Metabolic Engineering 64 (2021) 15–25

Contents lists available at ScienceDirect

Metabolic Engineering
journal homepage: www.elsevier.com/locate/meteng

Metabolic engineering of E. coli for pyocyanin production

Adilson José da Silva c, d, Josivan de Souza Cunha d, Teri Hreha b, Kelli Cristina Micocci e,
Heloisa Sobreiro Selistre-de-Araujo f, Blanca Barquera b, c, **, Mattheos A.G. Koffas a, b, c, *
a
Department of Chemical and Biological Engineering, Rensselaer Polytechnic Institute, Troy, NY, USA
b
Department of Biological Sciences, Rensselaer Polytechnic Institute, Troy, NY, USA
c
Center for Biotechnology and Interdisciplinary Studies, Rensselaer Polytechnic Institute, Troy, NY, USA
d
Department of Chemical Engineering, Federal University of São Carlos, São Carlos, São Paulo, 13565-905, Brazil
e
Center for the Study of Social Insects, São Paulo State University, Rio Claro, São Paulo, 13506-900, Brazil
f
Department of Physiological Sciences, Federal University of São Carlos, São Carlos, São Paulo, 13565-905, Brazil

A R T I C L E I N F O A B S T R A C T

Keywords: Pyocyanin is a secondary metabolite from Pseudomonas aeruginosa that belongs to the class of phenazines, which
Pyocyanin are aromatic nitrogenous compounds with numerous biological functions. Besides its antifungal and antimi­
Phenazines crobial activities, pyocyanin is a remarkable redox-active molecule with potential applications ranging from the
Pseudomonas aeruginosa
pharma industry to the development of microbial fuel cells. Nevertheless, pyocyanin production has been
Pathway balance
Vitreoscilla hemoglobin
restricted to P. aeruginosa strains, limiting its practical applicability. In this study, the pyocyanin biosynthetic
pathway was engineered for the first time for high level production of this compound in a heterologous host.
Escherichia coli cells harboring the nine-gene pathway divided into two plasmids were able to produce and secrete
pyocyanin at higher levels than some Pseudomonas aeruginosa strains. The influence of culture and induction
parameters were evaluated, and the optimized conditions led to an increase of 3.5-fold on pyocyanin accumu­
lation. Pathway balancing was achieved by testing a set of plasmids with different copy numbers to optimize the
expression levels of pyocyanin biosynthetic genes, resulting in a fourfold difference in product titer among the
engineered strains. Further improvements were achieved by co-expression of Vitreoscilla hemoglobin Vhb, which
relieved oxygen limitations and led to a final titer of 18.8 mg/L pyocyanin. These results show promise to use
E. coli for phenazines production, and the engineered strain developed here has the potential to be used in
electro-fermentation systems where pyocyanin plays a role as electron-shuttle.

1. Introduction biotechnological applications, pyocyanin can be easily reduced and

Microbial secondary metabolites are valuable sources of a wide
spectrum of molecules suitable for biotechnological applications.
Phenazines are pigmented natural products derived from the secondary
metabolism of a few bacterial and archaea genera, including Pseudo­
monas. This class of molecules is composed of nitrogen-containing aro­
matic rings with a rich diversity of substituents that confers them a vast
array of biological activities and bright colors.
Pyocyanin (PYO) is a blue-green phenazine produced by Pseudo­
monas aeruginosa. It has been described as a quorum sensing signaling
molecule (Dietrich et al., 2006), a virulence factor (Lau et al., 2004), and
an antimicrobial with activities against fungi and bacteria (Chincholkar
and Thomashow, 2013). Besides these biological roles and
oxidized, which makes this molecule an interesting target for electro­
chemical applications as, for instance, on biosensors development
(Jayaseelan et al., 2014), microbial fuel cells (Rabaey et al., 2005) or as
an electron mediator in electro-fermentation systems (Morrison et al.,
2018; Venkataraman et al., 2011). This latter application inspired us in
this work to build an E. coli strain able to produce pyocyanin for future
co-culture processes using electro-bioreactors, where one strain would
produce the electron shuttle pyocyanin and the other strain(s) would
carry on the production of a bioproduct of interest.
Although more than 150 natural phenazines have been described to
date (Chincholkar and Thomashow, 2013), this vast class of secondary
metabolites has been almost unexplored so far by the metabolic engi­
neering Community. Until now, the production of these molecules has

* Corresponding author. Department of Chemical and Biological Engineering, Rensselaer Polytechnic Institute, Troy, NY, USA.
** Corresponding author. Department of Biological Sciences, Rensselaer Polytechnic Institute, Troy, NY, USA.
E-mail addresses: barqub@rpi.edu (B. Barquera), koffam@rpi.edu (M.A.G. Koffas).

https://doi.org/10.1016/j.ymben.2021.01.002
Received 3 December 2019; Received in revised form 17 November 2020; Accepted 10 January 2021
Available online 14 January 2021
1096-7176/© 2021 Published by Elsevier Inc. on behalf of International Metabolic Engineering Society.
A.J. da Silva et al.
been studied in their natural producer strains, which limits their use as
many of these organisms are pathogenic or secrete only very small
amounts of such compounds. In this work, we present for the first time,
the engineering of a phenazine biosynthetic pathway in E. coli for het­
erologous production of pyocyanin.
P. aeruginosa contains two homologous core loci (phzA1B1C1­
D1E1F1G1 and phzA2B2C2D2E2F2G2) responsible for the synthesis of
phenazine-1-carboxylic acid (PCA) (Mavrodi et al., 2001). In most
phenazine producing bacteria, the phz operon is flanked by one or more
accessory genes that codify for enzymes that catalyze PCA modifications
giving rise to other phenazines (Chincholkar and Thomashow, 2013). In
the case of P. aeruginosa, the genes phzM and phzS are responsible for
converting PCA into PYO, and the gene phzH converts PCA into
phenazine-1-carboxamide (PCN) (Mavrodi et al., 2001).
Chorismate, the end product of the shikimate pathway, is the main
precursor for phenazines production (Fig. 1). The shikimate pathway
operates in microorganisms and plants for the biosynthesis of aromatic
compounds, and it is initiated by the condensation of erythrose-4-
phosphate (E4P) and phosphoenolpyruvate (PEP). In E. coli and many
other bacteria, this reaction is carried out by three isoenzymes (AroF,
AroG, and AroH). In Pseudomonas spp. this reaction is also performed by
the product of the phzC gene from the phenazines operon (Mentel et al.,
2009), making E4P and PEP part of the aromatic pathway. All the other
genes from the phz loci are involved in the conversion of chorismate to
PCA. However, some steps in this metabolic pathway are still not
completely understood.
In this work, we cloned the nine genes of the pyocyanin pathway
from P. aeruginosa PAO1 in E. coli using the ePathBrick vectors platform
(Xu et al., 2012). This is an excellent system for expressing complex
metabolic pathways in E. coli and proving the viability of producing
phenazines in heterologous hosts. In addition, the influence of key
process parameters and pathway balance on pyocyanin expression was
also evaluated to optimize its production.
2. Materials and methods
2.1. E. coli strain selection for pyocyanin production
Because PYO presents antibiotic activity against E. coli, we tested
nine strains (Table 1) to investigate if different strains would show
different sensitivities to pyocyanin. To determine the Minimum Inhibi­
tory Concentration (MIC) of pyocyanin against E. coli, a 100 μg/mL
pyocyanin stock solution (Sigma-Aldrich, USA) was two-fold serially
diluted into LB medium in 96-well micro-plates. Each well containing
200 μL of test solution was inoculated with 10 μL of properly diluted
Fig. 1. Pyocyanin biosynthetic pathway. E4P: Erythrose-4-phosphate; PEP: Phosphoenolpyruvate; DAHP: 3-deoxy-D-arabino-heptulosonate-7-phosphate; PCA:
phenazine-1-carboxylic acid; Pyr: pyruvate. Adapted from Mentel et al. (2009); Blankenfeldt (2013); Mavrodi et al. (2001).
16
Metabolic Engineering 64 (2021) 15–25
fresh cultures of each E. coli strain to reach 5 × 105 CFU/mL per well.
The plates were covered and incubated at 37 ◦ C, 225 rpm for 20 h, and
the bacterial growth was detected by adding 40 μL of 0.2 mg/mL p-
iodonitrotetrazolium chloride (INT) to each well followed by 30 min
incubation at the same conditions. Development of pink color indicated
bacterial growth, whereas clear colorless wells indicated growth inhi­
bition. MIC values were recorded as the lowest concentrations of pyo­
cyanin showing clear wells. The assay was performed in triplicates for
each strain.
2.2. Promoter replacement in the ePathBrick vectors
In the original set of vectors developed by Xu et al. (2012), gene
expression is controlled by the T7 promoter, requiring the use of strains
that contain an inducible T7 RNAP gene on the chromosome to be
functional (Studier and Moffatt, 1986). To broaden the applicability of
this metabolic pathway platform to any E. coli strain, the T7 promoter
was replaced in all five ePathBrick vectors by the lacUV5 promoter,
which is a mutated version of the E. coli lac promoter that is independent
of any activators. By using the lacUV5 promoter to control the expres­
sion of the pyocyanin pathway genes, any of the tested E. coli strains
would be ready to use as a host for PYO production, regardless of the
presence or not of the T7 RNAP gene.
To build the new set of plasmids, a sequence containing the lacUV5
promoter replacing the T7 promoter, together with the rrnB T1 termi­
nator, was ordered as a gBlock (Supplementary Table S1) from IDT
(Integrated DNA Technologies Inc, USA). The synthetic DNA was
amplified by PCR using the primers gB_lacUV5_F and gB_lacUV5_R
(Supplementary Table S1). The resulting amplicon was digested with
AvrII and SalI restriction enzymes (Fast Digest, Thermo Scientific) and
cloned into the respective sites in all five ePathBrick vectors. The con­
structs were transformed by heat shock into chemically competent E. coli
DH5α cells and verified by restriction digestion and Sanger sequencing
(Genewiz Inc, USA). The new plasmids containing the lacUV5 promoter
were named pACM5, pCDM5, pCOM5, pETM7, and pRSM4 (Table 1).
2.3. Construction of the pyocyanin biosynthetic pathway
For the heterologous production of pyocyanin in E. coli, a modular
approach was chosen to split the nine genes in the pathway into two
modules, where the first plasmid contains the genes for PCA production
from chorismate, and the second plasmid harbors the remaining genes
required for converting PCA into PYO.
To build both modules, the first step was to amplify by PCR (Accu­
zyme, Bioline) the genes phzA1, phzB1, phzC1D1E1, phzF1, phzG1, phzM,
A.J. da Silva et al. Metabolic Engineering 64 (2021) 15–25

Table 1 and phzS from P. aeruginosa PAO1 genomic DNA using the respective
Bacterial strains and selected plasmids used in this study. primers listed in the Supplementary Table S1. All the amplicons were
Strains and plasmids Description Reference or individually cloned into the MCS of pETM7 or pACM5 between NdeI and
source XhoI restriction sites, except for phzA which was introduced between
Strains NdeI and SpeI sites.
Pseudomonas For the first module, ePathBrick-directed gene assembly (Xu and
aeruginosa Koffas, 2013) was used, and the genes were assembled in the
PAO1 Wild-type Lab pseudo-operon configuration. The first step was performed by joining
collection
SpeI/ApaI digested pETM7_phzB with AvrII/ApaI digested pETM7_phzG,
Escherichia coli giving rise to pETM7_phzBG. Next, an iterative assembly was further
BL21(DE3) B strain with DE3, a λ prophage carrying Lab performed by joining phzA, phzF, and phzCDE using the pairs of enzymes
the T7 RNA polymerase gene and lacIq collection AvrII/BmtI for donor vectors and SpeI/BmtI for destination vectors,
BW25113 rrnBT14 ΔlacZWJ16 hsdR514 Lab
generating the construct pETM7_phzABGFCDE at the end, which was
ΔaraBADAH33 ΔrhaBADLD78 collection
DH5α Recombination-deficient strain Invitrogen named pE-PCA. The second module was constructed in a monocistronic
commonly used for cloning and plasmid configuration by ligating the shorter NheI/ApaI digested fragment from
propagation pACM5_phzM into the AvrII/ApaI digested pACM5_phzS, giving rise to
JCL299 BW25113 derivative (Tetr) ΔldhA ΔadhE Shen et al. pACM5_phzMS, named as pA-MS. All constructs were verified by re­
(2011)
striction digestion and gene sequencing (Genewiz Inc, USA).
ΔfrdBC Δpta
MG1655 Wild-type K-12 F– λ– ilvG– rfb-50 rph-1 Lab
collection
Nissle1917 Probiotic strain, O6:H1:K5 Lab 2.4. Pyocyanin pathway expression
collection
Top10F′ F plasmid containing lacIq and Tn10 Invitrogen
The constructs pE-PCA and pA-MS were co-transformed into chem­
(TetR, StrR) commonly used for cloning
and plasmid propagation ically competent E. coli QH4, resulting in strain 1M2L. After heat shock,
XL1-Blue F plasmid containing lacIq and Tn10 Strategene cells were recovered for 60 min at 37 ◦ C in Luria Bertani broth (LB
(TetR), commonly used for cloning and Lennox, Sigma) containing 10 g/L tryptone, 5 g/L yeast extract, and 5 g/
plasmid propagation L NaCl and plated on LB agar plates with 80 μg/mL ampicillin and 25 μg/
QH4 ATCC31884 derivative, ΔpheA, ΔtyrA Huang et al.
(2013)
mL chloramphenicol. An isolated colony was inoculated into 5 mL of LB
1L2M QH4 carrying pA-PCA and pE-MS This work medium and incubated overnight at 30 ◦ C and 225 rpm. A volume of 0.5
1L2M3H QH4 carrying pA-PCA, pE-MS, and pR- This work mL of this culture was transferred to 25 mL of LB medium in 250 mL
vhb baffled flasks and cells were incubated at the same conditions. After
1L2H QH4 carrying pA-PCA and pR-MS This work
reaching an OD of 0.6, the culture was induced by 1 mM IPTG and
1L2H3M QH4 carrying pA-PCA, pR-MS, and pE- This work
vhb samples were taken 4, 8, 12, and 24 h after induction. For PYO recovery,
1M2L QH4 carrying pE-PCA and pA-MS This work the supernatant was extracted with 1/3 (v/v) chloroform followed by
1M2L3H QH4 carrying pE-PCA, pA-MS, and pR- This work centrifugation. The organic layer was recovered and dried with N2 gas
vhb flow and the resulting powder was resuspended in phosphate buffer pH
1M2H QH4 carrying pE-PCA and pR-MS This work
1M2H3L QH4 carrying pE-PCA, pR-MS, and pA- This work
7.0. METODOLOGIA QUANTIFICAÇÃO
vhb Pyocyanin was quantified by HPLC according to Kern and Newman
(2014) with modifications. A C-18 column (Symmetry C18, 3.5 μm, 4.6 x
Plasmids 75 mm, Waters) was used on an integral HPLC system (Shimadzu). The
pACM4, pETM6, ePathBrick vectors Xu et al.
detector was set at 365 nm, and the column oven was kept at 30 ◦ C.
pRSM3, pCDM4, (2012)
pCOM4 Linear gradient elution with a flow rate of 0.6 mL/min using 0.1% TFA
pACM5 Derived from pACM4, PlacUV5, CmR, This work in water (solvent A) and TFA 0.1% in acetonitrile (solvent B) was set as
P15A ori follows: 15% B at time 0; 85% B at 15 min; 15% B at 17 min, with a total
pETM7 Derived from pETM6, PlacUV5, AmpR, This work run time of 22 min. A standard curve with known pyocyanin concen­
ColE1 ori
pRSM4 Derived from pRSM3, PlacUV5, KanR, This work
trations dissolved in TB medium was used for quantification.
RSF1030 ori
pCDM5 Derived from pCDM4, PlacUV5, StrR, This work 2.5. Optimization of pyocyanin production
CloDF13 ori
pCOM5 Derived from pCOM4, PlacUV5, KanR, This work
2.5.1. Influence of process parameters
CoLA ori
pA-PCA pACM5 containing phzA, phzB, phzG, This work After confirming PYO production by recombinant E. coli cells, we
phzF, phzC, phzD, and phzE genes attempted to optimize its production by evaluating the influence of
pE-PCA pETM7 containing phzA, phzB, phzG, This work medium composition, temperature, and induction time on cellular
phzF, phzC, phzD, and phzE genes growth and pyocyanin accumulation.
pA-MS pACM5 containing phzM and phzS genes This work
pE-MS pETM7 containing phzM and phzS genes This work
Initially, TB rich medium containing 12 g/L tryptone, 24 g/L yeast
pR-MS pRSM4 containing phzM and phzS genes This work extract, 0.17 M KH2PO4 and 0.72 M K2HPO4 was supplemented with 2%
pA-vhb pACM5 containing vhb50 gene This work glucose or glycerol and compared to the AMM defined medium (He
pE-vhb pETM7 containing vhb50 gene This work et al., 2015) supplemented with 2% glucose or glycerol, 0.3 mM
pR-vhb pRSM4 containing vhb50 gene This work
phenylalanine, and 0.2 mM tyrosine. Cells were cultured as described
Antibiotics were added when necessary at the following final concentrations: (section 2.4) and PYO production was evaluated by HPLC. TB medium
Ampicillin 80 μg/mL; Chloramphenicol 25 μg/mL; Kanamycin 50 μg/mL; supplemented with glycerol led to a higher final product titer and was
Streptomycin 50 μg/mL. chosen for all the following experiments.
Next, cellular growth and PYO formation were evaluated under three
different temperatures at the induction phase. Cells were grown in TB
medium with 2% glycerol, 225 rpm at 37 ◦ C prior to adding 1 mM IPTG
and the temperature was shifted to 20 ◦ C, 30 ◦ C, or kept at 37 ◦ C for 50 h
after IPTG induction. Samples for OD reading and PYO analysis were

17
A.J. da Silva et al.
collected throughout the cultivation.
The influence of the time of induction on pyocyanin production was
also analyzed. Cells were grown in TB with 2% glycerol, 225 rpm at
30 ◦ C, and early, mid, or late exponential phases of growth were tested
by adding 1 mM IPTG.
2.5.2. Effect of plasmids copy number
As a further step to improve PYO titer, we evaluated the effect of
plasmid copy number on the expression of each module. The sets of
genes phzABGFCDE and phzMS were subcloned into a low, medium, and
high copy number plasmids (pACM5, pETM7, and pRSM4, respectively).
The plasmid pE-PCA was digested with AvrII/BmtI restriction enzymes,
and the fragment containing the phz genes was ligated into AvrII/BmtI
digested pACM5, generating pA-PCA. The same procedure was per­
formed for the second module, where the plasmid pA-MS was digested
by ApaI/NheI, and the fragment containing the genes phzM and phzS was
ligated into ApaI/NheI digested pETM7 and pRSM4 vectors, generating
pE-MS and pR-MS, respectively. Despite many attempts, the construct
pRSM4_phzABGFCDE (pR-PCA) could not be obtained for unknown
reasons, possibly related to a high metabolic burden derived from this
high copy number plasmid carrying a large sequence of heterologous
genes. However, four combinations of the two modules were success­
fully constructed (Table 1), and the resulting strains were grown in TB
with 2% glycerol with proper antibiotics (Table 1), 225 rpm at 30 ◦ C and
induced at mid-log phase to check if the plasmid copy number of both
modules could affect pyocyanin accumulation.
2.5.2.1. Gene expression analysis. RT-qPCR was employed to evaluate
the expression of pyocyanin pathway genes in the four strains. Cells
grown in TB medium with glycerol 2% were induced with IPTG 1 mM
after 6 h of cultivation at 30 ◦ C, 225 rpm. An appropriate amount of
culture [OD600 × culture volume (mL) = 6] was sampled after 12 h of
cultivation. Cell pellets were immediately frozen in dry ice and stored at
-80 ◦ C.
Total RNA extractions were performed using TRIZOL reagent (Invi­
trogen) according to the manufacturer’s protocol. The RNA concentra­
tions and purity were measured by Nanodrop 2000 (Thermo Scientific).
Total RNA (1 μg) was treated with deoxyribonuclease I (DNAse I; Invi­
trogen) to remove genomic DNA. Treated RNA was reverse-transcribed
into cDNA with high-capacity cDNA Reverse Transcription kit (Applied
Biosystems).
The qRT-PCR was performed using a CFX 96 real-time PCR detection
system (Bio-Rad) at an annealing temperature of 60 ◦ C. The final reac­
tion volume of 10 μL contained 10 ng of cDNA, 300–400 nM of each
primer, and 5 μL of SsoFastTM Evagreen Supermix (Bio-Rad). Each
cDNA was analyzed in three technical replicates from three biological
replicates. Specific primers were designed using Primer 3 software
(bioinfo.ut.ee/primer3-0.4.0) and further synthesized by Exxtend
(Brazil) (Table S2). The results were normalized to rrsA RNA levels, and
the 1M2L sample was used as a control group. The relative gene
expression was calculated using the 2− ΔΔCt method (Livak and
Schmittgen, 2001). Statistical analysis was carried out by STATISTICA®
7.0 software (Stat Soft). The analysis of variance (ANOVA) one-way test
with Tukey’s multiple comparisons tests as post hoc was used for group
comparisons. All data are presented as means ± standard error (SEM),
and P values < 0.05 were considered statistically significant.
2.5.3. Effect of hemoglobin co-expression
Because oxygen participates in four reactions for pyocyanin forma­
tion, we hypothesized that improving the uptake of oxygen by the cells
in culture could benefit PYO production. To test this hypothesis, the
Vitreoscilla hemoglobin Vhb, which is a membrane protein facilitating
O2 transport, was co-expressed with the pyocyanin pathway. A mutated
version of the vhb gene, called vhb50, was PCR amplified from the
plasmid pA5c-tesA-vhb50-8fadR (Liu et al., 2017) acquired from
18
Metabolic Engineering 64 (2021) 15–25
Addgene (#86601) using the pair of primers vhb_F and vhb_R (Supple­
mentary Table S1). The amplicon was cloned between NdeI and XhoI
restriction sites of pACM5, pETM7, and pRSM4, resulting in a low,
medium, and high copy number plasmid for vhb50 expression, which
were named as pA-vhb, pE-vhb, and pR-vhb, respectively. Each of these
constructs was transformed into PYO producer strains 1M2H, 1L2H, and
1L2M/1M2L respectively, and the influence of Vhb expression on cell
growth and PYO production was evaluated and compared to those
observed in the absence of Vhb in shake flask cultures.
To further validate the influence of Vhb expression on PYO pro­
duction, strains 1M2L and 1M2L3H (1M2L carrying the plasmid pR-vhb)
were cultivated in a bioreactor. An aliquot of 0.5 mL of an overnight
culture was transferred to 25 mL of TB medium and further incubated at
30 ◦ C, 225 rpm in a baffled flask until reaching the OD of 1.0. Then 25
mL of broth were inoculated into 1 L of TB medium with glycerol 2% in a
bioreactor (Minifors, Infors). The culture was maintained at 30 ◦ C, and
the pH was controlled at 7.0 by the automatic addition of 5.0 M NaOH.
The agitation speed was kept constant at 800 rpm, and the airflow was
maintained at 0.5 L/min. Pyocyanin expression was induced by addition
of 1 mM of IPTG after 6 h of cultivation. Samples were taken periodically
to follow the biomass and pyocyanin accumulation over time.
3. Results and discussion
3.1. Strain selection for pyocyanin production
Pyocyanin is a known redox-active compound that can be used as an
electron mediator in electro-fermentation (EF) systems. In EF systems,
polarized electrodes are employed to unbalanced fermentation by
changing the medium redox balance (Moscoviz et al., 2016). These
systems include an electron transfer reaction between the microbes and
the electrodes, which can act either as electron sinks or electron donors.
Changes in external redox balance can affect intracellular NAD+/NADH
ratios, leading to profound metabolic changes that can positively impact
yields and productivities of a target bioproduct (Rabaey et al., 2011;
Choi et al., 2012; Liu et al., 2013; Li et al., 2019). Shewanella oneidensis
and Geobacter sulfurreducens are known examples of “electroactive”
microorganisms that can exchange electrons by directly interacting with
metal surfaces. In addition, most bacteria are able to use soluble sub­
stances as electron shuttles which can be oxidized or reduced by cells
and then be recycled electrochemically at the electrodes surfaces,
resulting in indirect electron transfer (Moscoviz et al., 2016). These
molecules act as electron mediators, and pyocyanin has a great potential
to be used in this way (Rabaey et al., 2005; Venkataraman et al., 2011).
In this context, having E. coli cells producing pyocyanin represents an
innovative approach for production of biomolecules under EF conditions
in a co-culture scheme. The in-situ PYO production could replace the
need for the addition of expensive electron mediators like neutral red.
The use of PYO for this purpose was shown by Venkataraman et al.
(2011) in a co-culture composed of pyocyanin producing cells of
P. aeruginosa PA14 and Enterobacter aerogenes cells producing 2,3-buta­
nediol. Because the use of human pathogenic organisms for industrial
processes may raise regulatory and safety issues, we reconstructed the
complete metabolic pathway for PYO synthesis in E. coli.
However, the natural pyocyanin toxicity towards E. coli raised the
need of having a suitable strain that could tolerate PYO concentrations
around 3–35 μg/mL, which are in the same range used for other electron
mediators like neutral red or methylene blue (Choi et al., 2012;
Sturm-Richter et al., 2015; Venkataraman et al., 2011).
Initial attempts were directed to improve E. coli capabilities to deal
with reactive oxygen species (ROS), which are associated with pyocya­
nin’s toxic effect (Hassan and Fridovich, 1980). Antioxidant enzymes,
superoxide dismutase and catalase, seem to play a role in protecting cells
from pyocyanin toxicity (Norman et al., 2004). However, although
mutant strains ΔsodA, ΔsodAΔsodB or ΔkatG showed to be more sensi­
tive to pyocyanin than the wild type strain (data not shown),
A.J. da Silva et al.
overexpression of any of the three E. coli superoxide dismutases (SodA,
SodB, or SodC) and both catalases (KatE or KatG) individually did not
improve the tolerance of cells to the presence of pyocyanin. Combined
overexpression of superoxide dismutase and catalase (SodA + KatE or
SodC + KatE) also resulted in no difference on MIC values for pyocyanin
in comparison to the wild type strain (data not shown). A last attempt to
improve cellular tolerance to increased ROS levels was made by over­
expression of the soxS gene. The product from this gene is a transcrip­
tional factor that activates cellular responses to oxidative stress (Duval,
2013) but even after overexpression of SoxS we could not detect any
increase in cellular resistance to pyocyanin effects (data not shown).
As an alternative strategy, we decided to evaluate the minimum
inhibitory concentration (MIC) of pyocyanin towards different E. coli
strains, looking for a naturally more resistant strain. Surprisingly, we
found a strong strain dependent toxicity response, with MIC values
ranging from 12.5 to 50 μg/mL among nine E. coli strains evaluated
(Fig. 2). DH5α, BL21(DE3), and XL1Blue strains were very sensitive to
pyocyanin and could only tolerate up to 6.25 μg/mL pyocyanin. On the
other hand, QH4 cells showed robust growth even at 25 μg/mL pyo­
cyanin, being inhibited only when the concentration was between 26
and 50 μg/mL pyocyanin. JCL299 and Nissle1917 strains showed
moderate growth at 25 μg/mL, while BW25113, MG1655, and Top’10
cells could tolerate up to 12.5 μg/mL pyocyanin.
E. coli QH4, the most resistant strain found among those evaluated,
was developed by Huang et al. (2013) for caffeic acid production, and
was chosen here for pyocyanin expression. This strain was derived from
E. coli ATCC31884 by knocking out the genes pheA and tyrA responsible
for converting chorismate into prephenate, which is a crucial precursor
for the aromatic amino acid biosynthesis. Although we could not
correlate these deletions to the higher resistance of this strain to pyo­
cyanin, avoiding the consumption of chorismate may be advantageous,
as chorismate is also the precursor for PYO biosynthesis.
3.2. Heterologous production of pyocyanin
More than 150 natural phenazines have been described to date
(Chincholkar and Thomashow, 2013). However, despite the vast po­
tential applications of these molecules, their production in heterologous
hosts under optimized conditions has not been explored so far. Previous
studies reported that mixed cultures of E. coli carrying regions of
P. aeruginosa genome containing the phz operon and accessory genes
Fig. 2. Pyocyanin MIC determination for different E. coli strains. Growth is detected by color development, and clear wells indicate inhibition. MIC value is defined as
the minimum concentration that inhibits cellular growth. Positive control shows cell growth in LB medium in the absence of pyocyanin.
19
Metabolic Engineering 64 (2021) 15–25
were able to produce small but detectable amounts of pyocyanin
(Mavrodi et al., 2001). Based on this report, we decided to employ
metabolic engineering tools to express and optimize the pyocyanin
synthesis pathway in E. coli.
Using the ePathBrick vectors platform (Xu et al., 2012), the nine
genes in the pyocyanin pathway were expressed in E. coli. This strategy
allowed the construction of two modules in parallel, saving time and
effort for cloning all the target genes. In addition, splitting the pathway
into two parts avoided large constructs that may be harder to handle and
difficult to transform. Lastly, having a separate module for PCA pro­
duction from chorismate, and another one for the accessory genes that
convert PCA into PYO, opened the possibility of producing other
phenazines just by replacing the second module by one that carries
genes for conversion of PCA into other derivatives.
Replacement of the strong T7 promoter by the lacUV5 promoter in
the ePathBrick vectors was required for using this expression platform in
QH4 E. coli cells, since this strain does not possess the T7 RNAP gene that
is necessary for proper T7 promoter function. Also, using a weaker
promoter for expression of metabolic pathways was shown to be ad­
vantageous in numerous examples, where higher levels of pathway en­
zymes overexpression do not correlate with higher metabolic fluxes
through the desired pathways (Jones and Koffas, 2016).
For PCA production (module I), all genes were assembled in a
pseudo-operon configuration, where each gene is preceded by a pro­
moter sequence, but all genes share the same single terminator after the
last cloned ORF. This allows the sequences closer to the terminator to be
transcribed more often than those with further distance from the
terminator; this results in incomplete transcription events by the RNA
polymerase when it leaves the DNA strand before encountering a
terminator sequence. Therefore, the pseudo-operon configuration may
result in different expression levels for each gene according to their
spatial positions. For PCA production, the genes phzCDE, which are
responsible for the commitment step of E4P and PEP to the shikimate
pathway (phzC) (Fig. 1) and for the first two reactions that channel
chorismate to the phenazines biosynthesis (phzE and phzD), were placed
right before the terminator, aiming to favor an increased expression of
these genes. Also, the amplification of phzC, phzD, and phzE for cloning
into pETM7 was done as a single amplicon since there may be a trans­
lational coupling as their stop and start codons overlap (Mavrodi et al.,
2001).
For the construction of module II carrying the genes phzM and phzS,
A.J. da Silva et al.
both sequences were cloned in monocistronic configuration, where each
gene has its own promoter and terminator. Fig. 3A illustrates the two
modules constructed for pyocyanin production. Cultivation of QH4 cells
transformed with both plasmids and induction by 1 mM IPTG led to
pyocyanin production. Fig. 3B shows the final extract from the culture
supernatant presenting the characteristic blue color of pyocyanin.
After confirming the success of heterologous expression of PYO
metabolic pathway in E. coli, the influence of medium composition on
yield was evaluated. Glucose and glycerol were tested as carbon sources.
We also compared a complex rich medium (TB) to a defined rich me­
dium (AMM) supplemented with aromatic amino acids. Growth and
PYO production were favored using TB medium supplemented with
glycerol, resulting in the production of 2.7 mg/L pyocyanin after 25 h of
induction. Growth on rich medium with glucose was affected by acetate
accumulation, leading to a final OD 30% lower compared to growth on
glycerol, and pyocyanin final titer of 1.1 mg/L. Using the defined AMM
medium, biomass formation was at least 5 times lower than in TB me­
dium, which directly affected PYO production. Concerning the carbon
sources, glycerol proved to be a better choice leading to 1.6 mg/L versus
0.8 mg/L pyocyanin when glucose was used to supplement the AMM
medium. Glycerol has been successfully used for production of many
bioproducts in E. coli, including aromatic compounds (Gottlieb et al.,
2014; Jones et al., 2016; Lee et al., 2017). Because of its high availability
as a byproduct from the biodiesel production, the current glycerol prices
have dramatically dropped, stimulating its use as a cheap carbon and
energy source for fermentation processes (da Silva et al., 2009). In
addition, when glycerol is used as a carbon source, the accumulation of
acetate decreases (Martínez-Gómez et al., 2012). The accumulation of
acetate in the culture has been proved to be deleterious for growth and
protein expression (De Mey et al., 2007). Thus, TB medium supple­
mented with 2% glycerol was chosen to be used for all the following
experiments.
3.3. Optimization of process parameters
The effect of temperature was analyzed on pyocyanin formation and
cell growth. A dual-temperature approach was employed, where the
temperature of 37 ◦ C was used to maximize growth prior to induction,
followed by a shift in temperature after IPTG addition to optimize re­
combinant enzyme expression. Temperatures of 20, 30, and 37 ◦ C were
tested for the induction phase (Fig. 4).
Incubation at 30 ◦ C favored both cellular growth and pyocyanin
production with the highest titers between 18 and 38 h of culture and a
concentration close to 5.0 mg/L. This yield is approximately 65% higher
Fig. 3. A) Schematic representation of the modules constructed for pyocyanin biosynthesis in E. coli. Genes for the conversion of chorismate into PCA were cloned in
pseudo-operon configuration in the first module. The second module contains the genes required to transform PCA into PYO cloned in monocistronic form. B)
Chloroform extraction from culture supernatants of QH4 cells carrying pE-PCA and pA-MS plasmids: (i) induced with 1 mM IPTG; (ii) non-induced (negative control).
20
Metabolic Engineering 64 (2021) 15–25
than the best values reached at 37 ◦ C.
After defining the best temperature, the time of induction was
evaluated. According to Donovan et al. (1996), induction optimization is
the most sensitive parameter of fermentation optimization. This has
been observed in different systems (Lim et al., 2015; Jones et al., 2015,
2016). Here, we tested the induction of PYO expression by addition of 1
mM IPTG at the early, mid, and late log phase. Results indicated that the
time of induction greatly affected PYO production. When induction was
done at mid-log phase, both biomass and pyocyanin accumulation were
the highest (Fig. 5). Early and late induction led to similar lower PYO
titers. Induction at mid-log phase led to an increase of 60% on PYO
production compared to the other conditions, reaching a titer of 9.0
mg/L. As shown in Fig. 5B, the different times of induction also affected
cell growth, suggesting that induction point optimization is intimately
involved with the allocation of cellular resources for both protein pro­
duction and cell growth (Wu et al., 2016).
3.4. Pathway balance
It is well known that the plasmid copy number plays a significant role
in the recombinant expression of proteins. However, the use of high
copy number plasmids may not be the best choice when it comes to
metabolic engineering applications (Jones and Koffas, 2016). Over­
producing all the enzymes in a pathway may introduce metabolic
burden to the cell resulting in decreasing reaction rates and cell growth
limitation (Wu et al., 2016).
A common strategy to deal with balancing expression of multiple
enzymes in a pathway is using plasmids with different copy numbers,
which drives expression of enzymes at different levels. To evaluate the
effect of the copy number of the plasmid used for PYO production, we
cloned the two modules in a set of different plasmids of the ePathBrick
vectors series with lacUV5 promoter: the low copy number pACM5
(10–12 copies per cell), the medium copy number pETM7 (around 40
copies per cell), and the high copy number pRSM4 (around 100 copies
per cell). Although the expression level tends to directly correlate with
the plasmid copy number, there are exceptions, as the same promoter
can behave distinctly in different genetic contexts (Xu et al., 2012).
Fig. 6 illustrates the four combinations tested for expression of both
modules under the influence of varying the copy number of the plasmids
used.
The four constructed strains with pyocyanin genes expressed from
varying plasmid copy numbers (Fig. 6) were evaluated for pyocyanin
production and cellular growth (Fig. 7).
The copy number of each module had a great impact on pyocyanin
A.J. da Silva et al. Metabolic Engineering 64 (2021) 15–25

Fig. 4. Efect of temperature of induction on pyocyanin production (A) and cellular growth (B). Strain 1M2L was grown in TB medium + 2% glycerol at 37 ◦ C prior to
induction with 1 mM IPTG. Pyocyanin production was induced at the early log phase. Error bars represent standard deviations from biological triplicates.

Fig. 5. Effect of induction point on pyocyanin production (A) and cellular growth (B). Strain 1M2L was grown in TB + 2% glycerol at 30 ◦ C. Pyocyanin production
was induced at the early, mid, or late log phase. Error bars represent standard deviations from biological duplicates.

production (Fig. 7). A fourfold increase of PYO titer was obtained by the genes, which was more pronounced in the worst-performing strain
strain 1M2H in comparison to the strain 1L2H. This result suggests that (1L2H), suggests that a similar expression level for all the pyocyanin
pyocyanin formation was favored when the first module was expressed genes may favor its production. Besides that, as the intermediate PCA
using the medium copy number plasmid. Also, a similar final yield was phenazine was not detected, it seems that the lower expression level for
observed for the strain 1M2L compared to the best producer strain the operon genes in the strains 1L2M and 1L2H had a major impact on
(1M2H), indicating that the expression of the first module using the the lower pyocyanin levels observed for these strains.
medium copy number plasmid is a better choice regardless of the
plasmid used for expressing the second module.
3.5. Effect of Vitreoscilla hemoglobin co-expression
Quantification of gene expression revealed that the best strains had a
similar expression profile for most of the pyocyanin pathway genes
Four reactions in the last steps of pyocyanin production are depen­
(Fig. 8), which correlates well with the similar final yields observed for
dent on oxygen availability (Fig. 1). To increase the ability of the cells to
these strains (1M2L and 1M2H).
use oxygen, we introduced into our PYO producer strains a third module
The strains 1L2M and 1L2H, which reached lower pyocyanin titers
carrying a modified version of the Vitreoscilla hemoglobin Vhb (Liu et al.,
and yields, showed a lower expression level for the phzABCDFG genes
2017).
than the control group (1M2L), as well as a much higher expression for
Vhb is a soluble microbial hemoglobin that can improve aerobic
the accessory genes phzM and phzS cloned in the second module (Fig. 8).
respiration under low oxygen availability. The heterologous expression
This unbalance between the expression of the operon and the accessory
of Vhb has been used for increasing protein or metabolites production,

21
A.J. da Silva et al. Metabolic Engineering 64 (2021) 15–25

Fig. 6. Schematic representation of four strains constructed to evaluate the influence of plasmid copy number on pyocyanin formation.

Fig. 7. Effect of plasmid copy number on pyocyanin production (A) and cellular growth (B). All strains were grown in TB + 2% glycerol at 30 ◦ C, and pyocyanin
production was induced at mid-log phase. Error bars represent standard deviations from biological duplicates.

Fig. 8. Quantification of relative gene expression for the strains expressing the pyocyanin genes using different plasmid copy numbers. Strain 1M2L was used as a
control group (gene expression set to 1.0 for all genes) for comparison purposes. Results from biological triplicates are expressed as means ± standard deviations.
Symbols – for groups 1L2H and 1L2M: * means different from phzE, # means different from phzM and phzS; for group 1M2H: & means different from phzA, * means
different from phzB, and # means different from phzC.

biomass accumulation, and toxic compounds degradation by different
species of microorganisms and plants (Zhang et al., 2007). In E. coli, Vhb
expression has been used for enhancing ethanol productivity (Arnaldos
et al., 2012) as well as for improving fatty-acids production (Liu et al.,
2017). Oxygen-requiring metabolic pathways have also been optimized
in plants by co-expression of Vhb, and facilitating oxygen diffusion by
Vhb expression appears to be an effective method for metabolite
productivity enhancement (Zhang et al., 2007).
A metabolic distribution flux analysis from fed-batch cultures of
E. coli expressing different levels of Vhb showed that this protein is able
to influence the redistribution of carbon fluxes in central metabolic
pathways. The main effects found when Vhb was included in cultures
were an increase of growth rate and yield, a larger oxygen consumption,
and a reduction of by-products formation (Tsai et al., 1996).

22
A.J. da Silva et al.
Transcriptomic analysis revealed that hundreds of genes can be affected
by Vhb expression in E. coli, especially those involved in energy meta­
bolism, central intermediary metabolism, and cell processes (Roos et al.,
2004). Although many studies and hypothesis have been published so
far, the exact mechanisms by which Vhb affects cellular metabolism is
still not completely understood.
Several point mutations in Vhb were also evaluated to investigate the
structure-function relations of this hemoglobin (Stark et al., 2012) and
to facilitate O2 uptake and biomass growth (Andersson et al., 2000). A
mutant version of Vitreoscilla Vhb (His36→Arg and Gln66→Arg) created
by Liu et al. (2017) showed better results compared to the wild type Vhb
for fatty acid production in E. coli. For the present studies, we used this
mutant version of Vhb, called Vhb50.
Co-expression of Vhb50 from a low copy number plasmid caused a
minor increase in pyocyanin titer (strain 1M2H3L compared to strain
1M2H, Fig. 9). However, expression of Vhb50 using the medium copy
number plasmid led to a three-fold improvement in PYO production
(strain 1L2H versus strain 1L2H3M, Fig. 9). In addition, the highest PYO
titers achieved in this work were observed when Vhb50 was expressed
from the high copy number plasmid (strains 1L2M3H and 1M2L3H,
Fig. 9). These results suggest that a proper balancing of Vhb expression
may be highly beneficial for metabolites production, as observed by
Zhang et al. (2007).
The higher PYO titers obtained by co-expression of Vhb50 indicate
that increasing the cell’s capability for oxygen uptake can be an effective
strategy to improve PYO accumulation. In addition, some evidence in­
dicates that Vhb can interact with the oxidative stress regulator OxyR
inducing the expression of several genes that mediate a protective effect
for cells under oxidative stress (Anand et al., 2010). This could alleviate
the toxic effects of pyocyanin related to ROS formation.
To further validate the positive influence of Vhb50 expression on
pyocyanin production, we ran a bioreactor culture for strains 1M2L3H
that carries the Vhb50 gene and 1M2L (control). Fig. 10 shows that when
the cells experienced a condition of oxygen limitation, strain 1M2L3H
expressing Vhb50 hemoglobin showed a superior pyocyanin production
compared to strain 1M2L (Fig. 10). This result supports the hypothesis
that the expression of Vhb50 improves the cell’s capability of oxygen
uptake, which ultimately favors the pyocyanin biosynthesis.
Fig. 9. Effect of Vitreoscilla hemoglobin co-expression on pyocyanin production. A third module containing the gene vhb50 was introduced into four PYO producing
strains to evaluate the influence of improved oxygen uptake on product formation. Error bars represent the standard deviations from biological duplicates.
23
Metabolic Engineering 64 (2021) 15–25
4. Conclusions
The complete pyocyanin biosynthetic pathway from P. aeruginosa
was successfully introduced in E. coli. The influence of medium
composition, temperature, and induction point were evaluated and
optimized. Together, the optimization of these parameters generated an
increase of 3.5-fold on PYO. Using a modular approach, where the whole
pathway was divided into two plasmids, allowed us to evaluate the effect
of plasmid copy number on pyocyanin production and cell growth. By
this strategy, a further balanced pathway expression was established,
resulting in an additional increase in pyocyanin titer. Finally, the co-
expression of a bacterial hemoglobin showed to be beneficial for pyo­
cyanin production by facilitating O2 uptake.
Under the optimized conditions presented in this study, an overall
increment of 5.4-fold was achieved with a final PYO titer of 18.8 mg/L
reached by the engineered E. coli strain 1M2L3H. This titer is higher than
the amounts produced by some P. aeruginosa strains under optimized
conditions (El-Fouly et al., 2015; Shen et al., 2014). Our results suggest
that the production of pyocyanin by E. coli reported here could effi­
ciently work as an electron shuttle for EF systems (Moscoviz et al.,
2016). Also, Rabaey et al. (2005) showed that 50 μM of pyocyanin in the
medium (~10 mg/L) can increase power output from 116 to 672 μW m-2
in a microbial fuel cell.
Finally, results from this work may guide future improvements on
pyocyanin production by E. coli cells, which may be possible through
further balancing the biosynthetic pathway by dynamic control
approaches.
Author statement
Adilson José da Silva: Conceptualization; Data curation; Formal
analysis; Funding acquisition; Investigation; Methodology; Project
administration; Resources; Validation; Visualization; Roles/Writing -
original draft; Writing - review & editing. Josivan de Souza Cunha:
Investigation; Methodology; Validation; Visualization; Roles/Writing -
original draft. Teri Hreha: Investigation; Methodology; Validation;
Visualization; Roles/Writing - original draft. Kelli Cristina Micocci:
Investigation; Methodology; Validation; Visualization; Roles/Writing -
original draft. Heloisa Sobreiro Selistre de Araujo: Investigation;
Methodology; Validation; Visualization; Roles/Writing - original draft.
Blanca Barquera: Conceptualization; Data curation; Formal analysis;
A.J. da Silva et al.
Fig. 10. Effect of Vhb50 expression on pyocyanin production during bioreactor cultures of strains 1M2L and 1M2L3H. (A) Cell growth, pyocyanin production and
dissolved oxygen concentration along time. (B) Pyocyanin yield comparison between the strains 1M2L and 1M2L3H.
Funding acquisition; Investigation; Methodology; Project administra­
tion; Resources; Validation; Visualization; Roles/Writing - original draft;
Writing - review & editing. Mattheos A. G. Koffas: Conceptualization;
Data curation; Formal analysis; Funding acquisition; Investigation;
Methodology; Project administration; Resources; Validation; Visualiza­
tion; Roles/Writing - original draft; Writing - review & editing.
Declaration of competing interest
There are no conflicts to declare.
Acknowledgment
The authors thank Dr. Thais Castral and Dr. Alyne Veroli for HPLC
analysis. Dr. Adilson J. Silva was supported by Fundação de Amparo à
Pesquisa do Estado de São Paulo (Processo FAPESP 2017/09695-2). This
work was supported in part by the Coordenação de Aperfeiçoamento de
Pessoal de Nível Superior – Brasil (CAPES) – Finance Code 001. This
work was funded in part by a grant from the National Science Founda­
tion (NSF-1616674) to B.B. H.S.S.A. would like to acknowledge funding
by FAPESP (2019/11437-7).
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.ymben.2021.01.002.
References
Anand, A., Duk, B.T., Singh, S., Akbas, M.Y., Webster, D.A., Stark, B.C., Dikshit, K.L.,
2010. Redox-mediated interactions of VHb (Vitreoscilla haemoglobin) with OxyR:
novel regulation of VHb biosynthesis under oxidative stress. Biochem. J. 426,
271–280. https://doi.org/10.1042/BJ20091417.
Andersson, C.I., Holmberg, N., Farrés, J., Bailey, J.E., Bülow, L., Kallio, P.T., 2000. Error-
prone PCR of Vitreoscilla hemoglobin (VHb) to support the growth of microaerobic
Escherichia coli. Biotechnol. Bioeng. 70, 446–455.
Arnaldos, M., Kunkel, S.A., Wang, J., Pagilla, K.R., Stark, B.C., 2012. Vitreoscilla
hemoglobin enhances ethanol production by Escherichia coli in a variety of growth
media. Biomass Bioenergy 37, 1–8. https://doi.org/10.1016/j.
biombioe.2011.12.048.
Blankenfeldt, W., 2013. The biosynthesis of phenazines. Microbial Phenazines:
Biosynthesis, Agriculture and Health. Springer, New York, pp. 1–18.
Chincholkar, S., Thomashow, L., 2013. Microbial Phenazines: Biosynthesis, Agriculture
and Health. Springer, New York.
Choi, O., Um, Y., Sang, B.-I., 2012. Butyrate production enhancement by Clostridium
tyrobutyricum using electron mediators and a cathodic electron donor. Biotechnol.
Bioeng. 109, 2494–2502. https://doi.org/10.1002/bit.24520.
da Silva, G.P., Mack, M., Contiero, J., 2009. Glycerol: a promising and abundant carbon
source for industrial microbiology. Biotechnol. Adv. 27, 30–39. https://doi.org/
10.1016/j.biotechadv.2008.07.006.
De Mey, M., De Maeseneire, S., Soetaert, W., Vandamme, E., 2007. Minimizing acetate
formation in E. coli fermentations. J. Ind. Microbiol. Biotechnol. 34, 689–700.
https://doi.org/10.1007/s10295-007-0244-2.
24
Metabolic Engineering 64 (2021) 15–25
Dietrich, L.E.P., Price-Whelan, A., Petersen, A., Whiteley, M., Newman, D.K., 2006. The
phenazine pyocyanin is a terminal signalling factor in the quorum sensing network
of Pseudomonas aeruginosa. Mol. Microbiol. 61 (5), 1308–1321. https://doi.org/
10.1111/j.1365-2958.2006.05306.x.
Donovan, R.S., Robinson, C.W., Glick, B.R., 1996. Review: optimizing inducer and
culture conditions for expression of foreign proteins under the control of the lac
promoter. J. Ind. Microbiol. 16, 145–154. https://doi.org/10.1007/BF01569997.
Duval, 2013. MarA, SoxS and rob of Escherichia coli – global regulators of multidrug
resistance, virulence and stress response. Int. J. Biotechnol. Wellness Ind. 2,
101–124. https://doi.org/10.6000/1927-3037.2013.02.03.2.
El-Fouly, M.Z., Sharaf, A.M., Shahin, A.A.M., El-Bialy, H.A., Omara, A.M.A., 2015.
Biosynthesis of pyocyanin pigment by Pseudomonas aeruginosa. J. Radiat. Res. Appl.
Sci. 8, 36–48. https://doi.org/10.1016/j.jrras.2014.10.007.
Gottlieb, K., Albermann, C., Sprenger, G.A., 2014. Improvement of L-phenylalanine
production from glycerol by recombinant Escherichia coli strains: the role of extra
copies of glpK, glpX, and tktA genes. Microb. Cell Factories 13, 96. https://doi.org/
10.1186/s12934-014-0096-1.
Hassan, H.M., Fridovich, I., 1980. Mechanism of the antibiotic action of pyocyanine.
J. Bacteriol. 141, 8.
He, W., Fu, L., Li, G., Andrew Jones, J., Linhardt, R.J., Koffas, M., 2015. Production of
chondroitin in metabolically engineered E. coli. Metab. Eng. 27, 92–100. https://doi.
org/10.1016/j.ymben.2014.11.003.
Huang, Q., Lin, Y., Yan, Y., 2013. Caffeic acid production enhancement by engineering a
phenylalanine over-producing Escherichia coli strain. Biotechnol. Bioeng. 110,
3188–3196. https://doi.org/10.1002/bit.24988.
Jayaseelan, S., Ramaswamy, D., Dharmaraj, S., 2014. Pyocyanin: production,
applications, challenges and new insights. World J. Microbiol. Biotechnol. 30,
1159–1168. https://doi.org/10.1007/s11274-013-1552-5.
Jones, J.A., Koffas, M.A.G., 2016. Optimizing metabolic pathways for the improved
production of natural products. Methods in Enzymology. Elsevier, pp. 179–193.
https://doi.org/10.1016/bs.mie.2016.02.010.
Jones, J.A., Vernacchio, V.R., Lachance, D.M., Lebovich, M., Fu, L., Shirke, A.N.,
Schultz, V.L., Cress, B., Linhardt, R.J., Koffas, M.A.G., 2015. ePathOptimize: a
combinatorial approach for transcriptional balancing of metabolic pathways. Sci.
Rep. 5, 11301. https://doi.org/10.1038/srep11301.
Jones, J.A., Vernacchio, V.R., Sinkoe, A.L., Collins, S.M., Ibrahim, M.H.A., Lachance, D.
M., Hahn, J., Koffas, M.A.G., 2016. Experimental and computational optimization of
an Escherichia coli co-culture for the efficient production of flavonoids. Metab. Eng.
35, 55–63. https://doi.org/10.1016/j.ymben.2016.01.006.
Kern, S.E., Newman, D.K., 2014. Measurement of phenazines in bacterial cultures. In:
Filloux, A., Ramos, J.-L. (Eds.), Pseudomonas Methods and Protocols. Humana Press,
New York, pp. 303–310.
Lau, G.W., Hassett, D.J., Ran, H., Kong, F., 2004. The role of pyocyanin in Pseudomonas
aeruginosa infection. Trends Mol. Med. 10, 599–606. https://doi.org/10.1016/j.
molmed.2004.10.002.
Lee, M.-Y., Hung, W.-P., Tsai, S.-H., 2017. Improvement of shikimic acid production in
Escherichia coli with growth phase-dependent regulation in the biosynthetic pathway
from glycerol. World J. Microbiol. Biotechnol. 33, 25. https://doi.org/10.1007/
s11274-016-2192-3.
Li, K., Xia, J., Mehmood, M.A., Zhao, X.-Q., Liu, C.-G., Bai, F.-W., 2019. Extracellular
redox potential regulation improves yeast tolerance to furfural. Chem. Eng. Sci. 196,
54–63. https://doi.org/10.1016/j.ces.2018.11.059.
Lim, C.G., Wong, L., Bhan, N., Dvora, H., Xu, P., Venkiteswaran, S., Koffas, M.A.G., 2015.
Development of a recombinant Escherichia coli strain for overproduction of the plant
pigment anthocyanin. Appl. Environ. Microbiol. 81, 6276–6284. https://doi.org/
10.1128/AEM.01448-15.
Liu, C.-G., Xue, C., Lin, Y.-H., Bai, F.-W., 2013. Redox potential control and applications
in microaerobic and anaerobic fermentations. Biotechnol. Adv. 31, 257–265.
https://doi.org/10.1016/j.biotechadv.2012.11.005.
Liu, D., Wan, N., Zhang, F., Tang, Y.J., Wu, S.G., 2017. Enhancing fatty acid production
in Escherichia coli by Vitreoscilla hemoglobin overexpression: enhancing fatty acid
A.J. da Silva et al.
production by VHb expression. Biotechnol. Bioeng. 114, 463–467. https://doi.org/
10.1002/bit.26067.
Livak, K.J., Schmittgen, T.D., 2001. Analysis of relative gene expression data using real-
time quantitative PCR and the 2(-Delta Delta C(T)) Method. Methods San Diego Calif
25, 402–408. https://doi.org/10.1006/meth.2001.1262.
Martínez-Gómez, K., Flores, N., Castañeda, H.M., Martínez-Batallar, G., Hernández-
Chávez, G., Ramírez, O.T., Gosset, G., Encarnación, S., Bolivar, F., 2012. New
insights into Escherichia coli metabolism: carbon scavenging, acetate metabolism and
carbon recycling responses during growth on glycerol. Microb Cell Fact 11, 46.
https://doi.org/10.1186/1475-2859-11-46.
Mavrodi, D.V., Bonsall, R.F., Delaney, S.M., Soule, M.J., Phillips, G., Thomashow, L.S.,
2001. Functional analysis of genes for biosynthesis of pyocyanin and phenazine-1-
carboxamide from Pseudomonas aeruginosa PAO1. J. Bacteriol. 183, 6454–6465.
https://doi.org/10.1128/JB.183.21.6454-6465.2001.
Mentel, M., Ahuja, E.G., Mavrodi, D.V., Breinbauer, R., Thomashow, L.S.,
Blankenfeldt, W., 2009. Of two make one: the biosynthesis of phenazines.
Chembiochem 10, 2295–2304. https://doi.org/10.1002/cbic.200900323.
Morrison, C., Heitmann, E., Armiger, W., Dodds, D., Koffas, M., 2018. Electrochemical
bioreactor technology for biocatalysis and microbial electrosynthesis. Adv. Appl.
Microbiol. 105, 51–86. https://doi.org/10.1016/bs.aambs.2018.07.001.
Moscoviz, R., Toledo-Alarcón, J., Trably, E., Bernet, N., 2016. Electro-fermentation: how
to drive fermentation using electrochemical systems. Trends Biotechnol. 34,
856–865. https://doi.org/10.1016/j.tibtech.2016.04.009.
Norman, R.S., Moeller, P., McDonald, T.J., Morris, P.J., 2004. Effect of pyocyanin on a
crude-oil-degrading microbial community. Appl. Environ. Microbiol. 70,
4004–4011. https://doi.org/10.1128/AEM.70.7.4004-4011.2004.
Rabaey, K., Boon, N., Höfte, M., Verstraete, W., 2005. Microbial phenazine production
enhances electron transfer in biofuel cells. Environ. Sci. Technol. 39, 3401–3408.
https://doi.org/10.1021/es048563o.
Rabaey, K., Girguis, P., Nielsen, L.K., 2011. Metabolic and practical considerations on
microbial electrosynthesis. Curr. Opin. Biotechnol. 22, 371–377. https://doi.org/
10.1016/j.copbio.2011.01.010.
Roos, V., Andersson, C.I.J., Bülow, L., 2004. Gene expression profiling of Escherichia coli
expressing double Vitreoscilla haemoglobin. J. Biotechnol. 114, 107–120. https://
doi.org/10.1016/j.jbiotec.2004.07.002.
Shen, C.R., Lan, E.I., Dekishima, Y., Baez, A., Cho, K.M., Liao, J.C., 2011. Driving forces
enable high-titer anaerobic 1-butanol synthesis in Escherichia coli. Appl. Environ.
Microbiol. 77, 2905–2915. https://doi.org/10.1128/AEM.03034-10.
25
Metabolic Engineering 64 (2021) 15–25
Shen, H.-B., Yong, X.-Y., Chen, Y.-L., Liao, Z.-H., Si, R.-W., Zhou, J., Wang, S.-Y.,
Yong, Y.-C., OuYang, P.-K., Zheng, T., 2014. Enhanced bioelectricity generation by
improving pyocyanin production and membrane permeability through sophorolipid
addition in Pseudomonas aeruginosa-inoculated microbial fuel cells. Bioresour.
Technol. 167, 490–494. https://doi.org/10.1016/j.biortech.2014.05.093.
Stark, B.C., Dikshit, K.L., Pagilla, K.R., 2012. The biochemistry of Vitreoscilla hemoglobin.
Comput. Struct. Biotechnol. J. 3, 1–8. https://doi.org/10.5936/csbj.201210002.
Studier, F.W., Moffatt, B.A., 1986. Use of bacteriophage T7 RNA polymerase to direct
selective high-level expression of cloned genes. J. Mol. Biol. 189, 113–130. https://
doi.org/10.1016/0022-2836(86)90385-2.
Sturm-Richter, K., Golitsch, F., Sturm, G., Kipf, E., Dittrich, A., Beblawy, S.,
Kerzenmacher, S., Gescher, J., 2015. Unbalanced fermentation of glycerol in
Escherichia coli via heterologous production of an electron transport chain and
electrode interaction in microbial electrochemical cells. Bioresour. Technol. 186,
89–96. https://doi.org/10.1016/j.biortech.2015.02.116.
Tsai, P.S., Hatzimanikatis, V., Bailey, J.E., 1996. Effect of Vitreoscilla Hemoglobin dosage
on microaerobic Escherichia coli carbon and energy metabolism. Biotechnol. Bioeng.
49, 139–150.
Venkataraman, A., Rosenbaum, M.A., Perkins, S.D., Werner, J.J., Angenent, L.T., 2011.
Metabolite-based mutualism between Pseudomonas aeruginosa PA14 and Enterobacter
aerogenes enhances current generation in bioelectrochemical systems. Energy
Environ. Sci. 4, 4550. https://doi.org/10.1039/c1ee01377g.
Wu, G., Yan, Q., Jones, A., Tang, Y.J., Fong, S.S., Koffas, M.A.G., 2016. Metabolic burden:
cornerstones in synthetic biology and metabolic engineering applications. Trends
Biotechnol. 34, 652–664. https://doi.org/10.1016/j.tibtech.2016.02.010.
Xu, P., Koffas, M.A.G., 2013. Assembly of multi-gene pathways and combinatorial
pathway libraries through ePathBrick vectors. In: Polizzi, K.M., Kontoravdi, C.
(Eds.), Synthetic Biology. Humana Press, Totowa, NJ, pp. 107–129. https://doi.org/
10.1007/978-1-62703-625-2_10.
Xu, P., Vansiri, A., Bhan, N., Koffas, M.A.G., 2012. ePathBrick: a synthetic biology
platform for engineering metabolic pathways in E. coli. ACS Synth. Biol. 1, 256–266.
https://doi.org/10.1021/sb300016b.
Zhang, L., Li, Y., Wang, Z., Xia, Y., Chen, W., Tang, K., 2007. Recent developments and
future prospects of Vitreoscilla hemoglobin application in metabolic engineering.
Biotechnol. Adv. 25, 123–136. https://doi.org/10.1016/j.biotechadv.2006.11.001.