Changes in Metabolism and Lipid Composition with Nitrogen Starvation and Recovery in a New Productive Strain of Neochlorella semenenkoi Using N15-Isotopic Labeling and HRMS

TL;DR

This study explores how a productive green algae strain adapts metabolically and in lipid composition during nitrogen starvation and recovery using isotopic labeling and mass spectrometry.

Contribution

The study introduces a new productive strain of Neochlorella semenenkoi and applies 15N isotopic labeling to reveal metabolic and lipid changes under nitrogen stress.

Findings

The algae strain utilizes ammonium acetate as a nitrogen source, primarily consuming nitrogen in the ammonium form.

Isotopic labeling revealed significant divergence between isotopic labeling and compound concentrations, suggesting isotopic analysis offers advantages over quantitative methods.

In vivo isotopic labeling helps identify compounds not found in standard mass spectrometry databases.

Abstract

Microscopic green algae are active producers of beneficial compounds, particularly those containing nitrogen. However, the metabolism of nitrogen-containing compounds is diverse and depends on the conditions of the nitrogen source. As a result, the approach to studying the metabolism of nitrogen-containing compounds becomes more complicated. This work demonstrates the metabolic changes in the high-productive green algae Neochlorella semenenkoi IPPAS C-1210 under conditions of nitrogen starvation and subsequent reintake, using high-performance liquid chromatography–mass spectrometry (HPLC–MS) with 15N isotopic labeling. The presented results include semi-quantitative chromatography–mass spectrometric analysis for 17 amino acids, a metabolomic profile of over 40 isotopically labeled compounds, an assessment of metabolic flux via isotopic incorporation, and an analysis of cellular lipid composition under varying growth conditions. The findings indicate that this strain can utilize ammonium acetate as a nitrogen source, consuming nitrogen in the ammonium form. The degree of isotopic labeling in compounds often diverged significantly from their quantitative changes (concentrations and chromatographic peak areas), suggesting that isotopic analysis may offer advantages over purely quantitative analysis for biological systems. Furthermore, in vivo biological isotopic labeling is shown to assist in identifying compounds absent from standard mass spectrometric databases.

Figures

Full text

1. Introduction

Even though Chlorella vulgaris—the first described species of the Chlorellaceae family—was discovered over a century ago, chlorella-like microalgae remain a vibrant subject of research, driven by modern analytical technologies [1]. Recent discoveries have revealed a wide diversity of these microorganisms in diverse ecosystems, including extreme environments like Antarctica, leading to the description of new genera and species [2,3,4,5]. Well-studied Chlorella strains have a broad range of practical applications, from medicinal extracts [6,7,8,9] to biofuel production [10]. However, similar research on newly discovered, related strains is less extensive but equally promising. One of these new interesting strains is Neochlorella semenenkoi IPPAS C-1210 [11], the focus of this study. This strain is notable for its metabolic flexibility, capable of mixotrophic and heterotrophic growth on organic carbon sources like acetate, and can utilize various nitrogen sources, including ammonium [11].

At the same time, it is well known that the production of certain metabolites and lipids strongly depends on the growth conditions of microalgae. For example, the production of biologically active antioxidants by Chlorella vulgaris doubles under the influence of Cu and Zn ions [12]. A significant effect on the metabolism and production of bioactive substances in Chlorella vulgaris is also exerted by a change in the trophic regime: mixotrophy and heterotrophy promote the growth and synthesis of lipids and saccharides by reducing expenses and energy for maintaining the physiological apparatus. This occurs by enhancing key metabolic pathways such as glycolysis, the TCA cycle, and PPP. In other words, the use of organic carbon (mixotrophy and heterotrophy) “accelerates” energy metabolism and leads to the accumulation of lipids and carbohydrates, while photoautotrophy is more focused on maintaining the photosynthetic apparatus and protein synthesis [13,14]. Changes in growth conditions (especially the addition of stress factors) also strongly affect lipid production in the Chlorella sorokiniana strain [15]. Moreover, the effect depends on the type of microalgae nutrition (photoautotrophic or heterotrophic) and the genetic characteristics of the strain, indicating that biodiversity will play a significant role in the peculiarities of metabolism.

Another key factor affecting the metabolism of Chlorella sp. strains is nitrogen starvation; in particular, it especially stimulates the accumulation of lipids in Chlorella vulgaris [16] Starch stored in cells during nitrogen deprivation is promptly degraded during the recovery phase. This indicates that starch serves as a source of carbon and energy to trigger the metabolic processes necessary for the resumption of cell growth and division [17] In certain cases, such as with the tropical strain of Chlorella sp. UMACC050, nitrogen deprivation results in diminished biomass yield and photosynthetic efficiency, while concurrently enhancing lipid content and altering fatty acid composition (increase in monounsaturated fatty acids, decrease in polyunsaturated fatty acids). Metabolic changes indicate that cells assimilate nitrogen from proteins and redistribute carbon into carbohydrates and lipids in response to nitrogen deficiency [18].

Here, we analyze both the lipid and metabolic profiles of a new strain of chlorellacean microalgae Neochlorella semenenkoi IPPAS C-1210 under nitrogen starvation and recovery. Moreover, our study has been supplemented with isotope labeling in vivo, which made it possible to better assess the rates of biosynthetic processes than measuring concentrations.

Isotopic labeling (^13^C) is an important tool for estimating metabolic fluxes (MF) in Chlorella protothecoides. It has been successfully utilized in plant research in conjunction with chromato-mass spectrometric analysis [19], and isotopic labeling (^13^C) is an important tool for quantifying metabolic fluxes (MF) in Chlorella protothecoides. ^13^C-MF analysis allowed us to compare cell metabolism under phototrophic and heterotrophic conditions, identify key metabolic pathways associated with lipid synthesis, and understand how different cultivation conditions affect these pathways [20]. A similar isotope labeling method has also been used for Tetradesmus obliquus to quantify fluxes in individual reactions, determine how individual fatty acids are synthesized (de novo or by modification), and identify enhanced glycolysis and PPP to provide substrates and reducing agents necessary for fatty acid synthesis [21]. Isotopic labeling is also used to study trophic relationships between microalgae and other organisms. For example, isotopic labels ^13^C and ^15^N make it possible to track and quantify the transfer of carbon and nitrogen from macroalgae to sea urchins at different temperatures, as well as to separate the effect of temperature on the assimilation of these two elements. This provides a more detailed understanding of trophic interactions in marine ecosystems and the effects of environmental changes [22] The C^13^ and N^15^ isotope labeling approach also makes it possible to comprehensively study carbon and nitrogen fluxes in sediments, as well as to determine the role of various organisms (microalgae and bacteria) in these processes and the influence of macroalgae on these fluxes [23].

At the same time, there has been little work on the use of isotopically labeled nitrogen for the analysis of algae metabolism under conditions of nitrogen starvation in recent years, and since the second half of the last century, the N^15^ isotope labeling method has been practically not used in conditions of microalgae growth on an isotopically labeled medium for the study of microalgae metabolites [24] That is why we decided to use N^15^ isotope labeling with the application of high-resolution mass spectrometry equipment to study changes in the metabolic and lipidomic profile of a new strain of green microalgae Neochlorella semenenkoi IPPAS C-1210 under nitrogen starvation and recovery.

2. Results

2.1. Analysis of the Metabolic Composition of Microalgae

The HPLC-MS analysis allowed us to identify at least 130 different compounds of polar metabolites (Supplementary Materials, Table S1) found in E IPPAS C-1210 under conditions of normal growth (before starvation), under conditions during nitrogen starvation, and after nitrogen addition (recovery). Among these metabolites, amino acids, which were measured semi-quantitatively and described below, sugars (such as sucrose), nitrogenous bases (such as adenine, uracil) and their derivatives (xanthine, cyclic ADP-ribose and other), enzyme cofactors (nicotinamide, riboflavin and other), and carboxylic acids, including components of the Krebs cycle (such as alpha-ketoglutaric acid, malic acid and succinic acid), were found.

All compounds were divided into functional groups depending on the metabolic pathway in which they are involved (Figure 1). The largest number of metabolites was found from the metabolic pathways as: fatty acid and lipid metabolism, glutamate family amino acid syntesis and carbohydrate metabolism (Figure 1A). Also large number of peptides (dipeptides) has been found, which are most likely fragments of proteins destroyed during homogenization and sample preparation. We could not attribute a known or unambiguous function to the 9 found metabolites (named «Unidentified»).

Figure 1B presents a heatmap of the total normalized peak areas of all metabolites within each pathway, which serves as a proxy for pathway-level metabolic activity or carbon/nitrogen flux. This visualization reveals a striking decoupling between metabolite diversity and pathway activity under nitrogen stress. Amino acid metabolism showed three distinct patterns.

Glutamate family and pyruvate family pathways exhibited a sharp decline (approximately 2-fold) under nitrogen starvation, followed by complete recovery upon nitrogen resupply. This pattern reflects the expected sensitivity of nitrogen-rich amino acid pools to nitrogen limitation.

In contrast, the aspartate family pathway was largely maintained during starvation (only a minor decrease) and showed a pronounced increase during recovery, exceeding pre-starvation levels. This suggests that aspartate-derived metabolism is selectively preserved under nitrogen stress, likely to sustain nucleotide biosynthesis and prepare for rapid cell division upon nitrogen availability.

The serine family pathway remained remarkably stable across all conditions, indicating a constitutive basal requirement for serine-derived metabolites.

Central carbon metabolism exhibited distinct responses to nitrogen manipulation. TCA cycle intermediates declined sharply under nitrogen starvation but not only recovered upon nitrogen resupply—they exceeded pre-starvation levels, suggesting a metabolic overshoot phenomenon. In contrast, glycolytic intermediates did not decline under starvation and instead showed a moderate increase, which became substantially more pronounced during recovery. Fatty acid and lipid metabolism was not suppressed during nitrogen starvation and increased during recovery, consistent with carbon redistribution toward membrane remodeling or triacylglycerol storage. Nucleotide metabolism pathways showed distinct dynamics: both purine and pyrimidine synthesis declined under starvation, but pyrimidines recovered fully, whereas purines showed only partial recovery. This differential response may reflect tighter regulation of purine biosynthesis under nitrogen limitation.

Pathways with very few detected compounds (e.g., AA conjugate/TCA conjugate, Carbohydrate Metabolism/Pyrimidine Synthesis) exhibited low total abundance and high variability, reflecting their specialized or transient metabolic roles. It is noticeable that the peak area of conjugated amino acids (AA), including the number of conjugated amino acids and Krebs cycle metabolites (AA conjugate/TCA conjugate) during starvation, decreases, which correlates with a decrease in the total peak area of the Krebs cycle metabolites (Citric acid cycle (TCA)). At the same time, during starvation, the total peak area of amino acid derivatives (AA derivative), such as choline, homo-L-arginine, hopantenic acid, and pentahomoserine, increases significantly. It is important to note that choline makes the main contribution to this total area of chromatographic peaks for a group of amino acid derivatives; at the same time, other nitrogen-containing amino acid derivatives behave differently. For choline, the maximum area is observed in the sample subjected to starvation (Figure 2). Pentahomoserine and homo-L-arginine accumulate under nitrogen starvation as well, but their maximum is observed after 20 h of nitrogen addition. Conversely, the peak area of hopantenic acid is minimal under starvation.

The aminated carbohydrate, meglumine, behaves in a similar way to choline in terms of accumulation during starvation (Figure 1B). However, no information could be found about its role in microalgae.

The total peak area in the samples under starvation is reduced relative to the samples before starvation in such groups of metabolites as amino acids synthesized from pyruvate, amino acids synthesized from glutamate, the shikimate biosynthetic pathway, as well as the synthesis of vitamins and cofactors. At the same time, the total peak area in such groups of metabolites as amino acids synthesized from aspartate, purine and pyrimidine biosynthesis, and amino acids synthesized from serine remains virtually unchanged or even increases.

It is interesting to note that, although the total peak area of the group of adenosine derivatives appears to decrease during starvation, this pattern is only seen for two of the three metabolites of this group: S-Adenosylmethionine and 5′-S-Methyl-5′-thioadenosine. In contrast, the peak area of Cyclic ADP-ribose increases during starvation (Figure 3).

To complement the pathway-level view presented in Figure 1B, we examined the behavior of individual metabolites within each functional group. A complete list of all detected compounds, their peak areas and fold changes is provided in Supplementary Table S1.

2.2. N15 Isotopic Labeling of Amino Acids Compared with Their Quantitative Analysis

In the samples of microalgae after starvation, where we added ammonium acetate with N15, we found the presence of an isotopic label (Figure 4). The amount of amino acids in all samples of microalgae extracts was estimated using calibration curves (Supplementary Materials, Figures S1–S17). The result of the quantitative analysis of 17 amino acids is shown in Table 1. How the amount of each amino acid changes under the three sampling conditions can be seen in Figure 5.

For most amino acids, the pattern of concentration changes looks like this: before nitrogen starvation, the amino acid content is maximal; during starvation, it decreases; and after nitrogen addition, it is restored. However, some of the amino acids in microalgae cells, after adding nitrogen after nitrogen starvation, are observed in even greater quantities than before starvation. This behavior is observed for alanine, phenylalanine, and tyrosine. In this case, phenylalanine and tyrosine are formed as a result of a single metabolic pathway (the Shikimate biosynthesis pathway), which is known to be activated after starvation (Figure 1B), if we analyze the total peak area of the metabolites included in it. However, alanine, which belongs to the group of amino acids synthesized from pyruvate, is the only amino acid whose maximum is observed precisely by adding nitrogen after starvation and not before starvation.

The Table 1 also shows the estimated percentage of labeling, calculated as the ratio of the change in amino acid concentration from the moment after starvation to the moment of nitrogen addition after starvation, and the percentage of real labeling calculated by the ratio of peak areas (+N^15^) to the total sum of peak areas N^14^ + N^15^. These data are very remarkable, since the percentage of isotopic labeling is very different from the simple percentage of the increment in amino acid content during recovery after starvation (expected percentage of labeling). For example, for glutamate, we observe an increase in concentration of only 6%, while isotopically labeled glutamate in microalgae samples reaches 91% 20 h after the addition of labeled nitrogen N^15^ to the culture under starvation.

This may be because the concentration of amino acids does not fully reflect all the biochemical processes occurring in cells, since amino acids are not only synthesized de novo but are also consumed by synthetic processes, embedding into proteins or forming other derived compounds. For example, glutamate is one of the first compounds that accumulate inorganic nitrogen into organic compounds and serves as the actual precursor of all nitrogen-containing compounds [25,26].

We combined amino acids by biosynthetic groups (Figure 6; Supplementary Materials, Table S2), relative to the known information about the biosynthetic families of amino acids [27], to then make assumptions about the activity of biosynthetic pathways by the percentage of labeling of amino acids in each biosynthetic family (Figure 6B). We performed a statistical analysis of the percentage of amino acid labeling: a one-factor analysis of variance (ANOVA) was performed to assess the effect of the biosynthetic group on the rate of amino acid synthesis. The ANOVA results did not show statistically significant differences in the rate of synthesis between different biosynthetic groups (F(3, 14) = 0.32, p = 0.81) (Figure 6A). All the presented biosynthetic groups show high average labeling percentages in the range from 88% to 93%. This indicates that the amino acids found in microalgae samples are actively synthesized under experimental conditions. Special attention should be paid to the biosynthesis of aromatic amino acids. Figure 5 shows a small standard deviation for the Shikimate biosynthesis pathway group, which may indicate that the synthesis of amino acids in this group is controlled to a greater extent than the synthesis of all other amino acids. Despite this, based on the available data, it is impossible to conclude that there are statistically significant differences between the groups. To confirm this hypothesis and identify possible differences, we plan to do a more detailed analysis.

2.3. Isotopic Labeling of Non-Targeted Compounds

At least 43 nitrogen-containing compounds were found in addition to the targeted search for amino acids and their derivatives of dipeptides (Supplementary Materials, Table S3).

We performed a similar analysis as for amino acids of the mean values and standard deviations of the percentage of isotopic labeling of metabolites (Figure 7). A one-factor analysis of variance (ANOVA) was performed to assess the effect of a group of metabolites on the percentage of labeling. The ANOVA results showed statistically significant differences in the percentage of labeling between different groups of metabolites (F(9, 21) = 2.74, p = 0.027).

The data show a significant difference in the synthesis activity of various biochemical groups. The groups associated with adenosine derivatives, purines, and amino acid conjugates exhibit very high and stable activity. The synthesis of carbohydrate derivatives, vitamins/cofactors, protective compounds, and amino acid derivatives proceeds at a lower rate and with greater variability.

The largest standard deviation for the group of protective compounds was obtained due to nicotinic acid, for which no isotopically labeled peak was detected. The highest concentration of this substance is observed in culture fluid samples after nitrogen addition after starvation, however, the functions of this substance and the pathways of metabolism in microalgae cells are unknown.

A low percentage of isotopic labeling is also observed in homo-L-arginine (14%) and carnitine (43%) compounds. The exact role of homo-L-arginine in chlorella cells and other related organisms is unknown, but this substance is being actively studied as a biologically active substance [28] and is also used in cosmetology as a photoprotector [29]. It may also play a protective role in cells and accumulate slowly during culture growth. Carnitine has a beneficial effect on the lipid metabolism of algae of the genus Chlorella, promoting the synthesis of pigments and transferring long-chain fatty acids from the cytoplasm to the mitochondrial matrix, where their oxidation occurs [30].

2.4. Lipid Analysis

Lipids were extracted from the Neochlorella semenenkoi IPPAS C-1210 samples based on the protocol described below. Untargeted UPLC-MS/MS analysis yielded 222 lipid species from 11 lipid classes, annotated by MS/MS spectra. Most of the annotated lipids were from the triacylglycerol (TG, n = 121), diacylglycerol (DG, n = 14), diacylglyceryl-(N,N,N-trimethyl)-homoserine (DGTS, n = 24), and lyso-diacylglyceryl-(N,N,N-trimethyl)-homoserine (LDGTS, n = 11) lipid classes (Figure 8A). Unfortunately, galactosyl derivatives of lipids are not widely represented. These lipids had relatively low intensities and were not confirmed by fragmentation; therefore, they were excluded from further analysis and discussion. On the level of fatty acid residues, the most abundant fatty acids were saturated and monounsaturated fatty acids 18:0, 18:1, 16:0, and 16:1, as well as polyunsaturated fatty acids 18:2, 18:3, and 16:3. The distributions of lipid classes and fatty acid residues intensities are represented in Figure 8B,C.

We found that under starvation conditions, the accumulation of glycerolipids such as triglycerides (TG) and diglycerides (DG) was activated, while the levels of phospholipids like diacylglyceryltrimethylhomoserine (DGTS and LDGTS) decreased (Figure 9A,B). Interestingly, the total lipid content increased during nitrogen starvation, indicating activation of fatty acid biosynthesis alongside the synthesis of storage lipids, primarily TG.

A notable observation involved unusual changes in the relative abundance of choline- and betaine-derived lipids. Under starvation, lyso-DGTS (LDGTS) content decreased, while lysophosphatidylcholine (LPC) increased, despite the fact that nitrogen is required for the synthesis of both classes of lipids. This correlated with elevated phosphocholine levels detected during starvation. Conversely, upon restoration of nitrogen availability, LPC levels decreased and LDGTS increased, suggesting a regulatory balance between these phospholipid classes.

The content of the Cer lipid class, which also contains nitrogen, decreased during starvation and increased with nitrogen availability, broadly correlating with betaine-containing phospholipid levels. Finally, the proportions of esterified fatty acids showed no significant differences across the three conditions (repeated measures ANOVA, p = 0.62).

Analogous to amino acids, we analyzed the incorporation of the ^15^N isotope label into nitrogen-containing lipid compounds. The distribution of ^15^N/^14^N isotopic intensities in lipid classes after growth in ^15^N-enriched medium is shown in Figure 9C. On average, 76% of all annotated lipids incorporated the label: 70% of DGTS, 78% of LDGTS, 76% of LPC, and 78% of Cer. The distribution of label incorporation ratios did not differ significantly among lipid classes (one-way ANOVA, p = 0.38).

In addition, some compounds in the lipid fraction were not found in the databases. However, we were able to presumably identify them by the inclusion of N^15^ isotopic label. For example, a compound with a mass of 607.29 in samples with the addition of isotopically labeled nitrogen acquired a mass of 611.28, which suggested that there were at least 4 nitrogen atoms in this molecule (Figure 10). By exact weight, we found a suitable candidate from the group of porphyrins derived from chlorophyll, Methyl pheophorbide A. While such compounds are commonly identified by their absorption spectra in spectrophotometry, this method does not provide information about functional groups or the exact molecular composition. Mass spectrometry, in contrast, enables precise structural characterization.

3. Discussion

Using isotopic labeling, we demonstrate that Neochlorella semenenkoi IPPAS C-1210 can effectively absorb ammonium acetate from the media and assimilate it into its metabolism.

To understand the processes of ammonium incorporation into amino acids, it is important to mention the main reactions involved in this process. In plants, the enzyme glutamine synthase (GS) plays a major role in this process [31]. This enzyme, together with glutamate synthase (glutamate-2-oxoglutarate aminotransferase, GOGAT), forms a small cyclic process, as a result of which two glutamate molecules are formed: one goes to synthesize other amino acids, and the second adds another ammonium, accumulating nitrogen bound to organic substances. The second rather controversial enzyme, possibly involved in the assimilation of ammonium, is glutamate dehydrogenase (GDH). Despite the fact that the role of GDH is being discussed, it is known that related strains of the Chlorella sp. have a gene for this enzyme, and its work can be induced by external factors. It is also known that in vivo, this enzyme functions in the synthesis of glutamate in the thermophilic strain Chlorella pyrenoidosa [30].

Thus, theoretically, both described ammonium inclusion systems (GS-GOGAT and GDH) can participate in the incorporation of an isotopic label from ammonium into amino acids (Figure 11). However, the percentage of isotopic labeling of those amino acids can give us some ideas about the ratio of these pathways in microalgae cells.

If we consider the incorporation of ammonium predominantly through GS, then we can expect that almost all isotopically labeled ammonium ^15^N is incorporated into glutamine. The inclusion of ammonium through GS and GDH will lead to a decrease in the percentage of glutamine labeling since some of the ammonium will be directly incorporated into glutamate. That is, if GDH is active and makes a more significant contribution to the incorporation of nitrogen into amino acids, then we can expect that the percentage of labeling of glutamate will be higher than the percentage of labeling of glutamine.

In our case, we see that the percentage of labeling is high for both glutamate (91%) and glutamine (98%), however, the percentage of labeling of glutamine is higher than that of glutamate, which indicates the great contribution of the GS-GOGAT system to the process of incorporation of ammonium into amino acids. At the same time, a large percentage of labeled glutamate suggests that the role of GDH in the synthesis of labeled glutamate cannot be excluded. It is important to note that we make several assumptions in this reasoning. Firstly, we assume that the synthesis of amino acids exceeds their breakdown, and the high level of isotopic label incorporation into amino acids indicates this. Secondly, we believe that for all subsequent reactions of amino acid and protein synthesis, the outflow of labeled and unlabeled amino acids is the same, and for subsequent enzymatic systems, the inclusion of an isotopic label does not affect their activity: that is, the ratio of the isotopic label does not change in subsequent reactions, but is caused only by the inclusion of the isotopic label. This is indicated by the presence of an isotopic label in amino acids of the glutamate family (proline, arginine, etc.) and amino acids that have received their amino group as a result of transamination, where glutamate is also the source of the amino group.

The inclusion of isotopically labeled nitrogen not only indicates the contribution of various biochemical processes in the cell, which cannot always be determined by measuring concentrations, but also correlates with the results of lipid and metabolite analysis.

As we noted previously, the amount of choline (based on the peak area) increases under starvation conditions (Figure 2A). This increase in choline levels correlates with changes in the lipid profile. During starvation, the content of both DGTS and lyso-DGTS decreased, while lysophosphatidylcholine (LPC) increased. It is known that DGTS are membrane lipids [32,33]. It is also known that for some other green microalgae, as well as plants, DGTS can serve as a source for the synthesis of other lipids, such as TAG [34]. The accumulation of LPC together with the accumulation of free choline may indicate that during the lipid remodelling, the choline precursors, released from DGTS and LDGTS lipids, are metabolized to free choline and then to LPC lipids. Our observations indicate that free choline and choline-containing lipids are not mobilized as a nitrogen source under nitrogen starvation in this strain. It is interesting to note that even though the amount of free choline after the addition of nitrogen after starvation decreased by about 10% (Supplementary Materials, Table S1), the percentage of choline N^15^ isotope labeling in samples with N^15^ was 66%. This percentage of labeling is less than the average value for different groups of amino acids, but at the same time, there are significantly greater concentration changes. Apparently, after adding nitrogen, the pool of nitrogen-containing membrane lipids is restored because, in both of these groups, we see a much larger proportion of labeled nitrogen-containing lipids relative to unlabeled ones.

A change in the amount of other compounds produced may also indicate their function. The peak area of 5′-S-Methyl-5′-thioadenosine and S-adenosylmethionine decreased under starvation conditions, but the peak area of cyclic ADP-ribose, on the contrary, increased during starvation. At the same time, the percentage of labeling of all these adenosine derivatives is close to 100%, which indicates a high degree of conversion of these substances. 5′-S-Methyl-5′-thioadenosine and S-adenosylmethionine are closely metabolically related, participating in methylation reactions and other metabolic processes, but cyclic ADP-ribose in the studied organisms is primarily a signaling molecule [35,36]. Unfortunately, the specific cascade of signaling mechanisms of cyclic APD-ribose has not been described for Chlorella and similar strains, but for the green algae Ulva, the role of cyclic ADP-ribose may be associated with calcium signaling and regulation of gene expression necessary for cell membrane repair under high light stress [37]. We assume that the accumulation of cyclic ADP ribose may be related to the signaling role of this molecule in conditions of nitrogen starvation, but this needs further research.

The TCA cycle intermediates, which declined sharply under nitrogen starvation, not only recovered upon nitrogen resupply but exceeded pre-starvation levels (Figure 1B), suggesting a metabolic overshoot phenomenon. This overshoot may reflect enhanced anaplerotic flux and increased demand for carbon skeletons to support rapid amino acid biosynthesis during recovery. Consistent with this, the accumulation of aminated carbohydrates (e.g., glucosamine derivatives, nucleotide sugars) points to increased demand for cell wall precursors and glycosylation substrates during growth resumption. Interestingly, several adenosine derivatives, including cyclic adenosine ribose, accumulated specifically during recovery. While their precise role in microalgae remains unclear, similar compounds have been associated with calcium signaling and stress responses in other eukaryotes. Metabolomics alone cannot resolve whether these shifts are driven by transcriptional or post-translational regulation; future studies integrating genomics, transcriptomics, and proteomics will be required to identify the molecular mechanisms underlying the adaptive strategies reported here.

4. Materials and Methods

4.1. Microalgae Strain and Selection of Growth Conditions

The non-axenic green algae strain Neochlorella semenenkoi IPPAS C-1210 was obtained from the Collection of Microalgae and Cyanobacteria IPPAS of the Institute of Plant Physiology, Moscow, Russia. Neochlorella semenenkoi IPPAS C-1210 is traditionally grown for experiments under sterile conditions on a BG-11 nutrient medium [38,39]. However, for the current experiment, we modified the medium so that NaNO_3_ was replaced with ammonium acetate with equimolar nitrogen content (17.6 mM). Before starting the main experiment, the strain was grown on a modified BG-11 medium with ammonium acetate as a nitrogen source and showed a good growth rate without changes in cell ultrastructure and pigmentation.The experiment was performed in three stages. First, the algae were grown grown for 5 days in a batch mode to increase biomass. The cultivation was performed inn sterile conditions on a BG-11 nutrient medium with ammonium acetate, with intensive CO_2_ supply (1.3–1.8%), constant illumination (300 µmol m^−2^ s^−1^ 300 mE), and constant temperature (32 °C) using a laboratory system for intensive cultivation). The nitrogen starvation lasted 3 days. Following the starvation period, the culture was divided into two parallel batches to initiate the labeling experiment. An appropriate amount of ammonium acetate (containing 17.6 mM N) was added to each batch: one batch received ammonium acetate containing ^14^N, while the other received ^15^N. All cultivation conditions were performed in three independent biological replicates.

The biomass for the experiment was collected at the three experimental conditions:

• (1)before starvation, after growing on the full medium for five days,
• (2)under nitrogen starvation, after incubation in nitrogen free medium for three days
• (3)in the recovery phase, 20 h after adding ammonium acetate as a nitrogen source (N^14^ as a control and N^15^ as an experiment).

At each of the three time points, the cells were lyophilized and then stored at -80 °C until further experiments. Biomass was collected at each time point during the exponential growth phase to ensure metabolic consistency across samples.

4.2. Sample Preparation of Polar Metabolites of Microalgae

For the extraction of algae samples, we used liquid-liquid extraction by Folch [40]. Briefly, the sample preparation was as follows: 200 µL of H_2_O and 320 µL of cold methanol (−20 °C) were added to 10 mg of lyophilized cells, followed by ultrasonication for 1 h to ensure homogenization. Then, 640 µL of cold chloroform was added to the same tube and incubated for 20 min in a 4 °C thermoshaker. Next, 300 µL of H_2_O was added and incubated for 10 min in a thermoshaker at 4 °C. The sample tubes were centrifuged at 2000× g for 5 min and 4 °C. After phase separation, the lower phase (nonpolar metabolites) was taken into a new tube, and 250 mL of CHCl_3_:MeOH (2:1, v/v) was added to the residues. Then, the contents of the tubes were centrifuged at 2000× g for 5 min at 4 °C, and the lower phase was taken back into the same tube. The upper phase was selected separately; it contains polar metabolites for analysis. Next, tubes with polar metabolites were dried in a vacuum concentrator without heating, and the contents were reconstituted in 80% acetonitrile. Next, this extract was sent for analysis on HPLC-MS.

4.3. Sample Preparation of Microalgae Lipids

The initial extraction steps coincide with the steps described above in the Section 4.2 and they were conducted in parallel. Thus, after liquid-liquid extraction, the lower phase was taken into a separate tube. The samples were then evaporated in a vacuum concentrator without heating and reconstituted in an acetonitrile:isopropyl alcohol 70:30 solution. Due to the extraction of mostly nonpolar lipids, this technique did not allow the extraction of a sufficient number of galactolipids for MS/MS analysis.

4.4. Preparation of Calibration Solutions of Standards for Semi-Quantitative Analysis

Amino acid reference materials were used to prepare calibration solutions: alanine (Sigma production, St. Louis, MO, USA), lysin (CDH productions, Clovis, CA, USA), glutamic acid (CDH productions, Clovis, CA, USA), serine (Sigma production, St. Louis, MO, USA), arginine (CDH productions, Clovis, CA, USA), aspartic acid (CDH productions, Clovis, CA, USA), phenylalanine (CDH productions, Clovis, CA, USA), proline (CDH productions, Clovis, CA, USA), glutamine (Panreact production, Chicago, IL, USA), cystein (Sigma production, St. Louis, MO, USA), leucine (Sigma production, St. Louis, MO, USA), isoleucine (Sigma production, St. Louis, MO, USA), ornitine (CDH productions, Clovis, CA, USA), asparagine (Panreact production, Chicago, IL, USA), hystidine (CDH productions, Clovis, CA, USA), tyrosine (CDH productions, Clovis, CA, USA), threonine (CDH productions, Clovis, CA, USA), treptophan (CDH productions, Clovis, CA, USA), glycine (Sigma production, St. Louis, MO, USA).

5 mg of each standard was diluted in 2 mL of H_2_O, and then the corresponding calibration points were prepared from this stock solution by sequential dilution. 250 µL of standards dissolved in water were dried using a vacuum concentrator without heating. The samples were reconstituted in 80% acetonitrile and analyzed using HPLC-MS.

4.5. HPLC-MS Analysis and Data Processing

To study the composition of metabolites HPLC–MS analysis was performed on a LTQ Orbitrap Velos system (Thermo Scientific Inc., Waltham, MA, USA) coupled with a Thermo Scientific Accela 1250 HPLC. Separation was done with Waters (Waters, MA, USA) Acquity UPLC Flush BEH HILIC column (2.1 × 150 mm, 1.7 μ) in the following gradient: 100% mobile phase B at 0–1 min, 100% to 10% mobile phase B at 1–20 min, 10% mobile phase B at 20–22 min, 10% to 100% mobile phase B at 22–25 min, 100% B at 25–33 min.

The flow rate was set at 0.2 mL min^−1^. The column temperature was maintained at 20 °C. Solutions of 0.1% formic acid with 0,176% ammonium formiate in acetonitrile and water were used as mobile phases A and B, respectively. MS analysis was performed in both ESI-positive and negative modes with ESI voltages of 4.5 and 3.5 kV. Data acquisition was performed both in full scan with high resolution (60,000) and DDA mode (Top 5) with stepped collision energy (10, 30, 45 V). The isolation window was set at 1 Da.

To study the composition of lipids HPLC–MS/MS analysis was performed on a LTQ Orbitrap mass spectrometer (Thermo Scientific, Waltham, MA, USA) coupled with a Thermo Scientific Accela 1250 HPLC system. Separation was carried out on an ACQUITY UPLC C18 column (2.1 × 150 mm, 1.7 µm, Waters, Milford, MA, USA). The mobile phase consisted of (A) acetonitrile/water (50:50, v/v) and (B) isopropanol/acetonitrile (90:10, v/v), both containing 10 mM ammonium formate and 0.2% formic acid. The flow rate was 150 µL/min, and the injection volume was 2 µL. The gradient program was as follows: 0–0.5 min, 30% B; 0.5–16 min, 30–100% B; 16–24 min, 100% B; 24–25 min, 100–30% B; 25–30 min, 30% B.

The mass spectrometer was operated in positive and negative ESI modes with data-dependent acquisition (DDA). Full scan range was m/z 200–2000 at a resolution of 70,000. The top 10 most abundant ions were fragmented using stepped normalized collision energy (20, 30, 40 eV). Lipid identification was performed using LipidSearch 5.0 and manual MS/MS validation.

4.6. Data Analysis

HPLC–MS data were processed with XCalibur 4.0 and Compound Discoverer 3.3 software (Thermo Scientific, Inc.) and with MS Dial [41,42]. A common routine, including peak picking, deconvolution, and deisotoping, was used to detect components. For primary components annotation, accurate masses and MS/MS spectra of detected components were searched against the mzCloud and MassBank [43] databases with either Compound Discoverer or with MS Dial 4.20 software. Mass accuracy was set at 10 ppm, and the annotation score threshold was set at 70.

For normalization, peak areas were normalized to the total ion current (TIC) of each sample to correct for technical variation. In addition, a pooled quality control (QC) sample was injected every 10 runs to monitor instrument stability and correct for signal drift; features with a relative standard deviation (RSD) > 30% in QC samples were excluded from further analysis.

All statistical analyses were performed using R 4.1.0. Group comparisons were evaluated by one-way ANOVA followed by Tukey’s HSD post-hoc test. Differences were considered statistically significant at p < 0.05.

4.7. Metabolite Annotation and Identification Confidence

Metabolite identification was performed according to the Metabolomics Standards Initiative (MSI) guidelines. Amino acids and nucleotides were confirmed at MSI Level 1 by comparing retention times and MS/MS spectra with those of authentic reference standards analyzed under identical instrumental conditions.

For all other metabolites, identification was based on MSI Level 2 using high-resolution MS/MS spectral matching against the MZcloud database. Collision energies were optimized to maximize fragment ion coverage. Only spectral matches with a mirror match factor ≥ 75% and at least three matching diagnostic fragment ions were accepted. All putative annotations were manually validated by an experienced operator.

To further reduce false positives, ^15^N isotopic labeling was used as an orthogonal confirmation tool. Compounds annotated as nitrogen-containing were verified by the presence of the expected mass shift (+0.997 Da per nitrogen atom) in the ^15^N-labeled samples. Compounds lacking this shift or exhibiting ambiguous isotopic patterns were excluded from further analysis.

5. Conclusions

In this article, we demonstrate a broad metabolomic and lipidomic analysis for a previously uncharacterized strain of microalgae, Neochlorella semenenkoi IPPAS C-1210. For the first time, we demonstrate that this strain can grow on a modified BG-11 medium, where ammonium acetate acts as a nitrogen source. In particular, we demonstrate that the isotopic label coming from ammonium acetate is successfully incorporated into the metabolites of this strain of microalgae.

We present data on the quantitative analysis of amino acids and also on the isotopic distribution in these amino acids. While the concentration change may be insignificant (for example, for glutamate, it is only 6%), the percentage of isotopic label inclusion is very high, which may indicate a high rate of amino acid metabolism under the described conditions, and may also suggest a major source of nitrogen incorporation into amino acids. We also demonstrate a wide range of non-targeted compounds found, which also include an isotopic label. The distribution of isotopic labeling often differs for other compounds from the percentage of labeling in amino acids and is usually lower.

We performed a broad lipidomic analysis, including an analysis of nitrogen-containing lipids. Lipidomic analysis was also performed for this microalgae for the first time. Our results show that under conditions of nitrogen starvation, a significant restructuring of the lipid composition occurs in cells. The accumulation of free choline, along with changes in the lipid profile (decrease in lyso-DGTS and increase in LPC), suggests that phosphatidylcholines (PC) play a key role in lipid remodeling aimed at the synthesis of reserve lipids such as TAG and DAG. The release of choline during the synthesis of reserve lipids is consistent with the observed accumulation of free choline. Interestingly, after the addition of nitrogen, the pool of nitrogen-containing membrane lipids is restored, as evidenced by a significant increase in the proportion of ^15^N-labeled choline, despite a slight decrease in its total concentration. This indicates the active incorporation of nitrogen into lipid metabolism after the removal of stressful conditions of nitrogen starvation.

By changing the peak area under different growth conditions, we can also assume some functions of the molecules. For example, we assume the signaling role of cyclo-ADP-ribose for signaling stress associated with nitrogen deficiency, but these assumptions need to be analyzed in more detail.