High-Spermidine-Producing Yeast Strain for Autophagy-Promoting Applications

Abstract

Polyamines, particularly spermidine, have emerged as key dietary factors with roles in cellular health, autophagy, and longevity. However, strategies for scalable production of polyamine-rich food ingredients remain limited. Here, we report the development of a high-spermidine-producing Saccharomyces cerevisiae strain, 3L63, obtained via ultraviolet mutagenesis of the K7 strain. This strain exhibited a 5.9-fold increase in the total polyamine content, with spermidine being the most abundant. A scalable fermentation system of up to 10⁴ L was established, yielding a dried yeast product that met food safety criteria. Whole-genome sequencing identified mutations in central metabolic pathways, including ARG3, and functional enrichment analysis suggested broad metabolic rewiring, supporting an enhanced biosynthetic capacity, including polyamines. Free amino acid profiling revealed higher arginine levels in 3L63 than in K7, which is consistent with its role as a polyamine precursor. The 3L63 yeast-derived product was enriched in essential amino acids and polyamines. Functionally, this strain promoted the proliferation of normal and senescent human dermal fibroblasts, and its autophagy-inducing activity exceeded that of equivalent concentrations of pure spermidine, suggesting synergistic effects of yeast-derived bioactive compounds. This study demonstrates a non-genetically modified, high-spermidine yeast strain as a promising functional food ingredient with potential applications in healthy aging.

1. Introduction

In recent years, global life expectancy has continued to rise, accompanied by a steady increase in the proportion of older adults in the population [1,2]. Aging is the most significant risk factor for numerous diseases, including cardiovascular disease, diabetes, neurodegenerative disorders, and cancer [3,4], and developing preventive strategies that target the aging process is considered a pressing challenge in both medical and nutritional sciences.

Various food-derived compounds with anti-aging potential have been identified [5,6]. Moreover, endogenous polyamines, such as putrescine (PUT), spermidine (SPD), and spermine (SPM), have attracted considerable attention because of their ability to promote cellular homeostasis and longevity [7,8]. SPD has demonstrated biological activities, such as anti-inflammatory effects [9], autophagy induction, and deceleration of age-related cellular decline [10]. These effects contribute to lifespan extension across diverse model organisms.

The naturally occurring polyamine content in conventional food sources varies widely but remains relatively low [11,12,13,14], making it difficult to intake sufficient polyamine solely through the diet. For instance, natto (fermented soybeans) contains approximately 190–680 nmol/g SPD [12]. The average daily intake in the Japanese population is estimated to be 90 μmol PUT, 74 μmol SPD, and 36 μmol mg SPM per day [14]. Polyamine levels tend to decline with age [10], suggesting the need for exogenous supplementation to sustain cellular function and health. Thus, there is a clear demand for high-polyamine ingredients that can be used as functional food supplements. Therefore, we applied selective breeding techniques to Saccharomyces cerevisiae sake yeast strains carrying polyamine biosynthetic genes to generate strains rich in SPD and SPM.

The core pathways of polyamine biosynthesis in yeast—such as those involving ornithine decarboxylase and S-adenosylmethionine (SAM)-dependent aminopropylation reactions—are well-characterized [15,16]. However, the specific genetic mutations and regulatory mechanisms responsible for enhanced polyamine accumulation remain largely unexplored. Considering growing public concern over genetically modified organisms (GMOs) [17] and the increasing preference for clean-label products among health-conscious consumers, we adopted a classical, non-GMO breeding strategy for strain development. This approach not only aligns with consumer and regulatory expectations but also improves the marketability and social acceptance of the resulting high-SPD yeast strain as a functional food ingredient. For this purpose, S. cerevisiae K7—originally isolated from sake brewing and known for its robust fermentation performance and long history of safe use in food production—was chosen as the parental strain [18].

Among the recognized anti-aging strategies, caloric restriction consistently extends lifespan across species, largely through enhanced autophagic activity [19,20]. Autophagy maintains cellular integrity by degrading and recycling damaged proteins and organelles, particularly under nutrient deprivation or stress [21]. Beyond cellular housekeeping, it contributes to infection resistance, metabolic regulation, and longevity. In the skin, autophagy supports fibroblast renewal, extracellular matrix maintenance, and stress resilience, thereby promoting a healthy appearance and delaying aging [22,23]. Nutrients that activate autophagy are therefore of particular interest as dietary strategies to counteract age-related decline. Among them, autophagy-inducing compounds such as SPD have been linked to multiple health benefits. In preclinical models, autophagy enhancement delays neurodegeneration by improving proteostasis and reducing toxic aggregates, protects cardiovascular function through removal of damaged mitochondria, and preserves skeletal muscle integrity by facilitating cytoplasmic turnover [24,25].

In this study, we aimed to develop a non-GMO S. cerevisiae strain with enhanced SPD production and to evaluate its potential as a functional ingredient for health-promoting applications. We hypothesized that adaptive strain development would yield a yeast variant with increased polyamine accumulation and beneficial biological activities, including support for cellular proliferation and autophagy. To test this, we established a scalable fermentation system and conducted a comprehensive evaluation of the strain’s biochemical and functional characteristics. This work was designed to address the growing demand for safe, food-derived autophagy-inducing nutrients and to explore the potential of polyamine-enriched yeast as a functional ingredient. The findings provide a foundation for the application of SPD-rich yeast in promoting skin health and broader human health benefits.

2. Materials and Methods

Nutritional composition analysis (Table S1), safety and quality evaluation (Tables S2 and S3), whole-genome sequencing and free amino acid profiling were outsourced to certified external laboratories, and analytical procedures were conducted in accordance with each laboratory’s internal quality standards. The methods are detailed in the Supplementary Information.

2.1. Yeast Strain Development

The parent yeast strain (S. cerevisiae K7) was cultured in Yeast Extract Peptone Dextrose (YEPD) medium containing 1% Bacto yeast extract, 2% Bacto peptone, and 2% glucose [26] at 30 °C for 16 h with 210 rpm shaking using shaking incubator (Takasaki Scientific Instruments Co., Ltd., Takasaki City, Japan). The culture was exposed to 45 W ultraviolet light at a dose sufficient to achieve ~1% survival and then diluted 10⁴–10⁵-fold with 50 mM potassium phosphate buffer (pH 7). Treated yeasts were spread onto agar plates containing the same nutrient composition and 2% agarose and incubated at 30 °C for 48 h. Colonies were screened for high polyamine content using an HPLC system with Shimadzu LabSolutions software version 5.97 SP1 (Shimadzu Corp., Kyoto, Japan). One mutant strain, designated as 3L63, demonstrated a significant increase in SPD content and was selected for further characterization and scaling up.

2.2. Yeast Fermentation and Drying

A high-polyamine-producing mutant yeast (3L63 strain) was cultured in a 3 L jar fermenter (Model MDL-3L, Marubishi Bioengineering Co., Ltd., Takamatsu City, Japan). The culture was maintained at 30 °C with agitation at 420 rpm for 4 days. The fermentation medium composed of 50 g/L sucrose, 3 g/L ammonium sulfate, 4 g/L potassium dihydrogen phosphate, 2 g/L dipotassium hydrogen phosphate, 0.4 g/L magnesium sulfate heptahydrate, 10 g/L Bacto yeast extract, 0.002 g/L biotin, 0.4 g /L calcium chloride dihydrate, 0.01 g/L manganese chloride tetrahydrate, 0.02 g/L zinc sulfate heptahydrate, 0.06 g/L ferric citrate hydrate, 0.002 g/L boric acid, 0.001 g/L ammonium molybdate tetrahydrate, 0.001 g/L potassium iodide, 0.001 g/L copper sulfate pentahydrate, 0.001 g/L cobalt chloride hexahydrate and 0.05 g/L sodium chloride (Japanese Patent 6663598). The yeast was cultured at 30 °C for 4 d with agitation at 420 rpm. After cultivation, the broth was centrifuged at 8000× g rpm, 4 °C for 10 min using high-speed refrigerated centrifuge (Hitachi, Ltd., Tokyo, Japan) to concentrate the yeast cells, which were then washed with water. The resulting mixture was then incubated at 60 °C for 30 min. The yeast was then spray-dried using a Büchi Mini Spray Dryer B-290 (BÜCHI Labortechnik AG, Flawil, Switzerland) at an outlet temperature of 95–100 °C to obtain dried yeast powder. The culture conditions were scaled up into a 30-L jar fermenter (Model MSJ-N2, Marubishi Bioengineering Co., Ltd., Takamatsu City, Japan) with the same medium composition and at the same temperature (30 °C). The 10⁴ L fermentation was carried out in a factory setting using the same medium composition, temperature, and duration as previously described for the smaller scale fermentations. The aeration and agitation conditions were adjusted according to the vessel scale.

2.3. Polyamine Quantification Through HPLC

Perchloric acid (60%), sodium carbonate, acetone, L-proline, toluene, and acetonitrile were purchased from Wako Pure Chemical Industries (Osaka, Japan). Dansyl chloride (10%) in acetone was purchased from the Tokyo Chemical Industry (TCI) (Tokyo, Japan). SPM phosphate hexahydrate, 1,7-diaminoheptane (TCI), SPD trihydrochloride (Sigma-Aldrich, St. Louis, MO, USA), and PUT dihydrochloride (Nacalai Tesque, Kyoto, Japan) were used as standards for SPM, SPD, and PUT, respectively. They were prepared in the same extraction solvents and processed as the samples to generate calibration curves.

The workflow for sample preparation and polyamine derivatization is shown in Figure S1. Powdered yeast samples (10–15 mg) were weighed into 15-mL tubes and mixed with 4 mL of extraction solvent containing 2 ppm 1,7-diaminoheptane (internal standard) in 5% perchloric acid. The samples were vortexed and incubated at 50 °C with shaking at 125 rpm for 30 min. After extraction, 1 mL of the supernatant was transferred to 1.5-mL microtubes and centrifuged at 12,000 rpm for 5 min. A 100-µL aliquot of the clear supernatant was diluted 5-fold with 400 µL of extraction solvent and placed into glass centrifuge tubes.

Subsequently, 250 µL of 18% sodium carbonate solution and 500 µL of 10 g/L dansyl chloride in acetone were sequentially added. The mixture was allowed to react at 45 °C for 1 h under light-shielded conditions. After cooling, 250 µL of 10% proline was added to quench the excess dansyl chloride, followed by a 10 min reaction. Subsequently, 2 mL of toluene was added for phase extraction. After mixing and centrifugation (2000 rpm, 5 min, 25 °C), 1 mL of the organic layer was collected and dried under a nitrogen atmosphere at room temperature (25 °C). The residue was re-dissolved in acetonitrile (0.5 mL) via ultrasonication for 40 min and then used for HPLC analysis.

Polyamines were separated through a C18 column (YMC-Pack ODS-A, 4.6 × 150 mm, 5 μm; YMC, Kyoto, Japan) and analyzed using an HPLC system with Shimadzu LabSolutions software Shimadzu Corp., Kyoto, Japan). The mobile phase consisted of acetonitrile and water (72:28, v/v) at a flow rate of 1.5 mL/min. The column was maintained at 40 °C, and the injection volume was 20 µL. Fluorescence was detected at excitation and emission wavelengths of 365 and 510 nm, respectively. The total run time was 30 min per sample.

2.4. Whole-Genome Sequencing, Mutation Detection, and Functional Enrichment Analysis

Genomic DNA was extracted from the selected 3L63 yeast strain using the Dr. GenTLE (from Yeast) High Recovery kit (Takara Bio, Shiga, Japan), according to the manufacturer’s protocol. Whole-genome sequencing and mutation analyses were outsourced to Hokkaido System Science Co., Ltd. (Sapporo, Japan), as described in the Supplementary Information. Genes with high and moderate annotations were identified from the results of the gene mutation analysis. Next, Gene Ontology (GO) and pathway enrichment analyses of the listed genes were performed using g:Profiler [27] (https://biit.cs.ut.ee/gprofiler (accessed on 17 July 2025)) and Kyoto Encyclopedia of Genes and Genomes (KEGG) Mapper [28] (https://www.genome.jp/kegg/mapper/ (accessed on 16 July 2025)) web tools, providing insight into their roles in metabolic networks and potential links to polyamine biosynthesis.

2.5. Free Amino Acid Analysis

For amino acid analysis, K7 and 3L63 yeast strains were cultured in YEPD medium composed of 10 g/L yeast extract, 20 g/L Bacto peptone, and 20 g/L glucose. Pre-cultures were prepared by inoculating each strain into 4 mL of YEPD medium in test tubes and incubating overnight at 30 °C. For the main culture, 500 mL of YEPD medium was dispensed into sterile 1-L Erlenmeyer flasks. Using an aseptic technique in a biosafety cabinet, pre-culture of each strain from one test tube was inoculated into triplicate flasks containing 500 mL of culture media. The cultures were incubated at 30 °C with shaking at 100 rpm for 20 h. After incubation, the yeast cultures were transferred into 500-mL centrifuge bottles and centrifuged at 8000 rpm for 10 min to pellet the cells. The supernatants were discarded, and the pellets from each strain were pooled and resuspended in 250 mL 0.9% sodium chloride solution. After centrifugation at 500× g, 25 °C for 15 min, the supernatant was discarded. Ethanol (200 mL) was added to the pellet to wash the cells, followed by centrifugation and removal of the supernatant. The final yeast pellets were dried under reduced pressure at 50 °C overnight and then air-dried at 25 °C for 3 d to obtain dried yeast samples for analysis.

Amino acid analysis was outsourced to IDEA Consultants, Inc. (Osaka, Japan), and the free amino acids in the K7, 3L63, and 3L63 yeast products were analyzed using HPLC. The sample preparation and analysis procedures are described in the Supplementary Information.

2.6. HDF Culture and Cell Proliferation Assay

We evaluated the effect of 3L63 yeast on the proliferation of HDFs provided by Kurabo Industries, Ltd. (Osaka, Japan). The cells were maintained in a culture medium (Dulbecco’s Modified Eagle Medium (DMEM) supplemented with 10% fetal bovine serum and 1% penicillin/streptomycin). For the proliferation assay, HDFs were seeded in 96-well plates and treated with different concentrations of 3L63 yeast in the control media. In a subset of experiments, to mimic aging conditions, the cells were treated with 400 μM hydrogen peroxide for 2 h (H₂O₂; Wako Pure Chemical Industries) to induce senescence by oxidative stress before adding 3L63 yeast at several concentrations. After 24 h of incubation, cell viability and proliferation were assessed using Cell Counting Kit-8 (Dojindo, Tokyo, Japan). The absorbance was measured at 450 nm using a microplate reader (Multimode Plate Reader ARVO X3; PerkinElmer, Shelton, CT, USA).

2.7. Autophagy Induction Analysis

2.7.1. CYTO-ID Staining for Autophagic Activity

HeLa cells were cultured in DMEM supplemented with 10% fetal bovine serum, 100 U/mL penicillin, and 100 µg/mL streptomycin. For the CYTO-ID autophagy detection assay (Enzo Life Sciences, Inc., New York, NY, USA), cells were seeded at a density of 3 × 10⁴ cells/well in a 24-well plate and incubated at 37 °C with 5% CO₂ for 24 h. On the following day, the culture medium was removed and replaced with 0.75 mL of fresh medium. Then, 0.25 mL of the test sample, prepared at four times the final concentration, was added to each well, resulting in a total volume of 1 mL. After incubation for 4.5 h at 37 °C and 5% CO₂, the cells were washed twice with 2 mL of 1× assay buffer. Subsequently, 0.5 mL of staining solution, composed of 2 µL CYTO-ID Green Detection Reagent 2 and 1 µL Hoechst 33342 Nuclear Stain diluted in 1 mL of culture medium, was added to each well. The cells were incubated in the dark at 37 °C for 30 min. After staining, the cells were washed once with 1 mL of 1× assay buffer and once with 1 mL culture medium. Finally, the medium was replaced with 0.5 mL of PBS, and the cells were imaged under a fluorescence microscope (Nikon Eclipse TC2000-S; Nikon, Tokyo, Japan) equipped with a digital camera (Nikon Ds-Ri2). The intensity of CYTO-ID fluorescence and number of nuclei were quantified using ImageJ software version 1.54g [29].

2.7.2. mRFP-GFP Tandem Fluorescent-Tagged LC3 (tfLC3) Assay

Yeast samples were treated with artificial digestion fluid (Disintegration Test Solution No. 1 containing 0.2% pepsin; Wako Pure Chemical Industries and Disintegration Test Solution No. 2 containing 0.28% pancreatin; Wako Pure Chemical Industries). An aqueous extract was prepared by soaking the samples in distilled water. The tfLC3 assay was performed by Human Metabolome Technologies, Inc. (Yamagata, Japan), following the method described in [30] and the Supplementary Information.

2.8. Protein Variant Modelling and Docking Simulation

The variant protein structure was predicted using protein structure modelling (SWISS-MODEL) via a homology modeling approached [31]. The parent strain K7 and variant strain 3L63 protein sequences were used as the input for structure prediction. The quality of the predicted models was evaluated based on GMQE and QMEAN scores provided by SWISS-MODEL.

Docking simulations were performed using AutoDock Vina through the CB-Dock2 platform [32]. This method automatically detects potentials binding cavities on the protein surface and carried out docking within the identified sites. For each simulation, CB-Dock2 provided the cavity volumes and the corresponding Vina score, which were used to evaluate the binding affinity and binding pocket characteristics. The best-ranked docking poses were selected on the lowest Vinal score and biologically relevant binding volume. These results were used for comparisons between the same proteins derived from the parent and variant strains. This comparison allowed an assessment of how the introduced mutation altered ligand-binding properties and potential functional differences between the two protein forms.

2.9. Statistical Analysis

Statistical analyses were performed using SigmaPlot 15.0 or Excel software. Data normality was assessed using the Shapiro–Wilk test, and homogeneity of variances was confirmed using the Brown–Forsythe test. One-way analysis of variance (ANOVA) was used for all pairwise multiple comparisons using Tukey’s honestly significant difference (HSD) test or the paired t-test for appropriate experiments.

3. Results

3.1. Physical Characterization of Newly Developed 3L63 Yeast

A high–polyamine-producing S. cerevisiae strain was generated through ultraviolet mutagenesis of the K7 strain. The selected mutant, designated 3L63, was scaled up to the commercial production level (Figure 1). The strain was successfully cultivated in a 30-L jar fermenter, and the resulting cells were washed, sterilized, and spray-dried to obtain an SPD-enriched yeast powder. Figure 1 shows the morphology of the yeast under a microscope and the appearance of the final powdered product. The dried yeast was off-white to light-yellow in color and insoluble in water.

The nutritional profiles of the 3L63 yeast-derived powders are summarized in

Table 1. Nutritional composition of 3L63 yeast powder (per 100 g).

Item	Measured Value
Moisture	4.5 g
Protein * (excluding Polyamines)	45.7 g (45.1 g)
Fat	10.0 g
Carbohydrate	29.9 g
Ash	9.9 g
Energy	392 kcal
Sodium	25 mg
Salt Equivalent	0.064 g

* Protein content calculated from nitrogen content.

. The parent S. cerevisiae K7 strain is food-safe, and the 3L63-derived powder passed acute oral toxicity and hygiene tests in accordance with the Japanese food safety standards (Table S2). Comprehensive quality assessments confirmed that the powder contains ≥2.0 mg/g of SPD, with a moisture content ≤6.0%, and no detectable foreign matter. Microbiological analysis showed total viable bacteria ≤300 CFU/g and fungi ≤100 CFU/g; coliforms were not detected. Heavy metals (Pb) and arsenic (As₂O₃) were within acceptable limits (≤20 ppm and ≤1 ppm, respectively). These results support the chemical and microbiological safety of these products in potential food and nutraceutical applications. The product specifications listed in Table S3 were established to ensure product safety and consistency. For the sub-end-product, the yeast was mixed with 15% α-cyclodextrin post-process, as described in the Supplementary Information.

3.2. Polyamine Content in Yeast Strains

To evaluate the polyamine-producing capacity of the newly developed yeast strain 3L63, we quantified the intracellular levels of three major polyamines: PUT, SPD, and SPM. The polyamine content of 3L63 was compared with that of the parent K7 strain and its commercial yeast product. To obtain a commercial product, the formulation was optimized for improved powder handling and consistent production, achieved through modifications to the culture conditions and addition of cyclodextrin. Although the underlying yeast strains were identical, these alterations resulted in distinct polyamine profiles.

The parent K7 strain exhibited low levels of SPD (0.51 mg/g) and SPM (0.45 mg/g), with PUT below the detection limit, resulting in a total polyamine content of only 0.96 mg/g (

Table 2. Polyamine content in dried yeast harvested from 3L63 and K7 strains.

Strain	Putrescine (mg/g)	Spermidine (mg/g)	Spermine (mg/g)	Total (mg/g)
K7	0.05 (±0.06)	0.50 (±0.02)	0.29 (±0.22)	0.84 (±0.18)
3L63	0.75 (±0.31)	3.80 (±0.38)	0.38 (±0.02)	4.93 (±0.66)
Commercial product *	0.39 (±0.05)	2.55 (±0.05)	0.28 (±0.06)	3.21 (±0.15)

Data indicate the mean ± standard deviation for K7 and 3L63 were obtained during the screening phase. * The commercial formulation was modified for enhanced powder handling and production stability by adjusting culture conditions and adding α-cyclodextrin (15% w/w).

and Figure 2a). In contrast, the 3L63 strain accumulated significantly higher levels of polyamines, with 1.07 mg/g PUT, 4.24 mg/g SPD, and 0.36 mg/g SPM, amounting to 5.66 mg/g. This reflects an approximately 5.9-fold increase compared with that of the parent strain.

The commercial yeast product contained 0.63 mg/g PUT, 2.26 mg/g SPD, and 0.19 mg/g SPM, with total polyamine content of 3.08 mg/g. The final product maintained a considerably higher polyamine content than K7 yeast. Additionally, we compared polyamine levels in 3L63 cells with those reported in natural food sources, such as natto (fermented soybeans) [12] and mushrooms [13] (Figure 2b). These findings demonstrate that 3L63 is a robust, high-polyamine-producing strain that is especially rich in SPD and is suitable for application as a functional ingredient in foods and supplements.

3.3. Whole-Genome Sequencing and Functional Enrichment Analysis

Whole-genome sequencing of S. cerevisiae 3L63 yielded 43,539,176 reads, of which 80.59% (35,089,430 reads) were successfully mapped to the reference genome of the K7 strain. Approximately 2.80% (1,220,805 reads) of the mapped reads were identified as duplicates, whereas 19.41% (8,449,746 reads) remained unmapped. After variant calling and annotation, genetic variants were classified based on their predicted functional impact. Only variants with high and moderate effects were selected for downstream analysis, including missense mutations, nonsense mutations, and frameshift variants, which are likely to alter protein function and are thus considered biologically meaningful.

To evaluate the biological implications of the identified variants, functional enrichment analysis was performed using a filtered gene list (860 genes) and the g:Profiler tool. A total of 55 significantly enriched biological process (BP) terms, 19 molecular function (MF) terms, and 26 cellular component (CC) terms were identified (adjusted p < 0.05) (Figure 3). The significant BP terms also included the regulation of metabolic processes (GO:0019222, adjusted p = 6.171 × 10⁻⁴), primary metabolic processes (GO:0044238, adjusted p = 6.171 × 10⁻⁴), and metabolic processes (GO:0008152, adjusted p = 1.070 × 10⁻³). This result suggests that the variants are involved in the regulation of core metabolic pathways and that these mutations may alter the ability of the yeast strain to synthesize, degrade, or regulate key metabolites. The detailed results are presented in Figure S2.

KEGG pathway mapping of the filtered gene list revealed hits for the pathways listed in Table S4. These pathways suggest potential rewiring of central metabolism and amino acid biosynthesis, which could affect cellular growth and stress adaptation.

Manual inspection of the gene list revealed five mutations in central metabolic pathway genes: SAM1, LEU2, ACO2, ARG3, and AGC1 (Tables S5 and S6). ARG3 is functionally associated with arginine metabolism, which occurs upstream of PUT and is a major polyamine. However, polyamine-specific GO terms (e.g., GO:0006596: polyamine biosynthetic process) were not significantly enriched, possibly because of the limited number of direct polyamine biosynthesis genes in the input list. To further investigate the potential impact of the ARG3 mutation, we compared the protein structures of ARG3 K7 and ARG3 3L63 using in silico modeling (Table S6). The analysis suggested that the mutation may alter the cavity volume for the ARG3 substrate ornithine and potentially enhance substrate binding, as indicated by improved docking scores from AutoDock Vina (Figure S3).

3.4. Free Amino Acid Profiles

Intracellular polyamine levels were quantified in K7 and 3L63 yeast strains to assess polyamine dynamics during cultivation (

Table 3. Selected free amino acid and polyamine data for S. cerevisiae K7 and 3L63 strain in flask cultivation, and 3L63 commercial product.

Amino Acid	K7 Strain (μmol/g)	3L63 Strain (μmol/g)	3L63 Product (μmol/g)
Aspartic acid (Asp)	4.96 (1.6%)	5.59 (2.0%)	4.73 (1.3%)
Glutamic acid (Glu)	134 (43.8%)	104 (36.7%)	43.9 (12.4%)
Glycine (Gly)	10.4 (3.4%)	13.8 (4.9%)	18.1 (2.6%)
Threonine (Thr)	9.39 (3.1%)	7.57 (2.7%)	10.5 (3.0%)
Arginine (Arg)	17.4 (5.7%)	28.3 (10.0%)	65.1 (18.4%)
Serine (Ser)	43.8 (14.3%)	43.5 (15.4%)	12.7 (3.6%)
Alanine (Ala)	14.9 (4.9%)	10.5 (3.7%)	38.9 (11.0%)
Valine (Val)	2.18 (0.7%)	1.83 (1.83%)	15.2 (4.3%)
Leucine (Leu)	2.21 (0.7%)	1.90 (0.7%)	19.4 (5.5%)
Lysine (Lys)	6.10 (2.0%)	12.9 (4.6%)	14.1 (4.0%)
Proline (Pro)	4.91 (1.6%)	2.21 (0.8%)	6.57 (1.9%)
Total amino acids	306 (100%)	283 (100%)	354 (100%)
Polyamine	K7 Strain (mg/g)	3L63 Strain (mg/g)	3L63 Product (mg/g)
Putrescine (Put)	0.04 (1.9%)	0.39 (9.4%)	0.45 (13.3%)
Spermidine (Spd)	1.62 (78.3%)	3.41 (81.8%)	2.60 (76.7%)
Spermine (Spm)	0.41 (19.8%)	0.37 (8.9%)	0.34 (10.0%)
Total polyamines	2.07 (100%)	4.17 (100%)	3.39 (100%)

Free amino acid and polyamine content in the K7 and 3L63 strains after overnight cultivation. Full data are presented in Table S7.

). The total polyamine content in K7 was 2.07 mg/g, comprising 0.04 mg/g PUT, 1.62 mg/g SPD, and 0.41 mg/g SPM. Consistently, the 3L63 strain exhibited a markedly higher polyamine level of 4.17 mg/g, consisting of 0.39 mg/g PUT, 3.41 mg/g SPD, and 0.37 mg/g SPM. This approximately two-fold increase was primarily attributed to the enhanced SPD accumulation.

Intracellular free amino acid profiles were examined during yeast cultivation. As an increase in free amino acids can result from protein degradation, ethanol was applied immediately after cultivation to fix the cells and suppress autolysis. The amino acid content of the final product was assessed using samples subjected to different cultivation times and sterilization processes. The results are presented in Table 3 and Table S7.

Overall, the total free amino acid content of the 3L63 strain (283 μmol/g) was slightly lower than that of the K7 parent strain (306 μmol/g). Despite this, several individual amino acids showed significant variation. Arginine nearly doubled in 3L63, increasing from 17.4 μmol/g (5.7%) to 28.3 μmol/g (10.0%), suggesting enhanced biosynthesis or accumulation in the new variant. Lysine also increased from 6.10 μmol/g (2.0%) in K7 to 12.9 μmol/g (4.6%) in 3L63. Conversely, glutamic acid decreased from 134 μmol/g (43.8%) in K7 to 104 μmol/g (36.7%) in 3L63, although it remained the most abundant amino acid in both strains. Minor differences were also observed in glycine (K7:10.4 μmol/g; 3L63:13.8 μmol/g), alanine (K7:14.9 μmol/g; 3L63:10.5 μmol/g), and serine (K7:43.8 μmol/g; 3L63:43.5 μmol/g), whereas the overall amino acid profile remained relatively consistent. Notably, the levels of branched-chain amino acids (leucine, isoleucine, and valine) were lower in 3L63 cells than in K7 cells.

A comparison between the mid-fermentation sample of 3L63 and the final 3L63-derived product revealed a substantial increase in total amino acids from 283 to 354 μmol/g. This suggests a concentration effect or selective enrichment during downstream processing. Arginine exhibited a considerable increase from 28.3 μmol/g (10.0%) to 65.1 μmol/g (18.4%), becoming a dominant amino acid in the final product. Alanine also increased sharply from 10.5 μmol/g (3.7%) to 38.9 μmol/g (11.0%). Significant increases were also observed in branched-chain amino acids, with valine increasing from 1.83 to 15.2 μmol/g, isoleucine from 1.17 to 13.6 μmol/g, and leucine from 1.90 to 19.4 μmol/g.

Glutamic acid decreased from 104 μmol/g (36.7%) to 43.9 μmol/g (12.4%), and serine from 43.5 μmol/g (15.4%) to 12.7 μmol/g (3.6%) during fermentation, indicating selective degradation or metabolic transformation. Additionally, the final product showed marked increases in aromatic and sulfur-containing amino acids, including phenylalanine (from 0.55 to 10.2 μmol/g), tyrosine (from 0.37 to 9.51 μmol/g), cystine (from 4.41 to 9.31 μmol/g), and methionine (from 2.79 to 6.10 μmol/g), thereby enhancing its nutritional value.

The total polyamine content shows a significant increase from the K7 strain to the 3L63 strain and the 3L63 product. The K7 strain contains a total of 2.07 mg/g of polyamines, which increases to 4.17 mg/g in the 3L63 strain and is slightly lower at 3.39 mg/g in the 3L63 product. This indicates that while the enhancement of polyamines in the 3L63 strain is substantial, there may be slight losses during the processing of the product. The PUT increases markedly from 0.04 mg/g (1.9%) in the K7 strain to 0.39 mg/g (9.4%) in the 3L63 strain, representing a significant enhancement. SPD is the most abundant polyamine in both the 3L63 strain (3.41 mg/g) and product (2.60 mg/g), accounting for the majority of the total polyamine content (81.8% in the 3L63 strain, 76.7% in the product). This high level of SPD is indicative of efficient biosynthesis pathways in the 3L63 strain. SPM levels exhibit a slight decrease when comparing the K7 strain (0.41 mg/g) and the 3L63 strain (0.37 mg/g), with an even lower concentration in the 3L63 product (0.34 mg/g). The reduction in SPM could be attributed to the metabolic allocation of resources toward the synthesis of PUT and SPD, which are often prioritized in yeast metabolism under certain growth conditions.

Overall, transformation from the parent yeast K7 to the 3L63 strain to the final product resulted in pronounced enrichment of essential amino acids and a notable increase in polyamine content, particularly SPD, underscoring the potential of this strain for developing nutritionally enhanced yeast-based products.

3.5. HDF Proliferation

We evaluated whether the proliferative ability of fibroblasts, vital for skin health, can be enhanced by the addition of SPD-enriched yeast. Our experiments confirmed that 3L63 yeast significantly promoted cell proliferation not only in normal fibroblasts but also in senescence-induced cells with 100 μg/mL yeast, equivalent to 1.72 μM SPD supplementation (Figure 4). These findings suggest that SPD-enriched 3L63 yeast may help restore the diminished proliferative capacity of aging fibroblasts and improve signs of skin aging.

3.6. Autophagy Induction by 3L63 Yeast

Autophagy activation was first assessed in HeLa cells using the CYTO-ID Autophagy Detection Kit (Figure 5a,b). Fluorescence intensity analysis revealed that treatment with 3L63 yeast (containing 0.86 μM SPD) significantly enhanced autophagic activity, reaching an average fluorescence intensity of 130% relative to the untreated control (p < 0.001). Notably, this effect was substantially greater than that observed with pure SPD treatment. Specifically, 100 μM pure SPD induced a modest increase to 113% of control levels, whereas 0.86 μM SPD showed only a slight elevation to 102%, neither of which were significant differences compared to the control group. These findings suggest a potential synergistic effect between SPD and other bioactive components present in the yeast matrix, which may stimulate autophagy more effectively than SPD alone.

To validate our observations regarding autophagy induction, a tandem fluorescent-tagged LC3 (tfLC3) assay was performed using HeLa cells (Figure 6a). The cells were treated with aqueous extracts and artificial digestion fluid-treated 3L63 yeast corresponding to 0.1, 1, and 5 μM SPD concentrations. Even at the lowest dose (0.1 μM SPD equivalent), a significant increase in autophagy was observed compared to that in the control, supporting the robust activity of the yeast product. Additionally, the levels of autophagy-related protein LC3-II were examined in HDFs using Western blot analysis under autophagic flux conditions (Figure S5). Although the result was based on a single experiment (no replicates were performed), the results consistently showed that 3L63 yeast containing 50 μM SPD induced higher LC3-II expression than pure SPD and starvation conditions, further supporting the superior autophagy-inducing capacity of the yeast-derived SPD formulation.

4. Discussion

4.1. Food Safety and Formulation Strategy

Yeast is considered GRAS (Generally Recognized as Safe) by regulatory authorities such as the U.S. FDA and has a long-established history of use in fermented foods, including bread, beer, and dietary supplements [33]. The parent strain used in this study, S. cerevisiae K7, was originally isolated from sake brewing and is well-known for its robust fermentation performance, high ethanol tolerance, and adaptability to various stress conditions. As a diploid homothallic strain, K7 has been extensively used in traditional Japanese food production. Genomic analyses have revealed that K7 harbors unique alleles related to flavor compound synthesis, nitrogen metabolism, and stress response, which contribute to its desirable brewing characteristics [18,34,35]. Furthermore, transcriptomic and proteomic studies have provided insights into K7’s metabolic regulation under fermentation conditions, thereby highlighting its potential as a platform for the development of functional food ingredients [36,37].

Based on this robust and well-characterized background, we developed a food-grade, non-GMO yeast strain (3L63) enriched in polyamines, particularly spermidine (SPD), using classical mutagenesis and selection strategies. Previous research has primarily focused on enhancing polyamine production through genetic engineering. For instance, Kim et al. (2017) engineered S. cerevisiae to overexpress SPD biosynthetic enzymes, achieving a yield of 224 mg/L SPD, while Qin et al. (2021) reported a production of 191 mg/L SPD using a metabolic engineering approach [38,39]. In comparison, our developed strain achieved a maximum yield of 146 mg/L SPD. Although this yield is lower than those reported in engineered strains, it demonstrates that significant SPD enrichment can be accomplished without genetic modifications.

Additionally, the 3L63 strain exhibited stable SPD accumulation under various nutrient and oxygenation conditions and was successfully scaled from shake flasks to 3-L and 30-L fermenters, ultimately exceeding industrial volumes of 10,000 L. To enhance powder handling and functionality for commercial readiness, we supplemented the strain with cyclodextrin [40,41]. While the engineered strains achieved higher production levels, they involve genetic modifications that may present challenges in regulatory approval and consumer acceptance for food applications. In contrast, our classical mutagenesis approach produced a non-GMO strain that complies with food-grade regulations and maintains market appeal. Although the absolute increase in polyamines is relatively modest compared to engineered systems, the novelty of our work lies primarily in the approach and market positioning rather than in the biochemical output.

Nevertheless, limitations of the mutagenesis approach remain. Random mutations make it challenging to precisely identify the genetic changes responsible for the enhanced SPD phenotype. Additionally, long-term evolutionary stability during repeated industrial propagation requires further validation. Despite these considerations, 3L63 represents a practical and regulatory-compliant strategy for producing SPD-enriched yeast, complementing prior metabolic engineering efforts reported [38,39]

4.2. Mechanism of Enhanced SPD Production

To elucidate the mechanisms underlying this phenotype, we combined whole-genome sequencing, amino acid profiling, and metabolic pathway interpretation. Whole-genome variant analysis revealed several mutations in genes associated with amino acid metabolism and central metabolic regulation, including ARG3, SAM1, LEU2, ACO2, and AGC1. Among these, ARG3, which encodes an ornithine carbamoyltransferase, plays a pivotal role in the urea cycle and catalyzes the conversion of ornithine and carbamoyl phosphate to citrulline, an essential step in arginine biosynthesis [42,43]. A moderate-impact missense mutation was identified in ARG3 of 3L63 strain. Structural modeling and molecular docking analysis revealed a reduction in the substrate-binding cavity volume (from 1580 to 1491 ų) and an improved Vina binding score (from −4.8 to −5.1), suggesting enhanced binding affinity of ornithine to the mutant enzyme. Although ornithine levels were not directly quantified in this study, this finding implies an improved conversion efficiency from ornithine to citrulline, supporting an increased arginine biosynthetic flux.

Free amino acid profiling further supported this hypothesis, showing significantly elevated levels of arginine, lysine, glycine, and aspartic acid in 3L63 cells, whereas methionine, leucine and glutamic acid levels remained unchanged or decreased. This increase in the upstream precursor pools likely fuels the native polyamine biosynthetic pathway in yeast; arginine is converted to ornithine via arginase (CAR1), followed by ornithine decarboxylation to PUT (SPE1), and then extended to SPD (SPE3) and SPM (SPE4) (Figure 7).

Although some organisms possess a direct arginine decarboxylation pathway that produces agmatine, S. cerevisiae lacks the enzymes required for this alternative route, such as arginine decarboxylase and agmatinase [15]. Therefore, the enhanced polyamine production in 3L63 is most plausibly attributed to the upregulation of the conventional ornithine decarboxylase pathway, facilitated by increased arginine availability and potentially improved flux regulation via ARG3. Additionally, functional enrichment analysis of the 3L63 variant genes revealed significant GO terms, including metabolic process, primary metabolic process, and regulation of metabolic process, suggesting that the mutation landscape in 3L63 may contribute to broad metabolic rewiring that supports enhanced polyamine biosynthesis.

Docking simulations for LEU2 and SAM1 were also conducted, as shown in the Supplementary Data (Table S6). The docking scores for SAM1 with methionine and LEU2 with their respective substrates showed no differences between the two strains, and our amino acid analysis similarly revealed no significant changes in methionine or leucine percentage. These findings suggest that mutations in LEU2 and SAM1 are unlikely to be major contributors to the increased spermidine levels observed in strain 3L63.

In summary, our findings support a hypothetical mechanism whereby mutations in key metabolic genes, such as ARG3, result in elevated intracellular arginine levels. This surplus arginine is channeled through the native ornithine decarboxylation pathway, ultimately leading to increased SPD production. Although our data suggest that the ARG3 mutation may contribute to elevated arginine levels and increased spermidine production, this conclusion is currently based on amino acid profiling and in silico docking. Further validation using metabolic flux analysis, transcriptomic profiling, and enzyme activity assays to directly validate the biochemical impact of the ARG3 mutation will provide deeper insights into the regulatory networks underlying polyamine overproduction.

4.3. Functional Benefits of SPD: Autophagy and Healthy Aging

SPD is a naturally occurring polyamine that has garnered significant attention owing to its role in promoting autophagy, maintaining cellular homeostasis, and supporting healthy aging [8,9,10,44]. Autophagy is a highly conserved intracellular recycling process that declines with age and is linked to the pathogenesis of numerous age-related diseases including neurodegeneration, cardiovascular dysfunction, and skin aging [22,23,24,25]. SPD reactivates autophagy through epigenetic and metabolic regulation, thus mimicking the effects of caloric restriction and inducing longevity [44]. This underlies the broad protective effect observed in preclinical and epidemiological studies [24].

Although no formal dietary reference intake has been established, a growing body of research derived from both animal and human studies supports a daily intake of approximately 15–30 mg SPD to promote longevity and reduce the risk of age-associated diseases [8,45,46]. For instance, Eisenberg et al. [8] reported that dietary SPD extended lifespan and improved cardiovascular health in mice, whereas Kiechl et al. [38] demonstrated inverse associations between SPD intake and mortality, as well as a lower incidence of cardiovascular events, in a human cohort study.

Achieving the longevity-promoting intake range through diet alone is challenging. Wheat germ (~24 mg/100 g), fermented soybeans (natto; ~12–20 mg/100 g), mushrooms (~5–10 mg/100 g), and green peas (~4–6 mg/100 g) are SPD-rich foods [12,13] that are typically not consumed in sufficient quantities. The average daily intake in Western and Asian populations is just 10–12 mg [11,46]. Our 3L63 yeast-derived product offers a potent and convenient dietary source of SPD, containing over 2.0 mg/g. Only 10 g/d provides approximately 20 mg, which is sufficient to bridge this dietary gap.

In addition to inadequate dietary intake, the body’s endogenous production of SPD declines with age, with blood concentrations averaging 0.63 μM at birth but falling to 0.25 μM by age 70 [47], a drop linked to cellular dysfunction and accelerated aging. This decline corresponds to ~0.28 mg less SPD in 5 L human blood. At 2.0 mg/g, only ~160 mg of 3L63 yeast powder would theoretically restore this youthful blood concentration. However, normalizing blood levels may not equate to replenishing whole-body stores or sustaining tissue autophagy.

For physiological and longevity benefits, the literature supports a daily intake of 15–30 mg SPD; at 2.0 mg/g, this corresponds to roughly 7.5–15 g of 3L63 powder. A practical target of 10–15 g/d (≈20–30 mg SPD) not only fills the typical dietary gap but also provides a margin for incomplete absorption, first-pass metabolism, and tissue distribution. This distinction between the minimal dose required to restore blood levels and the sustained intake required for systemic effects underscores the advantages of using 3L63 as a functional food ingredient for healthy aging. Notably, in a clinical study among 50–70-year-old men, daily supplementation with 300 mg of 3L63 yeast commercial product (together with 1 g BCAA given to both control and test groups) for 8 weeks led to significant increases in skeletal muscle index (SMI), muscle mass, and fat-free mass [48]. This suggests that even relatively low doses of 3L63 than theoretical estimation can confer measurable functional benefits in humans.

To further assess the biological relevance of the activity of 3L63-derived SPD, we performed a series of functional assays. Autophagy induction was observed in HeLa cells using a fluorescent assay, where the 3L63 yeast product (containing 0.86 μM SPD) significantly increased autophagy activity by ~130% relative to that observed in the control, exceeding the effect of pure SPD at 100 μM. Similarly, a tfLC3 assay revealed that even low concentrations of the 3L63 product (0.1–5 μM SPD) activated autophagic flux in cells. These results suggest a synergistic effect between SPD and other yeast-derived components, enhancing autophagy beyond that achieved with SPD alone.

Using HDFs, we demonstrated that the 3L63 yeast product promoted cell proliferation in both normal and senescent fibroblasts. Given that autophagy supports proteostasis, mitochondrial function, and resistance to oxidative stress, the enhanced proliferation of aging fibroblasts likely reflects the restoration of these core anti-aging mechanisms. These findings reinforce the role of SPD-induced autophagy as a central mediator of cellular rejuvenation, particularly in skin health, and highlight that 3L63 yeast is a promising candidate for anti-aging intervention via nutritional or topical routes. Additionally, in our previous study using C2C12 myoblasts, we showed that 25 μg/mL yeast-derived SPD promoted myogenic differentiation and enhanced muscle cell viability, indicating a role in supporting muscle growth [48]. Building on this, SPD-enriched 3L63 yeast may also confer functional benefits for skeletal muscle health—further underscoring its potential as a versatile anti-aging dietary ingredient. While autophagy induction was observed in cell lines; however, in certain experiments, the limited number of replicates may compromise the reliability of the functional interpretation. Similarly, while fibroblast proliferation assays yielded promising outcomes, they remain in the preliminary stages and require further validation. Collectively, these findings indicate that additional studies are necessary to substantiate the functional activity of the tested interventions.

Although our data established both biochemical potency and functional activity, in this study, genetic analysis of the yeast strain was limited to genomic information, and the incorporation of additional approaches such as gene expression profiling remains a subject for future investigation. Furthermore, the evaluation of autophagy induction was conducted only at the cellular level, and results from clinical trials are awaited to further validate its physiological relevance. Additionally, further studies are required to investigate the bioavailability, metabolic fate, and more clinical outcomes of 3L63-derived SPD. To fully elucidate the therapeutic potential of 3L63-derived SPD, additional studies are required to assess its bioavailability, metabolic fate, and clinical efficacy. Factors such as individual microbiome profiles, gut polyamine metabolism, and potential synergy with other nutrients should be considered in future human trials for optimal formulation development. To address these limitations, future studies should include controlled human trials to assess SPD bioavailability, pharmacokinetics, and clinical outcomes. Metabolomic analyses could elucidate the metabolic fate of SPD and identify individual factors, such as microbiome composition, that affect its efficacy. Additionally, investigating synergistic effects with other nutrients or bioactive compounds could inform optimal formulation strategies for functional foods or supplements. Incorporating these approaches will strengthen the translational relevance of 3L63-derived SPD and guide evidence-based product development.

5. Conclusions

The 3L63 SPD-enriched yeast strain developed in this study offers a novel and scalable approach for addressing the growing demand for natural polyamine supplementation. Owing to its high SPD content, GRAS status, scientific formulation, and bioactivity, it is a promising next-generation ingredient for functional foods, dietary supplements, and healthy aging applications. However, this study is limited to in vitro assays and compositional analysis; further mechanistic studies and human trials are needed to confirm its efficacy and safety in real-world applications.

6. Patents

Masanori Tamakoshi, Satoko Imaruoka, and Kazuto Ikemoto, Polyamine-Rich Yeast and Food and Beverage Composition Containing the Same, Japanese Patent No. 6663598, filed on 2 March 2018.