ABSTRACT
Diatoms are important phytoplankton and contribute greatly to the primary productivity of marine ecosystems. Despite the ecological significance of diatoms and the importance of programmed cell death (PCD) in the fluctuation of diatom populations, little is known about the molecular mechanisms of PCD triggered by different nutrient stresses. Here we describe the physiological, morphological, biochemical, and molecular changes in response to low levels of nutrients in the ubiquitous diatom Skeletonema marinoi. The levels of gene expression involved in oxidation resistance and PCD strongly increased upon nitrogen (N) or phosphorus (P) starvation. The enzymatic activity of caspase 3-like protein also increased. Differences in mRNA levels and protein activities were observed between the low-N and low-P treatments, suggesting that PCD could have a differential response to different nutrient stresses. When cultures were replete with N or P, the growth inhibition stopped. Meanwhile, the enzymatic activity of caspase 3-like protein and the number of cells with damaged membranes decreased. These results suggest that PCD is an important cell fate decision mechanism in the marine diatom S. marinoi. Our results provide important insight into how diatoms adjust phenotypic and genotypic features of their cell-regulated death programs when stressed by nutrient limitations. Overall, this study could allow us to better understand the molecular mechanism behind the formation and termination of diatom blooms in the marine environment.
IMPORTANCE Our study showed how the ubiquitous diatom S. marinoi responded to different nutrient limitations with PCD in terms of physiological, morphological, biochemical, and molecular characteristics. Some PCD-related genes (PDCD4, GOX, and HSP90) induced by N deficiency were relatively upregulated compared to those induced by P deficiency. In contrast, the expression of the TSG101 gene in S. marinoi showed a clear and constant increase during P limitation compared to N limitation. These findings suggest that PCD is a complex mechanism involving several different proteins. The systematic mRNA level investigations provide new insight into understanding the oxidative stress- and cell death-related functional genes of diatoms involved in the response to nutrient fluctuations (N or P stress) in the marine environment.
INTRODUCTION
Diatoms are eukaryotic unicellular microorganisms that make up one of the major groups of phytoplankton (1). They constitute up to 40% of annual primary production in the ocean (2) and 25% of global carbon fixation (3). Diatoms are widely distributed and affect biogeochemical cycles and global climate (1, 2, 4). Abiotic stresses caused by environmental changes (i.e., nutrient, light, CO2, oxygen, temperature, and osmotic) and biotic stresses (i.e., cell age, viral infection, bacterial growth, allelopathic effect, and grazing by heterotrophs) appear to influence the physiology and survival of diatoms, thereby influencing the flow of photosynthetically fixed organic matter (and associated elements) and primary productivity (5–7). It has been shown that conserved autocatalytic programmed cell death (PCD) in diatoms can be activated by adverse abiotic or biotic stress conditions, such as the limitation of N, P, Si, or Fe, the overproduction of exogenous aldehyde and sterol sulfate, and viral infection (7–18).
Phytoplankton PCD is a form of autocatalytic cell death involving an endogenous biochemical pathway that leads to apoptosis-like changes. The consequence of PCD in phytoplankton can influence the marine food web, vertical sinking flux, microbial loop, and marine biogeochemistry (7–9, 19). The apoptosis-like changes in phytoplankton PCD include chromatin condensation, DNA fragmentation, alterations in the plasma membrane (detected as the externalization of phosphatidylserine [PS] residues), induction of reactive oxygen species (ROS), activation of metacaspases, nuclear and cytoplasmic blebbing, compartmentalization of cell contents into apoptotic bodies, and cellular dissolution (8, 20, 21). These markers can be identified by multiple methods, and not all of these markers occur under each type of PCD (21). PCD and associated pathway genes have an integral grip on cell fate and have shaped the ecological success and evolutionary trajectory of diverse phytoplankton lineages (7).
Skeletonema marinoi is a ubiquitous diatom species that often forms massive blooms in many temperate coastal oceans, such as the Gullmar Fjord, Skagerrak-Kattegat, Baltic Sea, East China Sea, and Northern Adriatic Sea (Mediterranean Sea) (22–26). Previous studies on S. marinoi have focused more on its physiological responses to nutrient limitation (27), secondary metabolite production (28), biotic interactions with grazers (23, 24, 29), potential stress/cell death markers (30, 31), and responses to silicate limitation and sterol sulfates (StS) with PCD (14, 18). For example, silicate limitation could trigger oxidative stress and onset of the PCD process with apoptosis-like morphological changes (18). In addition, a high intracellular level of StS in S. marinoi could lead to ROS accumulation and, ultimately, PCD (14). Inhibition of StS biosynthesis could significantly delay the initiation of this cell death process and maintain active growth for several days (14). N and P stress could significantly inhibit the growth of S. marinoi (or other diatoms) and initiation of PCD (15, 17, 32–34). For example, Lin et al. (15) used a proteomics approach to evaluate the relative protein abundance in Thalassiosira pseudonana under nutrient limitations, and they found that N-limited cells had much stronger stress responses than P-limited cells. However, T. pseudonana and S. marinoi belong to different families (Thalassiosiraceae and Skeletonemataceae, respectively), and they may have different physiological and molecular responses under the same stress. For example, T. oceanica and T. pseudonana had different photoadaptational responses (35). Currently, there are no systematic mRNA-level data to show how the diatom S. marinoi responds to different nutrient stresses with PCD. Do diatoms respond to N and P deficiencies in the same way? Comparative transcriptomic analysis, reverse transcription-quantitative PCR (RT-qPCR), diagnostic biochemistry, and in vivo cell staining identified a suite of ROS- and PCD-related functional genes involved in acclimation to Fe and associated oxidative stress in T. pseudonana and silicate limitation in S. marinoi (13, 18). A molecular study on PCD responses in diatoms could deepen our understanding of oxidative stress- and cell death-related functional genes involved in the nutrient stress response, which occurs frequently toward the end of diatom blooms. It is unclear how diatom cells adjust their phenotypic and genotypic features, particularly in terms of their cell-regulated death process.
To understand how the death- and stress response-related genes of diatoms function during N and P stress, we investigated the expression of 13 genes that are related to oxidative stress and PCD in S. marinoi under nutrient-deficient conditions. These genes code for proteins such as aldehyde dehydrogenase (ALDH), glutathione synthase (GSHS), glycolate oxidase (GOX), heat shock protein 90 (HSP90), programmed cell death 4 (PDCD4), tumor susceptibility gene 101 (TSG101), death-specific protein (DSP), and metacaspases (MCs). Previous studies have demonstrated that these functional genes are related to the process of cell death and stress responses in diatoms (30, 31). Metacaspases are a family of caspase-orthologous proteins that are found in phytoplankton and play a key role in the process of initiation and execution of PCD (36–39). Six metacaspase genes have been identified in the diatom S. marinoi (18). By quantifying transcripts of the PCD-related genes, we hoped to gain a better understanding of the responses of the cosmopolitan bloom-forming diatom S. marinoi to nutrient stresses at the mRNA expression level.
In this study, we examined the morphological, physiological, and biochemical changes and mRNA expression of oxidative stress and cell death-related functional genes of S. marinoi in response to N or P deficiency with the intention of gaining new insight into the regulatory processes that lead to cell death in diatoms under nutrient stress.
RESULTS
Physiological responses to nutrient limitations.To generate a nitrogen-limited condition and maintain sufficient cell yield, the low-nitrogen treatment was started with a low nitrate concentration (88 μM). At day 4, the nitrate concentration in the low-nitrogen treatment was lower than 2 μM, which was considered the nitrogen-limited condition. In contrast, the concentration of nitrate was higher than 700 μM in the control at day 4. The cell density in the control (9.7 × 105 cells ml−1) was significantly higher than that in the low-nitrogen treatment at day 4 (P < 0.05). Therefore, the cells in the low-nitrogen treatment were nitrogen limited at day 4. Growth inhibition was observed in the S. marinoi cultures grown in low-nitrogen or low-phosphorus conditions after day 3 (Fig. 1A). The cell density was significantly higher in the control than the low-nitrogen treatment (P < 0.05 between days 3 to 9) and low-phosphorus treatment (P < 0.01 between days 3 to 11) (Fig. 1A). The maximum cell abundance in the low-nitrogen treatment (7.2 × 105 cells ml−1, day 3) was 65% lower than in the control (1.1 × 106 cells ml−1, day 5), and the maximum cell abundance in the low-phosphorus treatment (6.9 × 105 cells ml−1, day 5) was 63% lower than that in the control (Fig. 1A). After day 5, the density of S. marinoi cells in the control began to decline, while the cell counts in the low-nitrogen treatment and low-phosphorus treatment began to decline after day 3 and day 4, respectively.
Cell density (A) and concentrations of N-NO3− (B) and P-PO43− (C) in media when S. marinoi cells were cultured in control, low-nitrogen, low-phosphorus, resupplied-nitrogen, and resupplied-phosphorus conditions from day 1 to day 11. Statistically significant differences (P < 0.05) at given time points are denoted by asterisks (*; compared to the control) and pound signs (#; compared to the low-nitrogen/phosphorus treatment). The arrow represents the time point (the 9th day) when nutrients were resupplied to the cultures for supporting algal growth. The error bars represent the standard errors from triplicate measurements (mean ± standard deviations [SD]).
The nitrate concentration in the low-nitrogen treatment was nearly depleted at day 4 and then began to increase slightly until day 8 (Fig. 1B). Despite slow consumption, the nitrate level exceeded 500 μM in the low-phosphorus treatment and the control throughout the experiment (Fig. 1B). Furthermore, the consumption of nitrate in the low-phosphorus treatment was slower than in the control within 7 days (Fig. 1B). The same trend was found for the process of phosphate consumption. The concentration of phosphate decreased slowly in the low-nitrogen treatment compared with the control (Fig. 1C). With the addition of nitrate or phosphate at day 9, the cell densities in both low-nutrient treatments were significantly elevated (P < 0.05) (Fig. 1A). The cell density in the resupplied-nitrogen treatment was 2-fold higher than that in the low-nitrogen treatment at day 11 (Fig. 1A). The cell density in the resupplied-phosphorus treatment was 3.3-fold higher than that in the low-phosphorus treatment at day 11 (Fig. 1A). The growth of S. marinoi was inhibited in the N-limited treatment and P-limited treatment, and the growth increased quickly when the nutrients were resupplied.
Internal morphological characteristics.Representative images of the cells grown under different cultures are shown in Fig. 2. Cells displayed organized and compact organelles and intact nuclei, chloroplasts, mitochondria, dictyosomes, and cell membranes at day 2 in all the treatments and the control (Fig. 2). In the control culture, most organelles remained visible at day 4 (Fig. 2). After the 6th day, clear degradation of organelles was observed when the culture entered the stationary and decline phases, even though the cell membranes remained intact (Fig. 2).
Morphological changes in S. marinoi cells grown in control (A), low-nitrogen (B), and low-phosphorus (C) conditions. n, nucleus; c, chloroplast; m, mitochondria; d, dictyosome; v, vacuole. Scale bars, 1 μm.
The internal morphology of S. marinoi cells changed dramatically when the cells were grown under nutrient limitations (Fig. 2). Generally, the percentage of cells that underwent vacuolization was approximately 40% or even more under nutrient limitations. Under nutrient-limited conditions, the formation of vacuoles (vacuolization) was observed in S. marinoi cells. The nutrient-limited cells displayed cytoplasmic vacuolization and swollen mitochondria when the membrane and nuclei were intact. Most of the cells exhibited cytoplasmic vacuolization at day 6, and some appeared to be empty at day 8, although the cell membranes remained intact.
Expression of oxidation resistance- and death-related functional genes.We assessed the expression of S. marinoi metacaspase (SmMC) genes and other death-related functional genes using the same cell extracts. The expression level of target genes was normalized to the expression of the actin gene, and it was calibrated to the expression level observed in the control at day 2 to determine the difference between the treatments. Distinct expression levels were observed between the treatments and control in terms of the extent and pattern of these oxidation resistance- and death-related functional genes (Fig. 3).
The expression levels of 14 functional genes in S. marinoi in different treatments. (A) The expression levels of the PDCD4, DCD, TSG101, DSP, and GOX genes in S. marinoi; and (B) the expression levels of S. marinoi metacaspase genes in response to different media. Statistically significant differences (P < 0.05, compared to the control) at given time points are denoted by asterisks. Error bars represent the standard deviations of experiments performed in triplicate.
The PDCD4 gene was significantly upregulated in the low-nutrient treatments at day 4 compared to day 2 and decreased afterward (P < 0.05) (Fig. 3A). The same trend was found in the control when the culture entered the stationary phase (P < 0.05) (Fig. 3A). The expression of the PDCD4 gene was higher in the low-nutrient treatments than in the control during days 4 through 8 (P < 0.05) (Fig. 3A). Furthermore, the PDCD4 gene was upregulated in the low-nitrogen treatment in relation to the low-phosphorus treatment and control between days 8 and 10 (P < 0.05) (Fig. 3A). In the low-nutrient treatments, the TSG101 and DSP genes were upregulated from day 1 to day 10 and began to decrease at day 11 (Fig. 3A). The TSG101 and DSP genes showed increased expression in the low-nutrient treatments compared to the control between days 8 and 10 (P < 0.05) (Fig. 3A). Meanwhile, the expression level of the TSG101 gene increased in the low-phosphorus treatment compared to the low-nitrogen treatment during days 4 to 8 (P < 0.05) (Fig. 3A). N- or P-limited conditions triggered the upregulation of the TSG101 and DSP genes in S. marinoi, and the expression of the TSG101 gene was highly induced by P deficiency.
The GOX gene expression level was higher in the low-nutrient treatments than in the control between days 2 and 8 (Fig. 3A). Furthermore, the GOX gene was upregulated in the low-nitrogen treatment in relation to the low-phosphorus treatment and control between days 6 and 10 (P < 0.05) (Fig. 3A). The expression of the ALDH gene increased within 4 days in the low-nutrient treatments and control and decreased afterward (Fig. 3A); however, the ALDH gene was more upregulated in the control than in the low-nutrient treatments at day 4 (P < 0.05) (Fig. 3A). This likely occurred because the cells in the low-nutrient treatments experienced nutrient deficiency within 4 days and downregulated the ALDH gene. The HSP90 gene was upregulated in the low-nitrogen treatment compared to the low-phosphorus treatment and control between days 4 and 10 (P < 0.05) (Fig. 3A). The GSHS and HSP90 genes were upregulated from day 2 to day 4 in the low-nutrient treatments and control and then downregulated at day 6 (Fig. 3A).
Transcripts of the SmMC genes increased in the low-nutrient treatments compared with the control at day 4 (P < 0.05) (Fig. 3B), except that the expression of the SmMC5 gene did not show any significant variation during days 1 through 6. Most of these SmMC genes (SmMC1, SmMC2, SmMC4, and SmMC6) in the low-nutrient treatments were highly transcribed at day 4 compared to day 2 (P < 0.05) (Fig. 3B). The SmMC1 gene expression level in the low-nitrogen treatment was higher than those in the low-phosphorus treatment (day 6 and day 10) and control (days 4 to 10) (P < 0.05) (Fig. 3B). Furthermore, the SmMC2 gene was upregulated in the low-nitrogen treatment compared with the low-phosphorus treatment and control at day 4, day 8, and day 10 (P < 0.05) (Fig. 3B). In addition, expression levels of the SmMC4 and SmMC6 genes were higher in the low-phosphorus treatment than in the low-nitrogen treatment and control during days 2 to 4 (P < 0.05) (Fig. 3B). In addition, the SmMC4 and SmMC6 genes were upregulated in the low-nitrogen treatment compared to the low-phosphorus treatment and control at day 6, day 8 (SmMC6 only), and day 10 (P < 0.05) (Fig. 3B). N- and P-limited conditions could induce the upregulation of the SmMC genes in S. marinoi, and the six SmMC genes had differential expression responses.
Caspase 3-like protein activity.The activity of caspase 3-like protein in the low-nutrient treatments increased from day 2 to day 8 and decreased afterward (Fig. 4). Cells in the low-nutrient treatments displayed significantly higher caspase 3-like protein activity than those in the control after day 4 (P < 0.05) (Fig. 4). In addition, the low-nitrogen treatment showed a significantly higher level of caspase 3-like protein activity than the low-phosphorus treatment and control at day 8 (P < 0.05) (Fig. 4). Furthermore, the caspase 3-like protein activity was significantly lower in the nutrient repletion treatments than the low-nutrient treatments at day 11 (P < 0.05) (Fig. 4). Our results showed that S. marinoi cells increased the activity of caspase 3-like protein under nitrogen or phosphorus limitation and decreased caspase 3-like protein activity when nutrients were resupplied.
Temporal variation in caspase 3-like protein activity of S. marinoi in response to low-nitrogen, low-phosphorus, and resupplied conditions. Statistically significant differences (P < 0.05) at given time points are denoted by asterisks (*; compared to the control) and pound signs (#; compared to the low-nitrogen/phosphorus treatment). Error bars represent the standard deviations of experiments performed in triplicate. RFU, relative fluorescence units.
In vivo staining for PCD markers.The percentage of PS externalization (annexin V-fluorescein isothiocyanate [FITC] positive) in S. marinoi cells that were exposed to the low-nutrient conditions is shown in Fig. 5. There was a notable difference in the percentage of annexin V-FITC positive cells between the low-nutrient treatments and the control. The percentage of cells that were positively labeled with annexin V-FITC significantly increased in the low-nutrient treatments compared with the control during days 4 through 10 (P < 0.05) (Fig. 5). The highest percentages of positively labeled cells in the low-nitrogen treatment and low-phosphorus treatment were 2.6- and 2.2-fold higher than that in the control at day 6, respectively. Furthermore, the percentage of positively labeled cells in the control increased from day 1 to day 10 (Fig. 5). When cultures were replete with P or N, the growth inhibition stopped concomitantly with a decrease in cells with externalized PS residues. In general, the percentage of PS externalization in S. marinoi increased when the cells were in the stationary phase, decline phase, and N- or P-limited conditions.
In vivo detection of the proportion of annexin V-positive cells in S. marinoi in the different treatments. Statistically significant differences (P < 0.05) at given time points are denoted by asterisks (*; compared to the control) and pound signs (#; compared to the low-nitrogen/phosphorus treatment). The error bars represent the standard errors of triplicate cultures.
DISCUSSION
General physiological and morphological responses.For the low-nitrogen treatment, we amended a small amount of nitrate (88 μM) at the beginning of the experiment. Diatoms can consume this low amount of nitrate and encounter the nitrogen limitation stage in a few days. This kind of setup for the nitrogen limitation experiment has been used in previous studies in which the starting nitrate concentration ranged from 50 to 100 μM (15, 27, 40–44). The measurements of nitrate consumption and cell growth both indicate that the nitrogen-limited condition was reached at day 4 in the low-nitrogen treatment (Fig. 1A and B).
Fig. 1C shows a sharp decline in phosphate in the control; however, the phosphate concentration was still 1.5 μM in the control at the end of the experiment (day 11). The phosphate concentration in the control was not low enough to inhibit the growth of S. marinoi cells, which fulfilled the requirements of our experiments. The same pattern was shown in a study of dinoflagellate; Li et al. (45) started their experiment when the phosphorus concentration was 1 μM in the P-deprived experimental group, and the cell density continued to increase in the following 2 days. Their study suggests that a much lower phosphate concentration is needed to determine the impact of phosphate limitation. Alkaline phosphatase (AP) is widely used as a potential diagnostic marker for P stress in phytoplankton (46–50). The orthophosphate ion (Pi) threshold concentrations reported to induce AP are lower than 1 μM in many phytoplankton species, including Skeletonema costatum (lower than 0.5 μM) (50, 51). S. costatum turned out to be closely related to S. marinoi, and it can be the reference species (52). Therefore, the cells in the control did not experience phosphate limitation at the end of the experiment.
Due to N or P limitation, the cell density of S. marinoi in the low-nutrient treatments was significantly lower than that in the control after day 3 (P < 0.05) (Fig. 1A). N or P limitation can inhibit the growth of S. marinoi and decrease the ability to take up other nutrients (Fig. 1B and C). This result is in accordance with other studies that examined the fundamental cellular responses of diatoms in N- or P-limited conditions (15, 17, 53–56). In the control culture, the N and P were indeed rich at day 6, but the cell density peaked at day 5 and then decreased (Fig. 1A). These results showed that the algal culture entered the stationary and decline phases at day 6 in the control, and the cells encountered aging stress. However, algae in the low-nutrient treatments were stressed by nutrient limitations at day 4 or earlier, and the cultures had not yet reached the high cell density as the control; therefore, algae had not been stressed by aging. In addition, when cultures in the low-nutrient treatments were replete with P or N, inhibition of growth was stopped concomitantly with a decrease in the proportion of cells with everted PS residues and the caspase 3-like protein activities involved in PCD (Fig. 1A, Fig. 4 and 5). Therefore, the stresses leading to PCD were different in the control (aging) and in the low-nutrient treatments (nutrient limitation).
The S. marinoi cells grown in the low-nutrient treatments began to experience nitrogen limitation or phosphorus limitation and changed morphology after day 4 (Fig. 2). Cells in nutrient limitations showed the formation of vacuoles when they entered stationary and decline phases, a sign of PCD response, although the cell membranes remained intact. Other morphological changes, such as internal organelle degradation and cytoplasmic blebbing, were also found in S. marinoi cells under N or P deficiency (Fig. 2). These morphological changes were reported in T. pseudonana cells that were exposed to iron starvation (12, 57) and S. marinoi exposed to silicate limitation (18). We showed that both nutrient limitations and senescence could cause the morphological changes in S. marinoi cells that were commonly observed in the PCD responses.
Externalization of PS residues is a key feature of PCD (7, 9); T. pseudonana, Heterosigma akashiwo, and Emiliania huxleyi also exhibit this PCD hallmark (12, 21, 58). PS residues found on the interior layer of the lipid bilayer of a healthy cell membrane are actively translocated to the outer leaflet of the membrane under the process of PCD. In our study, low-nutrient treatments contained significantly more cells with externalized PS than the control during days 4 through 10. Furthermore, the percentage of positively labeled cells in the control increased from day 1 to day 10 (Fig. 5). In general, the proportion of cells with everted PS residue increased when the cells were in the stationary or decline phase and under N- or P-limited conditions. When cultures were replete with P or N, inhibition of growth was stopped concomitantly with a decrease in the proportion of cells with externalized PS.
The hallmarks of PCD, such as alterations in the plasma membrane (annexin V-positive cells) and nuclear and cytoplasmic blebbing, did not all occur under every type of PCD (21). Our results show that approximately 10% of S. marinoi cells became annexin V positive under nutrient limitations. This does not mean that only approximately 10% of diatoms underwent PCD, because part of the remaining 90% of cells may undergo other types of PCD without showing annexin V-positive results. The proportion of annexin V-positive cells is approximately 10% in some algae, for example, when H. akashiwo cells are under heat stress (21). Moreover, Franklin et al. (58) found that E. huxleyi showed fewer than 5% labeled cells throughout the nutrient depletion experiment (28 days), indicating that almost all cells of both cell types had intact plasma membranes over the period of study. Meanwhile, T. pseudonana had low numbers of labeled cells (<2%) until the stationary phase, whereupon the percentage of labeled cells rose rapidly to a maximum of 25% on the last sampling day (day 28). Our results were consistent with these previous studies and indicated that alterations in the plasma membrane did not occur under every type of PCD in marine diatoms.
Expression patterns of genes involved in PCD and oxidation resistance.To better understand the PCD processes in S. marinoi cells in response to nutrient limitations, we analyzed the transcriptional levels of several genes (TSG101, DSP, ALDH, GSHS, GOX, HSP90, and PDCD4) that are related to PCD and oxidation resistance (Fig. 3).
TSG101 is a eukaryotic protein responsible for cell growth and the differentiation process and acts as a negative growth regulator (59). The upregulation of the TSG101 gene was found in S. marinoi under N (days 4 to 10) or P (days 6 to 10) deficiency (Fig. 3A), and the TSG101 gene was also upregulated in S. marinoi under silicate limitation (31). Our results show that the TSG101 gene may act as a growth inhibitor in S. marinoi. The DSP gene had an expression response similar to that of the TSG101 gene (Fig. 3A). It was upregulated in the low-nutrient treatments from day 2 to day 10, and increased expression was observed in S. marinoi along the growth curve in the control (Fig. 3A). The DSP gene is localized in the chloroplast and appears to function as an important signal transducer and photosystem regulator (60). However, the DSP gene in S. marinoi was downregulated under silicate limitation and upregulated when the cells were in senescence (31). The expression pattern of the DSP gene in S. marinoi in response to N or P limitation was different from that in response to silicate limitation. In addition, the DSP gene in T. pseudonana was upregulated during Fe starvation, and it can facilitate higher levels of PS components and cyclic electron transport flow in diatom cells (60). DSP is responsible for the adaptive and death-related processes of T. pseudonana under Fe and light stresses (7, 60). Our results show that the DSP gene in S. marinoi appears to have dual roles in acclimation- and death-related functions.
The upregulation of antioxidant genes (ALDH, GSHS, and GOX) in S. marinoi during the exponential phase helped to remove the reactive aldehydes and ROS (Fig. 3A); however, antioxidant genes (ALDH, GSHS) were downregulated in the low-nutrient treatments when the stress decreased from day 6 to day 10 (Fig. 3A). ALDH is generally involved in aldehyde metabolism and amino acid catabolism, protecting cells from osmotic stress and detoxification reactions (30, 61). Glutathione, which is synthesized by GSHS, acts as an antioxidant, helping maintain intracellular redox conditions, detoxifying H2O2, and sensing environmental stress (62–64). Our study was consistent with the response patterns of these genes when S. marinoi was subjected to silicate limitation (18, 30). The reduction in expression of these genes suggests an energy-saving strategy for cells to avoid unnecessary overinvestment in the respective proteins (65). The downregulation observed in the low-nutrient treatments can be a reaction that allows restoration of homeostasis and survival of organisms (65).
While other antioxidant genes (ALDH, GSHS) were downregulated when the stress decreased from day 4 to day 10, GOX overexpression accounted for the antioxidant response to senescence and nutrient starvation. The glycolate-oxidizing enzyme plays a key role in photorespiratory carbon metabolism and is responsible for oxidative damage and reducing carbon loss (66). In addition, the process of photorespiration is also involved in nitrate reduction, amino acid metabolism, signal transduction, and stress resistance (66, 67). Our results suggest that the upregulation of the GOX gene may provide a specific protective antioxidant activity and allow S. marinoi to better cope with adverse conditions. The pattern was also found in S. marinoi cells under Si-limited conditions and in higher plants such as Arabidopsis thaliana and Oryza sativa (18, 30, 66, 67).
Between control and low-nutrient treatments, the status of S. marinoi cultures in the stationary phase was not the same. In the control, the phytoplankton cultures accumulated aged cells in the stationary phase but not under the stress of N or P limitation. Although the increased expression of DSP, GOX, metacaspases 2, and metacaspases 6 was found in the stationary phase in the control, death-related functional genes often show low expression levels (Fig. 3A and B). A previous study also found that metacaspases and some other cell death-related functional genes did not show significant variation in the stationary phase (31).
HSP90 is involved in maintaining regular cellular functions and protects cells against stress and apoptosis (68–70). PDCD4 is involved in abiotic stress responses in higher plants (71) and acts as a tumor-suppressing protein in animals (72). The levels of PDCD4 (days 8 to 10), GOX (days 6 to 10), and HSP90 (days 4 to 6, 10) transcripts in S. marinoi induced by N deficiency were relatively high compared to those induced by P deficiency and cell aging (Fig. 3A). In contrast, the expression of the TSG101 gene in S. marinoi showed a clear and constant increase under P limitation compared to N limitation and cell aging (Fig. 3A). Our results show that the expression patterns of genes involved in PCD and oxidation resistance (PDCD4, GOX, HSP90, and TSG101) were different when the S. marinoi cells were under different limitations (N and P).
Under adverse conditions, diatom cells suffer from oxidative stress, after which antioxidant capacities are induced. Once antioxidant capacity can no longer contain oxidative stress, PCD is initiated. The ALDH, GSHS, GOX, HSP90, and PDCD4 genes are involved in antioxidant effects, and they were upregulated when S. marinoi cells were under nutrient (N or P) limitation conditions or senescence. When the stress deteriorated, upregulation of the GOX gene allowed cells to better cope with adverse conditions. The TSG101 gene may act as a growth inhibitor in S. marinoi, and the DSP gene appears to have dual roles in acclimation- and death-related functions. Furthermore, the expression of some PCD-related genes (PDCD4, GOX, and HSP90) was high in S. marinoi during nitrogen deficiency, and the TSG101 gene induced by N deficiency was upregulated more during P limitation than N limitation. These findings demonstrate that differential expression patterns of genes involved in PCD and oxidation resistance in response to different stressful conditions could be found in S. marinoi.
Metacaspases and PCD.Multiple metacaspase genes have been identified in diatoms. Six metacaspase genes are present in T. pseudonana and S. marinoi, and five metacaspase genes are present in Phaeodactylum tricornutum (12, 18, 36). These metacaspases share DNA sequence similarity with metazoan “initiator” and “executioner” caspases (36). Our results reveal the response of S. marinoi to nutrient deficiency with the expression patterns of SmMC genes and caspase 3-like protein activity.
Cells under nutrient limitations displayed a significantly higher level of caspase 3-like protein activity than the control after day 4 (P < 0.05) (Fig. 4). Furthermore, the caspase 3-like protein activity in the low-nitrogen treatment was higher than that in the low-phosphorus treatment and control at day 8 (P < 0.05) (Fig. 4). These results indicate that N-limited and P-limited conditions could increase caspase 3-like protein activity in S. marinoi cells, especially in the N-limited condition. The increase in caspase 3-like protein activity observed in S. marinoi in the low-nutrient treatments further supports the enhanced caspase activity of diatoms observed in other stress conditions (12, 18). In addition, when the cultures were replete with P or N, growth was no longer inhibited, and caspase 3-like protein activity began to decrease.
Most of the SmMC genes (SmMC1, SmMC2, SmMC3, SmMC4, and SmMC6) were upregulated in the low-nutrient treatments at day 4 (P < 0.05) (Fig. 3B). However, the expression levels of the SmMC5 gene did not show a significant difference between the low-nutrient treatments and the control within the first 6 days (P > 0.05) (Fig. 3B). Diverse domain architecture and operon configurations in metacaspases suggest that they play roles not only in PCD but also in signaling, diverse enzymatic activities, and protein modification (7). Some metacaspases may function in housekeeping and/or stress response functions in the acclimation to specific stresses (12, 13). SmMC5 is unlikely to be responsible for initiating PCD in S. marinoi under nitrogen or phosphorus limitation, and it may have other functions, e.g., stress acclimation. Our results suggest that metacaspases could play diverse roles in S. marinoi cells under nutrient (N and P) limitation. In addition, SmMC6 was more specific to low-P limitation than low-N limitation. A previous study also suggested that differential activation and regulation of the PCD machinery were induced in the diatom by different stress factors (12).
The primary goal of this study was to understand the relationship between nitrogen limitation and phosphorus limitation with PCD in the diatom S. marinoi and to gain new insight into how diatom cells change their physiological and molecular characteristics, particularly in terms of programmed cell death. Primary nutrients such as phosphorus, nitrogen, and silicate are critical for the growth of phytoplankton. The cellular death processes of phytoplankton that mediate the bloom-to-postbloom transition play an important role in the biogeochemical cycle (9). PCD evolved in phytoplankton cells as a strategy for decreasing nutrient-related stress and removing aging and/or damaged cells from the population to ensure its competitiveness (7–9, 73, 74). Externalization of PS, intracellular vacuolization, high in vivo caspase 3-like activity, upregulation of SmMC genes, and oxidation resistance- and death-related genes of metacaspase-mediated PCD were observed in S. marinoi in response to nitrogen limitation and phosphorus limitation (Fig. 6). It is interesting that some PCD-related genes (PDCD4, GOX, and HSP90) in S. marinoi induced by N deficiency were relatively upregulated compared to those induced by P deficiency. In contrast, the expression of the TSG101 gene in S. marinoi showed a clear and constant increase during P limitation compared to N limitation. These findings suggest that differential PCD responses can occur for diatoms under different kinds of environmental stresses. Low N and P could induce PCD in the diatom S. marinoi and inhibit its growth. The effects of low N and P involved in PCD were quickly reversed by the resupply of nutrients, and the cells regained active growth. This means that PCD is an important mechanism to overcome the stress of nutrient limitation, which occurs frequently toward the end of diatom blooms.
The relationship of nutrient limitations and senescence with PCD in the diatom S. marinoi.
MATERIALS AND METHODS
Cultivation of the diatom.An S. marinoi strain was isolated from Jiaozhou Bay, China, and characterized based on its morphology and 18S rRNA gene sequence (18). S. marinoi was grown in f/2 medium (76) and not acclimated to reduced nitrogen or phosphorus medium prior to this study. Cultures were grown at 19°C under a 12-h light/12-h dark cycle with illumination at 100 μmol photons m−2 · s−1. Seawater was collected from the Yellow Sea, filtered through 0.22-μm-pore-size Millipore filters, and autoclaved. The concentrations of total nitrogen and total phosphorus in the seawater were 16.5 μmol liter−1 and 0.53 μmol liter−1, respectively. A starter culture of S. marinoi was grown to the mid-exponential phase and transferred three times. Diatom cells were harvested at the same time of day via filtration using 0.8-μm-pore-size Millipore filters. The harvested cells were inoculated into three different media: an f/2 medium (defined here as the control), a low-P f/2 medium containing 4 μmol liter−1 phosphate instead of 36.2 μmol liter−1, and a low-N f/2 medium containing 88 μmol liter−1 nitrate instead of 882 μmol liter−1. Each treatment was performed in triplicate, and all the cultures were grown in 1-liter sterile polycarbonate bottles. To better understand the physiological responses of S. marinoi to nutrient depletion and recovery, on the 9th day, when the nutrients were exhausted in the low-nutrient treatments, we took samples from each treatment and resupplied nitrate or phosphate to maintain the nutrient concentration at 882 μmol liter−1 nitrate and 36 μmol liter−1 phosphate.
To determine cell abundance, samples were fixed in Lugol’s fixative (at a final concentration of 2%) and cells were counted using a microscope (CX31; Olympus Corporation, Tokyo, Japan) with a hemocytometer. The concentrations of nitrate and phosphate in the cultures were measured on a daily basis using classic colorimetric methods. N-NO3− was analyzed using the copper-cadmium column reduction method (Bran+Luebbe method no. G-172-96, rev. 7), while P-PO43− was determined via the molybdenum blue method (Bran+Luebbe method no. G-175-96, rev. 8) (77).
Cellular observations with transmission electron microscopy.The internal cell morphology of the samples was observed and photographed with a Philips EM208 transmission electron microscope (Philips Scientifics, Eindhoven, the Netherlands). Cells were collected using 5-μm-pore-size polycarbonate filters and then fixed in 2 ml of glutaraldehyde fixative (final concentration 2.5%, pH 7.4) for 2 h. The cells were then rinsed once, suspended in 0.1 M phosphate-buffered saline solution (PBS, pH 7.4), and stored at 4°C prior to the next step. The fixed cells were rinsed three times for 15 min each in 0.1 M PBS (pH 7.4), postfixed for 3 h in 1% buffered OsO4, and washed three times with 0.1 M PBS (pH 7.4). After the supernatant was removed by centrifugation (4°C, 10,000 × g, 5 min), the pellets were dehydrated through a graded series of ethanol solutions and then placed in propylene oxide. The cells were then embedded in Epon 812. Embedded tissues were sectioned using a PowerTomo-XL ultramicrotome (RMC, Rochester, NY, USA), collected on 200-mesh copper grids, and stained using uranyl acetate and lead citrate. The stained sections were visualized and photographed using an H-7560 electron microscope (Hitachi, Tokyo, Japan).
RNA extraction and quantitative PCR analysis.Metacaspase and other oxidation resistance- and death-related genes were identified in the S. marinoi transcriptome (GenBank accession no. GSE77468 ), as previously described (18). The expression of target genes in triplicate samples was analyzed using RT-qPCR. RNA was extracted by use of TRIzol (Ambion, Life Technologies, Carlsbad, CA) according to the manufacturer’s protocol, and the quantity and purity of RNA were assessed using a Picodrop microliter UV/Vis (UV-visible) spectrophotometer (Picodrop, Cambridge, UK). Two hundred nanograms of cDNA was used as the template for the reverse transcription of all genes, which was performed using a One-Step genomic DNA (gDNA) removal and cDNA Synthesis SuperMix kit (TransGen Biotech, Beijing, China). Quantitative PCR was performed using an Applied Biosystems 7500 real-time PCR system (Thermo Fisher, MA, USA) in reaction mixtures containing FastStart universal SYBR green master (ROX) (Roche, Mannheim, Germany). All qPCR assays were carried out in triplicate, and the 20-μl qPCR mixture contained the following reagents: 10 μl of ROX, 0.3 μM each primer, 0.2 μg μl−1 bovine serum albumin, and 2.0 μl of cDNA. The PCR program consisted of a denaturation step at 94°C for 5 min; 35 cycles of 94°C for 30 s, 60°C for 30 s, and 72°C for 1 min; and a final extension step at 72°C for 5 min.
All primers used in this study are listed in Table 1. Primer efficiencies (E) were validated with serial dilutions of cDNA samples over a 100-fold range. Standard curves were generated with seven dilution points by using the cycle threshold (CT) value versus the logarithm of each dilution factor and E = 10−1/slope. In this study, primers with efficiencies ranging from 90% to 110% were chosen to quantify the corresponding genes. The relative expression levels of target genes were normalized to that of the actin gene and estimated with the 2−ΔΔCT method (75). The gene expression level observed in the control on the 2nd day was used as the control for each RT-qPCR analysis to determine differences among groups.
Primers used to monitor gene expression in S. marinoi
Determination of caspase 3-like protein activity.The caspase 3-like protein activity of each sample was analyzed using an EnzChek caspase 3 assay kit #2 (Invitrogen, WI, USA). Ten milliliters of each culture was filtered onto a 0.8-μm-pore-size Millipore filter. The filter was placed into 1.5-ml centrifuge tubes containing 1 ml of autoclaved seawater to desorb the S. marinoi cells. The cells were centrifuged at 4°C for 8 min (5,000 × g) and frozen immediately. The pellet was suspended in 200 μl of extraction buffer (100 mM Tris-HCl, 10 mM EDTA, and 100 mM NaCl; pH 7.4) followed by sonication in ice water using a sonicator (Scientz Biotechnology, Ningbo, China). Insoluble material and unbroken cells were removed by centrifugation (6,500 × g, 5 min, 4°C), and the supernatant containing the protein extract was collected and transferred to a new 200-μl microtube.
Fifty microliters of cell extract was incubated with the caspase substrate Z-DEVD-R110 (25 μM) for 2.5 h in the dark at room temperature. The appearance of fluorescent rhodamine 110 (R110) upon enzymatic cleavage of the nonfluorescent substrate Z-DEVD-R110 was subsequently assayed using a microplate reader (excitation [Ex], 496 nm; emission [Em], 520 nm). Negative controls (samples without the Z-DEVD-R110 substrate) and substrate-only controls (a mixture of activity buffer and Z-DEVD-R110 substrate) were also assessed. The cell extracts were preincubated for 30 min with 250 μM caspase inhibitor Ac-DEVD-CHO prior to the addition of substrate, and we confirmed that the fluorescence signals observed in the S. marinoi cell extracts resulted from the activity of caspase 3-like proteases.
In vivo cell staining and flow cytometry analysis.Cells were collected via centrifugation (6,000 × g, 5 min, 4°C) and then stained using annexin V (10 μl/100 μl cells; Invitrogen, WI, USA) to examine PS externalization (annexin V-FITC-positive cells). The detailed protocol was described elsewhere (18). The number of positively stained cells (out of a total of 20,000 cells) was determined at 520 nm after excitation with a 488-nm laser using a flow cytometer (BD Fortessa). The gating and data analysis were performed using FlowJo analytical software.
Statistical analyses.Statistical analyses were performed, and differences between treatments were compared at the same time point using a one-way analysis of variance (ANOVA) (P = 0.05). Tukey's post hoc tests were used to test the hypothesized differences. It was ensured that the differences were qualified, and then the statistical significance (P ≤ 0.05) was identified.
ACKNOWLEDGMENTS
This study was supported by the National Key Research and Development Program of China (2017YFC1404402), the National Natural Science Foundation of China (no. 41976133), and the Scientific and Technological Innovation Project of the Qingdao National Laboratory for Marine Science and Technology (2016ASKJ02).
We are grateful to Paul Rosenberger, Bo Wang, Xiaoli Jing, and Hui He for helpful comments and discussions during the writing of this paper. We thank Lingling Yang at Shandong Eye Institute, Qingdao, People’s Republic of China, for flow cytometry analysis; Jie Huang and Chen Li at Yellow Sea Fisheries Research Institute, Chinese Academy of Fishery Sciences, People’s Republic of China, for fluorometric determinations; and Jinshan Tan at Medical College of Qingdao University, Qingdao, People’s Republic of China, for transmission electron microscopy observations.
There is no conflict of interest among the authors of this article.
FOOTNOTES
- Received 23 October 2019.
- Accepted 15 November 2019.
- Accepted manuscript posted online 22 November 2019.
- Copyright © 2020 American Society for Microbiology.
