Introduction

Phosphorus (P) is an essential component for all life on Earth, and its function cannot be compensated by other nutrients. The majority of phosphates in nature exist as calcium phosphate apatite, and their supply solely depends on weathering and the dissolution of this ore, which has a very low solubility. The annual production of phosphate is approximately 40 million t by P2O5 conversion, which comes from roughly 140 million t of phosphate rock (Steen 1998). Approximately 82% of P ores are currently used in fertilizers and another 5% for livestock feed, with the remaining used for detergents and metal processing (Edixhoven et al. 2014). Currently, there are different views on global phosphate rock reserves (Steen 1998; Van Vuuren et al. 2010). In the worst-case scenario, it is estimated that all economically mined phosphate rock could be depleted by the 2060s if there is a 3% annual increase in the consumption of phosphate fertilizer (Steen 1998). Alternatively, about 40–60% of the current resource base will be extracted by 2100 (Van Vuuren et al. 2010). Thus, it is important to develop methods to recycle and reuse phosphorus from waste and wastewater within the next 50–80 years to reduce the current consumption of phosphorus by half and avoid these worst-case scenarios (Steen 1998).

The forms of P include inorganic and organic. Inorganic P is widely used in agriculture and other industries. Inorganic P compounds include orthophosphate (PO4-P), its deoxidized forms, such as phosphite (PO3-P), hypophosphite (PO2-P), phosphine (PH3-P), and polyphosphate, which is a polymer composed of pyrophosphate and tripolyphosphate. Among inorganic P compounds, phosphates, including orthophosphate and polyphosphate, are essential nutrients for plants, algae, and microorganisms. Algae can only use phosphate forms for their growth, and they accumulate P in their cells in the polyphosphate form. The majority of P ores are currently used in the phosphate form in the agriculture industry and in detergents. This results in a large release of phosphates into domestic and agricultural sewage and activated sludge into natural aquatic environments. Much attention has been given to measuring phosphate levels and monitoring water quality, survey, and control of algal phytoplankton dynamics. In addition, algal accumulation of phosphates has been extensively researched to evaluate their sustainable use of phosphates (De-Bashan and Bashan 2004).

Phosphite and hypophosphite components are used in several industries. Particularly, hypophosphite is used in plating, and phosphite oxidized from hypophosphite after plating is discharged as industrial waste. The phosphite that is discharged may be metabolized by some bacteria, including cyanobacteria that evolved mechanisms to acquire phosphite as a P source when phosphate is unavailable in natural habitats (Quinn et al. 2007; White and Metcalf 2007; Martinez et al. 2012; Stone and White 2012; Villarreal-Chiu et al. 2012; McGrath et al. 2013; Karl 2014; Horsman and Zechel 2017). In these bacteria, the phosphite oxidation enzyme NAD-dependent phosphite dehydrogenase (PtxD) catalyzes the oxidation of phosphite to orthophosphate and produces NADH from NAD+ (Metcalf and Wanner 1991; Imazu et al. 1998; Metcalf and Wolfe 1998; Costas et al. 2001; White and Metcalf 2004; Yang and Metcalf 2004; Wilson and Metcalf 2005; Martinez et al. 2012). In terrestrial plants, it was suggested that soil or the foliar application of phosphite could replace phosphate as a P source in avocado (Lovatt 1990). Phosphite products are sold in US and European markets as fertilizers, fungicides, and/or defense stimulators (Leymonie 2007). However, their fertilizing effects have been questioned for the following reasons: (1) phosphite has never shown any beneficial effect on the growth of healthy plants (Thao and Yamakawa 2009; Thao et al. 2009); (2) deleterious effects of phosphite were shown in phosphate-starved plants (McDonald et al. 2001; Ratjen and Gerendas 2009; Thao and Yamakawa 2009; Thao et al. 2009; Avila et al. 2012; Berkowitz et al. 2013; Manna et al. 2015); (3) phosphite inhibited phosphate uptake in a competitive manner (Danova-Alt et al. 2008); (4) phosphite functions as a fungicide and herbicide to reduce plant diseases, resulting in positive plant responses to phosphite (McDonald et al. 2001; Danova-Alt et al. 2008; Thao et al. 2009); and (5) although phosphite is presumed to be utilized as a phosphate fertilizer by soil microorganisms (McDonald et al. 2001), the indirect provision of phosphite by microbial oxidation is not an effective means as a P nutrition source compared with phosphate fertilizer (Thao and Yamakawa 2009). Thus, there is no convincing evidence for phosphite utilization in wild plants.

The utilization of phosphate by algae is not well understood. In Ulva lactuca, the growth rates of both phosphate-starved and phosphate-satisfied cells were reduced in the presence of 2 mM of phosphite, and it was concluded that U. lactuca were unable to metabolize phosphite (Lee et al. 2005). In addition, Chlamydomonas reinhardtii, Botryococcus braunii, and Ettlia oleoabundans were unable to use phosphite as a P nutrition source (Loera-Quezada et al. 2015). The high concentration of phosphite did not show any deleterious effect for C. reinhardtii in spite of growth suppression, because C. reinhardtii cells survived for a long time in the presence of high phosphite levels and recovered the capacity for cell division after their transfer to an orthophosphate medium. These reports strongly support the central concept that eukaryotic microalgae cannot use phosphite and can only directly use phosphate as a sole P source.

To the best of our knowledge, there is no report on the phosphite utilization of wild photosynthetic eukaryotes, including eukaryotic microalgae. Therefore, except for the possible utilization by some bacteria when phosphate is unavailable, the fate of the phosphite that is discharged as waste is largely unknown.

Chlorella species are a major algal model in plant biochemistry and physiology (Safi et al. 2014). Their high growth potential and production of useful compounds have contributed to the fields of algae industrial and environmental technologies, including carbon capture and utilization. C. vulgaris is a photosynthetic and mixotrophic species and grows in a variety of environments. Many studies have demonstrated the high potential of C. vulgaris for the treatment of different types of wastewater, including municipal, agricultural, and industrial. As C. vulgaris accumulates a considerable amount of phosphorus in its cells as polyphosphate, C. vulgaris is expected to recycle and reuse phosphate from wastewater. However, previous studies have focused on total phosphorus and phosphate and no study has investigated the phosphite utilization of this species (Safi et al. 2014). In this study, we investigated the availability of phosphite in C. vulgaris and three other green algae species, such as Coccomyxa subellipsoidea, Scenedesmus obliquus, and B. braunii. The results of this study clearly demonstrate that C. vulgaris and C. subellipsoidea are capable of directly metabolizing phosphite. Significantly, our study is the first to detail such dynamics in photosynthetic eukaryotes.

Materials and methods

Algal strains

Axenic strains of Chlorella vulgaris (NIES-2170), Coccomyxa subellipsoidea (NIES-2166), and Scenedesmus obliquus (NIES-2280) obtained from Microbial Culture Collection, National Institute for Environmental Studies, and an axenic strain of Botryococcus braunii (BOT-22), maintained in the Faculty of Life and Environmental Sciences and Algal Biomass and Energy System R&D Center (ABES), University of Tsukuba, were used in this study.

Preparation of medium

Chlorella vulgaris, C. subellipsoidea, and S. obliquus were cultured in P-free C medium, and B. braunii was cultured in AF-6 medium. The P-free C medium contained Ca(NO3)2.‧4H2O 150 mg L−1, KNO3 100 mg L−1, MgSO4.7H2O 40 mg L−1, vitamin B12 0.1 μg L−1, biotin 0.1 μg L−1, thiamine HCl 10 μg L−1, PIV metal solution 3 mL L−1, and tris (hydroxymethyl) aminomethane 500 mg L−1. The PIV metal solution contained Na2EDTA.2H2O 1000 mg L−1, FeCl3.‧6H2O 196 mg L−1, MnCl2.4H2O 36 mg L−1, ZnSO4.7H2O 22 mg L−1, CoCl2.6H2O 4 mg L−1, and Na2MoO4.2H2O 2.5 mg L−1. The pH was adjusted to 7.5. The P-free AF-6 medium contained NaNO3 280 mg L−1, NH4NO3 44 mg L−1, MgSO4.7H2O 60 mg L−1, CaCl2.2H2O 20 mg L−1, KNO3 10 mg L−1, Fe-citrate 4 mg L−1, citric acid 4 mg L−1, biotin 2 μg L−1, thiamine HCl 10 μg L−1, vitamin B6 1 μg L−1, vitamin B12 1 μg L−1, PIV metals 5 mL L−1, and MES 400 mg L−1. The pH was adjusted to 6.6. The original C medium contained 7.2 mg-phosphorus L−1 (Na2-glycerophosphate‧5H20 50 mg L−1), and the original AF-6 medium contained 3.2 mg-P L−1 (KH2PO4 10 mg L−1 and K2HPO4 5 mg L−1), but the P-free C and AF-6 medium were prepared in this study. A bacteria-check medium containing 1 g yeast extract and 2 g tryptone in 1 L of C medium or AF-6 medium was used to check bacterial contamination once a week throughout the experiment. The axenicity of the culture was determined by observation under a microscope.

Stock solutions of orthophosphate (NaH2PO4.2H2O 36.1 g L−1) with a P concentration that was 1000 times higher than the original C medium were prepared, and the pH was adjusted to 7.5. Stock solutions of phosphite (Na2HPO3.5H2O 50 or 500 g L−1) with a P concentration that was 1000 or 10,000 times higher than the original C medium were prepared, and the pH was adjusted to 7.5. Similarly, stock solutions of orthophosphate (16.1 g L−1) and phosphite (22.3 g L−1) with a P concentration that was 1000 times higher than the original AF-6 medium were prepared, and the pH was adjusted to 6.6. All P-component stock solutions were aseptically filter sterilized through a 0.2 μm cellulose acetate filter sterilized by ethylene oxide (DISMIC-25CS, Advantech Japan, Co., Ltd., Japan) on a clean bench. One milliliter of orthophosphate and phosphite stock solutions (1000 times higher) was added to the P-free C medium and P-free AF-6 medium, respectively, to obtain C medium containing 7.2 mg-P L−1 and AF-6 medium containing 3.2 mg-P L−1. Additionally, 0.05, 0.1, 0.5, 1, and 10 mL of the phosphite stock solution (10,000 times higher) were added to the P-free C medium, to generate C medium containing 3.6, 7.2, 72, and 720 mg-P L−1, respectively. These media were then sterilized through a 0.2-μm cellulose acetate filter. Media containing orthophosphate were named “PO4-P,” and media containing phosphite were named “PO3-P.” Autoclave-sterilized media were named “AS,” and filtered-sterilized media were named “FS.”

Preparation of P-starved algae cells

Algal cells were pre-cultured in PO4-P C or AF-6 medium until they reached the stationary phase. The pre-cultured cells were inoculated into P-free C or AF-6 medium and cultured until they reached the stationary phase, indicating the P-starved state of algal cells. These intracellular P-starved cells were washed several times with sterilized water by centrifugation to washout P from the cells. These P-starved cells were used for this study.

Measurements of growth and phosphoric acid concentrations

Growth was measured by determining the optical density at 680 nm (OD 680 nm) using a UV spectrophotometer (UV1800, Shimadzu, Japan). The concentrations of PO4-P and PO3-P in the media were measured using a capillary electrophoresis apparatus (G1600A, Agilent Technologies Japan, Ltd., Japan) before and after cultivation. Reagent grades of sodium orthophosphate and sodium phosphite were used as standards.

C. vulgaris NIES-2170 culture experiments

C medium was used to culture C. vulgaris. P-starved cells of C. vulgaris were grown in 500-mL Erlenmeyer flasks with 250 mL of PO4-P and PO3-P added C media and P-free C medium under continuous white fluorescent light (53–69 μmol photons m−2 s−1) or heterotrophically grown in 500-mL Erlenmeyer flasks with 250 mL of 20 g L−1 glucose-added to PO4-P and PO3-P C media and P-free C medium under dark conditions. To observe the effects of PO3-P concentrations on the growth of C. vulgaris, P-starved cells were grown in 500-mL Erlenmeyer flasks with 250 mL of PO3-P C media with 3.6, 7.2, 36, 72, and 720 mg-P L−1. P-free and PO4-P C media were used as a negative control and positive control, respectively. All of these experiments were performed in triplicate at 25 °C ± 1 °C on a shaker (NR-150, Taitec Corp, Japan) at 120 rpm. During these culture experiments, the amount of evaporated water was replenished regularly by adding sterile water.

Coccomyxa subellipsoidea NIES-2160, S. obliquus NIES-2280, and B. braunii BOT-22 culture experiments

C medium was used for C. subellipsoidea and S. obliquus, and AF-6 medium was used for B. braunii. P-starved cells of C. subellipsoidea and S. obliquus strains were grown in 500-mL Erlenmeyer flasks with 250 mL of PO4-P and PO3-P C media and P-free C medium on a shaker (NR-150, Taitec Corp, Japan) at 120 rpm. P-starved cells of B. braunii were grown in 90-mL tubes with 50 mL of PO4-P and PO3-P AF-6 media and P-free AF-6 medium and aerated with sterile air containing 1% (v/v) CO2 through a polytetrafluoroethylene filter (0.2 μm, Tokyo Roshi Kaisha, Ltd., Japan). All of these experiments were performed in triplicate at 25 °C ± 1 °C under continuous white fluorescent light (53–69 μmol photons m−2 s−1). During these culture experiments, the amount of evaporated water was replenished regularly by adding sterile water.

Statistical analysis

Results are expressed as the mean ± standard deviation (SD) of three replicates. Significant differences were determined by an analysis of variance (ANOVA) of the measured OD values. The effects of phosphoric acid concentrations before and after cultivation were also analyzed by ANOVA. Differences were considered significant when p < 0.05.

Results

Stability of orthophosphate and phosphite in medium

Orthophosphate and phosphite stabilities in algae-free media were analyzed for 180 days (6 months) by determining the change in orthophosphate and phosphite concentrations every several days in the first month (Fig. 1a and b) and every month for the next 5 months (Fig. 1c). We found that orthophosphate and phosphite concentrations did not change in the algae-free media (Fig. 1c). In addition, phosphoric acid was the only form found in each medium, and the levels of both orthophosphate and phosphite did not differ between AS and FS.

Fig. 1
figure 1

Orthophosphate (PO4-P) and phosphite (PO3-P) stabilities over 6 months in algae-free media. The phosphorus concentration of the PO4-P and PO3-P media was adjusted to 7.2 mg-P L−1. a PO4-P concentration measured every few days for the first month. b PO3-P concentrations measured every few days for the first month. c The PO4-P and PO3-P concentrations determined every month for 180 days (6 months). Data represent the mean ± standard deviation (SD) of three replicates. All bars in c are not statistically different from each other. AS autoclave sterilization, FS filter sterilization

Full size image

Growth and phosphite utilization of C. vulgaris NIES-2170

As mentioned above, the oxidation of phosphite to orthophosphate by the AS medium was not observed, but the possibility of oxidation occurrence was considered. Compared with autoclave sterilization, filter sterilization does not include treatments that cause the oxidation of phosphite to orthophosphate. The growth and phosphite utilization of this strain were investigated in the AS and FS media to confirm that there was no difference between the two media in terms of cultivation. In the P-free medium, the growth of C. vulgaris was not observed during the three-month culture period (Fig. 2a). However, this strain exhibited high growth rates and yields in the PO4-P media and reached the early stationary phase 20 days after inoculation. The algae aliquot was filtrated 90 days prior to the death phase, and orthophosphate concentrations in the filtrates were analyzed. However, no trace of orthophosphates was detected (Fig. 2b). In the PO3-P media, this strain showed reduced growth rates and yields than those in the PO4-P media. Specifically, growth was observed after a lag phase of 20 days and reached an OD680 of 0.3–0.5 after 180 days. The algae aliquot was filtered at 180 days and phosphite concentrations in the filtrates were analyzed. A 32–35% reduction (hereafter called the “utilization rate”) of phosphite in the media was detected (Fig. 2b). There were no significant differences in the growth and phosphate utilization rates between AS and FS media (Fig. 2a and b), and thus, FS media were used for all subsequent experiments. After 180 days of culture in the FS PO3-P medium, the strain was inoculated again into fresh phosphite medium to confirm the sustainability of growth. Similar to our previous results, growth was observed without a lag phase (Fig. 3a), and a 38% utilization rate of phosphite was detected (Fig. 3b).

Fig. 2
figure 2

Effects of sterilization methods. a Growth of C. vulgaris in P-free, PO4-P, and PO3-P media. The phosphorus concentration in PO4-P and PO3-P media was 7.2 mg-P L−1. Cell growth was determined during the 90 days of culture in PO4-P media and 180 days of culture in PO3-P media by measuring the optical density (OD). b Changes in P concentration in PO4-P and PO3-P media before and after the cultivation of C. vulgaris. After 90 days of culture, 100% of PO4-P was consumed. Data represent the mean ± SD of three replicates. Asterisks on bars indicate statistical significance, **p < 0.001, *< 0.005, ns not significant (p > 0.05). AS autoclave sterilization, FS filtered-sterilization

Full size image
Fig. 3
figure 3

Chlorella vulgaris in P-free, PO4-P, and PO3-P media. a Phototrophic growth after the inoculation of cells from the first cultures shown in Fig. 2. The phosphorus concentration in PO4-P and PO3-P media was 7.2 mg-P L−1. Cell growth was determined during the 80 days of culture in PO4-P medium and 180 days of culture in PO3-P medium by measuring the optical density (OD) at 680 nm. b Changes of P concentration in PO4-P and PO3-P media before and after the long-term cultivation of C. vulgaris. After 80 days in culture, 100% of PO4-P was consumed. c Heterotrophic growth of C. vulgaris in P-free, PO4-P, and PO3-P media under dark conditions. The phosphorus concentration in PO4-P and PO3-P media was 7.2 mg-P L−1. Cell growth was determined during 60 days of culture by measuring the optical density (OD) at 680 nm. d Changes in P concentration in PO4-P and PO3-P media before and after the long-term heterotrophic cultivation of C. vulgaris for 60 days. After 60 days in culture, 100% of PO4-P was consumed. Data represent the mean ± SD of three replicates. Asterisks on bars indicate statistical significance, ***p < 0.001, **p < 0.01, *p < 0.05. FS filter sterilization

Full size image

In order to completely exclude the possibility of oxidation due to oxygen species generated by photosynthesis, the growth and utilization of orthophosphate and phosphite of this strain were investigated under heterotrophic conditions for 2 months. The same results were obtained under these conditions (Fig. 3c and d). Specifically, no growth was observed in the P-free medium, whereas abundant growth and complete utilization of orthophosphate occurred in the PO4-P medium. In addition, a lower net growth and a 19% utilization rate of phosphite was observed in the PO3-P medium.

This strain was also observed for growth and utilization of phosphite in different concentrations of phosphite (3.6 mg-P L−1, 7.2 mg-P L−1, 36 mg-P L−1, 72 mg-P L−1, and 720 mg-P L−1) for 180 days. P-free medium and PO4-P medium (7.2 mg-P L−1) were used as a negative control and positive control, respectively. As with our previous experiments, this strain did not show any growth in the P-free medium during the 180 days but showed abundant growth in the PO4-P medium and reached an early stationary phase 20 days after inoculation. In the presence of different concentrations of phosphite, the growth of this strain increased as the phosphite concentration increased with maximum growth at 720 mg-P L−1, exhibiting a growth yield at 180 days that was similar to the PO4-P medium (Fig. 4a). In addition, the utilization rates of phosphite were 7, 39, 12, 9, and 13% in media with 3.6, 7.2, 36, 72, and 720 mg-P L−1, respectively (Fig. 4b).

Fig. 4
figure 4

Chlorella vulgaris at different concentrations of PO3-P: 3.6, 7.2, 36, 72, and 720 mg-P L−1. P-free medium and PO4-P medium were used as a negative control and a positive control, respectively. The phosphorus concentration in PO4-P medium was 7.2 mg-P L−1. a Growth in different PO3-P concentrations. Cell growth was determined during the 70 days of culture in PO4-P medium and 180 days of culture in PO3-P medium by measuring the optical density (OD) at 680 nm. b Changes in P concentration in PO4-P and PO3-P media before and after the long-term cultivation. After 70 days of culture, 100% of PO4-P was consumed. The vertical axis shows the P concentration divided to the left for initial concentrations of 3.6–72 mg-P L−1and right for initial concentrations of 720 mg-P L−1. Data represent the mean ± SD of three replicates. Asterisks on bars indicate statistical significance, ****p < 0.001, ***p < 0.005, **p < 0.01, *p < 0.05. FS filter sterilization

Full size image

Growth and phosphite utilization of other green microalgae

Growth and the utilization of phosphate and phosphite of C. subellipsoidea (NIES-2166), S. obliquus (NIES-2280), and B. braunii (BOT-22) were investigated in FS media. Coccomyxa subellipsoidea showed no growth in P-free medium for 60 days, but abundant growth in PO4-P medium, reaching an early stationary phase (> OD 2.0). The cells cultured in PO3-P medium grew much more slowly, reaching an OD of 0.2 after 2 months (Fig. 5a) and with a 17% utilization rate of phosphite (Fig. 5b). After 60 days in culture, the strain was inoculated again into fresh phosphite medium to confirm growth sustainability. Similar growth was observed (Fig. 5c), with a 26% utilization rate of phosphite (Fig. 5d). However, S. obliquus and B. braunii cells did not grow when cultured for 2 months (56–60 days) in the PO3-P media and P-free media, respectively. In contrast, their growth was abundant in the PO4-P media (Fig. 6a and c). No significant phosphite reduction was observed after 2 months in culture (Fig. 6b and d).

Fig. 5
figure 5

Coccomyxa subellipsoidea in P-free, PO4-P, and PO3-P media. The phosphorus concentration in PO4-P and PO3-P media was 7.2 mg-P L−1. Cell growth was determined during the 54 days in PO4-P medium and 60 days in PO3-P medium by measuring the optical density (OD) at 680 nm. a The first culture experiments using P-starved cells. b Changes in P concentration in the first cultures. c The second culture experiments using stationary phase cells from the first cultures. Data represent the mean ± SD. FS filter sterilization. d Change in P concentration in the second culture. Both b and d show changes in P concentration before and after the long-term cultivation of C. subellipsoidea for 54 days in PO4-P medium and 60 days in PO3-P medium. After 80 days in culture, 100% of PO4-P was consumed. Data represent the mean ± SD of three replicates. Asterisks on bars indicate statistical significance, **p < 0.001, *p < 0.01. FS filter sterilization

Full size image
Fig. 6
figure 6

Scenedesmus obliquus and Botryococcus braunii in P-free, PO4-P, and PO3-P media. The phosphorus concentration in PO4-P and PO3-P media was 7.2 mg-P L−1. a Growth of S. obliquus. Cell growth was determined during the 60 days of culture by measuring the optical density (OD) at 680 nm. b Changes in P concentration in PO4-P and PO3-P media before and after the long-term cultivation of S. obliquus for 60 days. After 60 days in culture, 100% of PO4-P was consumed. c Growth of B. braunii in P-free, PO4-P, and PO3-P media. Phosphorus concentration in PO4-P and PO3-P media was 3.2 mg-P L−1. Cell growth was determined during the 56 days in culture by measuring the optical density (OD) at 680 nm. Data represent the mean ± SD of three replicates. FS filter sterilization. d Changes in P concentration in PO4-P and PO3-P media before and after the long-term cultivation of B. braunii for 56 days. After 6 days of culture, 100% of PO4-P was consumed. Data represent the mean ± SD. Asterisk on bars indicates statistical significance, *p < 0.001, ns not significant (p > 0.05). FS filter sterilization

Full size image

Discussion

Utilization of phosphite

All strains used in this study were axenic; therefore, phosphite utilization due to bacterial oxidation was eliminated. Regarding the C. vulgaris strain NIES-2170, we carefully looked for orthophosphate contamination from the previous culture medium and from polyphosphates accumulated in algal cells. Because P-starved cells did not grow in the P-free medium that was used as the control in all experiments (see Figs. 2a; 3a, c; 4a; 5a, c; 6a, and c), no phosphate contamination occurred in the present experiments. Since it has been reported that the conversion of phosphite to phosphate occurred in the wet oxidation of phosphite at 180 °C, under an O2 pressure of 0.5–5 MPa, and between pH 6.05 and 7.01 (Fujita et al. 2006), we investigated whether phosphite was oxidized to orthophosphate in both AS and FS media within a 6-month period. However, the results clearly demonstrated that orthophosphate and phosphite were not oxidized in AS or FS media (Fig. 1a–c). Supported by the previous report that a large energy investment is necessary to chemically oxidize phosphite to phosphate (Liu et al. 2013), it is likely that phosphite never changed via physicochemical oxidation to orthophosphate under the normal culture conditions used in this study. In addition, when the cells were grown in P-free medium and PO3-P and PO4-P media under heterotrophic conditions, they showed growth and orthophosphate and phosphite utilization rates similar to those observed in photosynthetic conditions (Fig. 3c and d). These findings strongly suggested that phosphite was not oxidized to orthophosphate by oxygen species generated by photosynthesis and/or other unknown factors during the course of our study.

Therefore, our results clearly demonstrated that the C. vulgaris strain NIES-2170 was able to directly use phosphite as a P nutrient source. This resulted in increased growth as the concentration of phosphite increased with the highest utilization rate of 39% in 7.2 mg-P medium.

We also found that the C. subellipsoidea strain NIES-2166 and C. vulgaris strain NIES-2170 were able to use phosphite as a P nutrient source. In contrast, S. obliquus (NIES-2280) and B. braunii (BOT-22) did not exhibit growth and phosphite utilization in the PO3-P medium, whereas their growth and orthophosphate utilization were abundant in the PO4-P medium. Thus, S. obliquus and B. braunii were not able to use phosphite as a P nutrient source. Taken together, C. vulgaris strain NIES-2170 and C. subellipsoidea strain NIES-2166 can utilize phosphite as a P resource. To the best of our knowledge, this is the first study demonstrating phosphite utilization in eukaryotic algae and the first convincing report in photosynthetic eukaryotes.

PtxD, found in some bacteria, uses phosphite as a P source to catalyze the oxidation of phosphite to orthophosphate and reduce NAD+. Transgenic Arabidopsis, tobacco, and green alga C. reinhardtii lines expressing the ptxD gene from Pseudomonas stutzeri WM88 can use phosphite to achieve similar growth to those under orthophosphate conditions (López-Arredondo and Herrere-Estrella 2012; López-Arredondo and Herrera-Estrella 2013; Loera-Quezada et al. 2016). To identify PtxD-related sequences within the genome and transcriptome of C. subellipsoidea presented by Blanc et al. (2012), a BLAST search was conducted using the queries of bacterial PtxD sequences along with functional evidence (Martinez et al. 2012). As a result, several candidate sequences were suggested with identities of 29.4–50.6% (genome) and 30.2–54.2% (transcriptome). However, all candidate sequences had strong homology with eukaryotic D-3-phosphoglycerate dehydrogenase. PtxD-like enzymes were not found in C. subellipsoidea. It would be reasonable to presume some mechanism of intracellular phosphite oxidation in the C. subellipsoidea strain NIES-2166.

Future prospects

It has been suggested that P from (Fe,Ni)3P (schreibersite), which significantly influenced the geochemistry of P on the early Earth, reacts with water to form reduced P compounds, such as phosphite, hypophosphite, and phosphine. These reduced P forms may have been used as sources of P by early life on Earth (Pasek 2008). The existence of PtxD provides direct evidence that the utilization of reduced forms is a mechanism of P utilization in the present ecosystem. Prochlorococcus, a marine cyanobacterium, is a numerically dominant primary producer in the oligotrophic ocean that is able to use phosphite as a sole P source (Martinez et al. 2012). It was suggested that the phosphite utilization of this cyanobacterial strain is mediated by a phosphite dehydrogenase encoded by a ptxD gene cluster similar to that of other phosphite-utilizing heterotrophic bacteria. Phosphite utilization genes are present in diverse marine microbes, and their abundance is higher in species who live in low-P waters. In addition, Bisson et al. (2017) determined the phosphite- and hypophosphate-binding proteins from phosphite and hypophosphite ATP-binding cassette transporters of important globally abundant marine and terrestrial microorganisms. These two types of proteins are the initial encounters with phosphite or hypophosphite for assimilating these potentially important P species.

The present study demonstrated for the first time that the green algae C. vulgaris and C. subellipsoidea utilize phosphite as a P resource for their growth, though the intracellular phosphite oxidation mechanism is not yet clear. These species commonly inhabit lakes, reservoirs, and ponds. As previously mentioned, hypophosphite is currently used as a reducing agent in electroless nickel plating in the automotive, mechanical, electric, electronic, and semiconductor industries. The resulting liquid waste contains mainly inorganic phosphite (the oxidized by-product) at a level of approximately 1840 t-P, which was estimated in 2005 in Japan (Matsubae et al. 2015). In these conditions where phosphate is unavailable, the phosphite that is discharged as waste could be metabolized by these algae as well as phosphite-utilizing bacteria.

In the present study, the phosphite utilization rates of C. vulgaris and C. subellipsoidea were much lower than those of orthophosphate. It is possible that several algal species can use phosphite more efficiently as a P nutrient source given the vast genetic diversity of microalgae (Guiry 2012; Larkum et al. 2012). Identifying the ability of eukaryotic algae as well as heterotrophic bacteria and cyanobacteria, to use phosphite as a P source would increase the awareness that the utilization of the reduced P form by bacteria and microalgae is an important part of the global phosphorus redox cycle, similar to the geochemistry of P on the early Earth. In this context, a future study involving the analysis of the C. subellipsoidea NIES-2166 and the C. vulgaris NIES-2170 genomes is needed to determine phosphite dehydrogenase-related genes that are specific to these strains.

Conclusions

Here we report that C. subellipsoidea and C. vulgaris can utilize phosphite as a P resource, which is a novel finding in eukaryotic algae. In contrast, the other two chlorophyte algae, S. obliquus and B. braunii strains, could not grow in the phosphite medium. This finding could have important implications for the evolution and distribution of the phosphorus redox system among photosynthetic eukaryotes. Furthermore, the knowledge of phosphite utilization by eukaryotic algae will lead to a better understanding of the P cycle in aquatic environments.