Optimization of Heterotrophic Culture Conditions for the Microalgae Euglena gracilis to Produce Proteins

Euglena gracilis is one of the few permitted edible microalgae. Considering consumer acceptance, E. gracilis grown heterotrophically with yellow appearances have wider food industrial applications such as producing meat analogs than green cells. However, there is much room to improve the protein content of heterotrophic culture cells. In this study, the effects of nitrogen sources, temperature, initial pH, and C/N ratios on the protein production of E. gracilis were evaluated under heterotrophic cultivation. These results indicated that ammonium sulfate was the optimal nitrogen source for protein production. The protein content of E. gracilis cultured by ammonium sulfate increased by 113% and 44.7% compared with that cultured by yeast extract and monosodium glutamate, respectively. The manipulation of the low C/N ratio further improved E. gracilis protein content to 66.10% (w/w), which was 1.6-fold of that in the C/N = 25 group. Additionally, amino acid analysis revealed that the nitrogen-to-protein conversion factor (NTP) could be affected by nitrogen sources. A superior essential amino acid index (EAAI) of 1.62 and a balanced amino acid profile further confirmed the high nutritional value of E. gracilis protein fed by ammonium sulfate. This study highlighted the vast potency of heterotrophic cultured E. gracilis as an alternative dietary protein source.


Introduction
The global shortage of protein supply will be one of the biggest problems in the next 30 years, especially when the population is projected to exceed 9.1 billion, and the meat demand might rise to 455 million tons by 2050 [1,2].Recently, animal protein production was considered unsustainable because traditional animal agriculture is harmful to the environment, requiring large amounts of resources (e.g., land and water) and emitting many greenhouse gases [1,2].Therefore, the animal-free protein produced by legumes, cereals, and microbial species is increasingly popular and has shown increasing market share in recent years [3].
Microalgae have emerged as a promising bio-economy practical resource for plantbased protein production with the advantages of high protein content (up to 71%), less land requirement, no season limitation, and a higher growth rate than traditional crops [4,5].Euglena gracilis has several advantages to be applied as a single-cell protein resource among various microalgal species.For example, E. gracilis was approved as a dietary resource in Europe, and was also permitted to be used in China without the daily intake limitation [6].
Meanwhile, it contains 18 amino acids with a high proportion of essential sulfur amino acids, providing a healthy balance of essential amino acids to meet daily protein intake requirements [7,8].In addition, E. gracilis is highly digestible as it lacks a rigid celluloid cell wall [8].More importantly, E. gracilis can produce yellow-color biomass during their heterotrophic growth due to the degradation of secondary-endosymbiosis chloroplast [9], giving it wider application in the food industry such as meat analog [6].
The protein content of E. gracilis can reach up to 60% (w/w) and 47% (w/w) under mixotrophic and autotrophic culture conditions, respectively [10,11].However, E. gracilis grown heterotrophic mainly accumulates storage carbohydrates (paramylon), and the protein content could be less than 20% (w/w) [12].The low protein production of E. gracilis under heterotrophic conditions limits its wide application as a protein resource.
Nitrogen sources were one of the essential factors for microalgae protein biosynthesis [13].Chlorella sorokiniana prefers ammonium salt for protein production [14], while Scenedesmus obliquus prefers urea [15].The prevailing view is that sodium glutamate and ammonium salt can be utilized by E. gracilis [10,16], while nitrate and urea are not suitable substrates due to the lack of nitrate reductase and urease in E. gracilis [17].Nitrogen starvation is a commonly used strategy to promote lipid accumulation in microalgae, while a decrease in the C/N ratio was also found to be beneficial for protein accumulation by various microalgae such as C. sorokiniana [14], Porphyridium purpureum [18], and Dunaliella sp.[19].However, Regnault et al. (1990) claimed that the C/N ratio had little effect on protein content in E. gracilis [20].
It is necessary to systematically explore the optimal cultivation strategies for E. gracilis under heterotrophic conditions for improving its protein content.The influences of the nitrogen source and C/N ratio on protein accumulation in E. gracilis are not clearly studied.Meanwhile, initial pH and temperature have not yet been explored its effect on protein production in this species as well.In the present study, it is hypothesized that E. gracilis growing heterotrophically could have a protein content improvement by optimizing its cultivation conditions.The effects of the nitrogen source, temperature, initial pH, and C/N ratio on the protein content and protein yield of E. gracilis under heterotrophic conditions were determined.Other main cell compositions were also analyzed to further find the flux direction of the carbon skeleton under different cultivation conditions.The amino acid profile analysis was also conducted to evaluate its protein nutritional quality.

The Effects of Nitrogen Sources on the Protein Content and Protein Yield
As shown in Figure 1A,B, E. gracilis exhibited a yellow appearance under heterotrophic cultivation, while it was presented as green under the mixotrophic culture.The heterotrophic cultures only contained 0.051 ± 0.001 mg/g chlorophyll a and 0.019 ± 0.001mg/g chlorophyll b, much less than that of mixotrophic cultures, which contained 7.895 ± 0.381 mg/g chlorophyll a and 2.983 ± 0.110 mg/g chlorophyll b (Figure S1).
Five commonly used nitrogen sources were evaluated for protein production by E. gracilis.Like other microalgae, E. gracilis showed a disparity of protein content among different nitrogen sources (Figure 1C).E. gracilis cultured with ammonium sulfate showed the highest protein content of 52.26 ± 0.91%(w/w), which was significantly higher than that of the yeast extract group (24.53 ± 0.37%) and monosodium glutamate (MSG) group (36.11 ± 0.63%) (Figure 1C).In contrast, cells supplemented with urea and sodium nitrate not only had lower protein contents (<15%) but also lower protein yields (0.39 ± 0.03 g/L and 0.38 ± 0.02 g/L, respectively) (Figure 1D).This implied that nitrate and urea could not be assimilated effectively by E. gracilis to produce protein [21].Five commonly used nitrogen sources were evaluated for protein production b gracilis.Like other microalgae, E. gracilis showed a disparity of protein content am different nitrogen sources (Figure 1C).E. gracilis cultured with ammonium sulfate sho the highest protein content of 52.26 ± 0.91%(w/w), which was significantly higher t that of the yeast extract group (24.53 ± 0.37%) and monosodium glutamate (MSG) gr (36.11 ± 0.63%) (Figure 1C).In contrast, cells supplemented with urea and sodium nit not only had lower protein contents (<15%) but also lower protein yields (0.39 ± 0.03 and 0.38 ± 0.02 g/L, respectively) (Figure 1D).This implied that nitrate and urea could be assimilated effectively by E. gracilis to produce protein [21].
The total fatty acid content displayed a similar trend to the protein content across different nitrogen source treatments (Figure 1E).Interestingly, the sodium nitrate (69.3.60%) and urea (63.15 ± 2.51%) groups generated significantly higher total carbohyd content than other groups (Figure 1F), in contrast to the protein content trend obser Considering the low content of total fatty acid across the treatments, our results imp that the metabolic conversion between protein and carbohydrate was the main path in E. gracilis cultivated by different nitrogen sources.Overall, the nitrogen screening periment highlighted that ammonium sulfate was the potential optimal nitrogen so for protein biosynthesis, which yielded cells with a protein content of up to 52.26% (w The total fatty acid content displayed a similar trend to the protein content across the different nitrogen source treatments (Figure 1E).Interestingly, the sodium nitrate (69.86 ± 3.60%) and urea (63.15 ± 2.51%) groups generated significantly higher total carbohydrate content than other groups (Figure 1F), in contrast to the protein content trend observed.Considering the low content of total fatty acid across the treatments, our results implied that the metabolic conversion between protein and carbohydrate was the main pathway in E. gracilis cultivated by different nitrogen sources.Overall, the nitrogen screening experiment highlighted that ammonium sulfate was the potential optimal nitrogen source for protein biosynthesis, which yielded cells with a protein content of up to 52.26% (w/w).
The substantial variation in protein content among different nitrogen sources aroused our interest to further investigate the nitrogen assimilation pathway in E. gracilis.However, due to the high metabolic variability caused by the complex nature of yeast extract and the ineffectiveof E. gracilis to utilize urea and sodium nitrate, the nitrogen source metabolism of ammonium sulfate and MSG are mainly discussed.
It is widely recognized that ammonium assimilation is primarily related to the irreversible GS/GOGAT pathway and reversible GDH pathway (Figure S2) [22].The GDH pathway is considered a shunt in the whole ammonium assimilation pathway that could divert the carbon skeleton from nitrogen metabolism to carbon metabolism via the TCA cycle [23].Thus, it is speculated that the observed disparity in protein synthesis between ammonium sulfate and MSG groups may be attributed to the feedback inhibition as the following indications: (1) When using ammonium sulfate as a nitrogen source, ammonium is preferentially consumed along with 2-OG as a carbon skeleton via the nitrogen assimilation pathway and finally produce glutamic acid for further producing other amino acids as protein building blocks [22].(2) Since glutamic acid is the end-product of GDH, the appearance of a large amount of glutamate might conversely be catalyzed by GDH to form ammonium ions and 2-OG.However, 2-OG could be utilized as a carbon skeleton to produce carbohydrates via the TCA cycle.The feedback inhibition of glutamate may lead to the loss of carbon skeleton in the nitrogen assimilation procedure, resulting in a decrease in protein production in the MSG group compared to the ammonium sulfate group.Further investigation is needed to fully elucidate the mechanism underlying the protein content disparity caused by different nitrogen sources.

The Effects of Nitrogen Sources on the Amino Acid Profile
To evaluate the nutritional value and NTP of E. gracilis protein, the amino acid profiles under different nitrogen sources were determined.The samples fed with urea and sodium nitrate were not analyzed since these two nitrogen sources cannot be utilized by E. gracilis.As shown in Figure 2, E. gracilis protein had a balanced amino acids profile, including cysteine and methionine, which are typically insufficient in legumes, and lysine, which is the first limiting amino acid in cereals [2,24].Furthermore, glutamic acid, aspartic acid, and leucine were three major amino acids in E. gracilis protein.Glutamic acid served as a precursor of many amino acid anabolisms, such as transamination to form aspartic acid and alanine [25,26].This feature was consistent with the amino acid profile of six other microalgae species, including Chlorella vulgaris, Dunaliella bardawil, S. obliquus, Arthrospira maxima, Spirulina platensis, and Aphanizomennon sp.[27].The MSG group contained lower level of isoleucine (3.40 ± 0.13 g/100 g protein), leucine (6.50 ± 0.27 g/100 g protein), lysine (5.59 ± 0.23 g/100 g protein), phenylalanine (3.43 ± 0.13 g/100 g protein), threonine (3.74 ± 0.16 g/100 g protein), valine (5.46 ± 0.24 g/100 g protein), tryptophan (1.03 ± 0.02 g/100 g protein), glutamic acid (10.11 ± 0.41 g/100 g protein), and glycine (4.10 ± 0.16 g/100 g protein) than that of other two groups, but higher levels of tyrosine (4.81 ± 0.33 g/100 g protein) and cysteine (5.03 ± 0.39 g/100 g protein).The amino acid profiles were similar between the yeast extract and the ammonium sulfate group.As shown in Table 1, the ammonium sulfate group contained the highest amino acid content (47.53 ± 2.08 g/100 g sample), while the yeast extract group contained the lowest amino acid content (23.53 ± 1.24 g/100 g sample).This result corresponded to the result of protein content (Figure 1C).The proportion of total essential amino acid to total amino acid (E/T) was found to be about 42% for both the yeast extract group and the ammonium sulfate group, which were comparable to that reported for C. vulgaris protein [25].However, the E/T value was lower in the MSG group (37.02 ± 0.59%).The lower-E/T value suggested that MSG was not an ideal nitrogen source for protein production compared with ammonium sulfate and yeast extract, considering its nutritional value.
Table 1.Amino acid profile (g/100 g sample) of E. gracilis with different nitrogen sources under heterotrophic cultivation conditions and corresponding nitrogen-to-protein conversion factor (NTP). (MSG: monosodium glutamate, EAA: essential amino acid, NEAA: non-essential amino acid, AA: amino acid, E/T: ratio of total essential amino acid content to total amino acid content).

Yeast Extract
Ammonium comparable to that reported for C. vulgaris protein [25].However, the E/T value was lower in the MSG group (37.02 ± 0.59%).The lower -E/T value suggested that MSG was not an ideal nitrogen source for protein production compared with ammonium sulfate and yeast extract, considering its nutritional value.
.  Based on the amino acid profile, NTP was calculated since the conventional approach of using a versatile NTP value of 6.25 may lead to an overestimation of the protein content as the presence of non-protein nitrogen compounds, such as inorganic nitrogen, nucleic acids, and amino sugars [28].The k A values are similar among these three groups, ranging from 6.11 ± 0.01 to 6.17 ± 0. A previous study investigated the k P value of 10 different microalgal species, which ranged from 3.60 ± 0.27 to 4.99 ± 0.64 [28].The k P and NTP value of E. gracilis sample fed with ammonium sulfate were 4.87 ± 0.13 and 5.52 ± 0.06, respectively.The k P of E. gracilis protein was comparable with most of the reported species but were shown to be higher than Skeletonema costatum (3.82 ± 0.33), Prorocentrum minimum (3.88 ± 0.68), and Dunaliella tertiolecta (3.99 ± 0.48) [28].Recently, the NTP value of four industrial-applied microalgae (Scenedesmus sp., C. vulgaris, Phaeodactylum tricornutum, and Nannochloropsis sp.) were recalculated.The results showed that the NTP (ranging from 4.68 to 5.35) and k P (ranging from 3.00 to 4.43) values of these four microalgae were all lower than those obtained from the present study [29].This indicated that E. gracilis has a lower non-protein-nitrogen content than microalgae species [29].These findings suggested that heterotrophic cultivated E. gracilis, when fed with ammonium sulfate, could be considered a reliable and promising protein source.

The Effect of Initial pH on Protein Content and Protein Yield
In the present study, E. gracilis was found to grow well under an initial pH of 3.5-7.5 (Figure S3).However, it had poor growth under initial pH 9.5 and partial growth inhibition under initial pH 2 (Figure S3).The protein content remained consistent with the initial pH 2-5.5, which was higher than that of the initial pH 7.5 group (44.19 ± 1.03%) (Figure 3A).This finding was partly supported by a previous study on mixotrophic cultivated E. gracilis, which reported that an acidic condition was more suitable for protein accumulation, displaying the highest protein content at an initial pH of 3.5 with a gradual protein decreasing trend from initial pH 3.5 to initial pH 8.5 [30].In addition, the initial pH 2 group showed a lower protein yield (2.08 ± 0.08 g/L) than initial pH 3.5 (2.34 ± 0.11 g/L) and initial pH 5.5 (2.45 ± 0.02 g/L) (Figure 3B), suggesting that initial pH 3.5 and 5.5 were suitable initial pH conditions for protein production.However, compared to the initial pH of 5.5, the initial pH of 3.5 had an advantage in terms of anti-contamination.Therefore, the optimal initial pH for E. gracilis protein production was 3.5.

The Effect of Temperature on Protein Content and Protein Yield
The optimal growth temperature for most microalgae species is normally within the range of 15-30 °C [31].For E. gracilis, the optimal temperature for growth was found to be 29 °C [32] and 30 °C [33].As shown in Figure 3C, the protein content of E. gracilis displayed a non-significant difference between 28 °C (52.33 ± 1.41%) and 32 °C (53.05 ± 0.89%), which were higher than 25 °C (46.83 ± 0.70%) and 35 °C (44.29 ± 2.56%).In addition, the protein yield of E. gracilis cultured at 25 °C (2.32 ± 0.02 g/L), 28 °C (2.23 ± 0.04 g/L), and 32 °C (2.24 ± 0.10 g/L) was significantly higher than that cultured at 35 °C (1.61 ± 0.11 g/L) (Figure 3D), implying that E. gracilis was incapable of protein accumulation at 35 °C.E. gracilis showed the highest specific growth rate at 32 °C (Table S1).Consequently, it can be concluded that 32 °C was the most suitable temperature for protein production, as it resulted in a high protein yield and maximum specific growth rate.

Effect of C/N Ratio on Protein Content, Protein Yield, and Amino Acid Profile
To find out the impact of the C/N ratio on E. gracilis and further excavate its protein accumulation potential, the influence of the C/N ratio on protein biosynthesis in E. gracilis was studied.Figure 4A revealed a significant positive impact of the low C/N ratio for protein biosynthesis where the protein content of the C/N = 10 group was 1.6-fold of the C/N = 25 group.At an optimal C/N ratio of 10, E. gracilis showed a remarkably high protein content of 66.10 ± 1.12%, which is comparable to that of traditional microalgal protein sources such as C. vulgaris (51-58%) and A. maxima (60-71%) [4].Compared with the yeast extract group in Section 3.1 (Figure 1C), the optimal protein content of E. gracilis was increased by 169% under C/N = 10.This emphasizes the enormous influence of the nitrogen

The Effect of Temperature on Protein Content and Protein Yield
The optimal growth temperature for most microalgae species is normally within the range of 15-30 • C [31].For E. gracilis, the optimal temperature for growth was found to be 29 • C [32] and 30 • C [33].As shown in Figure 3C, the protein content of E. gracilis displayed a non-significant difference between 28 • C (52.33 ± 1.41%) and 32 • C (53.05 ± 0.89%), which were higher than 25 • C (46.83 ± 0.70%) and 35 • C (44.29 ± 2.56%).In addition, the protein yield of E. gracilis cultured at 25 • C (2.32 ± 0.02 g/L), 28 • C (2.23 ± 0.04 g/L), and 32 • C (2.24 ± 0.10 g/L) was significantly higher than that cultured at 35 • C (1.61 ± 0.11 g/L) (Figure 3D), implying that E. gracilis was incapable of protein accumulation at 35 • C. E. gracilis showed the highest specific growth rate at 32 • C (Table S1).Consequently, it can be concluded that 32 • C was the most suitable temperature for protein production, as it resulted in a high protein yield and maximum specific growth rate.

Effect of C/N Ratio on Protein Content, Protein Yield, and Amino Acid Profile
To find out the impact of the C/N ratio on E. gracilis and further excavate its protein accumulation potential, the influence of the C/N ratio on protein biosynthesis in E. gracilis was studied.Figure 4A revealed a significant positive impact of the low C/N ratio for protein biosynthesis where the protein content of the C/N = 10 group was 1.6-fold of the C/N = 25 group.At an optimal C/N ratio of 10, E. gracilis showed a remarkably high protein content of 66.10 ± 1.12%, which is comparable to that of traditional microalgal protein sources such as C. vulgaris (51-58%) and A. maxima (60-71%) [4].Compared with the yeast extract group in Section 3.1 (Figure 1C), the optimal protein content of E. gracilis was increased by 169% under C/N = 10.This emphasizes the enormous influence of the nitrogen source and C/N ratio for E. gracilis protein production.This result also confirmed the great potential of heterotrophic E. gracilis as a dietary protein source.However, the relationship between nitrogen availability and protein content was not always positively correlated.The C/N = 5 group displays lower protein content compared with the C/N = 10 group (Figure 4A), while their protein yields were at the same level (Figure 4B).A similar result was found in C. vulgaris, in which the decrement of the C/N ratio from 12 to 7 displayed a slight protein content drop [25].Additionally, the high nitrogen substrate concentration (ammonium sulfate at 4360 mg/L, C/N = 5) did not inhibit the growth of E. gracilis (Figure S4).The high ammonium toleration of E. gracilis further revealed the advantage of cultivation under an initial pH of 3.5 since the ratio of toxic NH 3 to NH 4 + decreased as the pH decreased [34].The acidic tolerance of E. gracilis is crucial in achieving high protein content when using ammonium sulfate as a nitrogen source.In addition, these results also suggested the potential of E. gracilis for acidic industries wastewater treatment, such as pharmaceutical industries, mining sites, and ammunition industries where ammonium concentrations are typically high (5-1000 mg/L) [34].
source and C/N ratio for E. gracilis protein production.This result also confirmed the great potential of heterotrophic E. gracilis as a dietary protein source.However, the relationship between nitrogen availability and protein content was not always positively correlated.The C/N = 5 group displays lower protein content compared with the C/N = 10 group (Figure 4A), while their protein yields were at the same level (Figure 4B).A similar result was found in C. vulgaris, in which the decrement of the C/N ratio from 12 to 7 displayed a slight protein content drop [25].Additionally, the high nitrogen substrate concentration (ammonium sulfate at 4360 mg/L, C/N = 5) did not inhibit the growth of E. gracilis (Figure S4).The high ammonium toleration of E. gracilis further revealed the advantage of cultivation under an initial pH of 3.5 since the ratio of toxic NH3 to NH4 + decreased as the pH decreased [34].The acidic tolerance of E. gracilis is crucial in achieving high protein content when using ammonium sulfate as a nitrogen source.In addition, these results also suggested the potential of E. gracilis for acidic industries wastewater treatment, such as pharmaceutical industries, mining sites, and ammunition industries where ammonium concentrations are typically high (5-1000 mg/L) [34].In Figure 4C,D, it could be found that the increment of the C/N ratio from 10 to 25 resulted in the carbon partitioning switch between carbohydrate (from 21.78 ± 1.61% to 43.40 ± 1.45%) and protein (from 66.10 ± 1.12% to 40.86 ± 0.93%), while the fatty acid content maintained low (<6%).This observation could also be visually confirmed by the TEM images, which showed an increase in paramylon granules in E. gracilis cells with an increase in the C/N ratio from 10 to 25 (Figure 5).Paramylon is a well-known storage carbohydrate of E. gracilis under aerobic cultivation conditions [35].Thus, the low C/N and appropriate nitrogen source under aerobic conditions favored protein accumulation in E. gracilis at the expense of carbohydrates.In Figure 4C,D, it could be found that the increment of the C/N ratio from 10 to 25 resulted in the carbon partitioning switch between carbohydrate (from 21.78 ± 1.61% to 43.40 ± 1.45%) and protein (from 66.10 ± 1.12% to 40.86 ± 0.93%), while the fatty acid content maintained low (<6%).This observation could also be visually confirmed by the TEM images, which showed an increase in paramylon granules in E. gracilis cells with an increase in the C/N ratio from 10 to 25 (Figure 5).Paramylon is a well-known storage carbohydrate of E. gracilis under aerobic cultivation conditions [35].Thus, the low C/N and appropriate nitrogen source under aerobic conditions favored protein accumulation in E. gracilis at the expense of carbohydrates.
Considering the amino acid pattern of protein-sourced plants (e.g., soybean, oat) changed with C/N ratio [36], the amino acid profile of E. gracilis under different cultured C/N ratios was also analyzed.It was only found significant differences in threonine between C/N = 5 (4.20 ± 0.44 g/100 g protein) and C/N = 10 (4.83 ± 0.06 g/100 g protein) and in tyrosine between C/N = 10 (4.33 ± 0.87 g/100 g protein) and C/N = 25 (3.21 ± 0.06 g/100 g protein), which implied that the variation of C/N ratio did not significantly alter amino acid profile of E. gracilis (Figure 6).The stable amino acid pattern under the variation of the C/N ratio in E. gracilis was different from C. sorokiniana, which had a dramatic increase in amino acid content derived from the TCA cycle as the decrease in the C/N ratio [37].Considering the amino acid pattern of protein-sourced plants (e.g., soybean, oat) changed with C/N ratio [36], the amino acid profile of E. gracilis under different cultured C/N ratios was also analyzed.It was only found significant differences in threonine between C/N = 5 (4.20 ± 0.44 g/100 g protein) and C/N = 10 (4.83 ± 0.06 g/100 g protein) and in tyrosine between C/N = 10 (4.33 ± 0.87 g/100 g protein) and C/N = 25 (3.21 ± 0.06 g/100 g protein), which implied that the variation of C/N ratio did not significantly alter amino acid profile of E. gracilis (Figure 6).The stable amino acid pattern under the variation of the C/N ratio in E. gracilis was different from C. sorokiniana, which had a dramatic increase in amino acid content derived from the TCA cycle as the decrease in the C/N ratio [37].Considering the amino acid pattern of protein-sourced plants (e.g., soybean, oat) changed with C/N ratio [36], the amino acid profile of E. gracilis under different cultured C/N ratios was also analyzed.It was only found significant differences in threonine between C/N = 5 (4.20 ± 0.44 g/100 g protein) and C/N = 10 (4.83 ± 0.06 g/100 g protein) and in tyrosine between C/N = 10 (4.33 ± 0.87 g/100 g protein) and C/N = 25 (3.21 ± 0.06 g/100 g protein), which implied that the variation of C/N ratio did not significantly alter amino acid profile of E. gracilis (Figure 6).The stable amino acid pattern under the variation of the C/N ratio in E. gracilis was different from C. sorokiniana, which had a dramatic increase in amino acid content derived from the TCA cycle as the decrease in the C/N ratio [37].The consistent amino acid pattern observed in different C/N ratio groups resulted in a stable NTP.The k P and NTP values were at a similar level among the five groups.The C/N = 5 group had a significantly higher k A value of 6.26 ± 0.07 than the C/N = 25 group of 6.17 ± 0.01 (Table 2).In addition, the EAAI of E. gracilis protein was 1.62 ± 0.05 of the C/N = 10 group, indicating superior protein quality [18].This was comparable with Chlorella sp.(EAAI = 1.67) and Spirulina sp.(EAAI = 1.63), which are the conventional microalgal protein source [38], while it is higher than D. salina protein (EAAI = 1.53), which is also a novel microalgal protein source [18].These results confirmed that the heterotrophic E. gracilis was a promising single-cell-protein resource with high protein content and nutritional value.

Microalgae Strain and Culture Conditions
Euglena gracilis Klebs (CCAP 1224/5Z) was purchased from the Culture Collection of Algae and Protozoa (CCAP, Oban, UK).Algal cells were maintained on an agar plate with modified Hutner's medium at 16 • C [39].For each experiment, a single colony of E. gracilis was inoculated into 20 mL modified Hutner's medium in a 50 mL Erlenmeyer flask and cultivated in shakers (150 rpm, 25 • C) in darkness for 5 days.The algal liquid was subsequently inoculated (7.5% v/v) into a 250 mL Erlenmeyer flask with a total volume of 100 mL and shaken for 48 h in the same condition for further use as a seed.

Optimization of Culture Conditions
Sequential optimization experiments of the nitrogen source (yeast extract, monosodium glutamate, ammonium sulfate, sodium nitrate, and urea), temperature (25 • C, 28 • C, 32 • C, 35 • C), initial pH (2, 3.5, 5.5, 7.5), and C/N ratio (5,10,15,20,25) were conducted in 250 mL Erlenmeyer flasks with a working volume of 100 mL.The initial biomass concentration was kept around 0.4 g/L for each inoculation.The initial glucose concentration was fixed at 10 g/L (except for the sodium glutamate group to maintain the consistent C/N).

Determination of Biomass Concentration
Five microliters of cell culture were sampled and washed twice before being filtered on weighted GF/C film (Whatman International, Maidstone, UK) and dried in the 80 • C vacuum dryer overnight to measure the biomass concentration.The samples were collected for further analysis 12 h after they reached the stationary phase.All the sample was washed three times with deionized water and freezing dried.

Protein Content Analysis
The crude protein content of the sample was measured using the Dumas method [40].Around 50 mg of the sample was wrapped in aluminum foil and combusted with highpurity oxygen gas at 900 • C. The nitrogen gas released from the sample was detected by a thermal conductivity detector (Haineng, Weifang, China).The crude protein content of E. gracilis was calculated with a conversion factor of 6.25 [41].

Photosynthetic Pigments Analysis
Chlorophyll a, chlorophyll b, and carotenoid content were determined using a spectrophotometer at the wavelength of 480, 646, and 662 nm [42].Pigment concentration (mg/L) was calculated using the following equation: where Ab 662 , Ab 646, and Ab 480 are the sample absorbances at 662, 646, and 480 nm, respectively.

Amino Acid Composition Analysis
The amino acid composition analysis was proceeded by acid hydrolysis except tryptophan.Around 50 mg of lyophilized sample was acid hydrolyzed by 6 M HCl at 110 • C for 24 h under vacuum conditions.The sample was cooled down and neutralized via saturated sodium carbonate until the pH reached 7. The hydrolyzed sample (10 µL) was mixed with 70 µL of AccQ-Tag buffer and 20 µL of derived AccQ-Tag reagent solution [43].The mixture was incubated at 50 • C for 10 min and then transferred to a micro-vial.The sample was analyzed using a Waters Alliance e2695 high-performance liquid chromatography (HPLC) system equipped with a Waters 2998 PDA detector and an AccQ-Tag amino acid column Nova-Pak C18, 4 µm (150 × 3.9 mm) (Waters, Milford, MA, USA).To analyze methionine and cysteine, samples were oxidized with performic acid before the acidic hydrolysis mentioned above.
For tryptophan analysis, 50 mg of the sample was hydrolyzed by 5 M LiOH at 110 • C for 24 h under vacuum conditions.After hydrolysis, the sample was cooled down and neutralized by 6 M HCl until the pH reached 7. The neutralized sample was filtered via filter paper and analyzed using a Waters ACQUITY ultra-performance liquid chromatography-mass spectrometry (UPLC-MS) equipped with a PDA detector and BEH C18 column (1.7 µm, 2.1 × 50 mm) [44].
The NTP was calculated according to the equation below [29].
where ΣEi, ΣDi, and N% are the summation of anhydrous amino acid, the summation of the nitrogen content of each anhydrous amino acid, and the nitrogen content of E. gracilis sample determined by the Dumas method using the conversion factor of 6.25, respectively [29,45].k A and k P are the upper bound and lower bound of the conversion factor, respectively [29].NTP, the average value of k A and k P , is considered the best conversion factor for practical use [45].

Total Fatty Acid Content Analysis
Twenty micrograms of lyophilized samples were methyl esterized with 1% (v/v) sulfuric acid in methanol and methylbenzene with heptadecanoic acid (C17:0, Sigma-Aldrich, St. Louis, MO, USA) as the internal standard.Samples were incubated in a 50 • C water bath overnight and analyzed using a gas chromatography-mass spectrometry (GC-MS) equipped with a DB-WAX column (30 m × 0.25 mm × 0.25 µm) (Agilent, Santa Clara, CA, USA) [47].

Total Carbohydrate Content Analysis
Ten micrograms of lyophilized samples were hydrolyzed by 2 mL of 12 M H 2 SO 4 for 1 h at 35 • C. Samples were then added to 10 mL of deionized water for further hydrolyzation at 98 • C for 1 h.The hydrolyzed samples were further determined using the phenol-sulfuric acid method [48].

Transmission Electron Microscopy (TEM)
The freshly prepared samples were washed with PBS buffer twice and fixed overnight with 2.5% glutaraldehyde buffer at 4 • C. The samples were washed with 0.1 M PBS buffer (pH 7.0) followed by 1-2 h fixing with 1% osmic acid.The osmic acid was then replaced by 0.1 M PBS buffer (pH 7.0) and gradient dehydrated by a graded series of alcohol.After being replaced with acetone, the samples were embedded with epoxy resin and ultrathin sectioned to 70-90 nm thickness using an ultramicrotome EM UC7 (Leica Biosystems, Solms, Germany).The ultrathin-sectioned samples were stained with acetic acid glaze and lead citrate and observed using TEM JEM1200EX (JEOL Ltd., Tokyo, Japan) operated at 120 kV.

Statistics Analysis
All experiments were conducted in triplicates.The results were expressed as mean ± standard deviation (SD).The statistical significance of the results was validated using a one-way analysis of variance (ANOVA) in the SPSS 26.0 software (IBM, Armonk, NY, USA); the significance level was set at p < 0.05.

Conclusions
In the present study, we demonstrated that ammonium sulfate could significantly enhance the protein biosynthesis ability of E. gracilis growing heterotrophically.The low cultural C/N ratio condition was favorable for protein production, and the protein content could reach up to 66.10% (w/w).In addition, the protein quality evaluation of the heterotrophic E. gracilis protein showed a high NTP factor (5.63) and high EAAI (1.62), further indicating its high nutritional value.These findings suggested that heterotrophic cultured E. gracilis has great economic potential as an alternative protein source due to its high protein content and nutritional value, acceptable color, cell wall-free nature, and dietary-approved status.

Figure 1 .
Figure 1.(A) Euglena gracilis cultured under heterotrophic and mixotrophic conditions.(B) L micrographs of E. gracilis under heterotrophic and mixotrophic conditions (Bars: 50 µm).(C) Pro content, (D) protein yield, (E) total fatty acid content, and (F) total carbohydrate content of E. gr cultured with different nitrogen sources under 32 °C/C/N = 17/initial pH 3.5.(YE: yeast extract, M monosodium glutamate); bars followed by the same letter are not significantly different (p < 0.0
Mar. Drugs 2023, 21, x FOR PEER REVIEW 6 of 13 of 5.5, the initial pH of 3.5 had an advantage in terms of anti-contamination.Therefore, the optimal initial pH for E. gracilis protein production was 3.5.

Figure 3 .
Figure 3. (A) Protein content and (B) protein yield of E. gracilis cultured with ammonium sulfate as nitrogen source under 25 °C/C/N = 17 and different initial pH.(C) Protein content and (D) protein yield of E. gracilis cultured with ammonium sulfate as nitrogen source/ initial pH 3.5/C/N = 17 and under different temperature conditions.Bars followed by the same letter are not significantly different (p < 0.05).

Figure 3 .
Figure 3. (A) Protein content and (B) protein yield of E. gracilis cultured with ammonium sulfate as nitrogen source under 25 • C/C/N = 17 and different initial pH.(C) Protein content and (D) protein yield of E. gracilis cultured with ammonium sulfate as nitrogen source/ initial pH 3.5/C/N = 17 and under different temperature conditions.Bars followed by the same letter are not significantly different (p < 0.05).

Figure 4 .
Figure 4. (A) Protein content, (B) protein yield, (C) total carbohydrate content, and (D) total fatty acid content of E. gracilis cultured with different C/N ratio under ammonium sulfate as nitrogen source, initial pH 3.5/32 °C, heterotrophic conditions.Bars followed by the same letter are not significantly different (p < 0.05).

Figure 4 .
Figure 4. (A) Protein content, (B) protein yield, (C) total carbohydrate content, and (D) total fatty acid content of E. gracilis cultured with different C/N ratio under ammonium sulfate as nitrogen source, initial pH 3.5/32 • C, heterotrophic conditions.Bars followed by the same letter are not significantly different (p < 0.05).

Table 2 .
NTP and EAAI of E. gracilis with different C/N under heterotrophic cultivation conditions.Row values with the same superscript letter display not significantly different (p < 0.05).