Probing rapid carbon fixation in fast-growing seaweed Ulva meridionalis using stable isotope 13C-labelling

The high growth rate of Ulva seaweeds makes it a potential algal biomass resource. In particular, Ulva meridionalis grows up to fourfold a day. Here, we demonstrated strong carbon fixation by U. meridionalis using 13C stable isotope labelling and traced the 13C flux through sugar metabolites with isotope-ratio mass spectrometry (IR-MS), Fourier transform ion cyclotron resonance mass spectrometry (FT-ICR-MS), 13C-nuclear magnetic resonance spectrometry (13C-NMR), and gas chromatography-mass spectrometry (GC–MS). U. meridionalis was first cultured in 13C-labelled enriched artificial seawater for 0–12 h, and the algae were collected every 4 h. U. meridionalis grew 1.8-fold (dry weight), and the 13C ratio reached 40% in 12 h, whereas 13C incorporation hardly occurred under darkness. At the beginning of the light period, 13C was incorporated into nucleic diphosphate (NDP) sugars in 4 h, and 13C labelled peaks were identified using FT-ICR-MS spectra. Using semiquantitative 13C-NMR measurements and GC–MS, 13C was detected in starch and matrix polysaccharides after the formation of NDP sugars. Moreover, the 14:10 light:dark regime resulted into 85% of 13C labelling was achieved after 72 h of cultivation. The rapid 13C uptake by U. meridionalis shows its strong carbon fixation capacity as a promising seaweed biomass feedstock.

Green tides of Ulva seaweeds are increasing with the eutrophication of the sea and have become a major environmental problem 1 . Rapid and vast Ulva growth disturbs the ecological system at the seashore. The growth dynamics and mechanisms of fast growth in Ulva have been studied to mitigate Ulva green tides [2][3][4] . At the same time, Ulva is an expected biomass source that can capture CO 2 and convert it into renewable biochemicals 5 . Filamentous Ulva meridionalis is a potential seaweed species that exhibits the fastest growth rate ever found in multicellular autotrophic plants 6,7 . U. prolifera, one of the most widely known edible Ulva species, doubles in a day at the optimised culturing conditions; in contrast, U. meridionalis grows over fourfold in a day. In general, many Ulva grow from winter to spring, whereas U. meridionalis grows in summer. U. meridionalis has a high rate of photosynthetic carbon fixation, a high frequency of cell proliferation, and rapid formation of algae bodies consisting of cell walls in high-temperature environments. The accumulated cell and cell wall components are important biochemical resources that include polysaccharides and rare sugars 5 , which are produced by Ulva through extraction and hydrolysis processes [8][9][10] . Land cultivation of Ulva has been extensively studied for stable seaweed production for both food and biomass applications 11 . The "germling cluster method" allows mass production of Ulva by clustering the algae bodies and cultivating them in high density floating Ulva clusters in a land tank. This method provides efficient exposure for Ulva growth by three-dimensionally stirring the clusters. Ulva are easily collected by a coarse filtration, which is much easier than harvesting microalgae biomass that requires flocculation and centrifugation. This efficient cultivation system enhances Ulva growth and captures more CO 2 .
Ulva suffers various environmental stresses such as frequent submergence into seawater and exposure to air by tides, high light intensity, and dryness at the intertidal zone. However, their photosynthetic systems are highly adapted to severe environments 12 . At the same time, Ulva are widely distributed worldwide and are a model macroalgae that are used for studying photosynthesis mechanisms 13 . Most macroalgae utilise bicarbonate ions (HCO 3 − ) as a carbon source because it is a major species of CO 2 that occurs at the pH and salt concentration Scientific Reports | (2020) 10:20399 | https://doi.org/10.1038/s41598-020-77237-1 www.nature.com/scientificreports/ levels of seawater 14 . Carbonic anhydrase shifts the equilibrium of HCO 3 − and CO 2 to enhance CO 2 uptake 15 . Several macroalgae, including Ulva, have a proton pump system that enhances the HCO 3 − uptake. Moreover, Ulva is equipped with a carbon concentrating mechanism (CCM). Beer et al. reported that the introduction of 14 C pulses into U. fasciata under light immediately produces 14 C labelled glycerate 3-phosphate with more than 90% of molecules labelled, and it was further transferred to other photosynthesis products such as sugars 16 . Since photorespiration is suppressed in Ulva, they concluded that Ulva have a carbon concentrating system in their chloroplasts along with their efficient HCO 3 − uptake mechanism 17,18 . The CCM maintains the CO 2 in cells at a sufficiently high level to saturate carboxylation by Rubisco in the Calvin cycle 17 . The CCM was also found in U. prolifera, which caused huge green tides in the Yellow Sea 19 . Ulva also exhibits high tolerance to strong light by non-photochemical quenching. Efficient photosynthesis can be maintained under strong light by protecting photosystems I and II from excess electrons (e − ) by trapping them in xanthophyll 20,21 .
Various polysaccharides such as storage, cell wall, and matrix polysaccharides are formed through biosynthetic pathways after carbon fixation in the Calvin cycle. Ulva bodies consist of cellulose (~ 10%), matrix polysaccharides (~ 30%), and starch (~ 10%) along with proteins (~ 20%) 8 . Ulva seaweeds are characterised by matrix polysaccharides (ulvan) that are heteropolysaccharides consisting of rhamnose (Rha), glucuronic acid (GlcA), iduronic acid (IdoA), and sulfates [22][23][24] . In some cases, mannose (Man), galactose (Gal), and arabinose (Ara) are found in ulvan. Various fine structures were reported including: Biosynthesis of ulvan occurs from nucleic diphosphate (NDP) sugar, similar to pectin biosynthesis in higher plants 23 . d-fructose-6-P supplied from photosynthesis and glucogenesis is converted to uridine diphosphate glucose (UDP-Glc) through several steps. The corresponding NDP sugars to each monosaccharide are produced through the interconversion of UDP-Glc. For instance, UDP-GlcA is converted from UDP-Glc or inositol and further converts to UDP-Xyl. Rha originates from UDP-L-Rha or dTDP-L-Rha, which are converted from UDP-Glc. Ulvan is synthesised by glycosyltransferase from the above-mentioned NDP sugars in the Golgi apparatus. Polysaccharides are then transported to the apoplast, and further assembling and modifications occur. However, the biosynthetic pathway of ulvan is yet not well understood although it is involved in the rapid growth of Ulva.
Isotope labelling has been used to analyse the kinetics of carbon uptake by photosynthesis and its transfer into metabolites. Percival et al. used the radioisotope 14 C to study carbon fixation of U. lactuca and its transfer into sugar metabolites 25 . They reported that 14 C incorporation in 80% aqueous ethanol extract or hot water extract was detected after culturing in a 14 C medium containing NaH 14 CO 3 for 10 min. Incorporation of 14 C into starch occurred earlier than ulvan. After the hydrolysis of ulvan, the radioactivity of Rha was higher than that of GlcA, indicating that Rha deposition occurs earlier than GlcA deposition. Recent advances in analytical instruments such as mass spectrometry (MS) and nuclear magnetic resonance (NMR) allow for a more comprehensive analysis of fluxes of metabolites and material exchanges between organisms by stable isotope labelling [26][27][28] . Natural 13 C distribution is only 1.07%, but concentrating 13 C makes structural analysis of metabolites by 13 C-NMR spectroscopy easier. High performance liquid chromatography electron spray ionization mass spectrometry (HPLC ESI-MS) and GC-MS analysis of 13 C-enriched metabolites enable tracing of the 13 C flux of each component. Further, FT-ICR-MS provides an accurate mass of metabolites that allows tracing of the metabolic foot print 29 .
This study demonstrated the behaviour of rapid carbon fixation in U. meridionalis and its transfer to polysaccharides through NDP sugars to understand the physiological mechanisms of their extremely high growth rate. U. meridionalis was cultured in enriched artificial seawater (EASW) 30-32 medium with 13 C-NaHCO 3 as the sole carbon source, and the temporal 13 C transfer to NDP sugars and polysaccharides (ulvan and starch) was analysed. Incorporation of 13 C into NDP sugars was analysed using FT-ICR-MS. Semiquantitative 13 C-NMR by inverse-gated-decoupling was used to observe 13 C incorporation into polysaccharides. GS-MS was used to trace 13 C uptake in monosaccharide moieties in polysaccharides after their acid hydrolysis. Finally, the flux of carbon fixation and its transfer into the cell wall polysaccharides was discussed in conjunction with the rapid growth rate of U. meridionalis. Figure 1A,B shows the changes in fresh weight and dry weight over time in light and dark conditions. The fresh weight shows linear increases over time, and it doubled in 24 h. U. meridionalis grew up to about 3.5-fold in a day at the same light and temperature condition (12:12 light:dark regime, 25 °C) 6 . In the present study, the growth rate of U. meridionalis was lower because closed flasks were used for its cultivation, without aeration, to prevent an exchange of 13 C with 12 C in the atmosphere. Therefore, diffusion of oxygen, which is an inhibitor of the carboxylase activity of Rubisco, might not be sufficient, resulting in a lower growth rate than the ideal one. Moreover, 13 C uptake can be slower than 12 C uptake because of its larger mass. The dry weight of U. meridionalis increased by 1.8 times, and the amount of oxygen generated reached 49 mL after 12 h in the light period ( Supplementary Fig. S1). In the dark period, the fresh weight increased linearly at the same rate as that under light, but the dry weight did not change. That is, photosynthetic metabolites accumulated in the light period to increase their dry weight. Meanwhile, water content increased during the dark period.

C uptake by U. meridionalis.
After powdering the dried sample, the amount of incorporated 13 C was measured by IR-MS. The 13 C uptake in the light period was 41.2%, which was consistent with a doubling of dry weight under light (Fig. 1C). On the other hand, 13 C uptake was hardly observed in the dark period, indicating that the carbon fixation of U. meridionalis occurs in a light-dependent manner (Fig. 1D). Then, the elemental composition was investigated by CHNS analysis. The amounts of C (27.8% to 30.2%) and H (4.7% to 5.5%) reflected entire organic compounds, including sugar chains and proteins (Fig. 1E,F). The amount of N (3.0% to 3.9%) was mainly attributed to protein Scientific Reports | (2020) 10:20399 | https://doi.org/10.1038/s41598-020-77237-1 www.nature.com/scientificreports/ and that of S (2.9% to 3.9%) was attributed to the sulphate groups of sugar chains such as ulvan. There was no change in the ratio of carbon, hydrogen, nitrogen, and sulfur in both light and dark periods throughout the cultivation. This suggests that the elemental composition of U. meridionalis is always constant regardless of the extensive accumulation of photosynthetic products during the light period. The carbon uptake by U. meridionalis was measured from the dry weight, and carbon content was obtained by CHNS analysis. The amount of carbon fixed per unit dry algae reached 16.8 mg C fixed g dry wt −1 h −1 (0-4 h), 15.7 mg C fixed g dry wt −1 h −1 (4-8 h), and 13.7 mg C fixed g dry wt −1 h −1 (8-12 h), which exceeded those of tubular and sheet-like Ulva sp. (5.16 mg C fixed g dry wt −1 h −1 on average) reported by Littler et al. 33 .   34 , and the incorporation of 13 C into the NDP sugar was measured by FT-ICR-MS. Low molecular weight compounds such as pigments were first extracted from U. meridionalis samples with 75% aqueous ethanol. The amount of ethanol extract used was consistent over time and was 37.7 ± 4.5% on average during both light and dark periods ( Supplementary Fig. S2). 75% aqueous ethanol extract was dissolved in aqueous NaHCO 3, and the soluble fraction was applied to ENVIcarb solid-phase extraction (SPE) cartridge. This column is suitable for the separation of highly polar compounds. Highly polar NDP sugars were adsorbed into the column, and then the NDP sugar was eluted with 25% acetonitrile containing triethylammonium acetate buffer (TEAA; 50 mM, pH 7) to recover NDP sugars. The NDP sugar was analysed by FT-ICR-MS equipped with ESI (Bruker solariX). The MS spectra were recorded in negative-ion mode since the NDP sugars are negatively charged. Table 1 shows the carbon number, monoisotopic mass, and m/z of 13 C-labelled UDP-pentose, UDPdeoxyhexose, UDP-hexose, and UDP-hexuronic acid as precursors of ulvan and starch. The monoisotopic mass is indicated as M 0 , and the mass changes from M 0 to M 0+15 , depending on the carbon number of each NDP sugar (carbon numbers = 14 or 15) by 13 C labelling, were analysed. For example, UDP-pentose with 14 carbons (M 0 = 535), which corresponds to the precursor of Xyl, exhibits M 0+14 when it is completely labelled with 13 C. Similarly, UDP-hexose and UDP-hexuronic acid correspond to precursors of Glc and Gal and GlcA, respectively. Figure 2 shows the temporal change in the FT-ICR-MS spectrum of 13 C-labelled NDP sugars. Peaks with m/z corresponding to UDP-pentose, UDP-hexose, and UDP-hexuronic acid were detected. On the other hand, UDPdeoxyhexose, which is considered to be a precursor of Rha, was not detected. Similarly, dTDP-Rha (M 0 = 547), which is the precursor of Rha in microorganisms, was not detected. In addition, m/z corresponding to UDP-Nacetyl-hexose was detected in both light and dark periods. However, the corresponding monosaccharide to UDP-N-acetyl-hexose has not been identified. Since the 13 C labelling may cause an overlap in M 0+x peaks of several compounds, FT-ICR-MS measurement of non-labelled NDP sugars was also performed using U. meridionalis cultured in a non-labelled medium under the same conditions ( Supplementary Fig. S3). The monoisotopic mass peaks corresponding to above NDP sugars were detected; thus, their presence was confirmed.
In Fig. 2C-H, the changes in the mass of each NDP sugar are extracted, and they were expressed as the ratio of the individual masses to the sum of M 0 to M 0+X (X = 14, 15). 13 C was rapidly incorporated into NDP sugars under light conditions. After 4 h of 13 C labelling, M 0+14 (UDP-pentose) and M 0+15 (UDP-hexose, UDP-hexuronic acid) were the major peaks followed by partially labelled peaks of M 0 to M 0+14 . The ratio of 13 C incorporated in UDP-hexose was constant at 4-8 h. After 12 h, the peak of M 0+15 increased; however, M 0+15 did not reach 100% even after 12 h. This suggested that Ulva also used non-labelled metabolites, which were accumulated before 13 C labelling, for production of NDP sugars. With UDP-hexuronic acid, the peak of M 0+15 at 12 h reached 70% of the total, suggesting that the UDP-GlcA was generated from 13 C-labeled UDP-Glc. On the other hand, UDPpentose showed lesser uptake of 13 C than that of UDP-hexose and UDP-hexuronic acid. 13 C incorporation into UDP-pentose was slow because the formation of UDP-Xyl is downstream of UDP-GlcA formation. During the dark period, 13 C uptake was much lower than that in the light period, but 13 C uptake still occurred to some extent. Peaks corresponding to M 0+1 to M 0+2 were observed and increased over time. Carbon fixation, also called dark reaction, proceeded even in the absence of light. However, carbon uptake was particularly low in the dark period.
Incorporation of 13 C in ulvan and starch. The transfer of 13 C to matrix polysaccharides and starch of U. meridionalis was investigated using 13 C-NMR. Polysaccharides were recovered through hot water extraction (90 °C, 1 h × 3) of residues after 75% aqueous ethanol extraction. The polymers in the hot water extract were recovered through ethanol precipitation. After freeze-drying, the incorporation of 13 C into the hot water extract was evaluated by IR-MS. This fraction contained starch and ulvan but was not purified further because the amount of extract was small. Figure 3A shows the 13 C ratio in the hot water extract. The incorporation rate of 13 C into polysaccharides was the same as that for the whole algae body because their slopes were almost the same. Then, the hot water extract was dissolved in a mixed solvent of D 2 O-DMSO-d 6 and analysed by 13 C-NMR. Inverse-gated-decoupling provided the quantitative 13 C spectra with 1 H decoupling with a sufficient relaxation time of the T 1 13 C 35 . However, in the present study, the repetition time was set to 10.69206 s since the T 1 13 C was long 36 . DMSO-d 6 contained in the solvent was used as an internal standard to determine the amount of 13 C uptake into each carbon-containing ulvan. Figure 3B,C shows the semiquantitative 13 C-NMR spectra of the hot water extract during the light and dark periods. A clear chemical shift was observed in the light period, and the intensity increased with time. These chemical shifts were similar to those of U. armoricana ulvan because there were similar signals for G1, Gg1, R'1, and R1 at around 100 ppm 23 . Chemical shifts in the range of 60 ppm-100 ppm were attributed to C-OH of sugar moieties, but they were difficult to analyse in detail because they overlapped with each other. Therefore, www.nature.com/scientificreports/ the signals at 17 ppm due to C-6 of Rha and 176 ppm due to the carboxyl group of C-6 of GlcA were analysed because they were completely separate from the other signals. A normalised intensity against those of 0 h is summarised in Fig. 3D. The transfer of 13 C to Rha and GlcA occurs later compared with the transfer of 13 C to the

Incorporation of 13 C in monosaccharides in ulvan and starch.
Further, 13 C incorporation at the monosaccharide level was analysed by GC-MS. Hot water extracts were hydrolysed with TFA (2 M), derivatised to alditol acetate, and then analysed by GC-MS. Figure 4A,B show the weight ratio of Rha, Xyl, Man, Glc, and Gal. Rha and Xyl were attributed to ulvan, while Glc was attributed to starch. Therefore, we focused on the 13 C/ 12 C ratios of Rha and Glc as representatives of ulvan and starch. The ratio of Rha was the highest in all conditions. On the other hand, Glc varied greatly depending on the cultivation time. In the light period, the Glc ratio increased from 0 to 8 h. This Glc increase was attributed to starch production during the light period. Then, the Rha ratio increased up to 12 h due to the deposition of ulvan and resulted in a relatively low Glc ratio. During the dark period, the Glc ratio gradually decreased with time. This is probably because the starch synthesised in the light period was consumed in the dark period. Therefore, the ratio of Rha and Xyl gradually increased as the ratio of Glc decreased.
The incorporation of 13 C into each monosaccharide was investigated in detail using EI-MS spectra. Figure 4C shows changes in typical fragmentation patterns of neutral sugars due to the introduction of 13 C. Replacement of the fragment ions with 13 C changes m/z from + 1 to + 4. Figure 4D-G show temporal changes in the fragmentation patterns of each neutral sugar in the light period. The m/z changes following the incorporation of 13 C are highlighted with bold lines and red letters. For example, in Rha, a fragment corresponding to m/z 217 began to www.nature.com/scientificreports/ migrate to 220 in 8 h. The peak intensity of 220 was 2.1 times higher than that of 217 at 12 h. The average 13 C/ 12 C ratio in Rha was about 1.3 after 12 h (Supplementary Table S2). On the other hand, the average 13 C/ 12 C ratio in Glc was 3.8 after 12 h, which was higher than that of Rha (Supplementary Table S2). This indicates that the vigorous starch biosynthesis in the light period increases the level of 13 C incorporation into Glc. On the other hand, ulvan biosynthesis represented by the 13 C/ 12 C ratio of Rha was slower than starch biosynthesis.

Discussion
This study performed 13 C labelling cultivation of U. meridionalis to understand the mechanisms by which they grow up to fourfold a day. We tracked the fresh and dry weight changes, 13 C uptake in whole algae, the formation of NDP sugars, matrix polysaccharides, and monosaccharides during the light and dark periods. Compared to the previous 14 C labelling of Ulva 25 , enriched 13 C was useful to track the carbon fixation into matrix and storage polysaccharides through NDP sugars by using IR-MS, FT-ICR-MS, 13 C-NMR and GC-MS. In the light period, www.nature.com/scientificreports/ photosynthesis promotes strong carbon fixation. Carbon fixed in the Calvin cycle is converted into UDP-Glc via fructose 6-phosphate and then UDP-GlcA and UDP-Xyl. Labelling of these NDP sugars already occurred at an early stage after the start of the light period. Meanwhile, the transfer of 13 C to matrix polysaccharides was delayed by the formation of NDP sugars and occurred more than 4 h after the start of the light period. Percival et al. studied carbon fixation in U. lactuca using radioactive 14 C as a tracer and found that sucrose was the earliest product, detected 10 min after labelling, followed by starch formation 25 . Then, ulvan was formed after a delay of about 3 h. Therefore, U. meridionalis has a similar sugar biosynthesis system to that of U. lactuca. However, the carbon fixation speed by U. meridionalis further exceeded the average primary productivity by other thin tubular and sheet-like seaweeds (mainly Ulva seaweeds) by 2.65-3.26-fold 33 .
The change in fresh weight during growth did not show a large difference between the light and the dark periods (Fig. 1A,B). This was also confirmed by the growth behaviour when cultured in the unlabelled medium ( Supplementary Fig. S5A). On the other hand, the change in dry weight differed greatly between the light and the dark periods. The dry weight increased clearly in the light period, but there was almost no change in the dark period. The similar behaviour was also observed with the unlabelled medium ( Supplementary Fig. S5B). This indicates that most of the carbon fixation products are obtained by photosynthesis during the light period, which is completely consistent with the 13 C uptake behaviour. Meanwhile, water content increased during the dark period. As a result, the fresh weight increased linearly in both light and dark periods.
Based on the above carbon uptake behaviours, we turned to prolong the light period to test if it is an effective method of accumulating more fixed carbon and promoting growth. Greater carbon fixation will be effective for biomass production and carbon dioxide fixation. Supplementary Fig. S4 shows the growth behaviour when the light:dark regime was varied from 12:12 to 24:0. Under a light:dark regime of 14:10, the daily growth rate increased, however, the growth rate did not increase when the light period was prolonged above 14 h. This suggests that the dark period is also important for U. meridionalis growth. According to Hiraoka et al., cell division should occur at least twice a day to grow fourfold a day 6 . According to Kuwano et al., the cell division cycle of U. partita (= Enteromorpha compressa in Kuwano et al.) is controlled by the diurnal cycle, and a dark period is required for 6 h 37 . During the dark period, the number of cells increase by cell division. That is, cell division occurs using photosynthetic metabolites produced under the light period. In fact, the percentage of Glc in the hot water extract-hydrolysate reduced in the dark period. This indicates that starch is consumed as an energy source for cell division. In addition, McArthur et al. reported that cell wall deposition occurs at the end of cytokinesis through light and electron microscopy of U. intestinalis during mitosis 38 . That is, the ulvan obtained during the light period also contributes to the formation of the cell wall, accompanying cell division during the dark period.
The optimal light:dark regime of U. meridionalis was the long-day condition of around 14:10. This is consistent with the fact that U. meridionalis was first discovered in summer and grows at high water temperatures 7 . Then, we examined how much carbon could be fixed by U. meridionalis using 13 C labelling under the light:dark regime of 14:10. A closed flask bubbling with N 2 gas was used to enable long-term 13 C labelling ( Supplementary Fig. S6). The same 13 C-labelled EASW mentioned above was added to the flask and filled with nitrogen gas so as not to contact the 12 C in the atmosphere. Nitrogen gas was circulated using a diaphragm pump to appropriately agitate the algal bodies. The 13 C medium was replaced every 24 h so that the medium was always fresh. Under these conditions, U. meridionalis grew three times each day (Fig. 5A). Uptake of 13 C at 72 h reached approximately 85%. This confirms the very high carbon fixation capacity of U. meridionalis (Fig. 5B).
In conclusion, the high growth rate of U. meridionalis is associated with its strong carbon fixation ability. Through 13 C labelling in the light period, the NDP sugar was rapidly labelled as the precursor of polysaccharides; then, 13 C was transferred to ulvan later. In contrast, carbon uptake hardly occurred in the dark period. Cell division was carried out in the dark period using the carbon fixed during the light period. For instance, the accumulated starch may be used as an energy source for cell division. Further, ulvan may be used for cell wall formation at the end of cell division. Under optimal long-day conditions, 90% of 13 C labelling was achieved within 72 h. We conclude that U. meridionalis exhibits a high carbon fixation capacity that contributes to its high growth potential.

Methods
Pre-cultivation of U. meridionalis. The pre-cultivation of U. meridionalis was conducted according to the procedure of Hiraoka et al. 6 . Namely, U. meridionalis stock was transferred to non-13 C labelled EASW (Supplementary Table S1) for 8 days [30][31][32] . The light irradiation was operated at 50 μmol m −2 s −1 under a light:dark regime of 12:12. The EASW medium was exchanged every second day. The flask size was gradually increased from 1 L (4 days) to 2 L (2 days) and 5 L (2 days) depending on the growth of U. meridionalis during pre-cultivation ( Supplementary Fig. S7). All cultivation were performed at 25 °C. The U. meridionalis were mixed using a stirring bar. Figure 6A,B show the schedule for 13 C-labelling cultivation and sampling of U. meridionalis. First, U. meridionalis, in a vegetative growth stage, was pre-cultured using non-labelled EASW under a 12:12 light:dark regime. U. meridionalis was transferred into a 13 C-labelled EASW medium containing 13 C-NaHCO 3 (99%, Cambridge Isotope Laboratories Inc., MA, USA). as the sole carbon source (Supplementary  Table S1). 13 C-labelled U. meridionalis was collected every 4 h under light (50 μmol m −2 s −1 ) and dark periods ( Fig. 6A-C). The medium was homogeneously mixed by a stirring bar. The apparent amount of O 2 generation was monitored from the amount of the gas emitted from the cultivation tank. After collecting the 13 C-labelled U. meridionalis, their fresh weight was measured, then immediately frozen, lyophilised, and powdered to measure the dry weight (Fig. 6D). The total amount of incorporated 13 C was measured by IR-MS (RMI-2, Hitachi, Ltd., Tokyo, Japan). The approximate elemental analysis was performed using the CHNS analyser (FlashEA 1112; Thermo Fisher Scientific Inc., MA, USA). Dried samples were further used for NDP sugar and polysaccharide analyses (Fig. 6D).  34 . Namely, the dried powders of U. meridionalis were extracted three times with 75% aqueous ethanol, and then the extracts were dried using an evaporator. The dried extracts were suspended in aqueous NaHCO 3 (10 mM) solution. The solubilised fractions were loaded on the ENVI-Carb SPE cartridge (Sigma-Aldrich Co. LLC., MO, USA). The SPE cartridge was successively washed with ultrapure water, TEAA buffer (50 mM, pH 7, Wako Pure Chemical Industries Ltd., Kyoto, Japan) solution, and 25% aqueous acetonitrile solution. The cartridge was finally eluted with 25% aqueous acetonitrile containing TEAA. The extracted NDP-sugars were analysed by the Bruker solariX FT-ICR-MS (Bruker Co., MA, USA) operated with ESI in negative ion mode with 20 μL min −1 of sample infusion.

NDP sugar analysis by FT
Ulvan analysis by semiquantitative 13 C-NMR and GC-MS. Ulvan was extracted from the residue following aqueous ethanol (75%) extraction with hot water (90 °C, 1 h × 3). Ulvan was recovered by ethanol precipitation and then freeze-dried. The 13 C uptake in ulvan was measured by IR-MS. Semiquantitative 13 C-NMR of 13 C-labelled ulvan was performed by the inverse-gated decoupling method using JEOL NMR ECA600 (150 MHz) 35,36 . Ulvan was re-solubilised in D 2 O-DMSO-d 6 (92:8) mixed solvent. NMR spectra were recorded at 40 °C with a repetition time of 10.69206 s. The chemical shift intensities were normalised to the DMSO-d 6 internal standard. The monosaccharide composition of ulvan was analysed by GC-MS (Agilent GC 8690 N and MS 5973; Agilent Technologies, CA, USA). Ulvan was hydrolysed into monosaccharides by aqueous trifluoro www.nature.com/scientificreports/ acetate (TFA; 2 M, 121 °C, 1 h). After removing TFA, the monosaccharides were derivatised to alditol acetate and analysed by GC-MS with a column of CBP-1 (Shimadzu GLC Ltd., Kyoto, Japan) operated at 140-160 °C (2 °C min −1 ).

Data availability
All relevant data are included in the main text and Supplementary Information.