Engineering of Synechococcus sp. strain PCC 7002 for the photoautotrophic production of light-sensitive riboflavin (vitamin B2)

Article abstract of DOI:10.1016/j.ymben.2020.09.010

Due to their capability of photosynthesis and autotrophic growth, cyanobacteria are currently investigated with regard to the sustainable production of a wide variety of chemicals. So far, however, no attempt has been undertaken to engineer cyanobacteria for the biotechnological production of vitamins, which is probably due to the light-sensitivity of many of these compounds. We now describe a photoautotrophic bioprocess to synthesize riboflavin, a vitamin used as a supplement in the feed and food industry. By overexpressing the riboflavin biosynthesis genes ribDGEABHT from Bacillus subtilis in the marine cyanobacterium Synechococcus sp. PCC 7002 riboflavin levels in the supernatant of the corresponding recombinant strain increased 56-fold compared to the wild-type. Introduction of a second promoter region upstream of the heterologous ribAB gene – coding for rate-limiting enzymatic functions in the riboflavin biosynthesis pathway – led to a further increase of riboflavin levels (211-fold compared to the wild-type). Degradation of the light-sensitive product riboflavin was prevented by culturing the genetically engineered Synechococcus sp. PCC 7002 strains in the presence of dichromatic light generated by red light-emitting diodes (λ = 630 and 700 nm). Synechococcus sp. PCC 7002 naturally is resistant to the toxic riboflavin analog roseoflavin. Expression of the flavin transporter pnuX from Corynebacterium glutamicum in Synechococcus sp. PCC 7002 resulted in roseoflavin-sensitive recombinant strains which in turn could be employed to select roseoflavin-resistant, riboflavin-overproducing strains as a chassis for further improvement.

Full text of DOI:10.1016/j.ymben.2020.09.010

Metabolic Engineering 62 (2020) 275–286
Contents lists available at ScienceDirect
Metabolic Engineering
journal homepage: www.elsevier.com/locate/meteng
Engineering of Synechococcus sp. strain PCC 7002 for the photoautotrophic
production of light-sensitive riboflavin (vitamin B2)
Benjamin Kachel, Matthias Mack*
Institute for Technical Microbiology, Department of Biotechnology, Mannheim University of Applied Sciences, 68163, Mannheim, Germany
A R T I C L E I N F O
A B S T R A C T
Keywords:
Due to their capability of photosynthesis and autotrophic growth, cyanobacteria are currently investigated with
regard to the sustainable production of a wide variety of chemicals. So far, however, no attempt has been un-
dertaken to engineer cyanobacteria for the biotechnological production of vitamins, which is probably due to the
light-sensitivity of many of these compounds. We now describe a photoautotrophic bioprocess to synthesize
riboflavin, a vitamin used as a supplement in the feed and food industry. By overexpressing the riboflavin
biosynthesis genes ribDGEABHT from Bacillus subtilis in the marine cyanobacterium Synechococcus sp. PCC 7002
riboflavin levels in the supernatant of the corresponding recombinant strain increased 56-fold compared to the
wild-type. Introduction of a second promoter region upstream of the heterologous ribAB gene – coding for rate-
limiting enzymatic functions in the riboflavin biosynthesis pathway – led to a further increase of riboflavin levels
(211-fold compared to the wild-type). Degradation of the light-sensitive product riboflavin was prevented by
culturing the genetically engineered Synechococcus sp. PCC 7002 strains in the presence of dichromatic light
generated by red light-emitting diodes (λ = 630 and 700 nm). Synechococcus sp. PCC 7002 naturally is resistant to
the toxic riboflavin analog roseoflavin. Expression of the flavin transporter pnuX from Corynebacterium gluta-
micum in Synechococcus sp. PCC 7002 resulted in roseoflavin-sensitive recombinant strains which in turn could be
employed to select roseoflavin-resistant, riboflavin-overproducing strains as a chassis for further improvement.
Synechococcus sp. PCC 7002
Photoautotrophy
Riboflavin
Vitamin B2
Roseoflavin
1. Introduction
a variety of chemicals of industrial relevance, including hydrogen gas
(Mona et al., 2020), ethanol (Kopka et al., 2017), isoprene (Chaves and
Most of the current bioindustrial processes are based on heterotro-
phic microorganisms that utilize classical (“first generation”) feedstocks
such as starch, glucose or fructose which are also used in the food- and
feed-industry. To avoid competition and to increase the social-ecological
sustainability of biotechnological processes, it is necessary to evaluate
novel host organisms that can thrive on alternative carbon sources. One
approach is to use photoautotrophic cyanobacteria to produce chemicals
from light and CO2, thus omitting carbohydrate feedstocks while
reducing greenhouse gas emission (Ducat et al., 2011). Marine cyano-
bacteria can be cultured in a wide range of salinity and the corre-
sponding biotechnological processes are thus independent of freshwater
resources. Compared to eukaryotic plants and algae, cyanobacteria grow
relatively fast and have a higher solar energy conversion efficiency
(Dismukes et al., 2008). Some cyanobacteria are naturally transformable
(Vioque, 2007) and toolboxes for genetic engineering are being devel-
oped at increasing speed (Ruffing et al., 2016; Santos-Merino et al.,
2019; Zess et al., 2016). Cyanobacteria have been engineered to produce
Melis, 2018), isobutanol (Miao et al., 2018a, 2018b), terpenes (Choi
et al., 2016; Davies et al., 2014), polylactide (Hirokawa et al., 2017;
Varman et al., 2013), poly-β-hydroxybutyrate (Singh and Mallick,
2017), sucrose (Ducat et al., 2012; Weiss et al., 2017) and amino acids
(Korosh et al., 2017).
Vitamins are essential compounds in human and animal diets. Their
demand is increasing globally in food, feed, cosmetics, chemical and
pharmaceutical industries (Acevedo-Rocha et al., 2019). Most vitamins
are currently produced by chemical synthesis using non-renewable
sources and often generating hazardous waste. Riboflavin (vitamin B2)
is a remarkable exception as it is now exclusively produced employing
biotechnological processes mostly based on the filamentous fungus
Ashbya gossypii (a natural riboflavin overproducer) and the
gram-positive bacterium Bacillus subtilis. In 2015, the riboflavin industry
reached a market volume of 9000 tons per year, most of which is used to
fortify animal feed (Schwechheimer et al., 2016). Riboflavin is synthe-
sized by bacteria and fungi in a seven-step enzymatic process (Fischer
* Corresponding author. Institute for Technical Microbiology Mannheim University of Applied Sciences, Paul-Wittsack-Str. 10, 68163, Mannheim, Germany.
E-mail address: m.mack@hs-mannheim.de (M. Mack).
https://doi.org/10.1016/j.ymben.2020.09.010
Received 13 July 2020; Received in revised form 9 September 2020; Accepted 19 September 2020
Available online 28 September 2020
1096-7176/© 2020 International Metabolic Engineering Society. Published by Elsevier Inc. All rights reserved.

B. Kachel and M. Mack
Metabolic Engineering 62 (2020) 275–286
and Bacher, 2005) (Fig. 1). The riboflavin biosynthetic genes and en-
zymes have been studied in a variety of different species and this in-
formation was used to engineer heterotrophic microorganisms such as
A. gossypii and B. subtilis (but also other microorganisms) for the pro-
duction of riboflavin. Photoautotrophic Synechococcus sp. PCC 7002 is a
naturally transformable, comparably fast-growing, unicellular, eryha-
line cyanobacterium that has been shown to be a promising host or-
ganism for biotechnological applications (Gangl et al., 2015; Ludwig
and Bryant, 2011; Wijffels et al., 2013). Synechococcus sp. PCC 7002 has
not yet been evaluated as a host for the production of vitamins which is
probably due to the light-sensitivity of many vitamins (Sheraz et al.,
2014). In the present work we demonstrate that a CO2-based process for
the bioproduction of a light-sensitive vitamin such as riboflavin is
possible using a genetically engineered cyanobacterium under special-
ized culture conditions.
2. Results
2.1. Synechococcus sp. PCC 7002 contains all genes necessary to
biosynthesize riboflavin
Seven enzymatic functions/genes are necessary to synthesize ribo-
flavin from GTP and ribulose-5′-phosphate, whereby one enzyme/gene
of the riboflavin biosynthetic (rib) pathway is still unknown (Fig. 1). In
Fig. 1. Biosynthesis of riboflavin and flavin cofactors in bacteria (Fischer and Bacher, 2005). Riboflavin is synthesized from one molecule guanosine-5′-triphosphate
(GTP) and two molecules ribulose-5-phosphate in a seven step enzymatic process (1.-7.). Riboflavin is further processed into the flavin cofactors flavin mono-
nucleotide (FMN) and flavin adenine dinucleotide (FAD) which are employed by flavoenzymes (8.-9.). Enzyme short names are according to the new genetic
nomenclature (Jankowitsch et al., 2012) whereby bifunctional enzymes are denoted e.g. as RibAB. Accordingly, RibAB functions both as a GTP cyclohydrolase II
(step 1, RibA) (EC 3.5.4.25) and a (3S)-3,4-dihydroxy-2-butanone 4-phosphate synthase (step 5, RibB) (EC 4.1.99.12). RibDG is a bifunctional 2,5-diamino-6-r-
ibosylamino-4(3H)-pyrimidinone 5′-phosphate deaminase (step 2, RibD) (EC 3.5.4.26) and a 5-amino-6-ribosylamino-2,4(1H,3H)-pyrimidinedione 5′-phosphate
reductase (step 3, RibG) (EC 1.1.1.193). In case of bifunctional enzymes the underline shows which enzyme function is active. RibH is a lumazine synthase (step 6)
(EC 2.5.1.78). RibE is a riboflavin synthase (EC 2.5.1.9) dismutating two molecules of VI to give IV and riboflavin (step 7). RibFC is bifunctional and acts as a
riboflavinkinase (RibF) (EC 2.7.1.26) and an FMN adenylyltransferase (RibC) (EC 2.7.7.2). The combined activities of rather unspecific phosphatases appear to
catalyze dephosphorylation of III to IV (step 4) (EC 3.1.3.104). A single, riboflavin pathway specific enzyme has not yet been described. Intermediates of the
riboflavin biosynthesis pathway are: I: 2,5-Diamino-6-(5-phospho-D-ribosylamino)pyrimidin-4(3H)one; II: 5-Amino-6-ribosylamino-2,4(1H,3H)-pyrimidinedione
5′-phosphate; III: 5-Amino-6-ribitylamino-2,4(1H,3H)-pyrimidinedione 5′-phosphate; IV: 5-Amino-6-ribitylamino-2,4(1H,3H)-pyrimidinedione; V: (3S)-3,4-Dihy-
droxy-2-butanone 4-phosphate; VI: 6,7-Dimethyl-8-D-ribityllumazine. Coloured arrows represent the riboflavin biosynthetic (rib) genes from Synechococcus sp. PCC
7002 studied in this work and show their location in the circular chromosome. Notably, two ribD genes (A2264 and A0498) are annotated in the genome of Syn-
echococcus sp. PCC 7002. (For interpretation of the references to colour in this figure legend, the reader is referred to the Web version of this article.)
276

B. Kachel and M. Mack
Metabolic Engineering 62 (2020) 275–286
some bacterial species two enzymes of the rib pathway are bifunctional
(RibAB and RibDG) whereas RibE and RibH in all known species are
monofunctional. To the best of our knowledge riboflavin biosynthesis
has not been studied in cyanobacterial species and to fill this gap the
following experiments were carried out. In the genome of Synechococcus
sp. PCC 7002 the gene SYNPCC7002_A0136 is annotated as ribH, SYN-
PCC7002_A0427 is annotated as ribAB, and SYNPCC7002_A2264 is an-
notated as ribE. The gene product of SYNPCC7002_A1465 is similar to
bifunctional RibDG from B. subtilis and thus the gene A1465 should be
annotated as ribDG (instead of ribD). In addition, the Synechococcus sp.
PCC 7002 genome contains a second putative ribD gene (SYN-
PCC7002_A0498) which is annotated as “ribD C-terminal domain”. The
translation product shows significant similarity to the C-terminal RibG-
part of B. subtilis RibDG and thus SYNPCC7002_A0498 probably encodes
a monofunctional riboflavin specific reductase (RibG). The rib genes in
Synechococcus sp. PCC 7002 are scattered over the chromosome (Fig. 1).
The following complementation experiments employing the riboflavin
auxotrophic Escherichia coli strains BSV11 (ΔribB), BSV13 (ΔribE),
BSV18 (ΔribA) and SI#78 (ΔribDG) were carried out to find out whether
the Synechococcus sp. PCC 7002 genes annotated as rib genes indeed
encode for enzymes involved in riboflavin biosynthesis. BSV11 (ΔribB)
was transformed with an expression plasmid containing SYN-
PCC7002_A0427 (ribAB). The resulting recombinant strain was grown
for 12 h in LB (without additional riboflavin) and the final OD600 value
was determined (Supplemental Fig. S1). The BSV11 strain transformed
with an empty plasmid was used as a control. In contrast to the control,
the strain expressing SYNPCC7002_A0427 grew to a final OD600nm > 1.0
suggesting that this gene indeed encodes for a RibB function. Accord-
ingly, the RibA function of SYNPCC7002_A0427 could be verified
employing BSV18 (ΔribA), the RibE function of SYNPCC7002_A2264
could be verified using BSV13 (ΔribE) and the RibDG function of SYN-
PCC7002_A1465 could be validated using SI#78 (ΔribDG). Notably,
SYNPCC7002_A0498 did not rescue the ribDG-deficient SI#78 strain
indicating that SYNPCC7002_A0498 does not encode a bifunctional
deaminase/reductase. We cannot exclude that SYNPCC7002_A0498
encodes a monofunctional reductase, although such enzymes have only
been identified in plants (Hasnain et al., 2013). Since a ribH-deficient
E. coli strain was not available for complementation experiments, the
RibH function of SYNPCC7002_A0136 was tested at the enzyme level.
Two different tagged versions (N-terminal- and C-terminal Strep-tag) of
the putative Synechococcus sp. PCC 7002 RibH were overproduced in
E. coli BL21(DE3), purified by affinity chromatography and tested using
a RibH enzyme assay (Haase et al., 2013; Sarge et al., 2015). Both RibH
versions produced the expected RibH reaction product 6,7-dimethyl-8--
ribityllumazine (Supplemental Fig. S2). Our experiments confirmed the
annotations with regard to the riboflavin biosynthetic enzymes in
riboflavin prototrophic Synechococcus sp. PCC 7002. A Synechococcus sp.
PCC 7002 transcriptome study showed that on a minimal growth me-
dium not containing riboflavin all of the above mentioned annotated rib
genes were expressed in all tested growth phases and under all tested
conditions (Ludwig and Bryant, 2011, 2012). Notably, SYN-
PCC7002_A0498 (monofunctional ribG) also was expressed and the
corresponding gene product may as well support riboflavin biosynthesis
in Synechococcus sp. PCC 7002. At the amino acid level the RibG do-
mains (reductase functions) encoded by SYNPCC7002_A0498 and SYN-
PCC7002_A1465 only share comparably weak similarities (ca. 26%) and
the genes probably were acquired independently. The presence of more
than one ribDG/ribG gene appears to not be unusual and also was
observed in other bacteria (Kissling et al., 2020).
upon illumination at wavelengths <500 nm photolysis at the ribityl
chain was reported to occur yielding the degradation products lumi-
chrome and lumiflavin (Sheraz et al., 2014) (Fig. 2). To evaluate the
stability of riboflavin under standard Synechococcus sp. PCC 7002
growth conditions (Vogel et al., 2017), riboflavin (50 μM) was dissolved
in the growth medium, exposed to white light (tubular fluorescent lamp
Philips Master TL-D, 18W/840, ~40 E) and the mixture subsequently
μ
was analyzed with regard to flavins at different time points (up to 72 h).
As expected, degradation of riboflavin (half-life 0.96 h) and concomitant
formation of lumichrome (but not of lumiflavin) were observed
(Fig. 2C). As it takes three weeks for Synechococcus sp. PCC 7002 to enter
the stationary growth phase in minimal medium at ambient CO2,
cultivation under white light would lead to degradation of riboflavin. It
was shown previously that Synechococcus sp. PCC 7002 can be cultivated
under dichromatic red light (λ = 630 nm and λ = 680 nm), specifically
stimulating the photosystem pigments phycocyanin and chlorophyll-a
(Bernstein et al., 2014; Bland and Angenent, 2016; Melnicki et al.,
2013). We used a similar approach and constructed a dichromatic LED
array containing diodes emitting at λ = 630 nm and λ = 700 nm. To test
the stability of riboflavin under these conditions, riboflavin (50
μM) was
incubated under dichromatic red LED light (15 E of λ = 630 nm, 10 μE
μ
of λ = 700 nm) at 32 ◦C for 42 days. Under these conditions the half-life
of riboflavin was calculated to be 501 days. For the control samples
incubated in the dark the half-life of riboflavin was calculated to be 911
days (Fig. 2C). The levels of lumichrome were <0.1% both in red light
and in the dark, while lumiflavin levels were below the detection limit.
Wild-type Synechococcus sp. PCC 7002 was cultivated for 25 days under
white light or dichromatic LED red light conditions. Both cultures
reached the stationary phase after approximately 20 days. Cells were
analyzed by light microscopy and no differences in cell size or shape
were observed. The higher OD values of the red light cultures can be
explained by the fact that the light intensities of the dichromatic LED red
light were higher compared to the white light source, when only the
photosynthetically active radiation is considered (Fig. 2D).
2.3. Characterization of cyanobacterial regulatory elements under red-
light conditions
Some of the strongest regulatory elements/promoters described for
cyanobacteria are the “cpc560 promoter” (Liu and Pakrasi, 2018; Zhou
et al., 2014) and the “psbA2 promoter” derived from Synechocystis sp.
PCC 6803 (Englund et al., 2016; Xu et al., 2011), both driving genes of
the photosynthetic apparatus. These regulatory elements are affected by
different light intensities and wavelengths (Conley et al., 1985; Eriksson
et al., 2000) and thus their performance had to be tested with regard to
our customized LED red-light conditions (see 2.2) using a reporter gene.
For the “psbA2 promoter” two versions were tested. The “long version”
(“psbA2L′′) consisted of a 493 nt DNA sequence present upstream of the
start codon of the highly expressed gene psbA2 encoding the photo-
system II D1 protein in Synechocystis sp. PCC 6803. The “short version”
(“psbA2s”) consisted of a 208 nt DNA sequence present upstream of the
start codon of psbA2. In addition, a DNA sequence 500 nt upstream of the
start codon of the SYNPCC7002_A2813 gene was tested as it was shown
to act as a moderately strong constitutive promoter in Synechococcus sp.
PCC 7002 (Ludwig and Bryant, 2011; Ruffing et al., 2016). We also
tested the pTrcTheo regulatory element in our expression analyses. This
element consists of an artificial E. coli trc-promoter in combination with
a theophylline-inducible riboswitch which previously has been used for
inducible gene expression experiments in Synechococcus elongatus PCC
7942 (Nakahira et al., 2013) and Synechocystis sp. PCC 6803 (Ohbayashi
et al., 2016). To examine the ability of the different regulatory regions to
drive expression of heterologous genes in Synechococcus sp. PCC 7002
the regions were ligated to a gene encoding a superfolder GFP (Pedelacq
et al., 2006) reporter protein and the resulting constructs were chro-
mosomally integrated at the previously described “neutral site 2” (NS2)
(Ruffing et al., 2016; Xu et al.) of Synechococcus sp. PCC 7002. The
2.2. Cultivation under red LED light enables growth of Synechococcus sp.
PCC 7002 and prevents degradation of riboflavin
Cyanobacterial species are routinely cultivated in the laboratory
under white light from a tubular fluorescent lamp which generates a
broad spectrum of wavelengths. Riboflavin is sensitive to white light and
277

B. Kachel and M. Mack
Metabolic Engineering 62 (2020) 275–286
Fig. 2. Stability of riboflavin and growth of Syn-
echococcus sp. PCC 7002 under white and red light
conditions. (A) Light intensity distribution of a stan-
dard white tubular fluorescent lamp (Philips TL-D
18W/840, grey) overlaid by the absorption spec-
trum of riboflavin in AA + medium (orange) and the
light intensity spectrum of our customized dichro-
matic red light LED-array (red). The flavins are not
exposed to light of wavelengths < λ = 500 nm. (B)
The major photo-degradation products of riboflavin
are lumichrome and lumiflavin which form upon
illumination < λ = 500 nm (Sheraz et al., 2014). (C)
Degradation of riboflavin (50 μM) in AA + medium
under white light, red light and in the dark. Forma-
tion of lumichrome occurred only when riboflavin
was incubated under white light conditions, but levels
remained at baseline for the samples incubated under
red light and dark conditions. (D) Synechococcus sp.
PCC 7002 wild-type grows to similar ODs under white
and red light conditions, respectively. WT: wild-type,
RF: Riboflavin, LC: Lumichrome. Lumiflavin was not
detected. (For interpretation of the references to
colour in this figure legend, the reader is referred to
the Web version of this article.)
resulting reporter strains were cultivated in 6-well plates to the sta-
tionary phase (21 days) and GFP fluorescence was determined (Fig. 3).
The strongest GFP fluorescence signal was found in the reporter strain
using cpc560 for reporter gene expression whereby a strong signal
already was found in the early exponential growth phase. The A2813
regulatory element was more active in the stationary phase being in line
with previously described results (Ruffing et al., 2016). Both regulatory
elements psbA2L and psbA2s yielded moderate reporter gene expression
both in the exponential as well as in the stationary growth phase. The
pTrcTheo element was as strong as A2813 upon induction with 3 mM
theophyllin, whereas the uninduced cultures generated fluorescence
similar to the wild-type control. Under the applied conditions the reg-
ulatory elements cpc560 and A2813 strongly stimulated reporter gene
expression and thus appeared to be the best choice for the following
overexpression experiments employing heterologous rib genes from
B. subtilis.
2.4. Overexpression of B. subtilis riboflavin biosynthesis genes in
Synechococcus sp. PCC 7002 leads to strongly enhanced riboflavin levels
It was previously shown for B. subtilis that overexpression of the
endogenous riboflavin biosynthesis operon ribDGEABHT led to high
levels of riboflavin in cultures of the corresponding recombinant strains
and that the catalytic rates of the rib enzymes are high enough to
establish an economic riboflavin production process (Hohmann and
Stahmann, 2010; Perkins et al., 1999). To prevent recombination that
might occur when introducing additional copies of the endogenous rib
genes into the chromosome of Synechococcus sp. PCC 7002, we decided
to utilize the B. subtilis rib genes to create a riboflavin-overproducing
Synechococcus sp. PCC 7002 strain. Accordingly, the B. subtilis rib
biosynthetic operon ribDGEABHT was coupled to the cyanobacterial
regulatory elements cpc560, psbA2s and A2813 which were shown to
support gene expression under the dichromatic red light conditions (see
2.3). A variety of different constructs (described in Fig. 4) was generated
to find out which genetic arrangement performs best with regard to
riboflavin overproduction. The resulting expression constructs were
chromosomally integrated at NS2 of Synechococcus sp. PCC 7002. As the
RibA-catalyzed GTP-cyclohydrolase II reaction was reported to repre-
sent a rate-limiting step in B. subtilis riboflavin production strains (Bir-
kenmeier et al., 2014; Perkins et al., 1999), one of the constructs
(psbA2s_ribDGEABHT) was modified to contain an additional internal
Fig. 3. Reporter gene experiments comparing the strengths of different
cyanobacterial regulatory elements in Synechococcus sp. PCC 7002.
Different regulatory/promoter regions were coupled to a reporter gene encod-
ing GFP and the resulting plasmid constructs were integrated into the chro-
mosome of Synechococcus sp. PCC 7002 at “neutral site 2 (NS2)”. The strains
were grown in AA + medium and fluorescence of the culture samples was
determined using a plate reader set to λ = 485 nm excitation and λ = 535 nm
emission. The pTrcTheo regulatory element was induced with 3 mM theophyl-
line at day 0.
278

B. Kachel and M. Mack
Metabolic Engineering 62 (2020) 275–286
Fig. 4. Overexpression of the Bacillus subtilis ribo-
flavin biosynthetic genes (rib genes) in Synechococcus
sp. PCC 7002 results in riboflavin-overproducing
strains. (A) Design of different plasmid constructs
for overexpression of the B. subtilis rib genes. All genes
were fused to a kanamycin resistance cassette (KmR)
and flanked by homology regions for integration into
the chromosome of Synechococcus sp. PCC 7002 (at
NS2). The constructs carry either the wild-type
B. subtilis ribDGEABHT genes in combination with
different promoters (broken arrows) (1), the psbA2s
promoter upstream of ribDG and an additional cpc560
promoter upstream of ribAB (2), or codon-adapted (c.
o.) rib genes. (B) Strain 2 was grown for 14 days and
the cell suspension was centrifuged. When compared
to Synechococcus sp. PCC 7002 wild-type (left sample)
the supernatant of
a strain overexpressing the
ribDGEABHT genes is yellow (right sample) due to the
produced riboflavin (which is yellow in its oxidized
form). The green pellet represents the chlorophyll-a
containing phototrophic cells. (C) Flavin levels in
the whole-culture lysate of genetically engineered
Synechococcus sp. PCC 7002 strains at day 13 of
growth. Highest riboflavin levels were observed for
the cpc560 variant (1a) and the variant containing
both psbA2s and cpc560 (2). Cultivation of the psbA2s
variant in white light yields mostly lumichrome
underlining the importance of growing riboflavin
production strains under red light. (D) Flavin levels in
the supernatant and pellet of the best-performing
Synechococcus sp. PCC 7002 strain containing
expressing construct 2. The supernatant contains
mostly uncharged riboflavin as it permeates passively
over the cell membrane. Error bars display one SD (n
= 3). WT: wild-type, RF: Riboflavin, LC: Lumichrome.
(For interpretation of the references to colour in this
figure legend, the reader is referred to the Web
version of this article.)
cpc560 promoter directly upstream of the ribAB gene resulting in
psbA2s_ribDGE_cpc560_ribABHT (Fig. 4A). Since codon-adaptation of
heterologous genes was found to enhance gene expression in cyano-
bacteria (Sebesta and Peebles, 2020; Zhou et al., 2016) we
codon-adapted the B. subtilis rib genes for expression in Synechococcus sp.
PCC 7002 (psbA2s_ribABDGEHT (c.o.)). To boost expression of ribAB in
this construct we rearranged the rib genes placing ribAB (encoding
rate-limiting enzymatic functions) immediately downstream of the
promoter region. In addition, a strain was generated overexpressing a
codon-adapted version of the B. subtilis ribAB gene (but not over-
expressing the other rib genes) (psbA2s_ribAB (c.o.)) (Fig. 4A). Following
validation of integration and complete segregation all strains were
cultivated in 6-well plates under LED red-light conditions (see Fig. 2).
Samples were taken from the liquid cultures and flavin levels in lysed
cell suspensions were analyzed by HPLC. In case of the strains containing
only one regulatory element to drive expression of the heterologous
B. subtilis genes ribDGEABHT, cpc560 apparently performed best as the
corresponding cultures produced more riboflavin when compared to the
cultures where psbA2s or A2813 drove gene expression. This was in line
with the results of the reporter gene expression experiments described in
the previous section. Unexpectedly, the strains overexpressing
codon-adapted rib genes (containing the expression construct psbA2s_-
ribDGEABHT (c.o.)) produced comparably small amounts of riboflavin
and we speculate that the mRNAs derived from the synthetic genes were
less stable. The highest flavin levels were generated by the strain which
contained two promoters to drive expression of the not codon-adapted
ribDGEABHT genes (expression construct psbA2s_ribDGE_cpc560_ri-
bABHT at NS2) (Fig. 4C). The strain overexpressing the B. subtilis rib
genes driven by a single psbA2s promoter (expression construct
psbA2s_ribDGEABHT at NS2) was also grown under white light condi-
tions and the fact that lumichrome but not riboflavin was present shows
that white light induced degradation of riboflavin is strong also under
the specific conditions of a liquid cell culture (Fig. 4C). Note that under
cell-free conditions a similar degradation of riboflavin was observed
(Fig. 2C). To analyse levels of the different flavin compounds in the
supernatant and cellular pellet of the recombinant strains, the strain
containing the expression construct psbA2s_ribDGE_cpc560_ribABHT and
generating the highest riboflavin levels was grown for 14 days. The cell
suspension was centrifuged and the supernatant and the cell pellet were
analyzed separately with regard to flavins. The data revealed that un-
charged riboflavin and lumichrome were mainly present in the super-
natant (and thus followed the concentration gradient), while charged
FMN mostly was retained within the cells (Fig. 4D). This is in line with
observations in B. subtilis and other riboflavin-overproducing microor-
ganisms where riboflavin is thought to passively permeate over the
cytoplasmic membrane when cytoplasmic levels are high enough
(Hemberger et al., 2011).
2.5. Analysis of flavin production of two riboflavin-overproducing
Synechococcus sp. PCC 7002 strains
For a more detailed characterization of growth and flavin production
Synechococcus sp. PCC 7002 wild-type and the two best-performing
riboflavin-overproducing strains containing the expression constructs
cpc560_ribDGEABHT and psbA2s_ribDGE_cpc560_ribABHT (see previous
section), respectively, were cultivated in 100 mL Erlenmeyer flasks for
38 days (and not in 6-well plates as in 2.4) under red-light conditions. In
case of the strain expressing cpc560_ribDGEABHT riboflavin levels in the
279

B. Kachel and M. Mack
Metabolic Engineering 62 (2020) 275–286
supernatant rapidly increased until day 10 but then remained constant
Fig. S3). These results strongly suggest that increased riboflavin levels in
the recombinant strains were indeed caused by overexpression of the
B. subtilis rib genes.
at 19.9 (±1.0) μM. In contrast, the strain carrying psbA2s_ribDGE_cp-
c560_ribABHT displayed an almost linear increase in riboflavin levels,
reaching a final concentration of 73.9 (±7.2) M (Fig. 5A). In this case
production increased 211-fold compared to the wild-type, which syn-
thesizes 0.34 (±0.04) M of the primary metabolite riboflavin. Under
μ
2.7. Monitoring production of heterologous B. subtilis riboflavin
biosynthesis enzymes RibDGEABHT in Synechococcus sp. PCC 7002 by
mass-spectrometry
μ
these conditions the major degradation product lumichrome represents
a fraction of 17% and 9% of total flavins for the strains expressing
cpc560_ribDGEABHT and psbA2s_ribDGE_cpc560_ribABHT, respectively
(Fig. 5B). It appears that increased riboflavin levels in the cultures
protects from degradation. No significant difference in growth was
observed for the riboflavin overproducing strains when compared to the
wild-type (Fig. 5C) being in line with the finding that the chlorophyll-a
levels in all strains are very similar (Fig. 5D).
The following experiments were carried out to validate that over-
expression of the B. subtilis rib genes in different recombinant Synecho-
coccus sp. PCC 7002 strains indeed leads to an increased amount of
heterologous riboflavin biosynthetic enzymes and that their activity
boosts riboflavin production in the photoautotrophic host. The soluble
proteins present in cell lysates of the riboflavin-overproducing strain
carrying the expression construct psbA2s_ribDGE_cpc560_ribABHT
(exponential growth phase) were separated by SDS-PAGE. The regions
corresponding to the expected apparent molecular masses of the
B. subtilis riboflavin biosynthesis enzymes were cut from the gel, proteins
within the gel piece were digested and masses of the resulting peptides
were determined by mass spectrometry. All B. subtilis rib enzymes were
detected with sufficient coverage, showing that the overproduced en-
zymes indeed were present (Supplemental Fig. S4). The amount of
B. subtilis enzymes present in the riboflavin-overproducing Synecho-
coccus sp. PCC 7002 strains carrying the expression constructs
cpc560_ribDGEABHT and psbA2s_ribDGE_cpc560_ribABHT, respectively,
were compared to the B. subtilis enzyme levels present in a strain car-
rying psbA2s_ribDGEABHT. When the additional promoter region cpc560
was present upstream of heterologous ribAB, RibAB enzyme levels were
4-fold higher when compared to a strain carrying only a single promoter
2.6. Monitoring expression of heterologous B. subtilis riboflavin
biosynthesis genes ribDGEABHT in Synechococcus sp. PCC 7002
To confirm expression of the heterologous B. subtilis riboflavin
biosynthesis genes in recombinant Synechococcus sp. PCC 7002, RT-PCR
was carried out using total RNA extracted from exponentially growing
Synechococcus sp. PCC 7002 wild-type (negative control) and two
riboflavin-overproducing strains carrying the expression constructs
cpc560_ribDGEABHT or psbA2s_ribDGE_cpc560_ribABHT (see previous
section). Expression of the endogenous gene SYNPCC7002_A0427
(encoding Synechococcus sp. PCC 7002 RibAB) was used as a positive
control. RT-PCR amplification using gene-specific primers revealed that
the B. subtilis genes ribAB, ribDGE, and ribHT are expressed, as specific
PCR products could be generated from total RNA samples (Supplemental
Fig. 5. Flavin production, growth and chlorophyll-a
levels of two different recombinant Synechococcus
sp. PCC 7002 strains overexpressing the Bacillus sub-
tilis riboflavin biosynthetic genes (rib genes) in com-
parison to the wild-type (WT). (A) Synechococcus sp.
PCC 7002 strains containing the expression constructs
cpc560_ribDGEABHT (“cpc560”) and psbA2s_ribDG-
E_cpc560_ribABHT (“psbA2s/cpc560”) (see Fig. 4A)
were cultivated in 100 mL Erlenmeyer flasks under
red-light conditions and flavin levels were deter-
mined in the supernatants. The psbA2s/cpc560 strain
continuously accumulates riboflavin, whereby the
cpc560 strain produces riboflavin mainly in the initial
phase of growth. (B) The main photo-degradation
product lumichrome is found in low levels in the
riboflavin-overproducing strains. (C) All Synecho-
coccus sp. PCC 7002 variants show similar growth
(OD730). (D) All strains show similar chlorophyll-a
levels indicating that riboflavin overproduction does
not negatively interfere with cyanobacterial physi-
ology. All data points are averages of triplicates, error
bars show one SD. The data were fitted using a fourth-
order polynomic non-linear regression in GraphPad
Prism. RF: Riboflavin, LC: Lumichrome. (For inter-
pretation of the references to colour in this figure
legend, the reader is referred to the Web version of
this article.)
280

B. Kachel and M. Mack
Metabolic Engineering 62 (2020) 275–286
to drive expression of rib genes. This also was in line with the drastically
be present in gram-negative bacteria (Gutierrez-Preciado et al., 2015).
Accordingly, Synechococcus sp. PCC 7002 was found to be
roseoflavin-resistant (Fig. 6A). To use the roseoflavin selection strategy
in Synechococcus sp. PCC 7002 the gene encoding the flavin transporter
pnuX from Corynebacterium glutamicum was chromosomally integrated at
NS2 of Synechococcus sp. PCC 7002 and expressed employing the regu-
latory element A2813. The resulting strain Synechococcus sp. PCC 7002
different riboflavin levels found in these strains (Supplemental Fig. S5).
2.8. Generation of a roseoflavin-sensitive Synechococcus sp. PCC 7002
strain and selection of riboflavin-overproducing mutants
The antibiotic roseoflavin is a structural riboflavin analog and in-
pnuX was sensitive to roseoflavin at a concentration <2 μM, whereas the
hibits growth of roseoflavin sensitive organisms such as B. subtilis (Otani
wild-type strain grew in the presence of 200 μM roseoflavin (Fig. 6A). To
et al., 1974). When 100 μL of a liquid B. subtilis stationary phase culture
select roseoflavin-resistant strains 2 × 108 cells of Synechococcus sp. PCC
is spread on a solid growth medium containing 100 μM roseoflavin, few
7002 pnuX were plated on AA + agar containing 100 μM roseoflavin and
colonies of spontaneous roseoflavin resistant strains appear after 2 days
of incubation (Matsui et al., 1982). Some of these strains contain mu-
tations which lead to an enhanced expression of riboflavin biosynthetic
genes and thus to riboflavin overproduction. Overproduction of ribo-
flavin in turn compensates the toxic effect of the antimetbolite rose-
oflavin. Roseoflavin is taken up by riboflavin transporters and thus only
these organisms are roseoflavin-sensitive which contain flavin uptake
systems (Hemberger et al., 2011). Riboflavin transporters appear to be
widespread in gram-positive bacteria such as B. subtilis, but seem not to
colonies appeared with a frequency of ca. 4 × 10ꢀ 6 (Fig. 6B). Some of
these apparently roseoflavin-resistant strains were grown in liquid cul-
tures in the absence of roseoflavin and showed increased riboflavin
production (5–11 μM per OD730) (Fig. 6C). Two roseoflavin-resistant but
not-riboflavin-overproducing strains (Fig. 6B) contained a mutation in
the start codon of pnuX and a premature stop codon in pnuX, respec-
tively, explaining their roseoflavin resistance. We noticed that levels of
riboflavin present in the supernatant of roseoflavin-selected strains
Fig. 6. A recombinant Synechococcus sp. PCC 7002
strain expressing the flavin transporter pnuX from
Corynebacterium glutamicum is sensitive to roseoflavin
and can be employed to select riboflavin-
overproducing mutants. (A) Synechococcus sp. PCC
7002 expressing C. glutamicum pnuX (“pnuX”) is sen-
sitive to roseoflavin (RoF), whereas the wild-type
(WT) tolerates high levels of roseoflavin. (B) Sponta-
neous RoF resistant mutants were selected on plates
containing 100 μM RoF. (C) Spontaneous RoF resis-
tant strains were grown in a liquid medium (red bars)
and some of them were found to overproduce ribo-
flavin. The parent pnuX-expressing strain (white bar;
n = 3) and wild-type Synechococcus sp. PCC 7002 (red
bar; n = 3) do not overproduce riboflavin (D) Cells of
a stationary phase liquid culture (L#1) of roseoflavin-
selected riboflavin-overproducing Synechococcus sp.
PCC 7002 strains (#48, #50, #51, #52, #55, #56,
#1new, #2new) were used to inoculate fresh liquid
cultures (L#2). The process was repeated three times
(L#3–5). For unknown reasons riboflavin levels of all
riboflavin-overproducing strains initially decreased
but then stayed constant at about 2.3
μM per OD730
(upper dotted line). The wild-type produces riboflavin
levels of about 0.08 μM per OD730 (lower dotted line).
(For interpretation of the references to colour in this
figure legend, the reader is referred to the Web
version of this article.)
281

B. Kachel and M. Mack
Metabolic Engineering 62 (2020) 275–286
grown in the absence of roseoflavin decreased over time in liquid cul-
tures inoculated with stationary phase cultures. Following five rounds of
2.9. Overexpression of the B. subtilis rib operon in a roseoflavin-selected
Synechococcus sp. PCC 7002 strain further increased riboflavin levels
inoculation the flavin levels remained at an average of 2.3
OD730, which is 28-fold higher than in the wild-type (0.08
μ
M flavins per
M) (Fig. 6C).
μ
It was shown previously for commercial riboflavin-overproducing
B. subtilis strains that mutants selected on solid growth media contain-
ing toxic antimetabolites such as roseoflavin serve as excellent parent
strains for further strain improvement (Perkins et al., 1999). One of the
roseoflavin-resistant, riboflavin-overproducing Synechococcus sp. PCC
7002 strains (strain #48; Fig. 6C) was employed for overexpression
experiments. Strain #48 was transformed with a plasmid containing the
expression construct psbA2s_ribDGE_cpc560_ribABHT fused to a chlor-
amphenicol resistance cassette and homology regions for chromosomal
integration at NS2. Transformants were selected on chloramphenicol
containing plates and only those strains could grow in which the pre-
viously integrated pnuX cassette was replaced by psbA2s_ribDGE_cp-
c560_ribABHT. This replacement was confirmed by PCR analysis. As a
control, a Synechococcus sp. PCC 7002 parent strain containing pnuX
(which had not been exposed to roseoflavin) was transformed with the
same construct. Growth and flavin production of the strains was
compared to the roseoflavin-selected Synechococcus sp. PCC 7002 strain
(Fig. 7). All strains were cultivated for 38 days in 100 mL Erlenmeyer
flasks (culture volume 30 mL). Growth was slightly impaired in the
roseoflavin selected Synechococcus sp. PCC 7002 strains expressing the
heterologous B. subtilis rib genes, when compared to the parent strain
The reduced flavin production can be explained by the fact that the
strains were not fully segregated and that the high riboflavin levels
impose a disadvantage to the corresponding overproducers. Apparently,
about 2.3 μM flavins per OD730 are tolerated by the cells under the
applied conditions.
Some of the riboflavin-overproducing, roseoflavin-resistant Syn-
echococcus sp. PCC 7002 strains as well as the wild-type and the pnuX-
expressing parent strain (which had not been exposed to roseoflavin)
were subjected to whole genome sequencing. The objective was to locate
the mutation(s) responsible for the riboflavin overproducing phenotype
and to identify novel targets for strain improvement in Synechococcus sp.
PCC 7002. All independently isolated roseoflavin-resistant riboflavin-
overproducing strains (but neither the wild-type nor the pnuX-express-
ing parent strain) showed a specific 4-nucleotide deletion at position nt-
14 of SYNPCC7002_A0225 causing a frame-shift in this gene. SYN-
PCC7002_A0225 codes for a short protein (141 aa) with unknown
function. It contains a tetratricopeptide repeat (TPR) motif, which in
some proteins mediates protein-protein interactions. Notably, the 4-
nucleotide deletion was not found in the roseoflavin-resistant strains
#53 and #54 that did not overproduce riboflavin (Fig. 6C). Primer-
based sequencing of the respective regions confirmed these findings
(Supplemental Fig. S6).
(Fig. 7A). This strain yielded highest riboflavin levels (41 μM), which
represents
a
2.2- and 1.3-fold increase compared to the
roseoflavin-selected parent strain and the rib gene expressing strain
based on the wild-type parent, respectively (Fig. 7B and C). The reduced
Fig. 7. A riboflavin overproducing roseoflavin-
resistant Synechococcus sp. PCC 7002 strain (#48;
see Fig. 6B) is used as a host for overexpression ex-
periments. (A) Growth curves of Synechococcus sp.
PCC 7002 wild-type, the recombinant Synechococcus
sp. PCC 7002 strain expressing the flavin transporter
pnuX from Corynebacterium glutamicum (“pnuX”), the
roseoflavin-selected pnuX-expressing strain #48
(“pnuXRoF”), an unselected strain containing the
expression construct psbA2s_ribDGE_cpc560_ribABHT
(see also Fig. 4 A and C) (“ribDGEABHT”), and #48
containing
the
same
expression
construct
(“RoFribDGEABHT”). (B) Riboflavin production is
highest for the roseoflavin-selected strain #48
expressing the B. subtilis rib genes. (C) Levels of
riboflavin, FMN and lumichrome in the supernatants
of the different strains at the end of growth. Ribo-
flavin levels are significantly higher in the RoFribDG-
EABHT strain (derived from strain #48), compared to
the other riboflavin-overproducing strains (p
0.005). RF: riboflavin, LC: lumichrome.
<
282

B. Kachel and M. Mack
Metabolic Engineering 62 (2020) 275–286
growth can be explained by the fact that riboflavin overproduction
consumes metabolic energy. Levels of the degradation product lumi-
chrome did not exceed 3% of total flavins for all
riboflavin-overproducing strains (Fig. 7C).
4.2. Culture conditions for Synechococcus sp. PCC 7002
Synechococcus sp. PCC 7002 was grown in standard AA + medium
(Vogel et al., 2017) containing 4
g/L vitamin B12 at 32 ◦C and 100 rpm.
μ
The cultures were grown at ambient CO2 (ca. 0.03%) and illuminated
with a custom LED-Array which contained 196 narrow-bandwidth LEDs
of 630 nm (5 mm, 20 mA, 2.1 V, Farnell GmbH, Germany) and 48
narrow-bandwidth LEDs of 700 nm (5 mm, 50 mA, 2.0–2.3 V, Roithner
Lasertechnik, Austria) mounted on a perforated scaffold (40 × 40 cm)
and placed approximately 30 cm above the shaking surface of the
3. Conclusions
In this study we describe the construction of a recombinant cyano-
bacterial strain for the photoautotrophic production of riboflavin
whereby the strains were grown on a simple mineral salt medium
without CO2-gassing. To the best of our knowledge, cyanobacteria have
never been engineered with the purpose of overproducing vitamins,
which is probably due to the known light-sensitivity of many of these
fine chemicals (Gazzali et al., 2016; Hemery et al., 2015). We show in
this work that this limitation can be overcome by cultivating
riboflavin-overproducing Synechococcus sp. PCC 7002 strains under
narrow-bandwidth dichromatic LED illumination at λ = 630 and λ =
700 nm. These wavelengths fully supported phototrophic growth of
Synechococcus sp. PCC 7002 as no growth reduction was found when
compared to white-light conditions. At the same time LED illumination
at λ = 630 and λ = 700 nm did not lead to degradation of riboflavin.
Heterotrophic bacteria such as B. subtilis have been engineered with
great success to overproduce riboflavin (Schwechheimer et al., 2016). A
B. subtilis strain containing similar modifications as now introduced in
Synechococcus sp. PCC 7002 was compared to our best performing
Synechococcus sp. PCC 7002 strain. Such a B. subtilis strain containing a
strong promoter upstream of the endogenous rib genes yielded a final
incubator. The light intensity of the LEDs was set to 15
to 10
μE (630 nm) and
μ
E (700 nm). Small scale cultures of Synechococcus sp. PCC 7002
wild-type and mutant strains were grown in 1 mL liquid AA + medium in
standard untreated 24-well plates (VWR, Germany) to an OD730 of ~2.0.
If cultivated in (untreated) 6-well plates (VWR, Germany), cultures were
diluted to an OD730 of 0.02 in 5 mL liquid AA + medium and grown at
◦
32 C and 100 rpm under red-light conditions. To compensate evapo-
ration water was added. The final volume of the cultures was measured
at the end of the growth experiment to normalize the data. For cultures
grown in 100 mL Erlenmeyer flasks, the pre-culture was diluted to OD730
of 0.02 in 35 mL AA + medium (liquid levels were about 0.5–1 cm while
shaking at 100 rpm).
4.3. Determination of flavins by high-performance liquid chromatography
(HPLC)
Flavins levels of culture supernatants were determined as follows.
Culture samples were centrifuged at 14,000 rpm for 10 min and the
supernatants were treated with 10% (v/v) of 50% (w/v) trichloroacetic
acid (TCA). The resulting mixtures were centrifuged at 14,000 rpm for
riboflavin concentration of 8
μ
M (3 mg Lꢀ 1) when grown to the sta-
tionary phase in LB (cultivation time 9 h; productivity 343
μ
g riboflavin
Lꢀ 1
h
ꢀ 1). Our best riboflavin-overproducing Synechococcus sp. PCC 7002
10 min, filtered (0.22 μ
m) and stored into lightproof HPLC vials at 4 ◦C
strain yielded higher stationary phase riboflavin levels of 74
μM (28 mg
L
ꢀ 1), however, the cultivation times were 38 days and thus the pro-
until measurement. To prepare cell-free extracts culture samples were
centrifuged at 2000×g for 5 min and lysed with 200 L of glass beads
(300 m), using the FastPrep device (MP Biomedicals, USA) at 6.5 m/s
ductivity was much lower (31
μ
g Lꢀ 1 hꢀ 1).
μ
B. subtilis riboflavin production strains engineered to express addi-
tional rib genes from strong promoters were shown to yield much higher
flavin levels, especially when cultivated in bioreactors under improved
conditions (Perkins et al., 1999). As expression of the heterologous rib
genes is very low in our cyanobacterial strains (Supplemental Fig. S7) we
assume that the flux through the riboflavin pathway can further be
increased e.g. by using endogenous plasmids with high copy numbers to
strongly enhance rib gene expression (Xu et al., 2011). Consequently, a
Synechococcus sp. PCC 7002 based riboflavin production process may
very well be competitive when costs for sugar, freshwater, energy and
CO2 emissions increase in the future. Overproduction of riboflavin most
likely interferes with the Calvin-Benson-Cycle, as ribulose-5-phosphate
(one of the riboflavin precursors; Fig. 1) is the precursor of ribulose-1,
5-bisphosphate, the substrate of the CO2-fixing enzyme ribulose-1,
5-bisphosphate carboxylase/oxygenase (EC 4.1.1.39). Considering this
possible limitation, future research effort could also be directed at en-
gineering edible cyanobacteria for vitamin production. Species of the
cyanobacterial genera Spirulina, Anabaena and Nostoc are already used
in fish and animal farming (Radhakrishnan et al., 2016; Venkataraman
et al., 1994), in which riboflavin is also used as a feed additive (Deng and
Wilson, 2003; Soliman and Wilson, 1992). Engineered Spirulina, Ana-
baena and Nostoc could serve as a vitamin-enriched source of nutrients
added to animal feed with little or no refinement.
μ
speed for 60 s for three consecutive rounds. The resulting extracts were
also treated with TCA prior to HPLC analysis (see above). Flavins levels
were also determined in cell suspensions which, prior to flavin analysis,
were subjected to FastPrep lysis (see above). HPLC analysis was per-
formed on a 1260 Agilent HPLC system using a Kinetex Biphenyl column
(Phenomenex, Germany) and a flow rate of 0.2 mL/min. The chroma-
tography was carried out at 50 ◦C. Flavins were separated using a
gradient elution program. Starting at 15% (v/v) methanol and 85% (v/
v) buffer A (10 mM formic acid, 10 mM ammonium formate; pH 3.7), the
methanol concentration was gradually increased to 32% (v/v) at t = 6.5
min. At t = 13.5 min the methanol concentration was gradually
increased from 32% (v/v) to 100% (v/v) at t = 20.1 min. Riboflavin,
lumiflavin, flavin mononucleotide (FMN) and flavin adenine dinucleo-
tide (FAD) were detected photometrically at λ = 445 nm, lumichrome
was detected at λ = 350 nm.
4.4. Molecular cloning
A list of all strains, plasmids and oligonucleotides used in this study is
given in the Supplemental Materials and Methods file in addition to
details with regard to the cloning procedures. Plasmids for superfolder
GFP reporter gene expression experiments in Synechococcus sp. PCC
7002 were created based on the high-copy plasmid pIDTsmart (Inte-
grated DNA Technologies, USA). For chromosomal integration of
expression cassettes by homologous recombination, flanking DNA se-
quences (750 bp) being part of “neutral site 2 (NS2) were introduced to
the expression constructs. NS2 is located between SYNPCC7002_A1202
and SYNPCC7002_A1203 (Ruffing et al., 2016). The flanking NS2 DNA
sequences were PCR-amplified from the Synechococcus sp. PCC 7002
genome using Phusion High-Fidelity DNA Polymerase (Thermo Scien-
tific, Germany). The kanamycin resistance cassette and the gene
encoding superfolder GFP (Pedelacq et al., 2006) were PCR-amplified
4. Methods
4.1. Chemicals
Roseoflavin was purchased from Chemos (Regenstauf, Germany). All
other chemicals were from Sigma-Aldrich Chemie GmbH (Munich,
Germany). Lumichrome and Lumiflavin were purchased from Cayman
Chemical (Ann Arbor, USA).
283

B. Kachel and M. Mack
Metabolic Engineering 62 (2020) 275–286
from the pUR-dualTag vector provided by the Wilde group (University
of Freiburg, Germany) and ligated to the NS2 homology fragments. The
regulatory regions/promoters cpc560, psbA2 and psbA2s driving GFP
expression were amplified from chromosomal DNA of Synechocystis sp.
PCC 6803, the regulatory region A2813 was amplified from the Syn-
echococcus sp. PCC 7002 genome, and the theophylline-inducible
pTrc_Theo promoter (Nakahira et al., 2013) was ordered as a synthetic
gene fragment. The plasmids used to create riboflavin-overproducing
Synechococcus sp. PCC 7002 strains are based on the medium-copy
vector pKT25 (Karimova et al., 1998). The gene coding for superfolder
GFP in the pIDTsmart based expression constructs was replaced by the
riboflavin biosynthesis genes ribDGEABHT amplified from B. subtilis
wildtype (strain 168) genomic DNA. For integration of the ribDGEABHT
operon into the roseoflavin-selected Synechococcus sp. PCC 7002 strains,
the chloramphenicol resistance cassette from pUC19_glgA1_cmR (AG
Wilde, University of Freiburg, Germany) was used to replace the kana-
mycin resistance cassette of the pKT25 derived expression construct. To
generate a roseoflavin-sensitive Synechococcus sp. PCC 7002 strain, the
pnuX gene (CYL77_00365) from Corynebacterium glutamicum ATCC
13032 was PCR-amplified and fused to psbA2L, a kanamycin resistance
cassette and the homology fragments for integration into NS2. Plasmids
for complementation experiments in E. coli were created by ligating the
genes SYNPCC7002_A0427 (ribAB), SYNPCC7002_A1465 (ribDG), SYN-
PCC7002_A0498 (ribD-C-terminal domain) and SYNPCC7002_A2264
(ribE) to the expression plasmid pGP380. SYNPCC7002_A0136 (ribH)
was cloned employing the StarGate® Acceptor Vector pPSG-IBA3 (IBA
GmbH, Germany). Two versions of this expression plasmid were con-
structed, either yielding a gene product with a C-terminal or an N-ter-
minal strep-tag.
4.7. Transformation of Synechococcus sp. PCC 7002
A liquid culture (1.5 mL) of Synechococcus sp. PCC 7002 was culti-
vated in AA + medium to an OD730 of ~1.0 under red-light conditions. A
total of 300 ng linearized plasmid was added to the culture which sub-
sequently was grown for 16 h under red-light without shaking. Cultures
were centrifuged at 2000×g for 5 min. The pellet was suspended in 200
μ
L AA+ and plated on AA + agar plates containing 100 μg/mL kana-
mycin. Plates were incubated under red light conditions until colonies
appeared. To initiate segregation, colonies were streaked on AA + agar
plates containing up to 300 μg/mL kanamycin. Integration and segre-
gation was confirmed using colony PCR. Genetically engineered strains
were cultivated in liquid cultures containing 100 μg/mL kanamycin to
an OD730 > 1.0 and stored in 5% DMSO at - 80 ◦C.
4.8. GFP fluorescence analysis
Culture samples were diluted in AA + medium to an OD730 of 0.4 and
100 μL was transferred to a black 96-well plate (Greiner Bio-One, Ger-
many). GFP fluorescence was measured using a GeniosPro plate reader
(Tecan Group Ltd., Switzerland) at λ = 485 nm excitation and λ = 535
nm emission.
4.9. Chlorophyll-a extraction
Chlorophyll-a extraction was performed as described previously
using ice-cold 100% MeOH (Ritchie, 2006). Absorbance was measured
at λ = 664 nm and chlorophyll-a concentration was calculated using the
coefficient Eλ = 12.9447.
4.10. RT-PCR
4.5. Complementation of riboflavin-auxotrophic strains
Total RNA was extracted from cultures grown to an OD730 of ~1.5
using the Quick RNA MiniPrep Plus Kit (Zymo Research Europe, Ger-
many). cDNA was synthesized from 100 ng total RNA using the Max-
ima™ H Minus cDNA Synthesis Master Mix (Thermo Scientific,
Germany) and random hexamer primers. Prior to cDNA synthesis the
Chemically competent cells of the riboflavin-auxotrophic E. coli
strains BSV11 (ΔribB), BSV13 (ΔribE), BSV18 (ΔribA) (Bandrin et al.,
1979) and SI#78 (ΔribDG) (Richter et al., 1997) were created using the
CaCl2 method (Seidman and Struhl, 2001). Competent cells were
transformed with the expression plasmids containing riboflavin
biosynthesis genes from either Synechococcus sp. PCC 7002 or B. subtilis.
Transformation mixtures were incubated overnight on LB plates con-
total RNA was treated with DNAse. 0.5 μL of the resulting cDNA was
used as a template for PCR in a 25 μL reaction using the Dream Taq DNA
polymerase (Thermo Scientific, Germany) and employing gene specific
primers.
taining 200
grown overnight at 37 ◦C, 220 rpm, in LB containing 100
ampicillin and 200 M riboflavin. The next day, cultures were used to
μ
M riboflavin and 100
μ
g/mL ampicillin. Cultures were
μ
g/mL
μ
4.11. Identification and quantitation of B. subtilis riboflavin biosynthesis
enzymes in cellular lysates using mass spectrometry
inoculate 2 mL LB either with or without 200 μM riboflavin. Cultures
were grown for 16 h at 37 ◦C, 220 rpm before measurement of the op-
tical density at λ = 600 nm.
Cells were lysed in PBS containing cOmplete™ protease inhibitor
(Roche, Germany) using a FastPrep device (MP Biomedicals, USA) at 6.5
m/s speed for 60 s for three consecutive rounds. Following centrifuga-
tion at 20,000×g (10 min) the supernatants of the culture lysates con-
4.6. Assay for the 6,7-dimethyl-8-ribityllumazine synthase RibH
A previously published assay to elucidate the role of different
phosphatases in riboflavin biosynthesis (Haase et al., 2013; Sarge et al.,
2015) was modified to test the 6,7-dimethyl-8-ribityllumazine synthase
taining 10 μg of total protein were denatured and proteins were
separated by SDS-PAGE. Bands were cut from the Coomassie R250-
stained gel, trypsinated and analyzed with an Ultimate 3000 liquid
chromatography system coupled to an Orbitrap mass spectrometer
(Thermo-Fischer, Bremen, Germany) as described in the Supplemental
Materials and Methods.
(RibH) from Synechococcus sp. PCC 7002. The assay contained 200
μ
M 3,
5-amino-6--
ribitylamino-2,4(1H,3H)-pyrimidinedion-5′-phosphate as substrates, as
well as 0.5 M of the purified phosphatase YcsE (EC:3.1.3.104) from
B. subtilis in 50 mM Tris-HCl (pH 8.0), 30 mM KCl, 5 mM MgCl2 and 1
mM DTT. Purified Synechococcus sp. PCC 7002 RibH (5 M) was added
to the assay mixture. As a positive control 5 M of the purified B. subtilis
RibH enzyme was used. All components were mixed in a total volume of
50 L EDTA (20 mM) was added and
L, incubated at 23 ◦C for 10 min 2
the reaction was continued at 37 ◦C for 20 min. The reaction was
4-dihydroxy-2-butanone-4-phosphate and 100
μM
μ
Importance
μ
μ
The light-sensitive vitamin riboflavin can be overproduced employ-
ing a photoautotrophic host.
μ
μ
Funding
stopped injecting 5.2 μL TCA (50% v/v) and the reaction product 6,
7-dimethyl-8-ribityllumazine was determined by HPLC as described
above. 6,7-dimethyl-8-ribityllumazine levels were determined using a
fluorescence detector (excitation λ = 410 nm; emission λ = 490 nm).
This work was funded by the state of Baden-Württemberg, Germany
“
¨
( Leistungsorientierte Forderung des akademischen Mittelbaus an
Hochschulen in Baden-Württemberg”).
284

B. Kachel and M. Mack
Metabolic Engineering 62 (2020) 275–286
Gazzali, A.M., Lobry, M., Colombeau, L., Acherar, S., Azais, H., Mordon, S., Arnoux, P.,
Baros, F., Vanderesse, R., Frochot, C., 2016. Stability of folic acid under several
parameters. Eur. J. Pharmaceut. Sci. 93, 419–430.
Author contributions
B. Kachel: Conceptualization; experimental design; data acquisition;
data analysis, visualization and interpretation; writing of the original
manuscript.
Gutierrez-Preciado, A., Torres, A.G., Merino, E., Bonomi, H.R., Goldbaum, F.A., Garcia-
Angulo, V.A., 2015. Extensive identification of bacterial riboflavin transporters and
their distribution across bacterial species. PLoS One 10, e0126124.
Haase, I., Sarge, S., Illarionov, B., Laudert, D., Hohmann, H.P., Bacher, A., Fischer, M.,
2013. Enzymes from the haloacid dehalogenase (HAD) superfamily catalyse the
elusive dephosphorylation step of riboflavin biosynthesis. Chembiochem 14,
2272–2275.
M. Mack: Conceptualization; supervision; review and editing of the
manuscript.
Hasnain, G., Frelin, O., Roje, S., Ellens, K.W., Ali, K., Guan, J.C., Garrett, T.J., de Crecy-
Lagard, V., Gregory 3rd, J.F., McCarty, D.R., Hanson, A.D., 2013. Identification and
characterization of the missing pyrimidine reductase in the plant riboflavin
biosynthesis pathway. Plant Physiol. 161, 48–56.
Declaration of competing interest
The authors declare that there are no competing interests.
Acknowledgements
Hemberger, S., Pedrolli, D.B., Stolz, J., Vogl, C., Lehmann, M., Mack, M., 2011. RibM
from Streptomyces davawensis is a riboflavin/roseoflavin transporter and may be
useful for the optimization of riboflavin production strains. BMC Biotechnol. 11,
119.
Hemery, Y.M., Fontan, L., Moench-Pfanner, R., Laillou, A., Berger, J., Renaud, C.,
Avallone, S., 2015. Influence of light exposure and oxidative status on the stability of
vitamins A and D(3) during the storage of fortified soybean oil. Food Chem. 184,
90–98.
Mass Spectrometry analysis for protein identification and quantita-
tion, as well as data analysis was performed at the Core Facility for Mass
Spectrometry and Proteomics at ZMBH, Heidelberg by Dr. Renata
Blatnik. Substrates and enzymes to carry out the RibH assay were kindly
provided by the group of Markus Fischer, Hamburg School of Food
Science, Institute of Food Chemistry, University of Hamburg, Germany.
Hirokawa, Y., Goto, R., Umetani, Y., Hanai, T., 2017. Construction of a novel d-lactate
producing pathway from dihydroxyacetone phosphate of the Calvin cycle in
cyanobacterium, Synechococcus elongatus PCC 7942. J. Biosci. Bioeng. 124, 54–61.
Hohmann, H.P., Stahmann, K.P., 2010. Biotechnology of riboflavin production.
Comprehensive Natural Products II Chemistry and Biology 7, 115–139.
Jankowitsch, F., Schwarz, J., Ruckert, C., Gust, B., Szczepanowski, R., Blom, J.,
Pelzer, S., Kalinowski, J., Mack, M., 2012. Genome sequence of the bacterium
Streptomyces davawensis JCM 4913 and heterologous production of the unique
antibiotic roseoflavin. J. Bacteriol. 194, 6818–6827.
Appendix A. Supplementary data
Supplementary data to this article can be found online at https://doi.
org/10.1016/j.ymben.2020.09.010.
Karimova, G., Pidoux, J., Ullmann, A., Ladant, D., 1998. A bacterial two-hybrid system
based on a reconstituted signal transduction pathway. Proc. Natl. Acad. Sci. U. S. A.
95, 5752–5756.
Kissling, L., Schneider, C., Seibel, K., Dorjjugder, N., Busche, T., Kalinowski, J., Mack, M.,
2020. The roseoflavin producer Streptomyces davaonensis has a high catalytic
capacity and specific genetic adaptations with regard to the biosynthesis of
riboflavin. Environ. Microbiol. 22, 3248–3265.
References
Acevedo-Rocha, C.G., Gronenberg, L.S., Mack, M., Commichau, F.M., Genee, H.J., 2019.
Microbial cell factories for the sustainable manufacturing of B vitamins. Curr. Opin.
Biotechnol. 56, 18–29.
Kopka, J., Schmidt, S., Dethloff, F., Pade, N., Berendt, S., Schottkowski, M., Martin, N.,
Duhring, U., Kuchmina, E., Enke, H., Kramer, D., Wilde, A., Hagemann, M.,
Friedrich, A., 2017. Systems analysis of ethanol production in the genetically
engineered cyanobacterium Synechococcus sp. PCC 7002. Biotechnol. Biofuels 10, 56.
Korosh, T.C., Markley, A.L., Clark, R.L., McGinley, L.L., McMahon, K.D., Pfleger, B.F.,
2017. Engineering photosynthetic production of L-lysine. Metab. Eng. 44, 273–283.
Liu, D., Pakrasi, H.B., 2018. Exploring native genetic elements as plug-in tools for
synthetic biology in the cyanobacterium Synechocystis sp. PCC 6803. Microb. Cell
Factories 17, 48.
Bandrin, S.V., Beburov, M., Rabinovich, P.M., Stepanov, A.I., 1979. [Riboflavin
auxotrophs of Escherichia coli]. Genetika 15, 2063–2065.
Bernstein, H.C., Konopka, A., Melnicki, M.R., Hill, E.A., Kucek, L.A., Zhang, S., Shen, G.,
Bryant, D.A., Beliaev, A.S., 2014. Effect of mono- and dichromatic light quality on
growth rates and photosynthetic performance of Synechococcus sp. PCC 7002. Front.
Microbiol. 5, 488.
Birkenmeier, M., Neumann, S., Roder, T., 2014. Kinetic modeling of riboflavin
biosynthesis in Bacillus subtilis under production conditions. Biotechnol. Lett. 36,
919–928.
Ludwig, M., Bryant, D.A., 2011. Transcription profiling of the model cyanobacterium
Synechococcus sp. strain PCC 7002 by next-gen (SOLiD) sequencing of cDNA. Front.
Microbiol. 2, 41.
Bland, E., Angenent, L.T., 2016. Pigment-targeted light wavelength and intensity
promotes efficient photoautotrophic growth of Cyanobacteria. Bioresour. Technol.
216, 579–586.
Ludwig, M., Bryant, D.A., 2012. Synechococcus sp. strain PCC 7002 transcriptome:
acclimation to temperature, salinity, oxidative stress, and mixotrophic growth
conditions. Front. Microbiol. 3, 354.
Chaves, J.E., Melis, A., 2018. Engineering isoprene synthesis in cyanobacteria. FEBS Lett.
592, 2059–2069.
Matsui, K., Wang, H.-C., Hirota, T., Matsukawa, H., Kasai, S., Shinagawa, K., Otani, S.,
1982. Riboflavin production by roseoflavin-resistant strains of some bacteria. Agric.
Biol. Chem. 46, 2003–2008.
Choi, S.Y., Lee, H.J., Choi, J., Kim, J., Sim, S.J., Um, Y., Kim, Y., Lee, T.S., Keasling, J.D.,
Woo, H.M., 2016. Photosynthetic conversion of CO2 to farnesyl diphosphate-derived
phytochemicals (amorpha-4,11-diene and squalene) by engineered cyanobacteria.
Biotechnol. Biofuels 9, 202.
Melnicki, M.R., Pinchuk, G.E., Hill, E.A., Kucek, L.A., Stolyar, S.M., Fredrickson, J.K.,
Konopka, A.E., Beliaev, A.S., 2013. Feedback-controlled LED photobioreactor for
photophysiological studies of cyanobacteria. Bioresour. Technol. 134, 127–133.
Miao, R., Xie, H., F, M.H., Lindblad, P., 2018a. Protein engineering of alpha-
ketoisovalerate decarboxylase for improved isobutanol production in Synechocystis
PCC 6803. Metab. Eng. 47, 42–48.
Conley, P.B., Lemaux, P.G., Grossman, A.R., 1985. Cyanobacterial light-harvesting
complex subunits encoded in two red light-induced transcripts. Science 230,
550–553.
Davies, F.K., Work, V.H., Beliaev, A.S., Posewitz, M.C., 2014. Engineering limonene and
bisabolene production in wild type and a glycogen-deficient mutant of
Synechococcus sp. PCC 7002. Front Bioeng Biotechnol 2, 21.
Deng, D.-F., Wilson, R.P., 2003. Dietary riboflavin requirement of juvenile sunshine bass
(Morone chrysops ♀×Morone saxatilis ♂). Aquaculture 218, 695–701.
Dismukes, G.C., Carrieri, D., Bennette, N., Ananyev, G.M., Posewitz, M.C., 2008. Aquatic
phototrophs: efficient alternatives to land-based crops for biofuels. Curr. Opin.
Biotechnol. 19, 235–240.
Miao, R., Xie, H., Lindblad, P., 2018b. Enhancement of photosynthetic isobutanol
production in engineered cells of Synechocystis PCC 6803. Biotechnol. Biofuels 11,
267.
Mona, S., Kumar, S.S., Kumar, V., Parveen, K., Saini, N., Deepak, B., Pugazhendhi, A.,
2020. Green technology for sustainable biohydrogen production (waste to energy): a
review. Sci. Total Environ. 728, 138481.
Nakahira, Y., Ogawa, A., Asano, H., Oyama, T., Tozawa, Y., 2013. Theophylline-
dependent riboswitch as a novel genetic tool for strict regulation of protein
expression in Cyanobacterium Synechococcus elongatus PCC 7942. Plant Cell Physiol.
54, 1724–1735.
Ducat, D.C., Avelar-Rivas, J.A., Way, J.C., Silver, P.A., 2012. Rerouting carbon flux to
enhance photosynthetic productivity. Appl. Environ. Microbiol. 78, 2660–2668.
Ducat, D.C., Way, J.C., Silver, P.A., 2011. Engineering cyanobacteria to generate high-
value products. Trends Biotechnol. 29, 95–103.
Ohbayashi, R., Akai, H., Yoshikawa, H., Hess, W.R., Watanabe, S., 2016. A tightly
inducible riboswitch system in Synechocystis sp. PCC 6803. J. Gen. Appl. Microbiol.
62, 154–159.
Englund, E., Liang, F., Lindberg, P., 2016. Evaluation of promoters and ribosome binding
sites for biotechnological applications in the unicellular cyanobacterium
Synechocystis sp. PCC 6803. Sci. Rep. 6, 36640.
Otani, S., Takatsu, M., Nakano, M., Kasai, S., Miura, R., 1974. Letter: roseoflavin, a new
antimicrobial pigment from Streptomyces. J. Antibiot. (Tokyo) 27, 88–89.
Pedelacq, J.D., Cabantous, S., Tran, T., Terwilliger, T.C., Waldo, G.S., 2006. Engineering
and characterization of a superfolder green fluorescent protein. Nat. Biotechnol. 24,
79–88.
Eriksson, J., Salih, G.F., Ghebramedhin, H., Jansson, C., 2000. Deletion mutagenesis of
the 5’ psbA2 region in Synechocystis 6803: identification of a putative cis element
involved in photoregulation. Mol. Cell Biol. Res. Commun. 3, 292–298.
Fischer, M., Bacher, A., 2005. Biosynthesis of flavocoenzymes. Nat. Prod. Rep. 22,
324–350.
Perkins, J.B., Sloma, A., Hermann, T., Theriault, K., Zachgo, E., Erdenberger, T.,
Hannett, N., Chatterjee, N.P., Williams Ii, V., Jr, G.A.R., Hatch, R., Pero, J., 1999.
Genetic engineering of Bacillus subtilis for the commercial production of riboflavin.
J. Ind. Microbiol. Biotechnol. 22, 8–18.
Gangl, D., Zedler, J.A., Rajakumar, P.D., Martinez, E.M., Riseley, A., Wlodarczyk, A.,
Purton, S., Sakuragi, Y., Howe, C.J., Jensen, P.E., Robinson, C., 2015.
Biotechnological exploitation of microalgae. J. Exp. Bot. 66, 6975–6990.
Radhakrishnan, S., Belal, I.E.H., Seenivasan, C., Muralisankar, T., Bhavan, P.S., 2016.
Impact of fishmeal replacement with Arthrospira platensis on growth performance,
285

B. Kachel and M. Mack
Metabolic Engineering 62 (2020) 275–286
body composition and digestive enzyme activities of the freshwater prawn.
Macrobrachium rosenbergii. Aquaculture Reports 3, 35–44.
Varman, A.M., Yu, Y., You, L., Tang, Y.J., 2013. Photoautotrophic production of D-lactic
acid in an engineered cyanobacterium. Microb. Cell Factories 12, 117.
Venkataraman, L.V., Somasekaran, T., Becker, E.W., 1994. Replacement value of blue-
green alga (Spirulina platensis) for fishmeal and a vitamin-mineral premix for broiler
chicks. Br. Poultry Sci. 35, 373–381.
Richter, G., Fischer, M., Krieger, C., Eberhardt, S., Luttgen, H., Gerstenschlager, I.,
Bacher, A., 1997. Biosynthesis of riboflavin: characterization of the bifunctional
deaminase-reductase of Escherichia coli and Bacillus subtilis. J. Bacteriol. 179,
2022–2028.
Vioque, A., 2007. Transformation of cyanobacteria. Adv. Exp. Med. Biol. 616, 12–22.
Vogel, A.I.M., Lale, R., Hohmann-Marriott, M.F., 2017. Streamlining recombination-
mediated genetic engineering by validating three neutral integration sites in
Synechococcus sp. PCC 7002. J. Biol. Eng. 11, 19.
Ritchie, R.J., 2006. Consistent sets of spectrophotometric chlorophyll equations for
acetone, methanol and ethanol solvents. Photosynth. Res. 89, 27–41.
Ruffing, A.M., Jensen, T.J., Strickland, L.M., 2016. Genetic tools for advancement of
Synechococcus sp. PCC 7002 as a cyanobacterial chassis. Microb. Cell Factories 15.
Santos-Merino, M., Singh, A.K., Ducat, D.C., 2019. New applications of synthetic biology
tools for cyanobacterial metabolic engineering. Front Bioeng Biotechnol 7, 33.
Sarge, S., Haase, I., Illarionov, B., Laudert, D., Hohmann, H.P., Bacher, A., Fischer, M.,
2015. Catalysis of an essential step in vitamin B2 biosynthesis by a consortium of
broad spectrum hydrolases. Chembiochem 16, 2466–2469.
Weiss, T.L., Young, E.J., Ducat, D.C., 2017. A synthetic, light-driven consortium of
cyanobacteria and heterotrophic bacteria enables stable polyhydroxybutyrate
production. Metab. Eng. 44, 236–245.
Wijffels, R.H., Kruse, O., Hellingwerf, K.J., 2013. Potential of industrial biotechnology
with cyanobacteria and eukaryotic microalgae. Curr. Opin. Biotechnol. 24, 405–413.
Xu, Y., Alvey, R.M., Byrne, P.O., Graham, J.E., Shen, G., Bryant, D.A., 2011. Expression
of genes in cyanobacteria: adaptation of endogenous plasmids as platforms for high-
level gene expression in Synechococcus sp. PCC 7002. Methods Mol. Biol. 684,
273–293.
Schwechheimer, S.K., Park, E.Y., Revuelta, J.L., Becker, J., Wittmann, C., 2016.
Biotechnology of riboflavin. Appl. Microbiol. Biotechnol. 100, 2107–2119.
Sebesta, J., Peebles, C.A., 2020. Improving heterologous protein expression in
Synechocystis sp. PCC 6803 for alpha-bisabolene production. Metab Eng Commun 10,
e00117.
Zess, E.K., Begemann, M.B., Pfleger, B.F., 2016. Construction of new synthetic biology
tools for the control of gene expression in the cyanobacterium Synechococcus sp.
strain PCC 7002. Biotechnol. Bioeng. 113, 424–432.
Seidman, C.E., Struhl, K., 2001. Introduction of plasmid DNA into cells. Curr Protoc
Protein Sci. Appendix 4, 4D.
Zhou, J., Zhang, H., Meng, H., Zhu, Y., Bao, G., Zhang, Y., Li, Y., Ma, Y., 2014. Discovery
of a super-strong promoter enables efficient production of heterologous proteins in
cyanobacteria. Sci. Rep. 4, 4500.
Sheraz, M.A., Kazi, S.H., Ahmed, S., Anwar, Z., Ahmad, I., 2014. Photo, thermal and
chemical degradation of riboflavin. Beilstein J. Org. Chem. 10, 1999–2012.
Singh, A.K., Mallick, N., 2017. Advances in cyanobacterial polyhydroxyalkanoates
production. FEMS Microbiol. Lett. 364.
Zhou, Z., Dang, Y., Zhou, M., Li, L., Yu, C.H., Fu, J., Chen, S., Liu, Y., 2016. Codon usage
is an important determinant of gene expression levels largely through its effects on
transcription. Proc. Natl. Acad. Sci. U. S. A. 113, E6117–E6125.
Soliman, A.K., Wilson, R.P., 1992. Water-soluble vitamin requirements of tilapia. 2.
Riboflavin requirement of blue tilapia. Oreochromis aureus. Aquaculture 104,
309–314.
286