Introduction

Harmful algal blooms (HABs) are widespread coastal phenomena that occur when accumulation of algal species have negative socioeconomic, public health or environmental impacts1. Among HABs, those caused by the globally distributed toxic species, Heterosigma akashiwo (Raphidophyceae) pose a significant threat to the global finfish and shellfish economy, with documented losses reaching into the millions of dollars2. Understanding environmental drivers that contribute to the proliferation of H. akashiwo is essential to preventing and mitigating blooms of this species.

A major factor in HAB formation is the availability of nitrogen both in terms of concentration and type3,4. Indeed, past research showed that H. akashiwo removes a substantial portion of nitrogen from coastal waters during blooms5, suggesting that nitrogen plays a role in H. akashiwo bloom initiation and maintenance. Although H. akashiwo can use both nitrate and ammonium6,7,8, there is evidence that H. akashiwo can also use nitric oxide (NO) and maintain positive growth with NO as a sole nitrogen source9,10,11. As a reactive nitrogen species, NO is involved as a signaling molecule in numerous plant and animal functions, including pathology and stress response12. NO also reacts with transition metals and proteins13 and is not considered a viable nitrogen source for eukaryotes due to its toxicity. However, concentrations of NO approaching 500 nM, with spikes exceeding 200 µM, have been documented at the oxic-anoxic transition zone of coastal marine sediments14,15 where it is produced as an intermediate in heterotrophic denitrification16. Consequently, NO may be accessible to H. akashiwo as it undergoes diel vertical migration17.

The potential for NO assimilation by H. akashiwo may be due to a unique hybrid nitrate reductase, NR2-2/2HbN (NR2) found in this species and at least one other raphidophyte9. Nitrate reductase (NR) catalyzes the rate-limiting reduction of nitrate to nitrite, with electrons provided by NAD(P)H18. Nitrite is transported into the chloroplast, where it is reduced to ammonium by nitrite reductase (NiR). Ammonium is then incorporated into amino acids through the glutamine synthetase/glutamine:2-oxoglutarate aminotransferase (GS/GOGAT) pathway. NR is highly regulated in plants and algae, both transcriptionally and post-transcriptionally, in part by the presence and relative proportions of nitrate and ammonium18,19,20,21,22. In some plants and algae, NR transcription and activity is upregulated by nitrate and repressed by nitrite and ammonium as well as downstream products of the GS/GOGAT pathway22.

In contrast to prototypical NR, however, NR2 in H. akashiwo includes a truncated hemoglobin (HbN) domain9. Similarities to mycobacterial HbN23 and truncated hemoglobins in other algal species24 suggest that NR2 converts NO to nitrate via nitric oxide dioxygenase (NOD) activity of the HbN domain9, which may then be reduced to nitrite via NR activity. In support of this hypothesis, NR in Chlamydomonas has been shown to associate with truncated hemoglobin (THB1) to convert NO to nitrate24. However, NR also participates in generation of NO in Chlamydomonas through partnership with amidoxime-reducing component/nitric oxide-forming nitrite reductase (ARC/NOFNiR)25. The generation of NO through this pathway plays a role in repression of NR by ammonium, and has been shown to repress NR transcription and enzyme activity as well as nitrate transport when both ammonium and nitrate are present26,27. Assimilation of externally sourced NO through NOD activity of NR2 in H. akashiwo may be inhibited in a similar manner, effectively preventing H. akashiwo from accessing this nutrient when ammonium is present.

The object of this study was to investigate the regulation of NO assimilation into biomass by H. akashiwo, and the effects of other nitrogen sources on this process. Specifically, this study tested the hypothesis that the assimilation of NO is repressed in the presence of ammonium. This may affect the ability of H. akashiwo to access this nitrogen source in coastal environments impacted by ammonium pollution, and limit the role of NO in bloom initiation and maintenance.

Results

For this investigation, Heterosigma akashiwo was acclimated to growth on nitrate (Experiment 1), a 50/50 mix of nitrate and ammonium (Experiment 2) or ammonium alone (Experiment 3) (see Fig. 1). Subcultures from each experiment were spiked with 15NO. Several parameters were measured to evaluate the response of H. akashiwo to addition of NO, including transcription of genes involved in nitrogen assimilation, NR enzyme activity, and assimilation of 15N-labelled NO into particulate nitrogen. Paired replicates also received either nitrate or ammonium at 10% ambient concentrations instead of NO as a means to verify responses by H. akashiwo to commonly encountered nitrogen sources.

Figure 1
figure 1

Experimental design. Heterosigma akashiwo was cultured with 100 µM nitrate (Experiment 1), 50 µM nitrate and 50 µM ammonium (Experiment 2), or 100 µM ammonium. Cultures (N = 4) were then subdivided into four treatments for a total of 16 subcultures for each experiment. Controls received nitrogen free medium. Nitrogen-spiked controls received 15N-labelled nitrate or ammonium at 10% of ambient concentrations (see Table 1), and the NO treatment was spiked with 100 µM 15N-labelled S-nitroso-N-acetylpenicillamine (SNAP) to generate 15N–NO.

Full size image

Gene expression

The expression of NR, GS, and GOGAT was normalized to the transcript abundance of GAP (Figs. 2 and S1). NR expression in the NO-spiked samples was significantly higher than the Control for each experiment at 15 min (Fig. 2A). Mean NR expression was highest in Experiment 1, and lowest in Experiment 3. At 60 min, the relative expression of NR remained significantly higher than Controls for Experiments 2 and 3 (Fig. 2B), but there were no significant differences between Control and NO treatment for Experiment 1 at 60 min. The large standard deviation between biological replicates at this time point was likely due to the highly dynamic gene expression activity across biological replicates in Experiment 1. A comparison between time points indicates that the expression of NR decreased over time for Experiment 1 and increased over time for Experiments 2 and 3.

Figure 2
figure 2

Gene expression analysis of Heterosigma akashiwo after spiking with NO. Relative gene expression of nitrate reductase (NR; A, B), glutamine synthetase (GS; C, D) and glutamine:2-oxoglutarate aminotransferase (GOGAT; E, F) at 15 min (A, C, E) and at 60 min (B, D, F) after spiking with NO for control and treatment cultures of Heterosigma akashiwo when acclimated to growth on 100 µM nitrate (Exp 1), 50 µM nitrate and 50 µM ammonium (Exp 2), or 100 µM ammonium (Exp 3). Expression was normalized to GAP transcript abundance. Error bars are + /− 1 standard deviation from the mean. Significant difference in relative expression of genes in the nitric oxide treatment compared to controls (*) is denoted.

Full size image

GS expression in the NO-spiked samples was not significantly different than the Controls for any experiment at 15 min, but was significantly lower than the Controls in Experiments 2 and 3 at 60 min (Fig. 2C, D). Similar to NR expression, GS expression was the highest in Experiment 1, and lowest in Experiment 3 at 15 min, while at 60 min, the mean GS expression was highest in Experiment 2 and lowest in Experiment 3. A comparison between time points indicates that the expression of GS relative to the Control decreased over time for all three experiments.

GOGAT expression in the NO-spiked samples at each time point was significantly higher than the Control in Experiments 2 and 3, but not significantly different from the Control in Experiment 1 (Fig. 2E, F). GOGAT expression in NO treatments at 15 min was similar across all three experiments. At 60 min, the mean expression of GOGAT was highest in Experiment 3 and lowest in Experiment 1. A comparison between time points indicates that the expression of GOGAT relative to Controls increased over time for all three Experiments (Fig. 2E, F).

Expression of NR, GS and GOGAT in nitrate- and ammonium-spiked treatments are shown in Supplementary Fig. S1 (available online). The amount of added nitrate and ammonium (10% of ambient) differed by treatment and by experiment, so that direct comparisons should be viewed with caution. Instead, the nitrate and ammonium treatments provided a qualitative assessment of H. akashiwo’s response to more commonly encountered nitrogen sources, and to provide context for its response to NO addition. NR and GS expression patterns and magnitude of expression relative to Controls when spiked with nitrate at both 15 and 60 min were similar to those of NO treatments. As expected, addition of ammonium inhibited NR expression relative to Controls at 15 min for all three experiments, while GS expression patterns when spiked with ammonium were similar to those of NO treatments at both time points. GOGAT expression at 15 min also followed the pattern of response observed for NO treatments. At 60 min, however, GOGAT expression for nitrate-spiked samples was lowest in Experiment 3, and lowest in Experiment 1 when spiked with ammonium.

Nitrate reductase enzyme assay

NR activity, measured at 2 h after spiking with NO, was up to 12 times higher than Controls, with no significant differences between NR activity between experiments (Fig. 3). NR activity was significantly higher than Controls in Experiments 1 and 3, but no significant differences were detected between Control and NO treatment in Experiment 2.

Figure 3
figure 3

Nitrate reductase enzyme activity in Heterosigma akashiwo after spiking with NO. Nitrate reductase (NR) enzyme activity 2 h after spiking with NO for Control and treatment Heterosigma akashiwo cultures when acclimated to growth on 100 µM nitrate (Exp 1), 50 µM nitrate and 50 µM ammonium (Exp 2), or 100 µM ammonium (Exp 3). Expression was normalized to protein content. Error bars are + /− 1 standard deviation from the mean. Significant difference in activity in the nitric oxide treatment compared to controls (*) is denoted.

Full size image

As with the gene expression analysis, direct comparisons between treatments or between experiments were not possible due to the differences in nitrogen added (10% ambient) for ammonium and nitrate. However, NR activity for nitrate- and ammonium-spiked cultures followed a pattern similar to NO treatments with highest activity in Experiment 1 and lowest activity in Experiment 2 (see Supplementary Fig. S2 online).

Nitrogen assimilation (15N)

15N uptake and assimilation was analyzed at 4 (T4) and 24 (T24) hours after spiking with 15N-labeled N. At T4, the 15N uptake rate (Fig. 4A) of the NO-spiked samples was significantly higher in Experiments 2 and 3 compared to Experiment 1 (p value < 0.05). Likewise, the 15N uptake per biomass (Fig. 4C) for NO-spiked samples was significantly higher in Experiments 2 and 3 compared to Experiment 1 at T4 (p value < 0.05).

Figure 4
figure 4

Assimilation of NO for Heterosigma akashiwo. Nitrogen uptake rate (A, B) and amount of nitrogen uptake per biomass (C, D) at 4 h (A, C) and 24 h (B, D) after spiking with 15NO for Heterosigma akashiwo cultures when acclimated to growth on 100 µM nitrate (Exp 1), 50 µM nitrate and 50 µM ammonium (Exp 2), or 100 µM ammonium (Exp 3). Error bars are + /− 1 standard deviation from the mean. Significant difference in the 15N-nitric oxide uptake rate and amount of 15N-nitric oxide uptake per biomass between experiments is denoted (a, b, c).

Full size image

The uptake rate of 15N at T24 was significantly different between all three experiments (Fig. 4B). The highest 15N uptake rate in NO-spiked samples occurred in Experiment 3 and the lowest was in Experiment 1. 15N-labelled NO uptake per biomass was highest in Experiment 3 and lowest in Experiment 1 at T24 (Fig. 4D). 15N uptake per biomass for NO-spiked samples was not significantly different between Experiments 2 and 3 (p value < 0.05).

Discussion

Stewart and Coyne9 described a unique enzyme, NR2-2/2HbN (NR2) in Heterosigma akashiwo, that includes a HbN domain, potentially coupling nitric oxide dioxygenase (NOD) with nitrate reductase (NR) activity to assimilate NO. The ability to use NO as a nitrogen source has also been demonstrated for this species10,11, which may provide H. akashiwo with an advantage over other algae in regions where NO is available. This study investigated the effects of competing nitrogen types (nitrate, ammonium or an equal mix of nitrate and ammonium) on the ability of H. akashiwo to use NO. Transcript abundance for key genes involved in nitrate reduction (NR) and ammonium assimilation into amino acids (GS and GOGAT), as well as NR activity and 15NO uptake rate and assimilation into biomass were measured. Overall this research demonstrated that NO stimulated transcription of NR and GOGAT, as well as NR activity, resulting in incorporation of 15NO into biomass by this alga.

Coyne28 showed that NR is constitutively expressed in H. akashiwo and upregulated by addition of nitrate. NR transcripts were detected under nitrogen starvation as well as in the presence of ammonium. Results presented here demonstrated that exposure to NO also increased expression of NR in H. akashiwo compared to Control cultures, regardless of the presence of other nitrogen sources (Fig. 2). Mechanisms involved in nitrogen uptake differ depending on nitrogen type: Nitrate and ammonium are moved into cells through inducible transporters29, while NO rapidly diffuses into cells30. Indeed, the highest expression of NR was measured at 15 min in cultures acclimated to growth on nitrate (Fig. 2A). Even when acclimated to ammonium as a sole source of nitrogen (Experiment 3), however, NR transcript abundance was significantly higher than Controls within 15 min after spiking with NO. Prior research showed that exposure to NO has a variable effect on NR transcription in plants and other algal species, with reports of increased transcript levels31, inhibition26 or no changes in transcript levels32. The mechanism for upregulation of NR expression by NO in H. akashiwo is not clear, but one possibility is that nitrate generated by NOD activity of constitutively expressed NR2 may have been sufficient to induce upregulation of NR. Comparison of NO spiked subcultures over time also supports the hypothesis that NR expression in response to NO is governed by regulatory mechanisms similar to cultures exposed to nitrate (see Supplementary Fig. S1 online). Interestingly, the magnitude of difference in the relative expression of NR in NO spiked subcultures decreased over time in cultures grown on nitrate alone (Experiment 1), indicating potential feedback inhibition from downstream products33,34. Glutamine, for example, has been shown to inhibit transcription of NR in Nicotiana plumbaginifolia35, while glutamine-derived ammonium downregulated NR in Dunaliella viridis36.

The increase in expression of NR after 60 min in NO-spiked subcultures acclimated to growth with ammonium (Experiment 3) or a mix of ammonium and nitrate (Experiment 2) may be due to alleviation of ammonium repression on transcriptional activity. In previous research, Coyne28 showed that NR transcripts increased after nitrate was added to H. akashiwo cultures that were grown with ammonium, suggesting that nitrate: ammonium concentrations play a role in regulating expression28. Likewise, nitrate produced from NO through NOD activity of NR2 may have increased intracellular nitrate: ammonium over time to overcome ammonium inhibition on NR transcription.

Unlike NR, transcription of the cytosolic form of GS examined here did not immediately respond after exposure of the culture to nitric oxide. The activity of cytosolic GS coordinates nitrogen and carbon use efficiency and plays a major role in the assimilation of inorganic nitrogen as well as nitrogen recycling within the cell37. In diatoms, transcripts of cytosolic GS increased with the onset of darkness38, suggesting that it may also participate in dark assimilation of nitrogen. Here, expression of cytosolic GS in H. akashiwo at 15 min after spiking with NO was not significantly different from Controls, regardless of ambient nitrate or ammonium concentrations (Fig. 2C). The subsequent decrease in GS transcript abundance for NO-spiked cultures at 60 min compared to Controls for all three experiments suggests feedback inhibition from products of N assimilation (Fig. 2D). While little is known about the regulation of GS transcription in eukaryotic algal species, results of this study are consistent with research on plants and bacteria, indicating that transcription responds to cellular nitrogen status and is downregulated in response to feedback inhibition by accumulation of organic nitrogen products within the cell.

Results of this study showed that GOGAT expression was upregulated by NO addition in cultures acclimated to growth on ammonium (Experiment 3) or ammonium and nitrate (Experiment 2; Fig. 2E, F). It has been shown that expression of GOGAT is responsive to ammonium concentrations39,40, so that the increase in GOGAT expression over time in the present study was likely due to accumulation of downstream products of nitrate reduction, including ammonium and glutamine39. However, there is little supporting evidence of GOGAT regulation in microalgae and additional studies are required to fully understand mechanisms involved in regulation of both GS and GOGAT expression in these species.

Gene expression provides insight on the transcriptional response to nitrogen input but assimilation of nitrogen is also controlled post-transcriptionally, through mechanisms regulating the activity of the enzymes. The effect of NO on the activities of enzymes in nitrogen assimilation pathways has been studied in only a few model species. In Chlamydomonas reinhardtii, NR activity as well as the uptake of ammonium and nitrate were inhibited by NO41. Inhibition of NR in C. reinhardtii was mediated by truncated hemoglobins (THBs) with homology with the HbN domain of H. akashiwo NR224. Although NR activity in H. akashiwo was previously found to be inhibited by ammonium42, the observed increase in NR activity with addition of NO in the current study was not significantly affected by the presence of ammonium (Fig. 3). It should be noted that NOD activity of NR2 was not measured here, and NR activity was used as a proxy for dioxygenase activity of NR2, converting NO to nitrate. Overall, results of the NR assay indicate that H. akashiwo responds to NO as it would to nitrate28, suggesting that increasing NO led to an increase in nitrate: ammonium concentrations to overcome repression of NR activity by ammonium.

While not measured here, less is known about the effects of nitrogen type on GS and GOGAT enzyme activities of algal species. Regulation of GS has been detailed primarily in plants37 and bacteria, including cyanobacteria43, where a broad array of N-containing biomolecules indicative of cellular N status participate in cumulative feedback inhibition of GS activity. The study of dinoflagellate Akashiwo sanguinea showed that GS activity was highest when grown in ammonium, but was still active in the presence of nitrate, and steadily increased until high glutamate concentrations were reached44. A study of phytoplankton in the Taiwan strait showed a positive relationship between GS activity and ammonium but noted that this relationship was not maintained in unialgal culture studies of Emiliania huxleyi or Dunaliella primolecta45. Information on the induction and repression of GOGAT activity in eukaryotic algae is also limited. As with NR, GOGAT activity is likely to be modulated by the presence of substrates (glutamine and 2-oxoglutarate) and products. The present study indicated a trend toward increasing GOGAT expression with time after spiking with NO (Fig. 2E, F), suggesting a similar increase in substrate and GOGAT activity. An increase in GOGAT activity with increased transcription is also supported by the incorporation of 15NO into cellular nitrogen as discussed below (Fig. 4).

The increased incorporation of 15NO into biomass by cultures acclimated to growth on ammonium (Experiments 2 and 3; Fig. 4) is consistent with results reported previously for this strain of Heterosigma, showing a preference for nitrate over ammonium for this species6. In addition, the lower NO uptake rate and assimilation in the presence of nitrate (Experiment 1) compared to cultures with ammonium present may also be due to competition at the active site on NR. Nitrate binds to the molybdenum-molybdopterin cofactor, Mo-MPT, on NR46. Unlabeled nitrate present in the growth medium for Experiment 1 (see Fig. 1) and 15N-labelled nitrate produced by the conversion of 15NO to nitrate may compete for the active site of NR, resulting in the reduced uptake rate and lower amount of NO assimilation in the presence of nitrate compared to Experiments 2 and 3, where unlabeled nitrate was lower or absent altogether.

Diffusion may play a role in reduced uptake rate of the NO at 24 h compared to 4 h (Fig. 4). Since NO rapidly diffuses into the cell29,30, the abundance of NO within the cell available for assimilation would be greatest immediately after spiking cultures. As 15NO was depleted over time, the ambient, unlabeled nitrogen in the form of either nitrate or ammonium would continue to be transported into the cell and assimilated into biomass, effectively diluting the signal from 15NO.

Conclusions

Results of this investigation do not support the hypothesis that uptake and assimilation of NO by H. akashiwo is inhibited by ammonium. Indeed, NO upregulated NR expression and resulted in an increase in NR activity regardless of the presence of other inorganic nitrogen sources, with an increase in 15NO uptake and assimilation in the presence of ammonium. Common regulatory mechanisms, such as accumulation of substrate and negative feedback inhibition by downstream products of nitrogen assimilation, may have affected the transcriptional and other regulatory responses to NO in a manner similar to that expected from nitrate.

A better understanding of factors that stimulate the formation of algal blooms will improve management of coastal ecosystems and inform the development of HAB prevention and mitigation strategies. The capacity to use NO as an alternate nitrogen source may provide H. akashiwo with an advantage over other species in areas with high concentrations of NO, such as coastal environments, or when other nitrogen sources are scarce. This study suggests that NO availability should be considered a factor in promoting toxic blooms of H. akashiwo.

Methods

H. akashiwo was acclimated to growth on nitrate, ammonium, or a 50/50 mix of nitrate and ammonium in continuous cultures. Subcultures of each biological replicate were then spiked with NO and evaluated for gene expression, nitrate reductase enzyme activity and the uptake and assimilation of NO into biomass.

Chemicals

15N-labeled S-nitroso-N-acetylpenicillamine (15N-SNAP, Berry & Associates/Icon Isotopes, Dexter, MI, US) was used to generate 15N-labeled nitric oxide. 15N-labelled S-nitroso-N-acetylpenicillamine (SNAP) is a donor molecule that generates 15N-nitric oxide47. There is 60% efficiency of conversion between 15N-SNAP and 15N-nitric oxide47. 15N-ammonium and 15N-nitrate were sourced from 15N-labeled ammonium chloride (Cat. 299251-1G, Sigma Aldrich, St. Louis, MO, US) and 15N-labeled potassium nitrate (Cat. 335134-1G Sigma Aldrich, St. Louis, MO, US).

Algal stock conditions

Stock cultures of Heterosigma akashiwo (National Center for Marine Algae and Microbiota, CCMP2393) were maintained in modified Enriched Seawater/Artificial Seawater (ESAW)48,49. Modifications to the ESAW included an adjusted salinity of 20 PSU, and the elimination of strontium chloride hexahydrate (SrCl2 6H2O) and sodium metasilicate nonahydrate (Na2SiO3 9H2O). The nitrogen source was varied depending on experiment (see Experimental Design). Cultures were incubated at 25 °C under a 12:12 h light: dark cycle with 130 μmol m−2 s−1 irradiance.

Continuous cultures

Continuous cultures (N = 4) were established from stock cultures acclimated to nitrate: ammonium concentrations for each experiment described below. Modified ESAW medium influx/outflux was set to maintain a cellular density of 100,000 cells mL−1 in 1.6 L volumes, with gentle stirring to maintain homogeneity. Cultures were incubated at 25 °C under a 12:12 h light: dark cycle with 130 μmol m−2 s−1 irradiance.

Experimental design

Three experiments were conducted with different nitrate to ammonium ratios, while keeping total nitrogen concentrations constant (Fig. 1). Experiment 1 cultures were grown in modified ESAW with 100 µM nitrate. Experiment 2 cultures were grown in modified ESAW with 50 µM nitrate and 50 µM ammonium. Experiment 3 cultures were grown in modified ESAW without nitrate and with 100 µM ammonium. Cultures were dosed with Kanamycin (50 µg/mL) to prevent uptake and incorporation of 15N-labelled substrate by bacteria50. Subcultures were created at 4 h after the start of the light cycle resulting in 4 replicates of 4 treatments for a total of 16 subcultures.

To verify responses to more commonly encountered nitrogen sources, the treatment subcultures were spiked with 15N-labeled nitrogen equal to 10% of ambient nitrogen source. A fourth subculture was treated with 15N-labeled SNAP at a concentration equal to total nitrogen concentrations of the continuous cultures (Table 1). Subcultures were incubated at 25 °C under a 12:12 h light: dark cycle with 130 μmol m−2 s−1 irradiance.

Table 1 Ambient nitrogen source for continuous cultures and 15N-nitrogen treatment spikes for each subculture in Experiment 1, 2 and 3. SNAP, S-nitroso-N-acetylpenicillamine.
Full size table

Gene expression (RT-qPCR)

Samples were collected for gene expression analysis using vacuum filtration onto 3 µm polycarbonate filters at 15 min and 60 min after spiking. Filters were immediately placed into RLT buffer from RNeasy Mini Kit (Qiagen, LLC Germantown, MD, US) and stored at − 80 °C until extraction. RNA was extracted using the RNeasy Mini Kit (Qiagen), treated with DNase I Amplification Grade (Thermo Fisher Scientific, Waltham, MA) as described in Coyne and Cary (2005), and reverse transcribed with random hexamers using SuperScript III First Strand Synthesis SuperMix Kit (Thermo Fisher Scientific). The resulting cDNA and a No-RT control reaction were diluted in LoTE [3 mM Tris–HCl (pH 7.5), 0.2 mM EDTA]. PCR was performed to verify reverse transcription and lack of contaminating DNA using 1 μL diluted template, 0.2 mM dNTPs, 2.5 mM MgCl2, 1X Taq polymerase buffer (Sigma Aldrich, St. Louis, MO, US), 0.25 units Jump-Start Taq Polymerase (Sigma Aldrich), 1 μg BSA, and 3 μM of GAP primers (Table 2) in a 20 μL reaction volume. Reactions proceeded for 35 cycles of 94 °C for 30 s, 56 °C for 30 s, and 72 °C for 1 min.

Table 2 Primer sequences and concentrations for qPCR.
Full size table

Primers for the cytosolic form of glutamine synthetase (GS) and glutamate synthase (GOGAT) were designed using the “Primer-BLAST” function in Basic Local Alignment Search Tool (BLAST)51 from National Center for Biotechnology Information (NCBI). Primers were designed from sequences in the 20130911 combined assembly for H. akashiwo CCMP2393 transcriptome, downloaded from the Moore Foundation Marine Microbial Eukaryote Transcriptome Sequencing Program52. PCR was conducted on cDNA with each primer pair and analyzed for the correct size fragment using gel electrophoresis. Primers were then optimized by qPCR and dissociation curves were evaluated to confirm specificity of primers.

Transcript abundance was measured by qPCR for GAP, NR, GS and GOGAT using an ABI 7500 Real-Time PCR System (Thermo Fisher Scientific). Diluted cDNA was used as template in triplicate 10 µL reactions. Each reaction consisted of 1 µL of template cDNA, 5 µL of SYBR Select Master Mix (Thermo Fisher Scientific), 2 µL sterile water, 1 µL each of forward and reverse primers (Table 2). qPCR cycling parameters were as follows: 50 °C for 2 min and 95 °C for 10 min, followed by 40 cycles of 95 °C for 15 s, 60 °C for 30 s and 72 °C for 1 min. A dissociation stage was added to assess the specificity of amplification. Average transcript abundance was evaluated by linear regression analysis of triplicate reactions. Transcript abundance for NR, GS and GOGAT was then normalized to GAP9,28.

Nitrate reductase activity

Samples were collected from subcultures by centrifugation at two hours after spiking. The supernatant was discarded and the cell pellet was immediately frozen in liquid nitrogen. Samples were stored at − 80 °C until analysis. The frozen cell pellets were resuspended in 200 mM KPi extraction buffer (27.2 g L−1 KH2PO4, 10 g L−1 KOH, adjusted to pH 7.9) on ice. The resuspended pellet was homogenized by sonication, and centrifuged to clarify at 4 °C. The supernatants were immediately assayed as follows: triplicate 100 µL aliquots of extract were each mixed with 100 µL of 500 mM KPi assay buffer (pH 7.9) and 25 µL of 2 mM NADH (Sigma-Aldrich). KNO3 (100 µM) was added to duplicate assays and sterile water was added to the third as a negative control reaction in place of KNO3. The reaction was incubated at room temperature for 15 min and 1 M zinc acetate was added to stop the reaction. The samples were centrifuged to pellet the zinc acetate and the supernatant was transferred to a new tube. The concentration of nitrite in the supernatant was measured colorimetrically by addition of 698 µM N-methylphenozonium methyl sulfate (Sigma-Aldrich). After incubation at room temperature for 20 min in the dark, 5 M HCl and 0.058 M sulfanilamide (Sigma-Aldrich) was added and the reaction was incubated at room temperature for another 5 min in the dark. 0.004 M N-(1-Naphthyl) ethylenediamine hydrochloride (Sigma-Aldrich) was then added and the reaction was incubated at room temperature for 10 min in the dark. Absorbance was measured at 543 nm using a Thermo Scientific Nanodrop 2000c (Thermo Fisher Scientific). NR activity was normalized to protein content of the cellular homogenate as determined using the Pierce BCA Protein Assay Kit (Pierce, Rockford, IL, US).

Nitrogen assimilation (15N)

Samples were collected at 4 and 24 h after spiking for particulate 15N analysis. Subsamples of each culture were vacuum filtered onto GF/F filters and dried at 60 °C for 24 h. Filters were folded into combusted tins and pelletized. Particulate 15N in pelletized samples was measured by mass spectroscopy at Boston University (Experiment 1) or UC Davis (Experiments 2 and 3). Nitrogen uptake per biomass (Vm, per hour) was calculated according to methods outlined in Dugdale and Wilkerson53 and the uptake rate of nitrogen (Rho, µM/hr) was calculated according to Middelburg and Nieuwenhuize54 and Dugdale and Wilkerson53.

Statistical analysis

Statistical analyses were performed in R using base functions and the “car” package. Gene expression and NR assay data were first evaluated using the Q Test55 and outliers (identified at 95% confidence level) were removed. A Shapiro test on the residuals from a linear model was used to assess normality. If the p value was not significant (p value > 0.05), we assumed normality. A Levene test was used to assess homogeneity of variance. If the p value is not significant (p value > 0.05), we assumed homogeneity of variance.

A one-way ANOVA with a post-hoc Tukey’s test was employed when the data were normal, and variance was homogeneous. A paired t-test was used when the data were normal, but not homogeneous. P values from the paired t-test were adjusted using the Benjamini and Hochberg Correction56,57. Data that were not normally distributed were log transformed. Significance was determined using one-way ANOVA analysis. If normality could not be corrected using a log transformation, a Kruskal Wallis and a post-hoc Pairwise Wilcox Sign Test was used. Statistically significant differences were defined by a p value < 0.05.