A Chloroplast-Localised Fluorescent Protein Enhances the Photosynthetic Action Spectrum in Green Algae

Green microalgae are important sources of natural products and are attractive cell factories for manufacturing high-value products such as recombinant proteins. Increasing scales of production must address the bottleneck of providing sufficient light energy for photosynthesis. Enhancing the photosynthetic action spectrum of green algae to improve the utilisation of yellow light would provide additional light energy for photosynthesis. Here, we evaluated the Katushka fluorescent protein, which converts yellow photons to red photons, to drive photosynthesis and growth when expressed in Chlamydomonas reinhardtii chloroplasts. Transplastomic algae expressing a codon-optimised Katushka gene accumulated the active Katushka protein, which was detected by excitation with yellow light. Removal of chlorophyll from cells, which captures red photons, led to increased Katushka fluorescence. In yellow light, emission of red photons by fluorescent Katushka increased oxygen evolution and photosynthetic growth. Utilisation of yellow photons increased photosynthetic growth of transplastomic cells expressing Katushka in light deficient in red photons. These results showed that Katushka was a simple and effective yellow light-capturing device that enhanced the photosynthetic action spectrum of C. reinhardtii.


Introduction
Microalgae have evolved over billions of years into a successful and widespread group of photosynthetic eukaryotes [1]. They are sources of natural products, including carotenoids and long-chain polyunsaturated acids, which are used as food additives and nutraceuticals [2,3]. Microalgae have attracted considerable interest as sources of biofuels [4][5][6] and sustainable lean-carbon production platforms for expressing vaccines and biopharmaceutical proteins [7,8]. Maximising biomass is key to the production of all high-value products expressed in microalgae [9]. At high cell densities and in deep water, biomass accumulation is limited by light availability [10]. Synthetic biology provides a route towards enhancing light-capture systems in microalgae to boost growth in limiting light [9]. Chlamydomonas reinhardtii provides a suitable algal model for synthetic biology projects by allowing the design of algal cells with new functionalities, for example nitrogen fixation, and pharmaceutical protein and vaccine production [7,8,[11][12][13].
Light penetration into dense algal cultures can be improved by reducing light-harvesting antennae sizes, which decreases chlorophyll content [14][15][16]. This improves algal biomass production at high light intensities but is a disadvantage in low light due to impaired light harvesting by short antennae [17]. In green algae and plants, photosynthetically active radiation extends from 400 nm to 700 nm. This provides opportunities to expand the expand the photosynthetic action spectrum by developing new light-harvesting mechanisms to capture photons with wavelengths below 400 nm (UV) and above 700 nm (far red) [16,18,19]. Sunlight at the Earth's surface contains more far-red photons than UV photons [19,20], providing an advantage for engineering far-red photon (701 nm to 750 nm) collection systems in terrestrial plants [15,[20][21][22]. Opportunities to enhance the photosynthetic action spectrum are not limited to far-red photons, which have been the focus of a number of studies [21,22]. Yellow photons (~570 nm to ~590 nm) are abundant in the photon flux spectrum of solar radiation at the Earth's surface [20] but correspond to a trough that lies between ~550 nm to ~600 nm in the absorption spectrum of green algae [23]. Importantly, yellow photons penetrate water to greater depths than far-red photons [15,24]. This makes the capture of yellow photons more attractive than harvesting far-red photons for expanding the photosynthetic action spectrum in aquatic green algae.
Here, we designed a yellow photon-capturing device based on a fluorescent protein to drive photosynthetic growth. The device was built and tested in C. reinhardtii. It was based on the observation that corals express fluorescent proteins capable of converting blue and green photons, which predominate in deep water, to longer wavelength photons which are scarce at these depths [25,26]. This light conversion to longer wavelengths may have a role in increasing photosynthesis in the symbiotic zooxanthellae present in corals [25,26], although other functions for fluorescence have been proposed [27,28].
We searched the palette of fluorescent proteins [29] for a suitable candidate capable of converting yellow photons to red photons. The far-red fluorescent protein Katushka, also known as TurboFP 635, is a 231 amino acid, 26 kDa protein that absorbs yellow photons with an excitation peak at 588 nm and emits red light with an emission peak at 635 nm [30]. These 588 nm excitation and 635 nm emission peaks overlap with the ~550 nm to ~ 600 nm trough and red peaks in the absorption spectrum of C. reinhardtii cells [23], respectively (illustrated by Figure 1). The high quantum yield of 0.34, brightness of 22.1, photostability and pH tolerance of Katushka [30] make it preferable to similar fluorescent proteins such as mPlum with 590 nm excitation and 649 nm emission peaks [31]. Crucially, Katushka fluorescence exhibits an eight-fold increase in brightness between 650 nm to 800 nm compared to mPlum [30], thereby providing more energy for chlorophyll absorption and photosynthesis. Expression in chloroplasts rather than the cytosol ensured that Katushka was located in relatively close proximity to the light-harvesting complexes in the thylakoid membranes of chloroplasts. This increases the efficiency of capture of red photons by chlorophyll. The use of Katushka as a light-capturing device for enhancing photosynthesis was validated by the observation that transplastomic algae expressing Katushka showed improved photosynthetic growth in yellow light compared to control strains.  [30] relative to the absorption spectrum (green) of C. reinhardtii cells (not to scale). The absorption spectrum of cells was normalised to the 680 nm peak.

Nucleic Acid Manipulations
Standard molecular biology and recombinant cloning procedures were followed [35]. Unless otherwise indicated, all PCR reactions were carried out in a 20 µL final volume using 0.6 U of Vent DNA polymerase (New England Biolabs, Hitchin, UK), 200 µM final concentration of deoxynucleotides (Bioline, London, UK), and 500 nM final concentration of primers (Sigma-Aldrich, Poole, UK). Primer sequences are listed in Supplementary  Table S1. Restriction enzyme digests used 1 µg of DNA under manufacturer's conditions (New England Biolabs, Hitchin, UK, or Roche, Welwyn Garden City, UK) except for Kpn I partial digestions, where 5 µg of DNA was used. Total DNA from Chlamydomonas was prepared according to Day et al. (1988) [36]. Sequences were visualised in silico using Vector NTI Advance v 11.5 (Invitrogen, Carlsbad, CA, USA) and Geneious Prime (Biomatters, Auckland, New Zealand). The wild-type (Acc no. BK0054) and CC-373 (MF083692) chloroplast genomes were aligned using MAFFT [37] in Geneious Prime using the default parameters to locate the deletions.

Synthesis of the CpKat Gene
The 231 amino acid Katushka fluorescent protein was back-translated in silico using vector NTI (Invitrogen, Carlsbad, CA, USA) with frequent codons in C. reinhardtii chloroplast genes [38]. A glycine GGT codon was inserted after the N-terminal methionine to create an Nco I site (Supplementary Figure S1). The resulting coding sequence, named cpKAT, was assembled by stepwise overlap PCR using thirty-two oligonucleotides of 34 to 38 nucleotides in length (oligonucleotides 783 to 815). Ten complementary nucleotides allowed assembly of dimers; 17 base overlaps between adjacent dimers enabled synthesis of 15 tetramers, which were mixed to assemble the complete gene. For dimers, reaction mixes were prepared to contain 90 mM final concentration of each of the two oligonucleotides and 1.25 mM final concentration of each deoxyribonucleotide in restriction enzyme buffer A (Roche Life Sciences, Burgess Hill, UK). Reaction mixes were denatured for two minutes at 96 • C and then cooled down to 37 • C. Next, 0.37 U of Klenow enzyme (Roche, Lewes, UK) was added to each mixture and incubated for 1 h at 37 • C and 16 h at 15 • C. Fifteen 100 to 103 bp long tetramers were then assembled from adjacent dimers by eight cycles of PCR using a 34 • C annealing step and Vent DNA polymerase (New England Biolabs). Finally, the full-length CpKat gene was assembled by 20 cycles of PCR amplification with BiomixRed master mix (Bioline) by combining all fifteen 100 to 103 bp long tetramers and increasing the concentration of external primers 782 and 815. The complete 699 bp CpKAT gene with TAA stop codon was purified by agarose gel electrophoreses and cloned into E. coli vector pGEM-T easy (Promega, Southampton, UK).

Construction of the pB10 Chloroplast Transformation Vector
C. reinhardtii chloroplast Bam10 fragment (7.6 kb) in pBR313 [39] from the Chlamydomonas Resource Center was used as a template to amplify the 2.5 kbp right targeting arm with primers 862 and 863, which was then digested with Xba I to Kpn I and cloned into pBluescript [40] to produce pBluescript-Bam10-Right. For the left targeting arm, a 3.5 kbp Bam HI to Pst I fragment of Bam10 was cloned into pUC18 [41]. The cloned left arm was digested with Bam HI and partially digested with Kpn I to obtain a 2.1 kbp fragment which was subcloned into pBluescript using linkers comprising annealed oligonucleotides 1123 and 1124, and annealed oligonucleotides 1125 and 1126. The left arm was excised with Sac I and Xba I and inserted into pBluescript-Bam10-Right to produce chloroplast transformation vector pB10 containing both targeting arms separated by a polylinker with unique Not I, Xba I, Apa I, and EcoRI sites for inserting foreign genes. A 297 bp rbcL 3 UTR region amplified from pUC-atpX-aadA [42] with primers 864 and 865 was cloned into pGEM-T easy. pB10-AKL was produced by cloning CpKat as an NcoI to PstI fragment into pUC-atpX to produce atpX-CpKat. The atpA promoter and 5 UTR and CpKat were excised from atpX-CpKat using EcoRI to PstI, mixed with the~300 bp rbcL 3 UTR excised with PstI and NotI, and ligated into EcoRIand Not I-restricted pB10. The psbD promoter and 5 UTR was isolated by PCR from CC-503 total DNA using primers 5199a and 5199b, cloned into pUC18, then excised as a HincII to NcoI fragment to replace the atpA 5 region excised with Eco RI and Nco I in pB10-AKL to form pB10-DKL. Hinc II and Eco RI ends were made blunt using Klenow enzyme and dNTPs.

Chloroplast Transformation
Chloroplast transformations were performed as described previously [43] using the Biolistic PDS-1000/He Particle Delivery System (Bio-Rad, Hemel Hempstead, UK) with 1 µM gold particles and conditions: 1100 psi rupture disks, 28 mm Hg vacuum, and 6 cm distance from the rupture disk to the lawn of cells on 9 cm Petri dishes containing minimal media. The pB10 constructs were transformed into strain CC-373 [43]. Bombarded plates were incubated at 25 • C with a light intensity of 200 µmol m −2 s −1 using cool white fluorescent lamps. Individual transformant colonies were picked after 2 weeks and serially propagated on minimal medium for six months.

PCR Analyses of Microalgae
DNA for PCR analyses was obtained as described by Cao et al. (2009) [44]. CpKat integration was verified by PCR using primers 784 and 1420. PCR products were sizefractionated by electrophoresis on tris-borate-EDTA 1% w/v agarose gels at 5 V cm −1 .

Protein Blot Analyses
Procedures for extracting proteins from algal cells, SDS-PAGE, transfer of size-fractionated proteins to blots, and visualisation of specific antibody-bound bands are described in Suarez et al. (2014) [45]. The anti-tRFP rabbit polyclonal antibody specific for Katushka was from Cambridge Bioscience (Cambridge, UK).

Measurement of Dissolved Oxygen
Dissolved oxygen was calculated using the modified Winkler method [46] according to Labasque et al. (2004) [47], measuring absorbance at 466 nm in a Synergy HT multimode plate reader (Biotek Instruments, Winooski, VT, USA). Transplastomic cell cultures were adjusted to 1 × 10 6 cells mL −1 density in minimal media during the dark-growth period in 100 mL shaking flasks. A total of 5 mL of culture were transferred to each of three 15 mL conical tubes and placed in 25 • C illuminated growth chambers with either 30 µmol m −2 s −1 cool white L36w/640 fluorescent lamps (Osram, Munich, Germany) or 30 µmol m −2 s −1 yellow (λ max 590) LED lamps (Ecolight Ltd., Chichester, UK). Cells were left for 6 h and then cultures were fixed by adding 60 µL of 2% w/v sodium azide. A total of 2 mL of fixed cultures were transferred to 6-well microtiter plates for absorbance readings. Experiments were performed in triplicate. Total chlorophyll was measured by spectrophotometric measurements at 647 nm and 664 nm, according to Ritchie (2006) [48]. All experiments were carried out in the dark, and solvents and extracts were kept at 4 • C until measurements were carried out in a Synergy HT multi-mode plate reader.

Algal Growth Conditions
For each construct, three transplastomic lines from colonies derived from independent transformation events were used. Each line was cultured in triplicate providing nine cultures (three biological replicates × three technical replicates). Transplastomic cells were inoculated to an initial optical density of 0.1 at 750 nm (Jenway 7315 spectrophotometer, Fisher Scientific, Loughborough, UK) in 6 mL minimal media in 14 mL Falcon culture tubes (Fisher Scientific) and incubated at 25 • C in continuous light. Growth of the non-transgenic C. reinhardtii strain CC-503 was in 100 mL of minimal medium. Cells were grown for two weeks and 200 µL samples were taken for OD 750 nm readings using a Synergy HT microplate reader (Biotek UK, Potton). Lamps used were cool white L36W/640 fluorescent lamps (Osram); blue (λ max 460), yellow (λ max 590) and red (λ max 660) LED lights (Ecolight Ltd., Chichester, UK). Light intensity received by each culture tube was measured using a SKP 200 photometer (Skye Instruments, Llandrindod Wells, UK).

Fluorescence and Spectral Analysis
Fluorescence measurements were performed in a Synergy HT microplate reader using 200 µL samples with a 590 nm/20 nm excitation filter and a 645 nm/40 nm emission filter. Cells growing in 200 µL of minimal media were bleached in 50 µM norflurazon (Sigma-Aldrich). Absorbance at 688 nm and 750 nm, and fluorescence (645 nm/40 nm) were measured hourly. All experiments were performed in triplicate. The absorption spectrum of C. reinhardtii CC-503 ( Figure 1) was obtained from 1 mL of 1 × 10 7 cells mL −1 grown in minimal medium using a Synergy HT microplate reader with a 1 nm bandpass filter and was similar to previous reports [23]. Absorption was normalised to the absorbance at 680 nm.

Design of Chloroplast Transformation Vectors
A synthetic Katushka coding region (CpKat) containing frequently used C. reinhardtii chloroplast codons was synthesised from oligonucleotides by overlap PCR. Transgenes were inserted downstream of the atpB gene in the large 4.5 kb intergenic region between the atpB gene in the single copy region and the rrnS gene in the large inverted repeat. This was carried out by cloning the 2.1 kb left and 2.5 kb right targeting arms from the 7.6 kb Bam 10 fragment ( Figure 2A) and inserting transgenes into the Kpn I site located at coordinate 159207 of the chloroplast genome [38]. The AKL and DKL expression cassettes (Supplementary Figure S1) contained the CpKat coding region bordered by the promoter and 5 UTR from the chloroplast atpA or psbD gene and a common 3 regulatory region from the rbcL gene ( Figure 2B). Chloroplast transformation vectors pB10-AKL and pB10-DKL were assembled by insertion of AKL and DKL expression cassettes into the linker sequence between the left and right targeting arms.

Isolation of CpKat Chloroplast Transformants Expressing Katushka
Selection of chloroplast transformants was based on repair of the deleted a in the non-photosynthetic CC-373 strain [33] with the wild-type atpB gene in p This selection regime allowed the isolation of photosynthetic transformants on media lacking acetate. Neither the vector nor the C. reinhardtii CC-373 strain u recipient for transformation contained antibiotic resistance genes. The absence o otic resistance genes avoids the regulatory concerns associated with the presence genes in transgenic algae [49]. The chloroplast DNA deletion in CC-373 [33] exte the inverted repeat ( Figure 2A). Crossover events in the left 2.1 kbp and right targeting arms of the pB10 vector ( Figure 2B) integrate the intact atpB gene and genes into the CC-373 chloroplast genome. These crossover events take place in se common to the pB10 vector and CC-373 genome that flank the CC-373 deletion respond to a 181 bp sequence in the left arm and 897 bp sequence in the right a hough the CC-373 deletion extends beyond the left Bam HI site in Bam 10, the left of the deletion contains a 181 bp sequence containing multiple ACACTTTATTTT repeats that is present in Bam 10 (location shown in Figure 2A). The empty pB1 and the two CpKAT vectors (pB10-AKL, pB10-DKL) were transformed into CC particle bombardment. Transplastomic lines from independent transformation were selected from separated colonies growing photosynthetically on minima lacking acetate. Three AKL and DKL lines and two pB10 lines were selected fo

Isolation of CpKat Chloroplast Transformants Expressing Katushka
Selection of chloroplast transformants was based on repair of the deleted atpB gene in the non-photosynthetic CC-373 strain [33] with the wild-type atpB gene in pB10 [43]. This selection regime allowed the isolation of photosynthetic transformants on minimal media lacking acetate. Neither the vector nor the C. reinhardtii CC-373 strain used as a recipient for transformation contained antibiotic resistance genes. The absence of antibiotic resistance genes avoids the regulatory concerns associated with the presence of these genes in transgenic algae [49]. The chloroplast DNA deletion in CC-373 [33] extends into the inverted repeat (Figure 2A). Crossover events in the left 2.1 kbp and right 2.5 kbp targeting arms of the pB10 vector ( Figure 2B) integrate the intact atpB gene and foreign genes into the CC-373 chloroplast genome. These crossover events take place in sequences common to the pB10 vector and CC-373 genome that flank the CC-373 deletion and correspond to a 181 bp sequence in the left arm and 897 bp sequence in the right arm. Although the CC-373 deletion extends beyond the left Bam HI site in Bam 10, the left junction of the deletion contains a 181 bp sequence containing multiple ACACTTTATTTT tandem repeats that is present in Bam 10 (location shown in Figure 2A). The empty pB10 vector and the two CpKAT vectors (pB10-AKL, pB10-DKL) were transformed into CC-373 by particle bombardment. Transplastomic lines from independent transformation events were selected from separated colonies growing photosynthetically on minimal media lacking acetate. Three AKL and DKL lines and two pB10 lines were selected for further analysis. These lines corresponded to the transplastomic algae transformed with the pB10-AKL, pB10-DKL, and empty pB10 vectors.
All transgenic lines were propagated continuously on minimal medium for six months to select transgenic chloroplast genomes containing the intact atpB gene. Integration of the CpKat transgene in CC-373 transformants was confirmed by PCR analysis using primer 784 located in CpKat and primer 1420 located in chloroplast DNA beyond the right targeting arm. Primer locations are shown in Figure 3A. The predicted PCR products of 3.7 kbp and 3.3 kbp were amplified from pB10-AKL and pB10-DKL DNA, respectively ( Figure 3B). The sizes of the PCR products resulted from the differing lengths of the 5 atpA and psbD regulatory sequences. No PCR products were detected using DNA from pB10 transplastomic lines and CC-373, which lacked the CpKat transgene ( Figure 3B).
Microorganisms 2022, 10, x FOR PEER REVIEW 7 of 1 right targeting arm. Primer locations are shown in Figure 3A. The predicted PCR product of 3.7 kbp and 3.3 kbp were amplified from pB10-AKL and pB10-DKL DNA, respectively ( Figure 3B). The sizes of the PCR products resulted from the differing lengths of the 5 atpA and psbD regulatory sequences. No PCR products were detected using DNA from pB10 transplastomic lines and CC-373, which lacked the CpKat transgene ( Figure 3B). Accumulation of the 26 kDa Katushka protein in AKL and DKL transplastomic line was verified by protein blot analysis. A Katushka-specific antibody detected a single band of the correct size in total protein from AKL and DKL lines fractionated by denaturing SDS-PAGE ( Figure 3C). The 26 kDa band co-migrated with recombinant Katushka over expressed in Escherichia coli ( Figure 3C, lane 1). No Katushka band was detected in th negative control ( Figure 3C, last lane) corresponding to pB10 transplastomic algae lacking the CpKat gene.  Figure 3D showed that 645 nm fluores cence was significantly higher in the AKL and DKL lines compared to the pB10 lin Accumulation of the 26 kDa Katushka protein in AKL and DKL transplastomic lines was verified by protein blot analysis. A Katushka-specific antibody detected a single band of the correct size in total protein from AKL and DKL lines fractionated by denaturing SDS-PAGE ( Figure 3C). The 26 kDa band co-migrated with recombinant Katushka overexpressed in Escherichia coli ( Figure 3C, lane 1). No Katushka band was detected in the negative control ( Figure 3C, last lane) corresponding to pB10 transplastomic algae lacking the CpKat gene.
Katushka fluorescence from cultures was detected by using an excitation wavelength of 590 nm and detecting fluorescence at 645 nm. Figure 3D showed that 645 nm fluorescence was significantly higher in the AKL and DKL lines compared to the pB10 line lacking the CpKat gene. This was consistent with correct folding and maturation of Katushka, which are required to form the active fluorescent tetramer [50].

Chlorophyll Absorption of Katushka Fluorescence
Chlorophyll capture of the red photons emitted by Katushka is an essential step for enhancing the photosynthetic action spectrum. This was investigated by removing chlorophyll which would be predicted to increase Katushka fluorescence. The chlorophyll content of cells was reduced with 50 µM norflurazon [45]. Cells were grown mixotrophically in TAP medium containing acetate [34] under continuous cool white illumination. Chlorophyll absorbance (688 nm), Katushka fluorescence (645 nm), and cell density (750 nm) were measured for a period of 72 h following the addition of norflurazon ( Figure 4).
shka, which are required to form the active fluorescent tetramer [50].

Chlorophyll Absorption of Katushka Fluorescence
Chlorophyll capture of the red photons emitted by Katushka is an essential step for enhancing the photosynthetic action spectrum. This was investigated by removing chlorophyll which would be predicted to increase Katushka fluorescence. The chlorophyll content of cells was reduced with 50 μM norflurazon [45]. Cells were grown mixotrophically in TAP medium containing acetate [34] under continuous cool white illumination. Chlorophyll absorbance (688 nm), Katushka fluorescence (645 nm), and cell density (750 nm) were measured for a period of 72 h following the addition of norflurazon (Figure 4).
Chlorophyll absorbance started to decrease after 24 h, dropping more rapidly from 32 to 44 h, and was not detectable after 72 h in all strains. At time zero, red 645 nm fluorescence was approximately six-fold higher from AKL and DKL transplastomic strains than the control pB10 strain ( Figure 4A-C), consistent with incomplete masking of Katushka fluorescence by chlorophyll in these strains. Red fluorescence peaked at 40 h in AKL and DKL strains when chlorophyll absorbance had dropped by over 50%. Red fluorescence then dropped rapidly in concert with the sharp drop in chlorophyll absorbance. These results were consistent with the presence of Katushka in its active state, releasing red photons that were absorbed by chlorophyll. Prolonged photo-bleaching damages chloroplasts, and the associated changes, including protein degradation, most likely account for the drop in Katushka fluorescence observed after 40 h. The low red autofluorescence in the pB10 strain lacking Katushka closely followed the pattern of decline of chlorophyll absorbance ( Figure 4A). OD750 nm readings were used to monitor changes in cell densities and showed comparable results for all three strains. Cell densities remained static for the first 40 h and then showed a relatively steady and moderate decrease in all three strains ( Figure 4A-C). Chlorophyll absorbance started to decrease after 24 h, dropping more rapidly from 32 to 44 h, and was not detectable after 72 h in all strains. At time zero, red 645 nm fluorescence was approximately six-fold higher from AKL and DKL transplastomic strains than the control pB10 strain ( Figure 4A-C), consistent with incomplete masking of Katushka fluorescence by chlorophyll in these strains. Red fluorescence peaked at 40 h in AKL and DKL strains when chlorophyll absorbance had dropped by over 50%. Red fluorescence then dropped rapidly in concert with the sharp drop in chlorophyll absorbance. These results were consistent with the presence of Katushka in its active state, releasing red photons that were absorbed by chlorophyll. Prolonged photo-bleaching damages chloroplasts, and the associated changes, including protein degradation, most likely account for the drop in Katushka fluorescence observed after 40 h. The low red autofluorescence in the pB10 strain lacking Katushka closely followed the pattern of decline of chlorophyll absorbance ( Figure 4A). OD 750 nm readings were used to monitor changes in cell densities and showed comparable results for all three strains. Cell densities remained static for the first 40 h and then showed a relatively steady and moderate decrease in all three strains ( Figure 4A-C).

CpKat Chloroplast Transformants Use Yellow Light for Photosynthesis
Photosynthesis was monitored by oxygen evolution [47]. Equal amounts of AKL, DKL, and pB10 cells in minimal medium were placed in cool white light or yellow light (λ max = 590) for six hours. Figure 5 shows the amount of dissolved oxygen found in the cultures. Similar dissolved oxygen concentrations were found in the AKL, DKL, and pB10 cultures exposed to cool white light ( Figure 5A). Under yellow light, clear differences were found between the AKL and DKL cultures expressing Katushka and the negative pB10 control cultures lacking CpKat ( Figure 5B). The results show that expression of Katushka in chloroplasts enables the utilisation of the energy in yellow photons to split water in photosystem II.

CpKat Chloroplast Transformants Use Yellow Light for Photosynthesis
Photosynthesis was monitored by oxygen evolution [47]. Equal amounts of AKL, DKL, and pB10 cells in minimal medium were placed in cool white light or yellow light (λmax = 590) for six hours. Figure 5 shows the amount of dissolved oxygen found in the cultures. Similar dissolved oxygen concentrations were found in the AKL, DKL, and pB10 cultures exposed to cool white light ( Figure 5A). Under yellow light, clear differences were found between the AKL and DKL cultures expressing Katushka and the negative pB10 control cultures lacking CpKat ( Figure 5B). The results show that expression of Katushka in chloroplasts enables the utilisation of the energy in yellow photons to split water in photosystem II.

Yellow Light Promotes the Growth of Transplastomic Lines Expressing Katushka
The influence of light conditions on growth of a control photosynthetic strain CC-503 lacking Katushka was determined ( Figure 6). Under the conditions tested the CC-503 strain grew best in a mixture of red and blue light compared to red or blue light alone.

Yellow Light Promotes the Growth of Transplastomic Lines Expressing Katushka
The influence of light conditions on growth of a control photosynthetic strain CC-503 lacking Katushka was determined ( Figure 6). Under the conditions tested the CC-503 strain grew best in a mixture of red and blue light compared to red or blue light alone. These results showed that under the blue light conditions used, addition of red light boosted cell growth. Although differences in cell sizes and division rates have been reported for cultures grown in blue light versus red light [51,52] the final cell densities of CC-503 cultures grown in our blue and red light conditions were not significant ( Figure 6). Growth of CC-503 in our standard cool white light conditions for propagating strains was reduced compared to combined blue and red light. The difference most probably reflects the lower energy   Figure 7 shows growth of the AKL, DKL and control pB10 transplastomic s three different light regimes. All strains grew to comparable final densities in co light ( Figure 7A), which contains blue and red photons. This shows that the pre Katushka was not advantageous when red light was provided. The result also s detectable growth penalty or metabolic load associated with the expression of K in chloroplasts. In yellow light (30 μmol m −2 s −1 ), the presence of Katushka had positive impact on the growth of the AKL and DKL strains relative to the pB1 ( Figure 7B). The final cell densities of AKL cultures were 4-fold, and DKL cultur higher than the pB10 cultures. The densities of the pB10 cultures had dropped f starting OD 750 nm values of 0.1. The growth of AKL and DKL cells can be expla the conversion of yellow photons, which are poor drivers of photosynthesis, to r tons that are efficiently captured and used for photosynthesis. When yellow light ( m −2 s −1 ) was combined with blue light (30 μmol m −2 s −1 ), growth of all strains was in relative to yellow light alone ( Figure 7B,C) illustrating the higher quality of b compared to yellow light for photosynthetic growth. However, the AKL and DK grew to higher densities than the pB10 strain in combined blue and yellow light 7C). This increased growth can be explained by the Katushka-mediated conversio low to red photons in AKL and DKL strains. The combined results show that en the photosynthetic action spectrum by using a device that can convert yellow ph  Figure 7 shows growth of the AKL, DKL and control pB10 transplastomic strains in three different light regimes. All strains grew to comparable final densities in cool white light ( Figure 7A), which contains blue and red photons. This shows that the presence of Katushka was not advantageous when red light was provided. The result also shows no detectable growth penalty or metabolic load associated with the expression of Katushka in chloroplasts. In yellow light (30 µmol m −2 s −1 ), the presence of Katushka had a clear positive impact on the growth of the AKL and DKL strains relative to the pB10 strain ( Figure 7B). The final cell densities of AKL cultures were 4-fold, and DKL cultures 3-fold higher than the pB10 cultures. The densities of the pB10 cultures had dropped from the starting OD 750 nm values of 0.1. The growth of AKL and DKL cells can be explained by the conversion of yellow photons, which are poor drivers of photosynthesis, to red photons that are efficiently captured and used for photosynthesis. When yellow light (30 µmol m −2 s −1 ) was combined with blue light (30 µmol m −2 s −1 ), growth of all strains was increased relative to yellow light alone ( Figure 7B,C) illustrating the higher quality of blue light compared to yellow light for photosynthetic growth. However, the AKL and DKL strains grew to higher densities than the pB10 strain in combined blue and yellow light ( Figure 7C). This increased growth can be explained by the Katushka-mediated conversion of yellow to red photons in AKL and DKL strains. The combined results show that enhancing the photosynthetic action spectrum by using a device that can convert yellow photons to red photons improved photosynthetic growth of algal cells in the absence of red light when yellow light was provided.

Discussion
The fluorescent protein Katushka provided a simple light-capturing devi hance the light spectrum for photosynthesis by converting yellow photons to red This was exemplified by expressing Katushka in C. reinhardtii chloroplasts, which in increased oxygen evolution and photosynthetic growth when cells were grow low light. Moreover, chloroplast-localised Katushka increased the growth of a grown in blue and yellow light. In white light, no differences in growth rate were o between cells containing or lacking Katushka. In this situation, the abundance of tons in white light would have masked those generated by Katushka fluorescenc results strongly support Katushka-mediated conversion of yellow to red photo mechanism responsible for promoting cell growth when red light was absent. Katushka is acting as an accessory protein to capture yellow photons for photos

Discussion
The fluorescent protein Katushka provided a simple light-capturing device to enhance the light spectrum for photosynthesis by converting yellow photons to red photons. This was exemplified by expressing Katushka in C. reinhardtii chloroplasts, which resulted in increased oxygen evolution and photosynthetic growth when cells were grown in yellow light. Moreover, chloroplast-localised Katushka increased the growth of algal cells grown in blue and yellow light. In white light, no differences in growth rate were observed between cells containing or lacking Katushka. In this situation, the abundance of red photons in white light would have masked those generated by Katushka fluorescence. These results strongly support Katushka-mediated conversion of yellow to red photons as the mechanism responsible for promoting cell growth when red light was absent. In effect, Katushka is acting as an accessory protein to capture yellow photons for photosynthesis. This synthetic biology approach involving redesigning algal cells to enhance the photosynthetic action spectrum complements other methods to boost photosynthesis, which include the use of submerged lighting [53] and fluorescent nanomaterials [54].
Katushka increases the number of recombinant fluorescent proteins expressed in C. reinhardtii [55]. Fluorescent proteins have been used mainly as reporters of gene expression and as protein tags for subcellular localisation [55][56][57]. As far as we are aware, this is the first use of a recombinant fluorescent protein to enhance the photosynthetic action spectrum. We used a coding sequence comprising frequently used chloroplast codons, a strategy that had previously been used to enhance accumulation of the green fluorescent protein in C. reinhardtii chloroplasts by 80-fold [57]. The CpKat expression cassettes were integrated close to the atpB gene in the intergenic region between the rrnS and atpB genes, which has been previously used to integrate foreign genes into chloroplast DNA [58]. We used the promoter and 5 UTR regions from the chloroplast atpA and psbD genes, together with the 3 region of the chloroplast rbcL gene, to express Katushka. These 5 regulatory regions from the highly expressed atpA and psbD genes were chosen because they have been shown to be effective in driving the expression of recombinant proteins in chloroplasts [42,58,59]. Both sets of regulatory elements led to the accumulation of Katushka in chloroplasts and improved growth of transplastomic lines when yellow light was present and red light was absent. Increasing the amount and timing of Katushka expression, to fluctuate in response to the availability of yellow and red light, provides an approach to further improve growth. This will require developing new synthetic regulatory elements, sensitive to red and yellow light, to replace the native chloroplast regulatory elements used in this work.
Red photons emitted by Katushka were absorbed by chlorophyll. This was shown by the increase in red fluorescence emitted from cultures following removal of chlorophyll using the bleaching herbicide norflurazon. Transfer of fluorescent energy from Katushka to chlorophyll was likely to involve radiative energy transfer, rather than the more efficient fluorescence resonance energy transfer (FRET) which requires very small intermolecular distances [60]. Inefficient energy transfer to chlorophyll resulted in escape of Katushkaemitted red photons from cells, which was readily detected in transplastomic cell cultures expressing Katushka. Reducing the distance travelled by red photons would require targeting and accommodating Katushka in the crowded thylakoid membranes in chloroplasts that contain the chlorophyll light-harvesting complexes [61]. The potential of fluorescent proteins as light-capturing devices for photosynthesis will benefit from the identification of new brighter and higher quantum yield fluorescent proteins with favourable excitation and emission wavelengths [29]. Here, we explored the use of yellow photons to drive photosynthesis but the approach is applicable to other underutilised wavelengths in photosynthetic active radiation, which varies between taxa [19]. Combining fluorescent proteins with different excitation and emission wavelengths will increase the wavelength range of photons that can be captured.

Conclusions
The results show that a simple light-capturing device, comprising the Katushka fluorescent protein expressed in chloroplasts, improves photosynthetic growth of C. reinhardtii in yellow light alone, or in a mixture of blue and yellow light. The device works by converting yellow photons to red photons that are absorbed by the light-harvesting chlorophyll complexes in thylakoids to drive photosynthesis. This demonstrates the potential of enhancing the photosynthetic action spectrum to improve algal biomass accumulation under conditions where red light limits growth. This proof-of-concept study based on fluorescence of a single protein supports the utility of developing synthetic light-capture systems for photosynthesis [16,[18][19][20]. Improving light-capture efficiencies at multiple wavelengths that are tunable to the environment is a desirable feature that will require more complex light-capture systems and advanced genetic engineering. This is likely to require multi-subunit complexes whose abundance is regulated by the quality and intensity of incident light.