Enhancing Nutritional Value and Sensory Quality of Spirulina (Arthrospira platensis) Through Preharvest Co-Cultivation with Yeast Saccharomyces cerevisiae

Abstract

Spirulina’s (Arthrospira platensis) use in food applications is limited by its dark color and sulfurous odor. This study aimed to develop a preharvest bioprocessing strategy using Saccharomyces cerevisiae co-cultivation to address these limitations. At a yeast/microalgae biomass ratio of 10:1000 with 5 g/L of glucose supplementation, co-cultivation for 24 h induced a rapid color transition from dark blue–green to light green and imparted “floral–fruity” aromas. Major bioactive compounds, including β-carotene, linoleic acid, and γ-linolenic acid, increased significantly, while volatile sulfur compounds were eliminated. Chlorophyll a and carotenoid contents rose by over two fold, reflecting enhanced photosynthetic efficiency. Mechanistic analyses revealed that yeast-derived acetic acid upregulated genes involved in flavor precursor biosynthesis and promoted biomass accumulation. This strategy integrates sensory improvement with nutritional enhancement, providing a sustainable approach for developing spirulina-based functional foods.

1. Introduction

Under the global carbon neutrality initiative, the development of novel food products that combine sustainability with nutritional benefits has emerged as a critical research focus in food science. Recognized by the FAO as a “future superfood”, spirulina (Arthrospira platensis and Arthrospira maxima) has attracted significant attention due to its rich content of high-quality protein (60–70%), polyunsaturated fatty acids including γ-linolenic acid, and natural pigments such as phycocyanin and β-carotene, all demonstrating important functional nutritional properties [1]. However, spirulina can hardly be incorporated into commercial foods in large quantities due to its undesirable dark blue–green color and seaweed/muddy odor. Currently most research on sensory quality improvement of spirulina biomass is focused on postharvest processing technologies, e.g., microencapsulation, fermentation, and masking.

Microalgae–microbial symbiosis is a common natural phenomenon that shapes the diversity and structure of aquatic ecosystems [2]. There is increasing interest in leveraging photoautotroph–heterotroph associations to enhance microalgae biomanufacturing [3]. Yeast is a particularly appealing edible microorganism with the potential to serve as a photoautotrophic partner due to its lack of endotoxins, high growth rate, and robustness to withstand alkaline pH levels [4]. Yeast can establish an obligatory symbiotic relationship with microalgae in co-culture systems by mutually benefiting each other through the exchange of metabolites (e.g., O₂, CO₂, and organic acids) [5,6]. The microalgae–yeast co-culture systems have been explored for applications in biodiesel production, wastewater treatment, chemical production, and aquaculture feed supply [7]. Co-culture strategies have been demonstrated to increase the yields of microalgae pigments and beneficial components/flavor precursors (e.g., lipids and carotenoids) [8,9]. Therefore, microalgae–yeast co-culture is a potential alternative strategy for improving the nutritional value and sensory quality of spirulina biomass at the preharvest stage.

Saccharomyces cerevisiae is a species of yeast instrumental in food processing (e.g., brewing, baking, and winemaking) and has been categorized as a Generally Recognized as Safe (GRAS) microorganism by the US Food and Drug Administration. However, there are few examples of co-cultivation of spirulina and S. cerevisiae. The present study established an efficient co-culture system comprising A. platensis and S. cerevisiae and compared color appearance, bioactive compound composition, aroma quality, and volatile flavor profile between monocultured and co-cultured A. platensis to assess nutritional and sensory acceptability as a future food. In addition, the plentiful organic acids generated by S. cerevisiae in alkaline medium were shown to be responsible for the stimulated β-carotene, linoleic acid, and γ-linolenic acid biosynthesis in co-cultured A. platensis.

2. Materials and Methods

2.1. Microorganism Strains and Pre-Culture Conditions

A. platensis (FACHB-902) was purchased from the Freshwater Algae Culture Collection at the Institute of Hydrobiology (FACHB), Chinese Academy of Sciences (Wuhan, China). Arthrospira remains the taxonomically validated classification per phylogenomic evidence (current study), whereas Limnospira remains a contentious proposal lacking consensus. Before co-culture experiments, A. platensis was cultured to the late-logarithmic phase in Zarrouk medium (Tables S1 and S2) in 5 L reactors. The 0.45 μm filtered compressed air was introduced at the bottom of the column to supply carbon and prevent cell aggregation.

S. cerevisiae CICC 1251 was obtained from the China Center of Industrial Culture Collection (CICC). Before co-culture experiments, S. cerevisiae was cultured in Yeast Extract Peptone Dextrose Agar (YPDA) at 28 °C for 48 h to the logarithmic phase, transferred to Zarrouk medium at a final concentration of 6−7 log CFU/mL, stored at 4 °C, and used within 12 h.

2.2. Co-Culture Experiments

The resuspension of S. cerevisiae was added to the logarithmic growth phase of A. platensis and co-cultivated in 5 L Zarrouk medium supplemented with 5 g/L glucose. The late-logarithmic phase pre-culture of A. platensis at the OD₅₆₀ of 0.8 ± 0.05 was inoculated with S. cerevisiae at yeast-to-microalgae biomass ratios of 1:1000, 10:1000, and 100:1000. During the co-cultivation, the co-cultures were kept in a constant-temperature oscillation incubator at 30 ± 5 °C under a light intensity of 100 μmol photons·m⁻²s⁻¹. The 0.45 μm filtered compressed air was introduced at the bottom of the column to supply carbon and prevent cell aggregation. All experiments were performed at least in triplicate.

Biomass was estimated by OD₅₆₀, direct cell counting, and the dry weight method. Throughout the cultivation phase, the OD₅₆₀ of a culture aliquot was measured using a UV-Vis spectrophotometer (Shimadzu, Kyoto, Japan). Cell counting was performed using a dilution series with a hemocytometer and an optical microscope (Nikon Ts2 FL, Nikon, Tokyo, Japan). To measure dry weight, a certain volume of culture aliquot (V) was filtered through a 0.45 μm Whatman GF/C glass-fiber filter, and after the filter paper was oven-dried, the dry weight was measured by subtracting the weight of the original filter paper (M1) from the weight of the paper after cell filtration (M2). The biomass concentration (g/L) was calculated using the following equation [10]:

$$\mathrm{Biomass} \mathrm{concentration}(\mathrm{g}/\mathrm{L})={\displaystyle \frac{{\mathrm{M}}_2-{\mathrm{M}}_1}{\mathrm{V}}}$$

(1)

Dissolved oxygen values were estimated using a JPBJ-610L DO meter (INESA Scientific Instrument, Shanghai, China).

2.3. Profiling of Yeast Extracellular Organic Acids

S. cerevisiae was cultured in Zarrouk medium supplemented with 5 g/L glucose at 30 °C for 36 h, followed by centrifugation at 6000× g for 10 min at 4 °C. The supernatant was then filtered and collected as the yeast culture supernatant, which was lyophilized for the profiling of organic acids. Organic acids were analyzed as previously reported with minor modifications [11]. Briefly, 0.9 g of the sample was mixed with 10 mL of ultrapure water with ultrasonic treatment for 30 min. The supernatant was filtered through a 0.22 μm filter and analyzed using an HPLC system (Agilent Series 1260, Agilent Technologies, Santa Clara, CA, USA) equipped with an Agilent C18 AQ analytical column (4.6 mm × 250 mm, 5 μm). Organic acids were quantified based on retention times and area of the standard mixture of the following 9 organic acids: oxalic acid, tartaric acid, malic acid, lactic acid, acetic acid, maleic acid, citric acid, succinic acid, and fumaric acid. A methanol-buffered salt solution (5:95, v/v) was used as the mobile phase with a flow rate of 1.0 mL/min.

2.4. Color Evaluation and Pigment Analysis

Microalgae cells were harvested at 24 h post-yeast inoculation for color evaluation and pigment analysis. The chromatism measurements of lyophilized microalgae powders were performed using a CR-400 colorimeter (Konica Minolta, Tokyo, Japan) equipped with a Commission Internationale de l’Éclairage (CIE) color system. The results are presented as L* (brightness), a* (+a, red; −a, green), and b* (+b, yellow; −b, blue).

The contents of Chlorophyll a (Chl. a) and carotenoids in A. platensis biomass were determined as previously described with some modifications [12]. Briefly, the microalgae were extracted using an 80% acetone solution (v/v) at 4 °C until almost colorless. The extract was then measured at wavelengths of 670 nm, 646 nm, and 470 nm. The concentrations of the pigments were calculated using the following equations:

ΔA = A _test − A _blank

(2)

$${\mathrm{C}}_{Chl. a} (\mathrm{mg}/\mathrm{L}) =12.25\times \mathsf{\Delta }{\mathrm{A}}_{663}-2.79\times \mathsf{\Delta }{\mathrm{A}}_{646}$$

(3)

$${\mathrm{C}}_{carotenoids} (\mathrm{mg}/\mathrm{L}) =5.05\times \mathsf{\Delta }{\mathrm{A}}_{470}-0.009\times C_{Chl. a}-9.22\times \mathsf{\Delta }{\mathrm{A}}_{646}+2.19\times \mathsf{\Delta }{\mathrm{A}}_{663}$$

(4)

Carotenoids in A. platensis biomass were profiled based on HPLC analysis as previously described [13]. Briefly, lyophilized microalgae samples were ground at 4 °C for 10 cycles of 60 s using a tissue grinding machine (Servicebio, Wuhan, China); this was followed by repeated extraction with methyl tertiary butyl ether until the biomass was almost colorless. After centrifugation at 6000× g for 10 min, the extract was dried with nitrogen gas, reconstituted in 500 μL of acetone, filtered through a 0.22 μm filter, and analyzed on an HPLC system equipped with a ZORBAX Eclipse XDB-C18 reverse phase column (4.6 mm × 250 mm, 5 μm; Agilent, USA).

2.5. Sensory and Instrumental Volatile Flavor Analysis

Microalgae cells were harvested 24 h post-yeast inoculation for sensory and instrumental volatile flavor analysis. Sensory evaluation of volatile flavor was conducted in a standardized sensory booth (ISO 8589) under controlled lighting and ventilation conditions [14]. The evaluation was performed by a team of panelists consisting of 8 males and 7 females, who were graduate students or faculty members from several different labs and were trained and selected according to the guidelines in ISO 5496, ISO 8586, and GB/T 39625 (China National Standard) [15,16,17,18,19]. These panelists had correction rates >80% in odor tests. Lyophilized microalgae samples (2 g) were kept in odorless clear SPME vials for 30 min at room temperature before being sniffed by the panelists. The free choice profiling (FCP) method was employed to generate an odor vocabulary. Sensory scores, ranging from 0 to 10, were assigned to microalgae samples based on the odor intensity.

The volatile compounds were characterized through headspace solid-phase microextraction (SPME) coupled with gas chromatography–mass spectrometry (GC-MS) as previously described [20]. To facilitate the volatilization of odorants, a saturated NaCl solution (~26%) was used to prepare the microalgae suspension (40 g/L), 2 mL of which was then transferred to a 20 mL SPME glass vial, and 20 μL of 2-methyl-3-heptanone (100 μg/g) was added as an internal standard. After a 20 min incubation at 60 °C with agitation at 300 rpm, the extraction needle (50/30 μm, DVB/CAR/PDMS 57348-U; Supelco, Bellefonte, PA, USA) was inserted for a 30 min headspace extraction of volatile compounds, followed by heating at 250 °C at the gas-injection port for 20 min. The volatile compounds were analyzed using an Agilent 7890A-5975C GC-MS system (Agilent, USA) equipped with an INNOWAX column (60 m × 0.25 mm, 0.25 μm; Agilent, USA). The carrier gas, helium, was injected at a flow rate of 1.2 mL/min. The temperature program was set as follows: 40 °C, holding for 5 min, ramping to 240 °C at a rate of 5 °C/min, ramping to 250 °C at a rate of 10 °C/min, and holding for 6 min. The mass scan range was set at 29–550 Da, and the electron energy was set at 70 eV. Three replicates were used for each microalgae sample in the SPME-GC-MS analysis.

2.6. Genome Sequencing and Transcriptomic Analysis

Microalgae cells were harvested 18 h post-yeast inoculation for genome sequencing and transcriptomic analysis. Dialysis tubing cellulose membrane (20 kDa pore size, 76 mm × 49 mm; Sigma-Aldrich, St. Louis, MO, USA) was employed to separate the cultures of A. platensis and S. cerevisiae before inoculating the S. cerevisiae at the beginning of the experiment, as previously described [21]. Total DNA was extracted from A. platensis cells using the QIAamp DNA Micro Kit (Qiagen, Redwood, CA, USA) and analyzed using agarose gel electrophoresis to assess DNA integrity and potential RNA/protein contamination. A NanoDrop One spectrophotometer (Thermo Scientific, Waltham, MA, USA) was used to determine DNA purity based on the OD₂₆₀/OD₂₈₀ ratio, and a Qubit 3.0 fluorometer (Thermo Scientific, Waltham, MA, USA) was used to quantify the DNA concentration. Subsequently, 2.5 μg of qualified DNA was used for library construction, which involved a series of steps, including purification with magnetic beads, fragmentation and end-repair, attachment of barcode labels, pooling of samples, and connection with sequencing adapters. The constructed library was loaded onto the R9.4 sequencing chip and sequenced using a PromethION sequencer (Oxford Nanopore Technologies, Oxford, UK) for 48–72 h. Raw data underwent quality control to remove low-quality and short reads. The high-accuracy Illumina data (Q₃₀ > 85%) were assembled using Unicycler 0.4.9 to build a high-quality genome skeleton, which was then combined with Nanopore data to form complete genome maps. Finally, the assembled genome was corrected using Pilon 1.23 to obtain a more accurate genome [22]. Prodigal 2.6.3 was used for protein-coding gene predictions [23]. tRNA gene prediction was conducted using tRNAscan-SE 2.0 [24]. rRNA gene analysis was conducted using Barrnap 0.7. For genome function annotation, the predicted gene sequences were aligned to GO and KEGG function databases using BLAST. (https://blast.ncbi.nlm.nih.gov/) The circlize R package (version 0.4.12) was used for the circular visualization of genome components and their relationships [25].

Total RNA was extracted from A. platensis cells using the QIAamp RNA Mini Kit (Qiagen, CA, USA). The NanoDrop One spectrophotometer was used to determine RNA purity based on the OD₂₆₀/OD₂₈₀ ratio, and the Agilent 2100 bioanalyzer (Agilent Technologies, CA, USA) was used to determine RNA integrity. Upon quality assessment of the total RNA samples, a cDNA library was constructed using a Maxima Reverse Transcriptase (Thermo Fisher Scientific, MA, USA) prior to sequencing using a sequencing by synthesis (SBS) method on a second-generation high-throughput sequencing platform. Prior to assembly, rigorous read filtering was carried out to ensure data quality. The resulting high-quality clean reads were utilized for de novo transcriptome assembly. CPC2 was used to identify unigenes with coding potential, and BLAST was employed for conduct sequence alignments against the Nr, Nt, SwissProt, KEGG, and GO databases, thereby obtaining annotation information for novel genes.

2.7. qRT-PCR Analysis

Total RNA was isolated from A. platensis cells using the TRIzol RNA extractor (Sangon, Shanghai, China) according to the manufacturer’s instructions. The quantity and purity of the isolated RNA were analyzed using a Nanodrop ND-1000 spectrophotometer (Thermo Scientific, USA). Then, purified RNA was converted to cDNA using the Maxima Reverse Transcriptase EP0743 (Thermo Scientific, USA). A quantitative real-time PCR assay was performed using the following cycles: 95 °C for 3 min, 45 cycles of 95 °C for 15 s, and 60 °C for 30 s. The 20 μL qRT-PCR reaction mixture comprised 1 μL cDNA, 10 μL Fast qPCR Master Mix, 0.5 μL of each PCR forward/reverse primer, and 8 μL ddH₂O. Melting curve analysis was performed to confirm the purity of the amplification product. All reactions were performed in triplicate, and the Cq values were recorded for each reaction. Gene expression was quantified using the 2^−ΔΔCT method.

2.8. Statistical Analysis

For each measurement, three independent experiments were conducted to generate a standard deviation. Statistical significance was analyzed using two-way ANOVA or a t-test with a multiple comparison Tukey test using the IBM SPSS program (Version 19.0, IBM SPSS, Armonk, NY, USA) and a level of p < 0.05.

3. Results and Discussion

3.1. Establishment of the Microalgae–Yeast Co-Culture System

To ensure production quality, commercial microalgae are typically harvested in the late-logarithmic phase, with sufficient biomass yield and relatively short cultivating duration [26]. In the present study, a preharvest culture of A. platensis in the late-logarithmic phase was inoculated with S. cerevisiae at three yeast-to-microalgae biomass ratios (1:1000, 10:1000, and 100:1000). The co-culture system was then supplemented with 5 g/L of glucose.

According to the kinetics of the OD₅₆₀ and biomass concentration following a co-cultivation duration of 36 h (Figure 1a), the biomass yield in the co-culture was greatly increased by the inoculation of S. cerevisiae at yeast-to-microalgae biomass ratios of 1:1000 and 10:1000 (p < 0.01), while excess amounts of S. cerevisiae, at an inoculation level of 100:1000, appeared to significantly decrease biomass production in the co-culture (p < 0.05).

As shown in Figure 1b, A. platensis proliferated throughout the co-cultivation period at inoculation levels of 1:1000 and 10:1000 but gradually died out of the co-culture at an inoculation level of 100:1000. Microscopic observation (Figure 1c) revealed cellular fragmentation of A. platensis at the inoculation level of 100:1000, confirming the dying out of A. platensis in the presence of excess yeast. Figure 1d illustrates the proliferation of S. cerevisiae throughout the co-cultivation period at all three inoculation levels. These results indicate survival stress induced in microalgae by excess yeast, which may explain the significantly lower levels of OD₅₆₀ in the co-culture at an inoculation level of 100:1000 compared to those at lower inoculation levels (Figure 1a). The decline in biomass observed at a 100:1000 ratio was primarily attributed to survival stress in A. platensis; however, the underlying mechanisms remain incompletely elucidated due to the paucity of comparative studies.

Based on the above results, a yeast-to-microalgae biomass ratio of 10:1000 was used to establish the preharvest co-culture system for A. platensis and S. cerevisiae. As shown in Figure 1e, compared to the monocultures of A. platensis and S. cerevisiae, the co-culture system had remarkably higher OD₅₆₀ and biomass concentration, indicating excellent biomass yield in the co-culture. As shown in Figure 1f, the level of dissolved oxygen in the S. cerevisiae monoculture gradually decreased throughout the 36 h of incubation, correlating with the aerobic catabolism of glucose. The level of dissolved oxygen in the A. platensis monoculture sharply decreased within 12 h of incubation and then remained at a constant level before rising gradually after 24 h of incubation, suggesting an increasing contribution of photosynthetic oxygen production after 12 h of incubation because of the exponential growth of microalgae (Figure 1e). In the co-culture system, the level of dissolved oxygen decreased more sharply than in the monocultures within 12 h of incubation. This was followed by a sharp increase over the subsequent 12 h of incubation and then a gradual increase until the end of incubation. This DO pattern was driven by the following metabolic shifts: (1) rapid respiration during the initial phase (0–12 h), fueled by S. cerevisiae proliferation under glucose supplementation and phototrophic oxygen release [27]; (2) DO recovery in the subsequent phase (12–24 h), mediated by A. platensis as the dominant photoautotroph, with biomass increase (Figure 1e) restoring oxygen production following glucose depletion.

3.2. Color and Compositional Variations in Co-Cultured Microalgae Biomass

Spirulina powders typically exhibit a dark blue–green color and have been reported to have appetite-suppressing properties [28]. However, the appearance of spirulina biomass is not appealing for many food applications. As shown in Figure 2b, A. platensis and S. cerevisiae in the co-culture can be separated by centrifugation based on their density differences. The lyophilized biomass of co-cultured A. platensis exhibited a noticeably lighter green color compared with the biomass observed under monoculture conditions (Figure 2a). This color has been reported to be favorable by consumers from the perspectives of general health interest and environmental concern [29,30]. As revealed by chromatism measurements (Figure 2c), co-cultivation increased the L* (brightness), −a (green), and +b (yellow) values of the microalgae biomass (p < 0.05), confirming the color change from the dark blue–green of the monocultured A. platensis to the light green of the co-cultured A. platensis (Figure 2a,b). The observed chromatic alteration, concomitant with increased chlorophyll a and carotenoid concentrations (Figure 2d), presumably represents a dilution phenomenon attributable to yeast-derived biomass accumulation or modified pigment organization in the co-cultured microalgae system.

As shown in Figure 2d,e, the co-cultured A. platensis had significantly higher levels of chlorophylls and carotenoids than the monocultured microalgae (p < 0.05), indicating that co-cultivation with yeast could induce the biosynthesis of chlorophylls and carotenoids. In fact, several previous studies reported the enhancement of microalgae biosynthesis of chlorophylls and carotenoids by co-cultivation with yeast or bacteria. A Chlorella–Saccharomyces co–culture system has been reported to enhance microalgal production of carotenoids [31]. The co-culture of Chlorella and wastewater-borne bacteria has been demonstrated to increase the concentrations of Chl. a, Chl. b, and carotenoids in microalgae biomass [8]. Co-culturing a green alga, Micractinium sp. GA001, with the endophytes Staphylococcus pasteuri PPE11 and Yersinia enterocolitica PPE118 has been shown to significantly increase the total chlorophyll content in microalgae biomass [9].

Figure 2e illustrates the changes in β-carotene and zeaxanthin contents during the in situ light co-cultivation of A. platensis and S. cerevisiae. Compared with monocultured A. platensis, the contents of both carotenoids increased significantly (p < 0.01), with β-carotene showing a particularly pronounced increase. As a precursor of terpenoid-derived flavor compounds [32], the increased β-carotene content resulting from co-cultivation may play an important role in enhancing the flavor profile of the microalgae biomass. Similarly, Figure 2f shows the changes in unsaturated fatty acid (UFA) content. Linoleic acid and γ-linolenic acid, the predominant UFAs in microalgae biomass, were significantly elevated during co-cultivation (p < 0.05). These UFAs serve as beneficial components/precursors for volatile flavor compounds through enzymatic or microbial oxidation and cleavage, contributing to the formation of aldehydes, ketones, and alcohols [33,34]. Collectively, the accumulation of these biochemical precursors suggests a synergistic enhancement in flavor-forming potential during co-cultivation.

3.3. Volatile Flavor Profile of Co-Cultured Microalgae Biomass

Spirulina biomass has some unpleasant odors, such as seaweed and muddy/earthy, which greatly restricts its incorporation into various foods [35]. In the present study, co-cultivation with yeast was found to result in a reduction in the intensity of the unpleasant microalgal odors of “seaweed” and “muddy”, as well as to confer pleasant aromas of “floral” and “fruity” to the microalgae biomass (Figure 3a). Conceptually, this is an interesting finding, as it shows, for the first time, that preharvest co-cultivation with yeast can improve the aroma quality of microalgae biomass.

In the present study, a total of 98 volatile compounds were identified in the monocultured and co-cultured microalgae using SPME-GC-MS, including 29 ketones, 17 aldehydes, 13 alcohols, 8 nitrogenous compounds, 5 phenols, 4 sulfur compounds, 3 organic acids, 2 furans, 1 ester, and 16 hydrocarbons (

Table 1. Volatile compounds detected in monocultured and co-cultured groups using GC-MS.

Classification	Volatile Compounds	CAS#	LRI	Concentration (μg/kg)
				Monoculture	Co-Culture
Ketones
A1	2-Butanone	78–93–3	894	0.305 ± 0.031	0.061 ± 0.002 ***
A2	2,3-Butanedione	431–03–8	971	0.286 ± 0.016	0.352 ± 0.015 *
A3	2-Methyl-3-Heptanone	13,019–20–0	1161	0.489 ± 0.009	10.25 ± 0.204 ***
A4	2-Heptanone	110–43–0	1177	1.82 ± 0.024	2.851 ± 0.042 ***
A5	3-Octanone	106–68–3	1249	0.896 ± 0.005	0.167 ± 0.003 ***
A6	6-Methyl-2-Heptanone	928–68–7	1232	0.285 ± 0.012	0.166 ± 0.003 ***
A7	2-Octanone	111–13–7	1281	1.155 ± 0.028	0.365 ± 0.037 ***
A8	1-Hydroxy-2-Propanone	116–09–6	1292	0.283 ± 0.006	ND
A9	2,2,6-Trimethyl-Cyclohexanone	2408–37–9	1314	0.964 ± 0.008	1.126 ± 0.06 *
A10	6-Methyl-5-Hepten-2-one	110–93–0	1332	1.525 ± 0.029	0.761 ± 0.068 ***
A11	2-Methyl-2-Hepten-4-one	22,319–24–0	1357	0.901 ± 0.015	0.305 ± 0.004 ***
A12	2-Nonanone	821–55–6	1384	2.039 ± 0.008	0.119 ± 0.001 ***
A13	3-Octen-2-one	1669–44–9	1402	ND	0.87 ± 0
A14	2-Decanone	693–54–9	1489	0.591 ± 0.009	0.591 ± 0.001
A15	1-(2-Methyl-1-Cyclopenten-1-yl)-Ethanone	3168–90–9	1589	0.299 ± 0.001	0.241 ± 0.007 ***
A16	4-Hydroxy-4-Methyl-Cyclohexanone	17,429–02–6	1601	25.174 ± 1.775	50.598 ± 0.329 ***
A17	Acetophenone	98–86–2	1648	0.21 ± 0.015	0.167 ± 0.022
A18	3-Acetyl-2-Octanone	27,970–50–9	1663	1.067 ± 0.017	1.744 ± 0.019 ***
A19	2,6,6-Trimethyl-2-Cyclohexene-1,4-Dione	1125–21–9	1691	0.999 ± 0.001	0.576 ± 0.005 ***
A20	1-(4-Methylphenyl)-Ethanone	122–00–9	1775	ND	0.216 ± 0.003
A21	Geranylacetone	3796–70–1	1849	0.807 ± 0.137	1.701 ± 0.04 ***
A22	trans-β-Ionone	79–77–6	1939	11.847 ± 0.125	29.688 ± 0.255 ***
A23	Epoxy-β-Ionone	23,267–57–4	1993	4.212 ± 0.01	13.234 ± 0.191 ***
A24	3,4-Dehydro-β-Ionone	1203–08–3	1999	ND	0.235 ± 0.004
A25	6,10,14-Trimethyl- 2-Pentadecanone	502–69–2	2121	1.017 ± 0.058	3.106 ± 0.077 ***
A26	(E)-4-Oxo-β-Ionone	27,185–77–9	2460	0.136 ± 0.004	0.296 ± 0.003 ***
A27	2-Pyrrolidinone	616–45–5	2027	1.55 ± 0.008	0.467 ± 0.027 ***
A28	Piperitenone Oxide	35,178–55–3	2144	ND	0.647 ± 0.006
A29	3-Ethyl-4-Methyl-1H-Pyrrole-2,5-Dione	20,189–42–8	2251	9.652 ± 0.042	21.381 ± 0.066 ***
Aldehydes
B1	Acetaldehyde	75–07–0	691	0.293 ± 0.011	0.272 ± 0.018
B2	2-Methyl-Butanal	96–17–3	907	0.812 ± 0.01	0.127 ± 0.005 ***
B3	3-Methyl-Butanal	590–86–3	911	1.3 ± 0.081	0.293 ± 0.05 ***
B4	Pentanal	110–62–3	972	0.41 ± 0.008	1.256 ± 0.045 ***
B5	Hexanal	66–25–1	1075	7.415 ± 0.094	14.336 ± 0.275 ***
B6	Heptanal	111–71–7	1180	0.9 ± 0.008	1.416 ± 0.069 ***
B7	Octanal	124–13–0	1283	ND	0.531 ± 0.009
B8	(E)-2-Heptenal	18,829–55–5	1320	ND	0.199 ± 0
B9	Nonanal	124–19–6	1388	0.622 ± 0.001	3.47 ± 0.22 ***
B10	(E)-2-Octenal	2548–87–0	1425	0.319 ± 0.001	0.601 ± 0.002 ***
B11	3-Furaldehyde	498–60–2	1454	0.117 ± 0.002	0.127 ± 0.002 **
B12	Benzaldehyde	100–52–7	1519	ND	0.127 ± 0.002
B13	(E)-2-Nonenal	18,829–56–6	1533	0.251 ± 0.007	1.04 ± 0.032 ***
B14	β-Cyclocitral	432–25–7	1621	5.31 ± 0.09	5.515 ± 0.004 *
B15	Safranal	116–26–7	1645	0.271 ± 0.009	0.259 ± 0.001
B16	3-Ethyl-Benzaldehyde	34,246–54–3	1707	0.054 ± 0.002	ND
B17	(E, Z)-2,4-Decadienal	25,152–83–4	1810	ND	0.165 ± 0.012
Alcohols
C1	Ethanol	64–17–5	925	1.854 ± 0.082	0.376 ± 0.011 ***
C2	1-Butanol	71–36–3	1137	12.617 ± 0.177	3.464 ± 0.216 ***
C3	1-Pentanol	71–41–0	1241	1.573 ± 0.006	ND
C4	1-Hexanol	111–27–3	1341	0.763 ± 0.121	0.397 ± 0.064 *
C5	1-Octen-3-ol	3391–86–4	1437	5.504 ± 0.085	4.424 ± 0.101 ***
C6	1-Heptanol	111–70–6	1442	0.442 ± 0.011	ND
C7	2-Ethyl-1-Hexanol	104–76–7	1476	2.418 ± 0.043	4.569 ± 0.767 *
C8	1-Octanol	111–87–5	1544	0.565 ± 0.012	0.147 ± 0.022 ***
C9	1-Nonanol	143–08–8	1646	0.52 ± 0.008	0.164 ± 0.02 ***
C10	Benzyl Alcohol	100–51–6	1862	1.818 ± 0.08	1.357 ± 0.039 ***
C11	Phenylethyl Alcohol	60–12–8	1898	0.196 ± 0.003	0.217 ± 0.011
C12	1-Decanol	112–30–1	1952	ND	0.156 ± 0.003
C13	Phytol	150–86–7	2575	ND	0.259 ± 0.009
Nitrogenous compounds
D1	N, N-Dimethyl-Methylamine	75–50–3	642	2.41 ± 0.325	5.367 ± 0.165 ***
D2	Pyrazine	290–37–9	1203	0.09 ± 0.002	ND
D3	2-Methyl-Pyrazine	109–08–0	1259	1.181 ± 0.066	0.162 ± 0.007 ***
D4	2,5-Dimethyl-Pyrazine	123–32–0	1316	1.784 ± 0.005	0.321 ± 0.001 ***
D5	2,6-Dimethyl-Pyrazine	108–50–9	1322	0.298 ± 0.002	ND
D6	2-Ethyl-6-Methyl-Pyrazine	13,925–03–6	1379	0.117 ± 0.003	ND
D7	Pyrrole	109–97–7	1501	0.072 ± 0.01	ND
D8	1-(1H-Pyrrol-2-yl)-Ethanone	1072–83–9	1958	0.078 ± 0.005	0.17 ± 0 ***
Phenols
E1	3-Methyl-4-Isopropylphenol	3228–02–2	1410	1.258 ± 0.002	0.736 ± 0.013 ***
E2	Butylated Hydroxytoluene	128–37–0	1903	0.057 ± 0.002	0.5 ± 0.016 ***
E3	Phenol	108–95–2	1986	0.204 ± 0.003	15.146 ± 0.012 ***
E4	P-Tert-Butyl-Phenol	98–54–4	2267	0.144 ± 0.001	1.358 ± 0.001 ***
E5	2,4-Di-Tert-Butylphenol	96–76–4	2286	0.369 ± 0.017	1.104 ± 0.078 ***
Sulfur compounds
F1	Methanethiol	74–93–1	677	0.069 ± 0.008	ND
F2	Dimethyl Disulfide	624–92–0	1063	0.3 ± 0.021	ND
F3	Dimethyl Trisulfide	3658–80–8	1375	0.114 ± 0.005	ND
F4	Dimethyl Sulfoxide	67–68–5	1568	0.069 ± 0	ND
Organic acids
G1	Acetic Acid	64–19–7	1435	1.319 ± 0.001	0.589 ± 0.001 ***
G2	Octanoic Acid	124–07–2	2037	0.774 ± 0.032	0.408 ± 0.058 ***
G3	n-Decanoic Acid	334–48–5	2248	ND	0.264 ± 0.003
Furans
H1	3-Methyl-Furan	930–27–8	857	0.102 ± 0.009	0.073 ± 0.005 *
H2	2-Pentyl-Furan	3777–69–3	1224	0.898 ± 0.026	2.368 ± 0.169 ***
Ester
I1	Dihydroactinidiolide	17,092–92–1	2358	4.982 ± 0.014	16.727 ± 0.223***
Hydrocarbons
J1	Decane	124–18–5	996	0.84 ± 0.041	0.279 ± 0.001 ***
J2	Undecane	1120–21–4	1089	4.533 ± 0.78	0.284 ± 0.003 ***
J3	1-(1-Cyclohexen-1-yl)-Ethanone	932–66–1	1115	ND	0.221 ± 0.001
J4	5-Ethyldecane	17,302–36–2	1129	0.551 ± 0.025	ND
J5	D-Limonene	5989–27–5	1191	ND	0.078 ± 0.006
J6	Dodecane	112–40–3	1196	12.567 ± 0.079	1.954 ± 0.038 ***
J7	Styrene	100–42–5	1251	1.153 ± 0.027	ND
J8	Tridecane	629–50–5	1297	6.323 ± 0.1	5.507 ± 0.087 ***
J9	2-Methyl-Tridecane	1560–96–9	1353	0.401 ± 0.007	0.409 ± 0.001
J10	3-Methyl-Tridecane	6418–41–3	1363	1.939 ± 0.003	2.193 ± 0.005 ***
J11	Tetradecane	629–59–4	1395	2.979 ± 0.517	5.455 ± 0.118 ***
J12	Pentadecane	629–62–9	1495	49.193 ± 0.109	30.276 ± 5.551 **
J13	Hexadecane	544–76–3	1595	39.117 ± 1.442	24.063 ± 0.051 ***
J14	Heptadecane	629–78–7	1705	421.347 ± 19.469	260.721 ± 4.31 ***
J15	Octadecane	593–45–3	1797	1.331 ± 0.017	ND
J16	(Z)-3-Heptadecene	1,000,141–67–3	1717	36.228 ± 1.801	12.014 ± 0.011 ***

Data are expressed as the mean ± standard deviations (n = 3). ND: not detected; LRI: linear retention index; CAS: Chemical Abstracts Service. The different significance values of p < 0.05, p < 0.01, and p < 0.005 (compared to the monoculture samples) are marked as *, **, and ***.

). Among them, norisoprenoids (e.g., trans-β-ionone, epoxy-β-ionone, and dihydroactinidiolide) significantly increased in the co-cultured microalgae (p < 0.05). Norisoprenoids are a diverse class of aromatic compounds contributing to the floral and fruity nuances of various grapes and wines [36]. Therefore, they might, at least in part, account for the pleasant microalgal aromas of “floral” and “fruity” imparted by co-cultivation with yeast. Notably, quantitative analysis of the OAVs (Table S3) demonstrated that β-ionone, epoxy-β-ionone, (E)-4-oxo-β-ionone, β-cyclocitral, safranal, and dihydroactinidiolide significantly exceeded thresholds, underscoring their dominant role in the enhanced floral and fruity aromas of co-cultured microalgae.

Cyanobacteria have been found to contain significant amounts of norisoprenoids [37], which are derived from the enzymatic oxidative cleavage of carbon–carbon double bonds in the polyene chains of carotenoids [38], and ionones are a typical class of carotenoid-derived norisoprenoids. In this study, several β-ionone derivatives (e.g., trans-β-ionone, epoxy-β-ionone, 3,4-dehydro-β-ionone, and (E)-4-oxo-β-ionone) were significantly increased in spirulina by co-cultivation with yeast (Table 1). β-Carotene is the flavor precursor of β-ionone and its derivatives, and its concentration in A. platensis was significantly increased by co-cultivation with S. cerevisiae (Figure 2e).

The C5–C9 aldehydes deliver desirable fresh–green/fruity to fatty–green/fruity aromas [39]. In the present study, several aldehydes, including pentanal, hexanal, heptanal, (E)-2-heptenal, and octanal, significantly increased in the co-cultured spirulina (p < 0.05) (Table 1). These C5–C9 aldehydes should contribute to the microalgal “fruity” aromas endowed by yeast co-cultivation. The odor-active aldehydes are generally derived from the non-enzymatic and enzymatic oxidative degradation of polyunsaturated fatty acids [40]. As mentioned above, co-culturing with S. cerevisiae significantly enhanced the accumulation of UFAs in microalgae biomass (Figure 2f), which can increase the flavor precursors for aromatic C5–C9 aldehydes.

Four malodorous volatile organic sulfur compounds (methanethiol, dimethyl disulfide, dimethyl trisulfide, and dimethyl sulfoxide) were identified in the monocultured A. platensis. These compounds were not detected in the co-cultured spirulina (Table 1). Cyanobacteria are known to emit alkane thiols (e.g., methanethiol) and methane sulfonic acids (e.g., dimethyl disulfide and dimethyl trisulfide) that have an intense putrid/muddy odor [41]. Co-culturing with S. cerevisiae seemed to eliminate these malodorous volatile organic sulfur compounds from the microalgal biomass, leading to a significant reduction in the “muddy” odor of spirulina (Figure 3a).

3.4. Acetic Acid as a Central Effector and Mediator in the Co-Culture System

Boosting the CO₂ fixation efficiency of microalgae under atmospheric conditions is a significant challenge, and several studies have reported that co-cultivation with S. cerevisiae can increase the atmospheric growth rate of certain microalgae (e.g., Chlamydomonas reinhardtii, Chlorella pyrenoidosa, and Chlorella vulgaris) [31,42,43]. In general, S. cerevisiae is an acidophilic organism that grows better under slightly acidic conditions (pH 4–6); however, the pH of the Zarrouk medium ranged from 9.2 to 11.0, which appears to be adverse to the growth of S. cerevisiae. Under alkaline medium conditions, S. cerevisiae has been reported to produce negligible ethanol and to fix CO₂ as bicarbonate and organic acids for intracellular pH regulation and medium acidification [44]. In the present study, S. cerevisiae grown in 5 g/L of glucose-supplemented Zarrouk medium was found to produce three types of organic acids—acetic acid, lactic acid, and fumaric acid—with acetic acid and lactic acid being predominant (Figure 3b). In the co-culture system, the concentration of acetic acid in the medium increased within 24 h to a maximal level of 0.45 g/L (Figure S1), which is approximately half of that detected in the medium of the monocultured S. cerevisiae. Thus, it seems that a significant amount of yeast-derived acetic acid was consumed by A. platensis in the co-culture system. The pH of the medium in the co-culture system decreased slightly from 10.49 to 10.32, suggesting that yeast-derived organic acids had a weak impact on the alkaline pH of the Zarrouk medium. In fact, alkaline pH is critical for preventing bacterial contamination during A. platensis cultivation. Our co-culture system is, thus, feasible in practical applications from the perspective of maintaining sterile conditions. Although microalgae have been documented to consume acetic acid mixotrophically [45,46], there have been no reports indicating that lactic acid and fumaric acid can also serve as additional sources of organic carbon.

To determine whether yeast-derived acetic acid was responsible for the improved color appearance and aroma quality of co-cultured spirulina, we investigated the effects of medium supplementation with acetic acid and yeast culture supernatant on the growth and volatile flavor profile of A. platensis. Acetic acid significantly enhanced microalgal biomass yield and aroma quality (p < 0.05), with improvements comparable to those induced by the yeast culture supernatant (Figure 3c,d). Considering that the medium’s pH (9.2–11.0) was not significantly affected by the addition of acetic acid during the cultivation period, acetate seems to be an effector for the yeast-induced changes in color appearance and aroma quality of spirulina (Figure S1).

Genome sequencing and transcriptomic analysis were performed to determine the transcriptional responses of A. platensis to yeast co-culture and acetate supplementation. The A. platensis FACHB-902 strain used in this study has a single circular chromosome of 6.44 Mb with a total sequence length of 6,437,174 bp and an average G+C content of 44.9%. No plasmid DNA sequences were found (Figure S2). The total RNA extracted from A. platensis FACHB-902 exhibited high integrity, and no contamination from S. cerevisiae was detected in the simulated co-culture separation (Figure S3). Figure S4 illustrates the relevant KEGG pathways and their corresponding gene counts enriched in the transcriptome of A. platensis FACHB-902. Figure 4a specifically illustrates the biosynthetic pathways of flavor precursors (β-carotene, linoleic acid, and γ-linolenic acid), including the fatty acid biosynthesis pathway and the methylerythritol phosphate pathway (MEP pathway). The subsequent cleavage of these precursors generates volatile flavor compounds with floral and fruity aromas (indicated by the boxes in the diagram). The internal control and gene primers for ribulose-1,5-bisphosphate carboxylase/oxygenase (RuBisCO), acetyl-CoA synthetase (ACS), 1-deoxy-D-xylulose 5-phosphate reductase (DXR), 1-deoxy-D-xylulose 5-phosphate synthase (DXS), geranylgeranyl diphosphate synthase (GGPPS), phytoene synthase (PSY), phytoene desaturase (PDS), acetyl-CoA carboxylase (ACCase), fatty acid synthase (FAS), acyl-CoA desaturase (ACD), and delta (6)-fatty-acid desaturase (FAD6) were designed based on the genome and transcripts (

Table 2. Primer sequences for real-time quantitative polymerase chain reaction of A. platensis FACHB-902.

Genes	Forward and Reverse Primers (5′ → 3′)
16S (internal control)	AP 16S-F	CGTAAACCTCTCCTCAGTTCAG
	AP 16S-F	GAACGGATTCACCGCAGTAT
Ribulose-1,5-bisphosphate carboxylase/oxygenase	RuBisCO-F	TTCTGCTTTGTTGCCTATCCG
	RuBisCO-R	ATCCAAATACGTTACCCACGA
Geranylgeranyl diphosphate synthase	GGPPS-F	ATCTGGAAGCTCAAAAGGCTAC
	GGPPS-R	CGCAACCACCGTTTCCAAA
Phytoene desaturase	PDS-F	AATTATATAGATCCGCTGCAT
	PDS-R	TCTTCACTGACATTATGGGGAC
1-Deoxy-D-xylulose 5-phosphate reductase	DXR-F	AGGCTCATTTTCTCTTTGGTT
	DXR-R	AACACAGAAGTATCCTGCAAC
1-Deoxy-D-xylulose 5-phosphate synthase	DXS-F	CTGTCTCCCCAATATGACCA
	DXS-R	ATTAATACCAGTCACCAGCAT
Phytoene synthase	PSY-F	GCTCTCGGTATTGCTAACCAG
	PSY-R	CCAGCGTTCATCTACTATGCC
Acetyl-CoA synthetase	ACS-F	TCGGTGATCTAATTCTAGCTG
	ACS-R	ATCGCATCAATAAACCCAT
Acetyl-CoA carboxylase	ACCase-F	GGAATGTTAAGCCTCATGCAA
	ACCase-R	CATGGCAAAACTAGCCGTCA
Fatty acid synthase	FAS-F	ATTCGGGGTATTGATCGCCC
	FAS-R	CTCCCCCATGTTCAACGTGA
Acyl-CoA desaturase	ACD-F	TATTTATGGCATTCCTCCACA
	ACD-R	GTAATTCCTAAGCCACCAGT
Delta (6)-fatty-acid desaturase	FAD6-F	GACACCGCTCATTTTCTCGGA
	FAD6-R	ATTCTTCTTTACGCCAGGGTT

F: forward; R: reverse.

).

In the present study, yeast co-cultivation significantly promoted the expression of genes responsible for the biosynthesis of beneficial components/flavor precursors (β-carotene, linoleic acid, and γ-linolenic acid) in A. platensis (Figure 4b). This enhancement coincided with the upregulation of RuBisCO expression over time, indicating that improved photosynthetic performance may underline the increased synthesis of these bioactive compounds.

Acetate has been extensively documented as an effective carbon source for the mixotrophic production of carotenoids and lipids in various microalgae species (e.g., Haematococcus, Chlorella, and Chlamydomonas) [47,48]. As an actively absorbed carbon source for microalgae, acetate is assimilated in the cytosol into acetyl-CoA and subsequently converted in the glyoxysome into succinate via the glyoxylate cycle [49], thereby fueling the tricarboxylic acid (TCA) cycle to enhance gluconeogenesis, fatty acid synthesis, and carotenoid synthesis [50,51]. Acetate assimilation has been shown to promote the biosynthesis of the phytyl side chain of chlorophylls by fueling the TCA cycle and the subsequent glycolytic pathway [52]. Collectively, yeast-derived acetic acid acts as a pivotal metabolic signal and mediator in the inter-organismal information exchange between A. platensis and S. cerevisiae throughout the co-cultivation process investigated in this study.

4. Conclusions

This study developed an innovative S. cerevisiae co-cultivation technique that simultaneously enhanced both the nutritional and sensory qualities of A. platensis. At a yeast-to-microalgae ratio of 10:1000, co-culture for 24 h transformed the color of the A. platensis biomass from dark blue–green to light green while imparting pleasant floral and fruity aromas. The process significantly increased beneficial components (β-carotene, linoleic acid, and γ-linolenic acid) while effectively eliminating undesirable sulfur compounds. Mechanistic analysis revealed that yeast-derived acetate regulates gene expression in A. platensis, promoting the synthesis of these valuable compounds. This preharvest bioprocessing strategy provides an effective solution for overcoming the limitations of A. platensis in functional food applications.