility.
We next sought to deter-
mine whether there are physiological correlates of PCA susceptibility or resistance
across taxonomically diverse bacteria. Given that phenazines are redox-active mole-
cules whose toxicity is thought to be related to ROS generation because of redox
cycling in the cell (38, 56), we hypothesized that susceptibility to PCA would be corre-
lated with higher redox-cycling rates. PCA oxidation occurs rapidly and abiotically in
the presence of oxygen (11), suggesting that the rate of PCA reduction would be the
primary driver of differences in redox-cycling rates across strains under oxic conditions.
Therefore, to test our hypothesis, we measured the rate of PCA reduction under anoxia
as a proxy for the true redox-cycling rate using a subset of strains that covered a range
of taxonomic groups and PCA susceptibility phenotypes.
For each tested strain, the PCA reduction rate was typically higher at pH 5.1 than at
pH 7.3 (Fig. 4A), matching the expectation that PCA uptake would tend to be faster at
a lower pH and the observation that most susceptible strains were more sensitive to
PCA at pH 5.1. The exceptions were: (i) strains belonging to the Firmicutes, which had
lower reduction rates at pH 5.1 in our assay despite being more susceptible to PCA
under this condition, and (ii)
Sphingomonas faeni
W4I17, which was more susceptible
to PCA at pH 7.3 than at pH 5.1 (Fig. 2A and B), matching the pH dependency of its
rate of PCA reduction. Notably, some of the tested Firmicutes often struggled to grow
at pH 5.1 even in the absence of PCA (Fig. S5). This observation might account for the
inhibitory effect of acidic pH on PCA reduction in these strains, given that a functional
metabolism would be required for the generation of cellular reductants, such as
NADH, that can indirectly or directly reduce PCA. For the strains of Firmicutes that
were able to grow at acidic pH (namely, B1I1 and B1I6), the reason for their lower PCA
reduction rate at pH 5.1 remains unclear.
We also examined whether there was any correlation between PCA reduction rates
and resistance levels across different strains at each pH. To do so, we plotted each
strain
’
s PCA reduction rate against the early stationary-phase snapshot ratio of growth
on PCA versus no PCA (Fig. 4B). We also calculated Spearman
’
s correlation coef
fi
cient
(
r
s
). Interestingly, there was a statistically signi
fi
cant negative correlation between
reduction rate and resistance at pH 7.3 (
r
s
=
2
0.70,
P
,
0.001) but not at pH 5.1 (
r
s
=
2
0.16,
P
= 0.4485). However, there were exceptions to the trend even at pH 7.3. For
example, while PCA severely and consistently inhibited the growth of the four fastest
PCA-reducing strains, the
fi
fth-fastest strain, W2I1 (
Pantoea agglomerans
), appeared to
be completely resistant to PCA in one biological replicate, though its growth was
mildly inhibited by PCA in another replicate (Fig. S7). In addition, strains with reduction
rates in the middle range of 0.2 to 0.4 nmol per optical density unit at 600 nm (OD
600
)
per min spanned the full range of resistance phenotypes, from a growth ratio of 0.12
(
Peribacillus simplex
B2I2) to 0.96 (
Phyllobacterium ifrigiyense
W4I11). Taken together,
these data suggest that a low reduction rate may help mitigate PCA toxicity but is nei-
ther necessary nor suf
fi
cient for resistance.
Effects of ef
fl
ux pump inhibitors on PCA susceptibility.
Given that the PCA reduc-
tion rate alone was insuf
fi
cient to account for the range of sensitivities to PCA, we
asked if ef
fl
ux pumps might serve as an important contributor to PCA resistance in
some strains. To address this question, we used two ef
fl
ux pump inhibitors (EPIs) in
combination: (i) reserpine, which is thought to target ef
fl
ux pumps in the major facilita-
tor superfamily (MFS) (57), and (ii) phenylalanine-arginine
b
-naphthylamide (PA
b
N),
which is thought to target resistance-nodulation-division (RND) ef
fl
ux pumps (58),
although it may also permeabilize the outer membrane of Gram-negative bacteria (59,
60). We tested these EPIs on a subset of the strains that were used in the PCA reduction
Perry and Newman
Applied and Environmental Microbiology
March 2022 Volume 88 Issue 6 e02320-21
aem.asm.org
10
Downloaded from https://journals.asm.org/journal/aem on 05 April 2022 by 131.215.249.7.
assay. We excluded the
Bacillus
strains due to their tendency to settle on the bottom
of the wells during growth in the plate reader, which confounded the OD
600
readings.
Preliminary experiments revealed that different strains were differentially sensitive to
the EPIs themselves with growth inhibition evident even in PCA-free controls. We
therefore determined for each strain the maximal EPI concentration that, when the
two EPIs were used separately, did not markedly reduce the growth of PCA-free con-
trols (up to a maximum dose of 20
m
g/mL for reserpine and 50
m
g/mL for PA
b
N)
(Table S2). Interestingly, for many strains, the combination of both EPIs remained
highly toxic even after this optimization step, particularly at pH 7.3, possibly indicating
synergistic toxicity or redundant roles of different ef
fl
ux pumps in preventing the intra-
cellular accumulation of toxic metabolic intermediates (Fig. S9). The toxicity appeared
to be pH-dependent, suggesting that the protonation state of the EPIs might in
fl
uence
their ef
fi
cacy. In these cases, it was not possible to assess the impact of ef
fl
ux pump in-
hibition on PCA sensitivity. However, for the remaining strains, three types of responses to
the EPIs emerged. For one group of strains, treatment with both EPIs in combination did
not affect their resistance to PCA (Fig. 5A and B). However, we cannot distinguish whether
this is because ef
fl
ux does not contribute to PCA resistance in these strains or because the
FIG 4
PCA reduction rates of selected strains at pH 5.1 and pH 7.3. (A) PCA reduction rates of representative strains from each
phylum at pH 5.1 and pH 7.3. The values represent the mean of three biological replicates and error bars represent the standard
deviation. Reduction rates were normalized to OD
600
values. (B) PCA reduction rates versus each strain
’
s growth ratio (
1
PCA/
2
PCA)
at early stationary phase. The growth ratios are the same values calculated and used in Fig. 2.
Phenazine Resistance in Wheat Rhizosphere Community
Applied and Environmental Microbiology
March 2022 Volume 88 Issue 6 e02320-21
aem.asm.org
11
Downloaded from https://journals.asm.org/journal/aem on 05 April 2022 by 131.215.249.7.
tested EPIs did not effectively target the relevant ef
fl
ux pumps. For the second group of
strains,theEPIsenhancedthetoxicityofPCA,whichwouldbeexpectedifef
fl
ux is an im-
portant mechanism of resistance (Fig. 5C). Finally, for the third group of strains, the EPIs
counterintuitively appeared to mitigate the toxicity of PCA, or vice versa (Fig. 5D).
Although unexpected, the latter phenomenon indirectly suggests that PCA interacts with
ef
fl
ux systems in these strains. For example, treatment with either PCA or the EPIs might
upregulate the expression of certain ef
fl
ux pumps, leading to decreased toxicity relative to
treatment with either type of toxin alone. Alternatively, given that some pumps in the
MFS family are thought to be bidirectional (61), MFS pumps may promote uptake rather
than the export of PCA in certain strains.
FIG 5
Effects of ef
fl
ux pump inhibitors on PCA susceptibility. (A and B) Growth curves of strains for
which treatment with a combination of the ef
fl
ux pump inhibitors (EPI) reserpine and PA
b
N did not
affect susceptibility to PCA at pH 5.1 (A) or pH 7.3 (B). (C) Growth curves of strains for which
treatment with reserpine and PA
b
N potentiated the toxicity of PCA. (D) Growth curves of strains for
which treatment with reserpine and PA
b
N improved growth in the presence of PCA (strain W4I17) or
vice versa (strains W1I16 and W4I9-1). In all panels, data points are the mean of three biological
replicates and the shaded ribbon represents the standard deviation. The concentrations of reserpine
and PA
b
N used for each strain can be found in Table S2.
Perry and Newman
Applied and Environmental Microbiology
March 2022 Volume 88 Issue 6 e02320-21
aem.asm.org
12
Downloaded from https://journals.asm.org/journal/aem on 05 April 2022 by 131.215.249.7.
DISCUSSION
In this study, we have characterized pro
fi
les of resistance to an agriculturally rele-
vant phenazine across taxonomically diverse bacteria isolated from a wheat
fi
eld where
phenazine producers are indigenous. We also examined potential physiological corre-
lates of phenazine resistance in a subset of these strains. Our
fi
ndings establish a basis
for inferring whether intrinsic resistance is a factor that affects how phenazine produc-
tion shapes bacterial communities in the rhizosphere. It will now be possible to test
data-driven predictions regarding which strains, species, or even phyla are most likely
to be affected by phenazine-mediated antibiosis, even though the mechanistic under-
pinnings of phenazine susceptibility or resistance are likely multifactorial and appear
to differ across species.
One of our more notable
fi
ndings is that the previously reported pH dependency of
PCA toxicity varies across bacterial taxonomic groups. For most Actinobacteria and
Firmicutes, sensitivity to PCA was higher at pH 5.1 than at pH 7.3, indicating greater
toxicity at the lower pH as expected according to the pKa of PCA. On the other hand,
nearly all tested Proteobacteria were completely resistant to 100
m
M PCA regardless of
pH, at least down to pH 5.1. Consequently, at pH 5.1, PCA resistance phenotypes
largely correlated with phylum and, more broadly, a Gram-positive versus Gram-nega-
tive divide, which has previously been reported for the toxicity of another phenazine,
pyocyanin (33). The Gram-positive versus Gram-negative divide at pH 5.1 is perhaps
not surprising because a signi
fi
cant proportion (
;
14%) of PCA in solution is proto-
nated at this pH and, therefore, would not be repelled by the negatively charged cell
wall. Under this condition, the outer membrane of Gram-negative species presumably
presents an additional barrier to the entry of PCA, helping to limit the intracellular
accumulation of the toxin in the same manner as for numerous other antibiotics (62).
Less expected, however, was the within-phylum and even within-species variation in
PCA resistance phenotypes at pH 7.3 among certain taxonomic groups. Interestingly,
several
Streptomyces
strains exhibited at least some PCA-dependent growth inhibition
at pH 7.3, even though many
Streptomyces
species can produce their toxic phenazines
(32, 48). Given that the major limit on PCA toxicity at circumneutral pH is thought to
be its ability to enter cells, the phenotypic variability at pH 7.3 may indicate the pres-
ence of transporters or channels capable of phenazine uptake in some PCA-sensitive
Actinobacteria. Alternatively, it is possible that some PCA-sensitive strains locally acidi-
fi
ed the growth medium, which was only weakly buffered (1.4 mM K
2
HPO
4
).
Importantly, despite the general Gram-positive versus Gram-negative divide in their
sensitivity to PCA at pH 5.1, possessing an outer membrane is not a leakproof shield
against PCA toxicity. Proteobacteria as a group were more resistant to PCA at pH 5.1
compared to strains of Bacteroidetes, most of which exhibited increased lag in the
presence of PCA, even though both phyla comprise Gram-negative bacteria. Intriguingly,
one major difference between these clades is that Proteobacteria generally utilize ubiqui-
none as an electron carrier during aerobic growth (63), while the Bacteroidetes genera
screened in this study (
Chitinophaga
,
Chryseobacterium
,and
Pedobacter
)utilizemenaqui-
none (64
–
66). Menaquinones have a lower reduction potential compared to ubiquinones
(67). The standard reduction potential of menaquinone is still higher than that of PCA
(
2
74 mV compared to
2
177 mV) (52, 67), indicating that PCA likely is not reduced by
menaquinone. Nevertheless, this difference with ubiquinone raises the possibility that PCA
may interact differently and, perhaps more readily with the aerobic respiratory electron
transport chain of menaquinone utilizing Bacteroidetes strains compared to Proteobacteria,
thereby generating more ROS and/or interfering with the generation of ATP. Interestingly,
another study has shown that the reduced form of a different phenazine with a low reduc-
tion potential, neutral red (3-amino-7-dimethylamino-2-methylphenazine), can directly trans-
fer electrons to menaquinone, bypassing the proton-pumping NADH dehydrogenase com-
plex that normally transfers electrons from NADH to menaquinone and
“
short-circuiting
”
the
electron transport chain. We hypothesize that a similar phenomenon may occur with PCA in
strains that rely on menaquinone. In future studies, this hypothesis could be tested by
Phenazine Resistance in Wheat Rhizosphere Community
Applied and Environmental Microbiology
March 2022 Volume 88 Issue 6 e02320-21
aem.asm.org
13
Downloaded from https://journals.asm.org/journal/aem on 05 April 2022 by 131.215.249.7.
performing
in vitro
experiments with puri
fi
ed menaquinone, ubiquinone, and reduced PCA
to directly test whether quinones oxidize the latter and, if so, determine whether the kinetics
differ between menaquinone and ubiquinone; by measuring ROS production and steady-
state ATP pools in selected strains of Bacteroidetes and Proteobacteria both in the presence
and absence of PCA; and by forcing species of Proteobacteria (such as
Escherichia coli
)that
utilize both ubiquinone and menaquinone in different branches of their electron transport
chains to rely only on the latter (for example, by deleting the genes for ubiquinone biosyn-
thesis) followed by reassessing their sensitivity to PCA to see if there was an effect.
Beyond the speci
fi
c strains screened in this study, the risk assessment of phenazine-
producing biocontrol strains, and the understanding of phenazine biology in general,
would bene
fi
t greatly from the development of a platform for the prediction of phena-
zine resistance phenotypes from genomic or phylogenetic information. The results of
this study already indicate that, depending on the environmental pH, phylogenetic in-
formation may be of limited utility for the prediction of PCA resistance given the varia-
tion in phenotypes for Actinobacteria at circumneutral pH. However, this variability
may also hold the key to identifying genome-based predictive markers of phenazine
resistance. Given the taxonomic and phenotypic diversity of our strain collection,
whole-genome sequencing of our isolates followed by comparative genomics could
potentially reveal such markers.
In summary, this work has laid the groundwork for rectifying a major gap in studies
of how the introduction of a phenazine-producing biocontrol strain affects rhizosphere
bacterial communities. Previous studies have lacked information about the baseline
prevalence of resistance in the native communities (34
–
37), and, in the absence of
such information, it is impossible to determine whether a negative result (lack of
change in the rhizosphere community) re
fl
ects a high prevalence of resistance to PCA
that is particular to the studied community versus a general lack of toxicity of PCA to
most bacteria or fundamental abiotic constraints on the antibacterial activity of PCA in
the rhizosphere (e.g., limited diffusion, adsorption to soil particles, etc.). Distinguishing
between these scenarios is key to assessing the risk of unwanted side effects in rhizo-
sphere communities upon the application of phenazine-producing biocontrol strains.
In addition, recent studies have demonstrated that phenazines produced by the
opportunistic pathogen
Pseudomonas aeruginosa
can promote bacterial tolerance and
resistance to clinical antibiotics (42, 68
–
70) and that these effects can extend to other
opportunistic pathogens that are resistant to phenazines (42). Thus, understanding the
prevalence and genetic determinants of resistance to phenazines may have implica-
tions for agriculture, human medicine, and beyond as we continue to uncover new ec-
ological roles for these multifaceted bacterial metabolites.
MATERIALS AND METHODS
Isolation of bacteria from wheat rhizosphere and bulk soil samples.
Three types of samples were
collected from a nonirrigated wheat
fi
eld at Washington State University
’
s Lind Dryland Research Station
on August 9, 2019: wheat plants and surrounding soil, bulk soil from in between the planted rows, and
bulk soil from a
“
virgin
”
hillside site that has never been farmed. The wheat had been harvested a few
weeks before sample collection. All samples were immediately stored on ice in clean plastic bags and
subsequently at 4°C for 4 days until processing. Rhizosphere soil samples were obtained by shaking the
wheat roots until only 1 to 2 mm of tightly adhering soil remained followed by excising the roots at the
crown with a sterile razor blade. The roots of 2 to 3 plants per replicate were placed in 50 mL conical
tubes with 30 mL of sterile deionized water, vortexed at top speed for 1 min, and treated in an ultrasonic
water bath for 1 min to dislodge bacteria from the roots. Bulk soil samples (1 g per sample) were proc-
essed in the same manner. Large soil particles were allowed to settle to the bottom of the tubes on the
benchtop, and 100
m
L each of a 10-fold dilution series of the supernatants was spread onto 0.1
TSA
plates containing 50
m
g/mL nystatin to inhibit fungal growth. The plates were incubated upside down
at room temperature in the dark and monitored for the appearance of new colonies over a week.
Colonies that appeared morphologically distinct in each sample were picked and restreaked on 0.1
TSA until visually pure cultures were obtained. Multiple representatives were also picked for the most
common colony types to account for strain variations that might not be apparent to the eye. Once the
streaks yielded uniform single colonies, the isolates were inoculated into 1.5 mL of 0.1
tryptic soy broth
(TSB) in 5 mL polycarbonate culture tubes and incubated at 30°C with shaking at 250 rpm. After 1 to
3 days of incubation, depending on when the cultures became turbid, 0.5 mL of each culture was mixed
Perry and Newman
Applied and Environmental Microbiology
March 2022 Volume 88 Issue 6 e02320-21
aem.asm.org
14
Downloaded from https://journals.asm.org/journal/aem on 05 April 2022 by 131.215.249.7.
with 0.5 mL of 50% glycerol and stored at
2
80°C. Some cultures never became turbid under these
growth conditions but yielded viable frozen stocks.
Species identi
fi
cation by 16S rRNA gene sequencing.
Single colonies or patches of morphologi-
cally pure streaks were picked and resuspended in 20
m
L sterile nuclease-free water. Colony PCR was
performed using GoTaq Green Master Mix (Promega) in 50
m
L reactions (1
m
L of cell suspension) accord-
ing to the manufacturer
’
s instructions. For putative
Streptomyces
(isolates that formed hard colonies
rooted in agar, often with aerial hyphae), the thermocycling protocol was modi
fi
ed to include a 10 min
initial heating step at 95°C (compared to 2 min for other samples). The primers used were 27F
(AGAGTTTGATCMTGGCTCAG) and 1492R (TACGGYTACCTTGTTACGACTT) (71). The PCR products were
run on a 1% agarose gel to verify the presence of a single band at the expected size (
;
1500 bp) fol-
lowed by puri
fi
cation with the Monarch PCR and DNA Cleanup kit (New England Biolabs). The puri
fi
ed
products were submitted for Sanger sequencing at Laragen, Inc. using the same 27F and 1492R primers.
The resulting forward and reverse sequences were aligned using MAFFT (
https://mafft.cbrc.jp/alignment/
server/
) and subjected to BLAST against the NCBI 16S rRNA sequences database. For a few strains, full-length
sequences could only be obtained following PCR on genomic DNA extracted using the DNeasy Blood and
Tissue kit (Qiagen) following the manufacturer
’
s instructions for Gram-positive bacteria. In addition, for two
strains (B1I5 and B2I8), the forward sequence consistently appeared to contain multiple products. We pre-
sumed that this was due to multiple primer binding sites or another sequencing artifact rather than mixed
cultures as the corresponding sequences from the other direction were clean. In these cases, only the clean
sequence was submitted to BLAST. The 16S rRNA gene sequences and closest BLAST hits for each strain are
provided in Table S1.
PCA resistance screen.
The optimized version of the PCA resistance screen was performed with four
conditions: 0.1
TSA (15 g/liter agar plus 3 g/liter tryptic soy broth no. 2 from MilliporeSigma) with or
without 100
m
M PCA at pH 7.3 or pH 5.1 (adjusted with HCl). PCA was purchased from Princeton
BioMolecular Research and dissolved in
fi
lter-sterilized 14 mM NaOH to make 10 mM stock solutions.
The PCA stock solution or solvent control (14 mM NaOH) was added at 1% vol/vol to autoclaved molten
0.1
TSA. We veri
fi
ed that this addition did not noticeably alter the pH of the medium by pipetting ali-
quots onto pH test strips. Subsequently, 1 mL of the medium was pipetted into each well in 24-well
polystyrene Cellstar cell culture plates (Greiner Bio-One). The plates were allowed to set and dry with
the lids off in a biological safety cabinet for 20 to 30 min followed by storage upside down with the lids
on at room temperature in the dark for 2 days before use.
Cell suspensions for inoculation in the screen were prepared in one of two ways. First, individual
strains were inoculated into 5 mL TSB cultures in glass culture tubes and incubated at 25°C with shaking
at 250 rpm. Strains that grew overnight were then diluted to an OD
600
of 0.05. Certain strains did not
grow well under this condition, including strains of
Agromyces
,
Streptomyces
,
Paenibacillus
, and
Neobacillus
. For these, some replicates were prepared by directly scraping cells from streaks grown on
0.1
TSA plates and resuspending the cells in 200
m
L TSB with pipetting and brief vortexing at top
speed. The OD
600
of these cell suspensions was adjusted to 0.05, except for
Streptomyces
due to their
fi
la-
mentous nature and tendency to clump. Instead, visible cell clumps were allowed to settle to the bot-
tom, and the inocula were taken from the visually clear upper portion. Subsequently, 10
m
L of each cell
suspension was pipetted onto the agar in a single well in the 24-well plates. Three adjacent wells per
condition were inoculated with each cell suspension, representing technical replicates. After the spots
dried, the plates were incubated at room temperature upside in the dark for up to 7 days. Every 24 h,
the plates were imaged in color at 600 dpi with an Epson Perfection V550 Photo
fl
atbed scanner. For
strains that were prepared both from liquid cultures and by resuspension from plates, we did not
observe any consistent effect on the resulting PCA susceptibility phenotypes.
Image analysis and quanti
fi
cation of growth.
Images from the scanner were analyzed using Fiji
(72). Circular regions of interest (ROIs) were drawn around each culture spot and the mean gray value
was measured for each ROI. We also measured the mean gray values of equivalent ROIs in the wells of
blank, uninoculated plates for each condition. The latter values were averaged across each 24-well blank
plate to give the
“
background
”
gray value, which was then subtracted from the mean gray values of the
culture spots. The resulting numbers were reported as the metric for growth. Importantly, while this
method generally worked well for comparisons across conditions within each strain, there are a few cav-
eats. First, this metric underestimated growth for strains that produced a dark pigment. Second, growth
was dif
fi
cult to quantify for a few strains that grew as nearly transparent colonies. Finally, this metric is
not very sensitive to low levels of growth. Nevertheless, for most strains, this approach captured visible
differences in growth across the four conditions in the screen.
Metagenomics-based estimation of the relative abundance of PCA producers.
DNA was extracted
from 250 mg of each soil sample using the DNeasy PowerSoil kit (Qiagen). Illumina metagenomic DNA
libraries were prepared and sequenced (150 bp single-end reads) at the Millard and Muriel Jacobs
Genetics and Genomics Laboratory at the California Institute of Technology. The resulting sequence
data were processed according to the pipeline described by Dar et al. (2020) (48).
PCA reduction assay.
Selected strains were grown in triplicate overnight in 5 mL TSB cultures at
25°C with shaking at 250 rpm. The cells were then concentrated by serially centrifuging 1 mL aliquots
at 9000 g for 2 min in microcentrifuge tubes, washed with 0.1
TSB (pH 7.3), and standardized to OD
600
= 8 in 0.1
TSB (pH 7.3). Subsequently, 100
m
L of each cell suspension was transferred into a 96-well
plate, and the plate was moved into a Coy anaerobic chamber (5% hydrogen/95% nitrogen headspace).
Using plasticware and media that had been passively degassed in the chamber for at least 3 days, elec-
trochemically reduced PCA was added to a
fi
nal concentration of 100
m
M in 190
m
L of 0.1
TSB (pH 5.1
or pH 7.3) per well in a black clear-bottom 96-well plate. A standard curve was also prepared using
Phenazine Resistance in Wheat Rhizosphere Community
Applied and Environmental Microbiology
March 2022 Volume 88 Issue 6 e02320-21
aem.asm.org
15
Downloaded from https://journals.asm.org/journal/aem on 05 April 2022 by 131.215.249.7.
reduced PCA at 1
m
M, 10
m
M, 50
m
M, and 100
m
M. Cells were added to each well (except the standard
curve) to an OD
600
of
;
0.4, and the plate was immediately transferred to a Biotek Synergy 5 plate reader
maintained at 30°C. OD
600
and
fl
uorescence readings (360 nm ex/528 nm em) were taken every 5 min
with continuous shaking on the
“
medium
”
setting. Reduction rates were derived from the slope of a
linear regression
fi
t to the visually linear range of the data using the function lm in R (see Fig. S8 for
an example). The rates were converted to nmol/OD
600
/min by relating the
fl
uorescence units to the
standard curve of reduced PCA concentrations and then dividing by the initial OD
600
reading for
each replicate. No appreciable change in OD
600
was observed throughout the experiment for any of
the tested strains.
Ef
fl
ux pump inhibition growth curves.
Reserpine was purchased from Sigma-Aldrich and dissolved
in dimethyl sulfoxide (DMSO) to make a 20 mg/mL stock solution. PA
b
N was purchased from
MedChemExpress and dissolved in distilled water to make a 50 mg/mL stock solution. Selected strains
were grown in triplicate overnight in 5 mL TSB cultures at 25°C with shaking at 250 rpm, washed once
with 0.1
TSB (pH 7.3), and inoculated to an initial OD
600
of 0.05 in 160
m
L of 0.1
TSB (pH 5.1 or pH 7.3)
in a 96-well plate. Each well contained either a combination of reserpine and PA
b
N (see Table S2 for the
concentrations used for each strain) or solvent control (DMSO at 0.1%
fi
nal concentration) along with ei-
ther 0 or 100
m
M PCA. The wells were topped with 50
m
L of autoclaved mineral oil to prevent evapora-
tion, and growth was monitored with OD
600
readings every 15 min in a Biotek Synergy 5 plate reader set
to 30°C with continuous shaking on the
“
medium
”
setting.
Data availability.
The metagenomic sequencing data generated in this study were deposited in the
Sequence Read Archive (SRA) under accession
PRJNA521160
.
SUPPLEMENTAL MATERIAL
Supplemental material is available online only.
SUPPLEMENTAL FILE 1
, PDF
fi
le, 2.2 MB.
SUPPLEMENTAL FILE 2
, XLSX
fi
le, 0.03 MB.
ACKNOWLEDGMENTS
We thank Linda Thomashow and David Weller of the USDA Agricultural Research
Service for providing the wheat rhizosphere and soil samples used in this study, and the
members of the Newman lab for helpful feedback throughout the process of designing
and analyzing the results of the PCA resistance screen. We also thank Zevin Condiotte
for assistance with the initial rounds of testing the design of the PCA resistance screen
and Richard Horak for assistance with extracting genomic DNA.
This material is based upon work supported by the National Science Foundation
Graduate Research Fellowship under Grant no. DGE-1745301. This work was also
supported by grants to D.K.N. from the ARO (W911NF-17-1-0024), NIH (1R01AI127850-
01A1), and the Resnick Sustainability Institute at Caltech.
We also thank the Doren Family Foundation for their support.
We declare no con
fl
ict of interest.
REFERENCES
1. Granato ET, Meiller-Legrand TA, Foster KR. 2019. The evolution and ecol-
ogy of bacterial warfare. Curr Biol 29:R521
–
R537.
https://doi.org/10.1016/j
.cub.2019.04.024
.
2. Bérdy J. 2012. Thoughts and facts about antibiotics: where we are now
and where we are heading. J Antibiot (Tokyo) 65:385
–
395.
https://doi
.org/10.1038/ja.2012.27
.
3. Aminov RI. 2010. A brief history of the antibiotic era: lessons learned and
challenges for the future. Front Microbiol 1:134.
4. Hutchings MI, Truman AW, Wilkinson B. 2019. Antibiotics: past, present
and future. Curr Opin Microbiol 51:72
–
80.
https://doi.org/10.1016/j.mib
.2019.10.008
.
5. Manyi-Loh C, Mamphweli S, Meyer E, Okoh A. 2018. Antibiotic use in agri-
culture and its consequential resistance in environmental sources: poten-
tial public health implications. Molecules 23:795.
https://doi.org/10.3390/
molecules23040795
.
6. Davies J, Davies D. 2010. Origins and evolution of antibiotic resistance.
Microbiol Mol Biol Rev 74:417
–
433.
https://doi.org/10.1128/MMBR
.00016-10
.
7. Sengupta S, Chattopadhyay MK, Grossart H-P. 2013. The multifaceted
roles of antibiotics and antibiotic resistance in nature. Front Microbiol 4:
47.
https://doi.org/10.3389/fmicb.2013.00047
.
8. Aminov RI. 2009. The role of antibiotics and antibiotic resistance in na-
ture. Environ Microbiol 11:2970
–
2988.
https://doi.org/10.1111/j.1462
-2920.2009.01972.x
.
9. Davies J, Spiegelman GB, Yim G. 2006. The world of subinhibitory antibi-
otic concentrations. Curr Opin Microbiol 9:445
–
453.
https://doi.org/10
.1016/j.mib.2006.08.006
.
10. Davies J. 2006. Are antibiotics naturally antibiotics? J Ind Microbiol Bio-
technol 33:496
–
499.
https://doi.org/10.1007/s10295-006-0112-5
.
11. Wang Y, Newman DK. 2008. Redox reactions of phenazine antibiotics
with ferric (hydr)oxides and molecular oxygen. Environ Sci Technol 42:
2380
–
2386.
https://doi.org/10.1021/es702290a
.
12. McRose DL, Newman DK. 2021. Redox-active antibiotics enhance phos-
phorus bioavailability. Science 371:1033
–
1037.
https://doi.org/10.1126/
science.abd1515
.
13. Glasser NR, Saunders SH, Newman DK. 2017. The colorful world of extrac-
ellular electron shuttles. Annu Rev Microbiol 71:731
–
751.
https://doi.org/
10.1146/annurev-micro-090816-093913
.
14. Glasser NR, Kern SE, Newman DK. 2014. Phenazine redox cycling enhan-
ces anaerobic survival in
Pseudomonas aeruginosa
by facilitating genera-
tion of ATP and a proton-motive force. Mol Microbiol 92:399
–
412.
https://
doi.org/10.1111/mmi.12566
.
Perry and Newman
Applied and Environmental Microbiology
March 2022 Volume 88 Issue 6 e02320-21
aem.asm.org
16
Downloaded from https://journals.asm.org/journal/aem on 05 April 2022 by 131.215.249.7.
15. Linares JF, Gustafsson I, Baquero F, Martinez JL. 2006. Antibiotics as inter-
microbial signaling agents instead of weapons. Proc Natl Acad Sci U S A
103:19484
–
19489.
https://doi.org/10.1073/pnas.0608949103
.
16. Dietrich LE, Price-Whelan A, Petersen A, Whiteley M, Newman DK. 2006. The
phenazine pyocyanin is a terminal signalling factor in the quorum sensing
network of
Pseudomonas aeruginosa
. Mol Microbiol 61:1308
–
1321.
https://
doi.org/10.1111/j.1365-2958.2006.05306.x
.
17. Demain AL, Fang A. 2000. The natural functions of secondary metabolites.
Adv Biochem Eng Biotechnol 69:1
–
39.
https://doi.org/10.1007/3-540-44964
-7_1
.
18. Davies J, Ryan KS. 2012. Introducing the parvome: bioactive compounds
in the microbial world. ACS Chem Biol 7:252
–
259.
https://doi.org/10
.1021/cb200337h
.
19. Handelsman J, Stabb EV. 1996. Biocontrol of soilborne plant pathogens.
Plant Cell 8:1855
–
1869.
https://doi.org/10.1105/tpc.8.10.1855
.
20. Haas D, Défago G. 2005. Biological control of soil-borne pathogens by
fl
u-
orescent pseudomonads. Nat Rev Microbiol 3:307
–
319.
https://doi.org/10
.1038/nrmicro1129
.
21. Tyc O, Song C, Dickschat JS, Vos M, Garbeva P. 2017. The ecological role
of volatile and soluble secondary metabolites produced by soil bacteria.
Trends Microbiol 25:280
–
292.
https://doi.org/10.1016/j.tim.2016.12.002
.
22. Mavrodi DV, Mavrodi OV, Parejko JA, Bonsall RF, Kwak Y-S, Paulitz TC,
Thomashow LS, Weller DM. 2012. Accumulation of the antibiotic phena-
zine-1-carboxylic acid in the rhizosphere of dryland cereals. Appl Environ
Microbiol 78:804
–
812.
https://doi.org/10.1128/AEM.06784-11
.
23. Thomashow LS, Weller DM, Bonsall RF, Pierson LS. 1990. Production of
the antibiotic phenazine-1-carboxylic acid by
fl
uorescent
Pseudomonas
species in the rhizosphere of wheat. Appl Environ Microbiol 56:908
–
912.
https://doi.org/10.1128/aem.56.4.908-912.1990
.
24. Bergsma-Vlami M, Prins ME, Raaijmakers JM. 2005. In
fl
uence of plant spe-
cies on population dynamics, genotypic diversity and antibiotic produc-
tion in the rhizosphere by indigenous
Pseudomonas
spp. FEMS Microbiol
Ecol 52:59
–
69.
https://doi.org/10.1016/j.femsec.2004.10.007
.
25. Mazzola M, Cook RJ, Thomashow LS, Weller DM, Pierson LS. 1992. Contri-
bution of phenazine antibiotic biosynthesis to the ecological competence
of
fl
uorescent pseudomonads in soil habitats. Appl Environ Microbiol 58:
2616
–
2624.
https://doi.org/10.1128/aem.58.8.2616-2624.1992
.
26. Yu JM, Wang D, Pierson LS, Pierson EA. 2018. Effect of producing different
phenazines on bacterial
fi
tness and biological control in
Pseudomonas
chlororaphis
30-84. Plant Pathol J 34:44
–
58.
https://doi.org/10.5423/PPJ
.FT.12.2017.0277
.
27. Thomashow LS, Weller DM. 1988. Role of a phenazine antibiotic from
Pseudomonas
fl
uorescens
in biological control of
Gaeumannomyces grami-
nis
var.
tritici
. J Bacteriol 170:3499
–
3508.
https://doi.org/10.1128/jb.170.8
.3499-3508.1988
.
28. Haas D, Keel C. 2003. Regulation of antibiotic production in root-colo-
nizing
Peudomonas
spp. and relevance for biological control of plant
disease. Annu Rev Phytopathol 41:117
–
153.
https://doi.org/10.1146/
annurev.phyto.41.052002.095656
.
29. Mazurier S, Corberand T, Lemanceau P, Raaijmakers JM. 2009. Phenazine
antibiotics produced by
fl
uorescent pseudomonads contribute to natural
soil suppressiveness to Fusarium wilt. ISME J 3:977
–
991.
https://doi.org/
10.1038/ismej.2009.33
.
30. Mazzola M, Fujimoto DK, Thomashow LS, Cook RJ. 1995. Variation in sen-
sitivity of
Gaeumannomyces graminis
to antibiotics produced by
fl
uores-
cent
Pseudomonas
spp. and effect on biological control of take-all of
wheat. Appl Environ Microbiol 61:2554
–
2559.
https://doi.org/10.1128/
aem.61.7.2554-2559.1995
.
31. Costa KC, Bergkessel M, Saunders S, Korlach J, Newman DK. 2015. Enzy-
matic degradation of phenazines can generate energy and protect sensi-
tive organisms from toxicity. mBio 6:e01520-15
–
e01515.
https://doi.org/
10.1128/mBio.01520-15
.
32. Turner JM, Messenger AJ. 1986. Occurrence, biochemistry and physiology
of phenazine pigment production. Adv Microb Physiol 27:211
–
275.
https://doi.org/10.1016/S0065-2911(08)60306-9
.
33. Baron SS, Rowe JJ. 1981. Antibiotic action of pyocyanin. Antimicrob
Agents Chemother 20:814
–
820.
https://doi.org/10.1128/AAC.20.6.814
.
34. Kozdrój J, Trevors JT, van Elsas JD. 2004. In
fl
uence of introduced potential
biocontrol agents on maize seedling growth and bacterial community
structure in the rhizosphere. Soil Biol Biochem 36:1775
–
1784.
https://doi
.org/10.1016/j.soilbio.2004.04.034
.
35. Bankhead SB, Landa BB, Lutton E, Weller DM, Gardener BBM. 2004. Mini-
mal changes in rhizobacterial population structure following root
colonization by wild type and transgenic biocontrol strains. FEMS Micro-
biol Ecol 49:307
–
318.
https://doi.org/10.1016/j.femsec.2004.04.005
.
36. Roquigny R, Novinscak A, Léger G, Marcoux N, Joly DL, Filion M. 2018.
Deciphering the rhizosphere and geocaulosphere microbiomes of potato
following inoculation with the biocontrol agent
Pseudomonas
fl
uorescens
strain LBUM223. Phytobiomes 2:92
–
99.
https://doi.org/10.1094/PBIOMES
-03-18-0013-R
.
37. Gagliardi JV, Buyer JS, Angle JS, Russek-Cohen E. 2001. Structural and func-
tional analysis of whole-soil microbial communities for risk and ef
fi
cacy test-
ing following microbial inoculation of wheat roots in diverse soils. Soil Biol
Biochem 33:25
–
40.
https://doi.org/10.1016/S0038-0717(00)00110-3
.
38. Hassan HM, Fridovich I. 1980. Mechanism of the antibiotic action pyocya-
nine. J Bacteriol 141:156
–
163.
https://doi.org/10.1128/jb.141.1.156-163.1980
.
39. Voggu L, Schlag S, Biswas R, Rosenstein R, Rausch C, Götz F. 2006. Microe-
volution of cytochrome
bd
oxidase in Staphylococci and its implication in
resistance to respiratory toxins released by
Pseudomonas
. J Bacteriol 188:
8079
–
8086.
https://doi.org/10.1128/JB.00858-06
.
40. Perry EK, Newman DK. 2019. The transcription factors ActR and SoxR dif-
ferentially affect the phenazine tolerance of
Agrobacterium tumefaciens
.
Mol Microbiol 112:199
–
218.
https://doi.org/10.1111/mmi.14263
.
41. WolloscheckD,KrishnamoorthyG,NguyenJ,ZgurskayaHI.2018.Kineticcon-
trol of quorum sensing in
Pseudomonas aeruginosa
by multidrug ef
fl
ux
pumps. ACS Infect Dis 4:185
–
195.
https://doi.org/10.1021/acsinfecdis.7b00160
.
42. Meirelles LA, Perry EK, Bergkessel M, Newman DK. 2021. Bacterial
defenses against a natural antibiotic promote collateral resilience to clini-
cal antibiotics. PLoS Biol 19:e3001093.
https://doi.org/10.1371/journal
.pbio.3001093
.
43. Noto MJ, Burns WJ, Beavers WN, Skaar EP. 2017. Mechanisms of pyocya-
nin toxicity and genetic determinants of resistance in
Staphylococcus aur-
eus
. J Bacteriol 199:e00221-17.
https://doi.org/10.1128/JB.00221-17
.
44. Khare A, Tavazoie S. 2015. Multifactorial competition and resistance in a
two-species bacterial system. PLoS Genet 11:e1005715.
https://doi.org/10
.1371/journal.pgen.1005715
.
45. Meirelles LA, Newman DK. 2018. Both toxic and bene
fi
cial effects of pyo-
cyanin contribute to the lifecycle of
Pseudomonas aeruginosa
. Mol Micro-
biol 110:995
–
1010.
https://doi.org/10.1111/mmi.14132
.
46. Liu GY, Nizet V. 2009. Color me bad: microbial pigments as virulence fac-
tors. Trends Microbiol 17:406
–
413.
https://doi.org/10.1016/j.tim.2009.06
.006
.
47. Lau GW, Ran H, Kong F, Hassett DJ, Mavrodi D. 2004.
Pseudomonas aerugi-
nosa
pyocyanin is critical for lung infection in mice. Infect Immun 72:
4275
–
4278.
https://doi.org/10.1128/IAI.72.7.4275-4278.2004
.
48. Dar D, Thomashow LS, Weller DM, Newman DK. 2020. Global landscape
of phenazine biosynthesis and biodegradation reveals species-speci
fi
c
colonization patterns in agricultural soils and crop microbiomes. Elife 9:
e59726.
https://doi.org/10.7554/eLife.59726
.
49. Brisbane PG, Janik LJ, Tate ME, Warren RF. 1987. Revised structure for the
phenazine antibiotic from
Pseudomonas
fl
uorescens
2-79 (NRRL B-15132).
Antimicrob Agents Chemother 31:1967
–
1971.
https://doi.org/10.1128/AAC
.31.12.1967
.
50. Cezairliyan B, Vinayavekhin N, Grenfell-Lee D, Yuen GJ, Saghatelian A,
Ausubel FM. 2013. Identi
fi
cation of
Pseudomonas aeruginosa
phenazines
that kill
Caenorhabditis elegans
. PLoS Pathog 9:e1003101.
https://doi.org/
10.1371/journal.ppat.1003101
.
51. Smiley RW, Cook RJ. 1973. Relationship between take-all of wheat and rhi-
zosphere pH in soils fertilized with ammonium vs. nitrate-nitrogen. Phy-
topathology 63:882
–
890.
https://doi.org/10.1094/Phyto-63-882
.
52. Price-Whelan A, Dietrich LEP, Newman DK. 2006. Rethinking
“
secondary
”
metabolism: physiological roles for phenazine antibiotics. Nat Chem Biol
2:71
–
78.
https://doi.org/10.1038/nchembio764
.
53. Séveno NA, Morgan JAW, Wellington EMH. 2001. Growth of
Pseudomonas
aureofaciens
PGS12 and the dynamics of HHL and phenazine production
in liquid culture, on nutrient agar, and on plant roots. Microb Ecol 41:
314
–
324.
https://doi.org/10.1007/s002480000104
.
54. Tagele SB, Lee HG, Kim SW, Lee YS. 2019. Phenazine and 1-undecene pro-
ducing
Pseudomonas chlororaphis
subsp.
aurantiaca
strain KNU17Pc1 for
growth promotion and disease suppression in Korean maize cultivars. J
Microbiol Biotechnol 29:66
–
78.
https://doi.org/10.4014/jmb.1808.08026
.
55. LeTourneau MK, Marshall MJ, Cliff JB, Bonsall RF, Dohnalkova AC, Mavrodi
DV, Devi SI, Mavrodi OV, Harsh JB, Weller DM, Thomashow LS. 2018. Phen-
azine-1-carboxylic acid and soil moisture in
fl
uence bio
fi
lm development
and turnover of rhizobacterial biomass on wheat root surfaces. Environ
Microbiol 20:2178
–
2194.
https://doi.org/10.1111/1462-2920.14244
.
Phenazine Resistance in Wheat Rhizosphere Community
Applied and Environmental Microbiology
March 2022 Volume 88 Issue 6 e02320-21
aem.asm.org
17
Downloaded from https://journals.asm.org/journal/aem on 05 April 2022 by 131.215.249.7.