Hormesis effects of phosphorus on the viability of Chlorella regularis cells under nitrogen limitation

Microalgae strain, cultivation, and protocols

As a typical microalga, C. regularis var. minima (FACHB-729) was used in this study, which was purchased from the Freshwater Algae Culture Collection of the Institute of Hydrobiology in Wuhan, China. Prior to the experiments, C. regularis was purified via the streaking plating method and preserved on agar jelly at 4 °C. Before the experiment, C. regularis was activated using classic BG11 medium under photoautotrophic condition [24]. As an inoculum of the experiment, the activated C. regularis cells were collected at the logarithmic phase (20 days of phototrophic cultivation) via centrifuging (6000×g for 5 min at 4 °C), and then washed and re-suspended three times with 0.9% sterile physiological saline. After the above procedures, C. regularis was inoculated into the modified BG11 media, which resulted in an initial cell density of 1.5 × 10⁷ cell mL⁻¹ for further heterotrophic at 28 °C, 160 rpm, in the dark (ZWY-240, Zhicheng Shanghai, China). During the experiments, all sampling operations were accomplished in a clean bench (SW-CJ-1FD, Airtech, Suzhou, China).

Real wastewater with high P concentration also has high levels of organics [21, 25]. Accordingly, glucose was selected to simulate organic carbon for microalgae heterotrophic cultivation [11, 26]. In this study, 10 g L⁻¹ glucose (sterilized using sterile 0.22 mm filter) was mixed to the BG11 mediums to create heterotrophic cultivation. Nitrogen limitation was controlled by adding 300 mg L⁻¹ NaNO₃. The form of nitrogen was identical to the BG11 medium; however, its content was only 20% of the classic BG11 [4]; therefore, this condition was named nitrogen limitation (N_lim). Furthermore, a series of phosphorus (PO₄³⁻–P) levels were proposed to elaborate the effects of P concentration (mg L⁻¹), 5.4, 25, 45, 150, and 250 mg L⁻¹. The 5.4 mg-P L⁻¹ was equal to the P concentration of the classic BG11 medium; therefore, it was used as control protocol (N_limP5.4). The others were called N_limP25, N_limP45, N_limP150, and N_limP250 in the following. Additionally, MgSO₄, NaCO₃, CaCl₂, citric acid, ammonium ferric citrate, Na₂EDTA, H₃BO₃, MnCl₂, ZnSO₄, CuSO₄, CoCl₂, and NaMoO₄ were also supplemented as nutrients and trace elements according to a previous study [11]. The medium was autoclaved at 121 °C for 30 min, and the final pH was about 7.1 adjusted with sterile HCl and NaOH solution and no precipitation appeared in the medium.

Growth profiles and nutrient consumptions

Microalgal growth was evaluated via cell density [10], obtained with an optical microscope (BX53, Olympus, Japan) coupled with a hemocytometer. The results were determined based on the averages of at least three repetitions. The cell growth rate was calculated according to Eq. (1):

$${\text{Cell growth rate }}\left( {{\text{cell }}\left( {{\text{mL}}\;{\text{d}}} \right)^{ - 1} } \right) = (X_{ 2} - X_{ 1} )/\left( {t_{ 2} - t_{ 1} } \right)$$

(1)

where X₁ and X₂ (cell mL⁻¹) represent the cell density at times t₁ and t₂, respectively; t₁ and t₂ represent the initial and final point within the linear portion of Fig. 1a.

Organics, phosphorus, and nitrogen consumptions were of particular concern, which were represented by the changing of chemical oxygen demand (COD), dissolved inorganic nitrogen (DIN), and dissolved inorganic phosphorus (DIP) in this work. These were determined via spectrophotometry using a water quality analyzer and the provided reagent kits (5B-3C V8, Lian Hua Technology, China) according to the manufacturer’s instructions. Prior to measurements, the samples were filtered with 0.45 µm cellulose acetate membranes to remove any suspended residues and biomass.

Intracellular storage products: lipid, protein, and starch

To measure protein and starch, the cells were pretreated via ultrasound at 1500 W, − 4 °C for 30 min to break the cell-walls. Then, the solution was filtered with a 0.45 µm cellulose acetate membrane to remove fragments. The filtrate was ready for intracellular protein and starch evaluation. A folin-phenol assay kit (Dingguo Changsheng Biotechnology, Beijing, China) was employed for protein analysis [29]. For the hydrolysis of starch, 30% perchloric acid was added to the filtrate and stirred at 25 °C for 15 min, the extracts were analyzed with the anthrone method [30].

Mitochondrial activity assay

The mitochondrial activity was determined with the 3-(4,5-dimethyl-2-thiazolyl)-2,5-diphenyl-2H-tetrazolium bromide (MTT) assay kit (Beyotime, Beijing, China) according to a previous publication [31]. Briefly, the cells were seeded to a 96-well plate (5 × 10⁵ cells mL⁻¹) and were incubated with fresh air ventilated (5% CO₂), at 37 °C. After 48 h, 10 µL MTT (5 mg mL⁻¹) was added to the wells for another 4 h of incubation. The absorbance of the culture was measured at 570 nm using a microplate reader (RT-6000, Rayto, China).

Morphology and ultrastructure of cells

Microalgal cell morphology was observed via optical microscopy (BX53, Olympus, Japan) and scanning electron microscopes (SEM) (XL-30 FE-SEM, FEI Co., USA). The ultrastructure of the cell was observed with a transmission electron microscope (TEM) equipped with an energy dispersive X-ray spectrometer (EDX) (JEM 1200EX, JEOL, Japan). Prior to SEM and TEM observation, the cells were pretreated by a series of multiple fixative and dehydrated procedures [32, 33]. Before TEM–EDX analysis, the dehydrated cells should be embedded in epoxy resin (Epon812, Shell Chemical, USA) for cutting (Microtome, Leica UCT, Germany) and ultrathin sections were further stained with uranyl acetate [34].

³¹P nuclear magnetic resonance spectroscopy (NMR)

The P-storage compounds in microalgal cells were extracted via the ice-cold HClO₄ method [14, 35], and then 10% D₂O was added to provide a field-frequency lock. The ³¹P NMR spectra were acquired at 400 MHz with a 5 mm probe using an NMR spectrometer (Vnmrs-300, Varian Co., USA). The acquisition time was 16 h at room temperature. MestReNova software 11 was used to analyze the peaks in the spectra, and the P compounds were identified by their chemical shifts based on the previously published spectra [35–37].

X-ray photoelectron spectroscopy (XPS)

The binding energies of N1s and P2p were analyzed to investigate the character of the intracellular compounds via XPS (ESCALAB 250, Thermo, USA). The sample was the lyophilized powder of microalgal cells, which was obtained via the above-mentioned freeze-drying method. The beam source type was Al K Alpha, and it was operating at 250 W, a voltage of 15 kV, a current of 15 mA, an energy of 1486.71 eV; the sampling spot size was 500 µm in diameter and an electron takeoff angle of 50° was used. The binding energy was calibrated with reference to C1s at 284.8 eV in the spectra analysis with XPS Peak 4.1 software [38].

Statistical analysis

P concentration effects were investigated based on triplicate cultivations, and then the statistically significances were analyzed using one-way ANOVA method. The mean values were compared using a least significant difference (LSD) test. Differences were significant when P < 0.05. All statistical analyses were performed using the SPSS software package (IBM SPSS statistic 20.0).

Effect of P concentrations on C. regularis growth

The cell density of C. regularis was ~ 4.70 × 10⁷ cell mL⁻¹ at the stationary stage when P supply was regular, N_limP5.4 (refers to the classic BG11), see Fig. 1a. The cell density was about 76% of that in NP5.4 [18], indicating that nitrogen limitation inhibited cell growth. Microalgal growth was stimulated when P was slightly excessive, i.e., the cell density increased by 10.2% for N_limP45 (P < 0.05). A considerable number of previous studies reported that such a stimulation was caused by storage of excess P, under either autotrophic or heterotrophic cultivations [4, 14, 39, 40], where nitrogen was either limited or unlimited [41]. Excessive P induced improvements on supplying phospholipid, genetic materials, and energy for cell division [42, 43]. However, further increasing of the P supply could result in microalgal growth inhibition. The cell density decreased for N_limP150 and decreased as much as 38.8% for N_limP250, compared to the control (N_limP5.4) (P < 0.05). In addition, the growth rate in Fig. 1b intuitively shows that the effects of P depended on its concentration (P < 0.05). These results indicate that C. regularis growth was promoted when P was at an appropriate level; however, a large excess of P had a negative impact on the cells. This is a typical manifestation of hormesis [44, 45].

Glucose, nitrogen, and phosphorus uptake profiles were represented by COD, DIN, and DIP changing versus time, respectively, see Additional file 1: Fig. S1 and Fig. 1c. The organic and nutrient consumptions were consistent with the growth profiles: stimulated by low P concentrations (P ≤ 4.49 × 10⁻⁷ mg cell mL⁻¹) and slowing down at high P levels (P ≥ 14.97 × 10⁻⁷ mg cell mL⁻¹). It is worth mentioning that P (PO₄³⁻–P) was completely consumed within 4 days when P was in slight excess (less than 45 mg L⁻¹); in contrast, a large amount of P (PO₄³⁻–P) remained for both N_limP150 and N_limP250 (Fig. 1c). These results implied that the capability of P storage reached a threshold, above which, P may cause microalgal cell damage.

Effects of P concentration on lipid productivity

It seemed that P supply affected the carbon flow of the microalgal glucose metabolism. The lipid content enhanced by 22.9% for slightly excessive P (N_limP45) (P < 0.05). However, lipid synthesis was significantly suppressed when C. regularis cells were inhibited by a large excess of P (N_limP250), whereas the lipid content decreased by 15.9% in comparison to the control (N_limP5.4) (P < 0.05). For the cell density (Fig. 1a), the highest lipid productivity of 354.38 mg (L d)⁻¹ was achieved for N_limP45, which increased by 39.3% in comparison to the control (P < 0.05). However, the lipid productivity of C. regularis for N_limP250 was only 47.3% of N_limP45 and decreased by 34.2% in comparison to control (P < 0.05). Accordingly, the intracellular protein and starch contents both positively correlated with P supply, as shown in Fig. 2a.

It has previously been reported that P enhances both microalgal lipid synthesis and production. P-replete favored the enzymes of the up-regulation of the lipid synthesis pathway [40] and could enhance the biomass production by activating phosphofructokinase and pyruvate kinase related glycolysis [46]. Here, for the first time, the toxicity of a large excess of P was addressed in microalgae. Excessive P dosage does not always increase the lipid production of microalgae under nitrogen limitation.

Hormesis mechanism

Effect of P concentration on cell viability

Mitochondrial activity was used to study the microbial viability [31, 47], since more than 95% energy is generated in mitochondria, which are the “powerhouses” of the cell [48]. In this work, the trend of mitochondrial activity again supported P hormesis deduction, as shown in Fig. 3. Mitochondrial activity was stimulated and consequently increased by 25.0% (P < 0.05), when P was only slightly excessive (N_limP45). However, it decreased as much as 71.3% for N_limP250 (P < 0.05), indicating dysfunction of mitochondria due to the toxicity of the large excess of P.

Cell morphology and ultrastructure

Cell morphology and ultrastructure further corroborated the hormesis mechanism of P. C. regularis cells of N_limP45 had a round shape, integrated structure, and intact organelles (Fig. 4a–c). All of these characteristics were identical to control cells. In contrast, the cells became folded for N_limP250, and several cells enlarged, reaching up to two- to three-fold of the size of the control (Fig. 4d, e). TEM images further showed that the cell walls thinned and even detached. Such damage of the plasma membrane and the following plasmolysis caused the described cell surface folding [49]. In addition, several organelles, such as mitochondria, were disordered when P was in large excess (Fig. 4f). These signs indicate targeted P poisoning of microalgal cells.

Intracellular P distribution and the role of polyphosphates

P distributions in C. regularis cells were different between stimulated (N_limP45) and inhibited (N_limP250) cell growth. P was mainly found in the intracellular region for N_limP45 (4.13% in weight), while it was located at the cell periphery for N_limP250 (5.4% in weight), see Fig. 5. P in C. regularis formed poly-Ps (small black spots in TEM) in the cells for N_limP45, which were at the resonances in − 5 to − 7 and − 17 to − 22 ppm [35, 36] in the ³¹P NMR spectra (Fig. 5). The poly-Ps also convinced by the peaks at 129.4 eV and 137.1 eV [50] in P 2p XPS spectra (Fig. 5). Poly-P served as an energy pool to stimulate cell division and metabolism [37].

However, the forms of poly-P became different when the excess poly-P was stored for C. regularis (N_limP250). The poly-Ps located near the cell periphery, which likely damaged both plasma membrane and cell wall (Fig. 5). In addition, mitochondria were disordered, as the peaks (− 11.5 ppm) of ATP and NADH [35] almost disappeared in ³¹P NMR spectra for N_limP250 (Fig. 5). This indicated that excess poly-Ps damaged the energy production process, which was consistent with the results on growth inhibition and mitochondrial activity decrement. Consequently, the excess poly-P accumulation disintegrated the thylakoid membranes and resulted the cell lysis and cell death at the end [51]. Moreover, excess poly-P was able to form dinucleoside polyphosphates with similar structure to ATP and other essential mononucleotides; therefore, it was also toxic for cells and caused DNA damage in the intra-S phase [52].

Excess poly-P could bind to intracellular components and inhibit cell viability [18]. There were several new peaks at − 7.7 ppm and − 19.8 ppm in the ³¹P NMR spectra, which indicated that the new form of P appeared for N_limP250 (Fig. 5). This was also confirmed by the peak at higher binding energy of 139.8 eV in the P 2p XPS spectra (Fig. 5), which represented the binding compounds of poly-P to intracellular components [18]. Furthermore, a new peak of protonated amide-N in protein at 402.3 eV [53, 54] appeared for N_limP250 (Fig. 6). It seems that the excess poly-P bonded to the proteins and resulted protonated amide-N, which affected the binding of the protein–ligand [55]. Thus, a major constraint in protein breakage and large cytoplasmic crevices appeared, followed by interfered interactions of protein subunits [55]. The changes on channel and transporter proteins induced the across membrane process of ions and disordered small molecules [56]. In this way, membrane permeability was damaged and membrane proteins were disrupted [57].