Bioconversion of Deproteinized Cheese Whey to Metabolites by Understudied Cryptococcus-Related Yeasts: Characterization and Properties of Extracted Polysaccharides

Abstract

Microbial bioconversion of agro-industrial by-products into high-value-added metabolites such as polysaccharides or lipids serves a dual purpose: mitigating environmental pollution through waste reduction and supporting the development of novel bioproducts. In this study, a non-conventional, poorly studied Cryptococcus albidus strain was initially assessed for its ability to grow on semi-defined media containing lactose, glycerol, or glucose under three distinct nitrogen availability conditions at C/N equal to 20, 80, and 160 mol/mol in shake flask cultures. The goal was to evaluate biomass production and synthesis of valuable metabolites under these conditions. C. albidus demonstrated robust growth on all commercial carbon sources, particularly under nitrogen-rich conditions, producing more than 25.0 g/L of microbial biomass with a high intracellular polysaccharide content (>45%, w/w). Additionally, mannitol production was detected in cultures with glycerol and glucose (9.1 and 13.1 g/L, respectively), especially after nitrogen depletion. Subsequently, C. albidus and a Cutaneotrichosporon curvatus strain were batch-cultivated using pretreated secondary cheese whey (SCW) as a carbon-rich waste substrate. When cultivated on SCW, both yeast strains partially metabolized lactose and produced polysaccharide-rich biomass, dominated by β-glucans (>29% of total biomass), compounds known for their functional and bioactive properties. The cellular polysaccharides (cPS extracted from C. albidus exhibited cytotoxic effects against cancer cells, suggesting their potential use as biological response modifiers. In contrast, the cPS from C. curvatus did not affect cell viability, indicating their promise as ingredients for applications in the food, feed, pharmaceutical, or cosmetic sectors.

1. Introduction

Microorganisms have been important bioconverters of carbon sources, such as the cheese whey-derived lactose, the biodiesel-derived crude glycerol, or sugars from waste streams or hydrolysates, among others, to value-added metabolites, such as polysaccharides and lipids [1,2,3,4,5]. The valorization of low-cost industrial by-products through microbial biotechnology provides a dual benefit: reducing the environmental impact of these waste streams while enabling the production of high-value bioproducts with important properties. Deproteinized secondary cheese whey (SCW) has been studied in recent years as a substrate for microbial cultivation, and its utilization represents a promising example of circular and sustainable economy implementation in practice [6,7,8,9,10]. According to the literature, SCW has been employed, among other applications, for the production of microbial storage metabolites, such as lipids [11,12,13,14,15], polysaccharides [12,16,17,18,19], and single-cell protein [20,21].

Referring to yeast species capable of producing lipids and/or polysaccharides, strains belonging to the genus Cryptococcus spp. have been studied for their ability to valorize industrial by-product streams and produce metabolites of interest. Notable examples include Cryptococcus curvatus (also known as Cutaneotrichsporon oleaginosum, Candida curvata, or Apiotrichum curvatum) [12,18,22,23,24], C. albidus (also known as C. terricolus) [25,26,27], C. (Papiliotrema) laurentii [11,12,19] and C. podzolicus [28,29,30], among others, which have been extensively studied in the development of bioprocesses. These yeasts belong to the Basidiomycota phylum and are characterized by spherical to elongated cells, being reproduced asexually. The colonies they form on solid nutrient media are creamy in texture and predominantly white in color, although in some cases, the production of carotenoids imparts distinctive pigmentation to certain strains. Most species are surrounded by a capsule, a thick layer primarily composed of polysaccharides, which is why Cryptococcus spp. is often referred to as “the sugar yeast” [31]. Strains of the genus C. albidus (also known as Naganishia albida and previously classified as C. terricolus) have been studied in the literature regarding their ability to produce and accumulate lipids, utilizing both commercial and waste-derived substrates, achieving single-cell oil production exceeding 20% (w/w) of the dry cell biomass [25,27,32]. Studies have been carried out using various strategies, including batch, fed-batch, continuous, and continuous with cell recycling culture systems [26,27,32,33], under conditions of nitrogen limitation or excess, as well as across a range of temperatures, carbon and/or nitrogen sources [27,32,33,34,35,36]. Of particular significance is the observation that C. albidus was capable of accumulating lipids regardless of the C/N ratio of the medium [32,33].

Microbial lipids produced by yeasts have been extensively studied, with those rich in oleic acid being considered putative candidates for the production of various final products in the food, biofuel (biodiesel), cosmetics, or oleochemical sectors [37,38,39,40,41]. A key factor in triggering secondary metabolism—leading to lipid accumulation in oleaginous microorganisms—is the limitation of an essential nutrient, most commonly nitrogen, combined with the simultaneous availability of a carbon source (certain strains of C. albidus appear to be exceptions to this rule, as reported in the international literature and mentioned in the previous section). Concerning that, cultivation parameters—particularly the carbon-to-nitrogen (C/N) molar ratio—play a pivotal role in determining the metabolic direction of carbon flow during microbial culture [42,43]. Microbial polysaccharides (PSs) are structurally diverse biopolymers with broad industrial and biomedical potential due to their functional properties. These PSs are categorized as either cellular (cPS), stored intracellularly or integrated into the cell wall, or extracellular (EPS), often secreted into the medium or forming protective capsules [44,45]. Among the most studied intracellular PSs are β-glucans—primarily located in the yeast cell wall—which have attracted considerable attention for their potent biological activities. Rich in β-1,3/1,6-glucan linkages, these polysaccharides are known as biological response modifiers, capable of modulating immune responses, inhibiting cancer growth and metastasis, and protecting against oxidative stress [46,47,48]. Their efficacy is influenced by structural features such as branching degree, molecular weight, solubility, and triple-helix conformations. Applications of β-glucans span pharmaceuticals, food and feed additives, cosmetics, and even agriculture, where they reduce antibiotic use in livestock and act as antiviral agents in plants [49]. The intracellular polysaccharide content in certain yeast strains makes them promising sources of bioactive compounds for high-value applications, underscoring the need for further research into their composition, extraction, and functional potential [46].

Based on the abovementioned information, the aim of the present study was the valorization of Mizithra secondary cheese whey, using non-conventional and poorly studied yeast strains for the production of value-added metabolites, including polysaccharides and lipids. Initially, a C. albidus strain was screened on various carbon sources and carbon-to-nitrogen molar ratios to evaluate the effects of these cultivation parameters on yeast growth and metabolite synthesis. Subsequently, this strain and a Cutaneotrichosporon curvatus strain were batch-cultivated in bioreactor, using pretreated SCW as the sole substrate, aiming to promote biomass growth and metabolite production. The extracted lipids and cellular polysaccharides were subsequently analyzed and characterized. Special focus was given to the polysaccharide fractions, which were further investigated to assess potential bioactivities, including antioxidant properties and cytotoxic effects.

2. Materials and Methods

2.1. Microorganisms & Screening Media

Cryptococcus albidus var. kuetzingii NRRL Y-6965 and Cutaneotrichosporon curvatus NRRL YB-775 regeneration precultures were carried out in semi-defined media containing yeast extract at 10 g/L (Nitrogen ≈ 10%, w/w) (Condalab, Madrid, Spain), peptone at 10 g/L (Nitrogen ≈ 10%, w/w) (MC024, LAB M Ltd., Lancashire, UK) and either lactose (Himedia Laboratories, Pvt. LTd, Maharashtra, India), glycerol (Merck KGaA, Darmstadt, Germany), or glucose (Carlo Erba, Cornaredo, Italy) at 10 g/L, depending on the primary carbon source in the subsequent main cultures. 250-mL (Erlenmeyer) flasks containing 50 ± 1 mL of the medium (Working volume—Vw = 20%, v/v) were used for the precultures, after proper sterilization at 121.1 °C for 20 min in autoclave. The medium was aseptically inoculated using a cryovial containing the strain, and incubated in an orbital shaker (Lab-Line, Melrose Park, IL, USA) at 28 ± 1 °C with an agitation rate of 180 ± 5 rpm. The inoculation of the subsequent main cultures was carried out during the exponential growth phase of the cells.

C. albidus Y-6965 growth and metabolite production were evaluated under various nitrogen contents (i.e., C/N = 20, 80, 160 mol/mol), using three different main carbon sources (lactose, glycerol, or glucose) at an initial concentration (S₀) equal to 60 ± 2 g/L. Nitrogen was derived half from yeast extract (organic source) and the other half from ammonium sulfate (Penta Chemicals, Prague, Czech Republic—inorganic source containing ≈21%, w/w of N). The concentrations of both sources were adjusted appropriately to achieve the desired C/N ratios (C/N = 20, 80, 160 mol/mol). Minerals and salts of analytical grade were also added (g/L): KH₂PO₄ 7.00, Na₂HPO₄ 2.50, MgSO₄×7H₂O 1.50, MnSO₄×H₂O 0.06, ZnSO₄×7H₂O 0.02, CaCl₂×2H₂O 0.15, FeCl₃×6H₂O 0.15, according to Papanikolaou et al. [50] and the mixture was diluted in distilled water. The media were then transferred to 250-mL flasks (Vw = 20%, v/v), sterilized, aseptically inoculated (using 1 mL of the preculture- ≈ 2%, v/v), and incubated, at the conditions described previously. The pH was maintained within the range of 5.5–6.0 using 5 M NaOH or HCl solutions.

2.2. Bioreactor Cultures on Secondary Cheese Whey

The strain C. albidus Y-6965, as well as the strain C. curvatus NRRL YB-775 (for more information about its screening experiments see [19]) were batch-cultivated on pretreated secondary cheese whey (from the Laboratory of Dairy Research, Agricultural University of Athens, 11855 Athens, Greece), following the successive cheese and Mizithra whey cheese making processes, as described by Vasilakis et al. [12]. The SCW was microfiltered (0.22 μm pore size—Polycap AS 36 Capsule Filter, Whatman^TM Cytiva, Maidstone, UK), the coagulated proteins were removed, as well as the medium was sterilized by microfiltration, avoiding autoclave sterilization. Lactose concentration was 56.7 ± 5.3 g/L, FAN = 51.6 ± 6.9 mg/L, Total Kjeldahl Nitrogen = 0.97 ± 0.09 g/L, density = 1.030 ± 0.013 g/mL and pH = 6.3 ± 0.2, according to assays described below (see ‘Section 2.3. Analytical Methods’). A sterilized jacketed bioreactor (Labfors Infors HT, Basel-Landschaft, Switzerland) of 2.0 L total volume (Vw = 1.0 L—50%, v/v, of total volume) was used for the bioprocess and the pretreated (microfiltered) SCW was aseptically added without adding extra nutrients nor minerals and salts. 100 mL of the exponentially grown lactose-based preculture were used as inoculum of 900 mL of SCW, under aseptic conditions. Regeneration preculture was containing only pretreated SCW and the process was carried out as described previously (see ‘Section 2.1. Microorganisms & Screening Media’). The bioreactor culture was inoculated when the cells were at the exponential growth phase. The cultivation conditions included incubation temperature at 28 ± 1 °C, stirring at 600 ± 5 rpm, aeration at 1.5 vvm and all were maintained constant throughout the cultivation.

2.3. Analytical Methods

During screening experiments, flasks were periodically moved out of the orbital shaker, while samples were periodically taken from the bioreactors and the cells were harvested through centrifugation at 15,000× g and 4 °C for 10 min (Hettich Universal Centrifuge (Model 320-R, Merck KGaA, Darmstadt, Germany)]. The precipitate was washed twice with distilled water and freeze-dried to a constant weight at −45 °C using a VirTis Freezemobile 12SL (The Virtis Company, Gardiner, NY, USA). Dry biomass concentration (X, g/L) and lipids (L, g/L) were gravimetrically quantified, after proper extraction in the latter case, in presence of a 2:1 (v/v) mixture of chloroform/methanol organic solvents and solvents evaporation. Thin Layer Chromatography was conducted using n-hexane/diethyl ether/glacial acetic acid (70:30:1, v/v/v) as mobile phase in silica-gel 60 aluminum sheets, while fatty acids’ (FA) profile was determined via gas chromatography analysis after fatty acid methyl esters (FAME) derivatization and the results were expressed as g FA/100 g of total FA or %, w/w). Kjeldahl analysis in a KjeltekTM 8100 Distillation Unit (Foss A/S, Hillerød, Denmark) was used to quantify total Kjeldahl nitrogen (TKN, g/L). Total cellular polysaccharides (cPS) were determined through DNS assay after proper chemical acid hydrolysis. For more details concerning the aforementioned procedures, see [12]. β-Glucans content (g/100 g of dry biomass-X) was determined based on the β-glucan enzyme assay kit (Yeast and Mushroom, K-YBGL) of Megazyme^® (County Wicklow, Ireland), following the manufacturer’s instructions [51]. The supernatant (culture broth) was collected and analyzed as follows. Residual lactose (S_R, g/L) concentration was determined through HPLC using a Waters Alliance 2695 HPLC system (Waters Corporation, Milford, MA, USA) equipped with RI (2414 Refractive Index) detector and an Aminex HPX-87H column (Bio-Rad Laboratories, Hercules, CA, USA) working at 60 °C, with 0.005 M H₂SO₄ as the mobile phase (0.5 mL/min). The molecules were detected and quantified based on standard curves. The determination of free amino nitrogen (FAN, mg/L) was conducted based on the ninhydrin photometric method. The consumed concentrations of carbon sources (S_CON, g/L) or of free amino nitrogen (FAN_CON, mg/L) were calculated as the difference between the initial and residual concentrations: S_CON = S₀ − S_R and FAN_CON = FAN₀ − FAN_R, respectively. For more details see Vasilakis et al. [12].

All flask cultures were performed in triplicate and bioreactor cultures in duplicate (biological replicates). Mean values and standard deviations are performed for consumed carbon sources (S_CON, g/L) and free amino nitrogen (FAN_CON, mg/L). The same applies to the produced lipids (L, g/L), mannitol (MANN, g/L), cellular polysaccharides (cPS, g/L), and biomass (X, g/L) and all were calculated based on three technical replicates. In each culture, the yield Y_X/S (g total dry biomass produced per g substrate consumed), Y_L/S (g/g), Y_cPS/S (g/g), and Y_MANN/S (g/g) were determined. The contents (on dry biomass) for lipids (K_L/X, g/g), and polysaccharides (K_cPS/X, g/g), as well as the productivities for biomass (P_X, mg/L/h), lipids (P_L, mg/L/h), mannitol (P_MANN, mg/L/h), and cellular polysaccharides (P_cPS, mg/L/h) were calculated, according to Papanikolaou et al. [50]. Each experimental point in tables and figures present the mean value and standard deviation of the independent determinations. Data were plotted using Kaleidagraph 4.0.3.0 (Synergy Software 1988–2006).

2.4. Isolation of PS-Rich Extracts and Characterization of Isolates

Freeze-dried yeast biomasses of C. albidus Y-6965 and C. curvatus NRRL YB-775, derived from bioreactor cultures on SCW, were subjected to hexane treatment for lipids removal and subsequent high-pressure homogenization (Constant Cell Disruption Systems, Daventry, UK) for the isolation of crude extracts. The biomass was defatted with hexane (10% w/v) under stirring for 24 h, dried, then resuspended in water (10% w/v) and homogenized at 2.5 kbar (×4 cycles). After centrifugation (15,000× g, 4 °C, 10 min), the supernatant was collected and filtered through a 0.45 μm filter (Whatman™ Cytiva, Maidstone, UK). Deproteinization was performed using a 5:1 chloroform/n-butanol mix (Sevag method), followed by 2–4 centrifugation cycles (15,000× g, 4 °C, 10 min) to remove coagulated proteins. The clarified supernatant was mixed with cold ethanol (1:4), was chilled (4 °C, 24 h), and centrifuged (15,000× g, 4 °C, 10 min) [51]. The polysaccharides-rich extracts (namely 6965-cPS and 3594-cPS) were isolated after precipitation and lyophilization.

ATR-FTIR analysis was carried out and the spectra of the samples were recorded using a Perkin Elmer Spectrum-Two spectrometer. The spectrometer was equipped with a Diamond ATR compartment (Perkin Elmer, Hopkinton, MA, USA). Spectrum 10 software was used to analyze the spectra (v.10.5.1.581). Each recording was an average of 32 scans in the mid-infrared region (4000 and 400 cm⁻¹) at a resolution of 4 cm⁻¹. The recorded spectra were ATR-corrected using a refractive index for diamond crystal of 1.5, to be comparable to the available spectral libraries. In addition to crystal correction, the spectra were smoothed by Savitzky–Golay algorithm (smoothing window of 7 points with zero polynomial), linearly baseline corrected and normalized by the mean using The Unscrambler X v.10.5 software (CAMO software, Oslo, Norway).

Polysaccharide monomer composition was determined using a modified version of the NREL protocol [52]. In specific, freeze-dried samples underwent acid hydrolysis with 10 mL of 1.0 M H₂SO₄ at 100 °C for 1 h. After hydrolysis, the solutions were neutralized to pH 5–6 using CaCO₃ and centrifuged (15,000× g, 4 °C, 10 min). The resulting supernatants were filtered through 0.02 µm membranes (Whatman™ Cytiva, Maidstone, UK) and analyzed by HPLC with RI detector (Waters Alliance 2695 system, Waters Corporation, Milford, MA, USA). Sugars were quantified using a Shodex SP0810 column (Showa Denko K.K., Tokyo, Japan) at 60 °C with ultrapure water as the mobile phase (0.6 mL/min). Uronic acids were analyzed using an Aminex HPX-87H column (Bio-Rad Laboratories, Hercules, CA, USA), as described previously Identification of monomers was based on their retention times, compared to standards (glucose, galactose, mannose, xylose, glucuronic acid), and quantification was performed via calibration curves. Results are expressed as the weight percentage of each monomer in the total polysaccharide content and represent mean ± standard deviation from three independent measurements.

The antioxidant activity of the polysaccharide extracts (at a concentration of 10 g/L) was assessed using the Ferric Reducing Antioxidant Power (FRAP) assay, according to Benzie & Strain [53]. This method evaluates the ability of the extracts to reduce ferric-tripyridyltriazine (Fe³⁺–TPTZ) to its ferrous form (Fe²⁺–TPTZ). Antioxidant capacity was expressed as Trolox equivalents (µM). All values presented in the tables represent the mean ± standard deviation of three independent measurements.

2.5. Cytotoxicity Testing on Caco-2 and HT29-MTX Cell Lines

Caco-2 cells (human colorectal adenocarcinoma cells) and HT29-MTX cells (a mucus-producing HT29 subclone) were cultured in high-glucose DMEM (Biosera, Cholet, France) supplemented with 10% FBS (Gibco ThermoFisher Scientific, Waltham, MA, USA), 10 U/mL L-glutamine, 100 U/mL penicillin, 100 μM non-essential amino acids, 100 μg/mL streptomycin, and 1 mM sodium pyruvate (Biosera, Cholet, France). Cells were maintained at 37 °C in a humidified 5% CO₂ atmosphere and subcultured with trypsin (PAN-Biotech, Bavaria, Germany) before reaching confluence. Media were refreshed every 2–3 days, and cells up to passage 25 were used. Cytotoxicity of cellular polysaccharide (cPS) samples was assessed using the MTT assay (Cayman Chemical, Ann Arbor, MI, USA), which measures mitochondrial conversion of MTT to formazan by metabolically active cells [54]. cPS samples were dissolved in distilled water at 4 mg/mL and stored at 4 °C until use. Caco-2 and HT29-MTX cells were seeded in 96-well plates at 5 × 10⁴ and 3 × 10⁴ cells/well, respectively, in 100 μL medium and allowed to adhere overnight. After washing with PBS (Takara Bio, Shiga, Japan), cells were treated with 150, 300, or 600 μg/mL cPS for 24 h, based on prior polysaccharide studies (Vasilakis et al., 2025a).

2.6. Statistical Analysis

SPSS for Windows statistical package software (version 22.0.0) was used for the statistical analysis of all presented results. Kolmogorov–Smirnov test was used to test normality and all means were compared using one-way or two-way ANOVA followed by Tukey’s post hoc test (p < 0.05).

3. Results

3.1. Growth of C. albidus on Different Carbon Sources and Various Nitrogen Conditions

C. albidus NRRL Y-6965 was shake-flask batch-cultivated on commercial lactose-, glycerol-, and glucose-based media (S₀ ≈ 60 g/L), to investigate its ability to assimilate each different carbon source, to grow, and produce metabolites, under various C/N molar ratios (i.e., 20, 80, and 160 mol/mol). The results from endpoints of all cultures (and intermediate time point in case of glucose-based culture at C/N = 20 mol/mol, where mannitol was produced and subsequently assimilated) are recorded in

Table 1. Data from C. albidus Y-6965 after batch cultivation on lactose, glycerol, or glucose (S₀ ≈ 60 g/L and C/N = 20, 80, 160 mol/mol). The table presents the results for each carbon source and each C/N molar ratio at the endpoint of the shake-flask cultures, as well as at an intermediate time point, when mannitol production and subsequent assimilation were observed on glucose-based culture at C/N = 20 mol/mol.

Carbon Source	C/N (mol/mol)	Time (h)	SCON (g/L)	X (g/L)	L (g/L)	cPS (g/L)	MANN (g/L)	YX/S (g/g)	YL/S (g/g)	YcPS/S (g/g)	YMANN/S (g/g)	KL/X (g/g)	KcPS/X (g/g)	PX (mg/L/h)	PL (mg/L/h)	PcPS (mg/L/h)	PMANN (mg/L/h)	FANCON (mg/L)
Lactose	20	200	61.4 a ± 0.2	28.0 a* ± 0.4	0.8 a ± 0.1	12.3 a* ± 0.4	n.d.	0.46 b ± 0.00	0.01 b ± 0.01	0.20 ab* ± 0.01	-	0.029 b ± 0.004	0.455 a ± 0.005	140.0 a ± 2.0	4.0 ab ± 0.5	61.3 a ± 2.2	-	377 a ± 10
	80	200	32.0 b ± 0.3	13.4 b ± 0.2	0.9 a ± 0.0	5.6 b ± 0.1	n.d.	0.42 c ± 0.01	0.03 ab ± 0.00	0.18 b ± 0.00	-	0.063 a ± 0.005	0.410 b ± 0.022	67.0 b ± 1.3	4.3 a ± 0.0	28.1 b ± 0.5	-	103 b ± 5
	160	200	17.2 c ± 0.3	9.2 c ± 0.2	0.7 a ± 0.0	3.8 c ± 0.1	n.d.	0.53 a* ± 0.03	0.04 a ± 0.00	0.22 a* ± 0.01	-	0.076 a ± 0.002	0.410 b ± 0.023	46.0 c ± 0.6	3.5 b ± 0.0	18.9 c ± 0.3	-	48 c ± 3
Glycerol	20	160	59.6 a ± 0.1	28.2 a* ± 0.3	2.1 a* ± 0.2	12.7 a* ± 0.2	0.0 c ± 0.0	0.47 a ± 0.01	0.04 a ± 0.00	0.21 a* ± 0.01	0.00 c ± 0.00	0.075 c ± 0.007	0.451 b ± 0.011	176.0 a ± 2.1	13.1 a* ± 1.3	79.4 a ± 1.2	0.0 c ± 0.0	408 a ± 17
	80	338	61.4 a ± 0.3	13.9 b ± 0.2	1.8 a* ± 0.1	7.3 b ± 0.1	5.3 b ± 0.3	0.23 b ± 0.01	0.03 a ± 0.00	0.12 b ± 0.00	0.09 b ± 0.00	0.129 b ± 0.010	0.523 a* ± 0.017	41.1 b ± 0.6	5.3 b ± 0.3	21.6 b ± 0.3	15.6 b ± 1.0	112 b ± 8
	160	450	42.1 b ± 0.4	9.5 c ± 0.1	1.5 b ± 0.0	5.2 c ± 0.0	9.1 a ± 0.2	0.23 b ± 0.00	0.04 a ± 0.00	0.12 b ± 0.01	0.22 a* ± 0.00	0.158 a* ± 0.002	0.550 a* ± 0.003	21.1 b ± 0.2	3.3 c ± 0.0	11.6 c ± 0.0	20.3 a ± 0.4	53 c ± 4
Glucose	20	68	61.8 ± 0.2	17.6 ± 0.2	0.9 ± 0.0	8.8 ± 1.1	4.6 ± 0.1	0.28 ± 0.01	0.01 ± 0.01	0.14 ± 0.02	0.07 ± 0.01	0.051 ± 0.001	0.500 ± 0.069	258.5 * ± 3.3	13.2 * ± 0.0	129.4 * ± 16.4	67.1 * ± 2.0	320 ± 10
		195	61.8 a ± 0.2	25.5 a ± 0.3	1.3 b ± 0.1	12.5 a* ± 0.3	0.0 c ± 0.0	0.41 a ± 0.01	0.02 a ± 0.01	0.20 a* ± 0.01	0.00 c ± 0.00	0.051 c ± 0.005	0.490 a ± 0.018	130.7 a ± 1.6	6.7 b ± 0.5	64.1 a ± 1.5	0.0 c ± 0.0	380 a ± 21
	80	155	60.5 a ± 0.1	15.0 b ± 0.4	1.7 a ± 0.1	6.6 b ± 0.1	7.0 b ± 0.3	0.25 b ± 0.01	0.03 a ± 0.00	0.11 b ± 0.00	0.12 b ± 0.00	0.113 b ± 0.010	0.440 b ± 0.018	96.8 b ± 2.6	11.0 a ± 0.6	42.6 b ± 0.6	45.2 b ± 1.9	100 b ± 6
	160	345	61.1 a ± 0.3	9.0 c ± 0.1	1.5 ab ± 0.1	3.8 c ± 0.0	13.1 a* ± 0.2	0.15 c ± 0.00	0.03 a ± 0.01	0.06 c ± 0.00	0.21 a* ± 0.01	0.167 a* ± 0.013	0.422 b ± 0.005	26.1 c ± 0.3	4.3 c ± 0.3	11.1 c ± 0.0	37.9 a ± 0.7	46 c ± 5

n.d.: not-detected; -: undeterminable; Different letters (a–c) indicate statistically significant differences according to one-way ANOVA and Tukey’s test (p < 0.05) among the final results obtained for the three different C/N ratios tested, for each carbon source used separately (the final values of the dependent variables (e.g., X, L, etc.) derived from the independent variables—C/N ratios of 20, 80, and 160 mol/mol—were compared exclusively within each substrate type—first for lactose-based cultures only, then for glycerol-based cultures only, and finally for glucose-based cultures only); The statistically highest values per dependent variable (per column) among different conditions tested (carbon-to-nitrogen ratio and carbon source—two-way ANOVA) are presenting with an asterisk (*), serving as an approach to highlight only the most significant outcomes per column.

. The kinetics derived from lactose- (see Supplementary Materials—Figure S1), glycerol- (see Supplementary Materials—Figure S2) and glucose-based (Figure 1) cultures are also presented, including carbon source and free amino nitrogen (FAN) assimilation, as well as dry biomass, lipid, and mannitol (when detected) production.

Under nitrogen-rich conditions (C/N = 20 mol/mol) in lactose-based media, cultivation of C. albidus resulted in the production of 28.0 g/L dry biomass after 200 h, following complete consumption of both the carbon and nitrogen sources (FAN_CON = 377 mg/L) (Figure S1a). The dry biomass contained 45.5% (w/w) polysaccharides (cPS = 12.3 g/L) and 2.9% (w/w) lipids (L = 0.8 g/L). Biomass productivity reached 140 mg/L/h, while the productivities of polysaccharides and lipids were 61 and 4 mg/L/h, respectively, at the end of the cultivation. At higher C/N ratios (80 and 160 mol/mol), the microorganism exhibited slow lactose consumption after nitrogen depletion (FAN_CON = 103 and 48 mg/L, respectively) (Figure S1b,c). Cultivations were terminated at 200 h, as lactose concentration stabilized, and approximately 53% (S_CON = 32.0 g/L) and 29% (S_CON = 17.2 g/L) of the initial lactose had been assimilated for the 80 and 160 mol/mol conditions, respectively. Dry biomass levels reached 13.4 g/L (P_X = 67 mg/L/h) and 9.2 g/L (P_X = 46 mg/L/h), containing 5.6 g/L (K_cPS/X = 41.0% w/w; P_cPS = 28 mg/L/h) and 3.8 g/L (K_cPS/X = 41.0% w/w; P_cPS = 19 mg/L/h) of polysaccharides, for the two nitrogen-limited conditions, respectively. Lipid content increased moderately to 6.3% (w/w) (L = 0.9 g/L) and 7.6% (w/w) (L = 0.7 g/L), with corresponding productivities around 4 mg/L/h in both cases.

Strain’s cultures under nitrogen-rich conditions (C/N = 20 mol/mol) in glycerol-based media resulted in the production of 28.2 g/L of dry biomass after 160 h, following complete consumption of both the carbon and nitrogen sources (FAN_CON = 408 mg/L) (Figure S2a). The dry biomass contained 45.1% (w/w) polysaccharides (cPS = 12.7 g/L) and 7.5% (w/w) lipids (L = 2.1 g/L). Biomass productivity reached 176 mg/L/h, while the productivities of polysaccharides and lipids were 79 and 13 mg/L/h, respectively, at the end of the cultivation. Under C/N ratios of 80 and 160 mol/mol, the microorganism fully consumed glycerol only at the lower ratio (80 mol/mol), albeit at slower rate compared to nitrogen-rich conditions, with complete consumption occurring after 338 h. In the case of C/N = 160 mol/mol, the culture was terminated at 450 h, as glycerol concentration stabilized, and approximately 70% (S_CON = 42.1 g/L) of the initial glycerol had been assimilated. The available FAN was depleted within the first 20 h (FAN_CON = 112 and 53 mg/L, respectively), in both cases (Figure S2b,c). The resulting dry biomass reached 13.9 g/L (P_X = 41 mg/L/h) and 9.5 g/L (P_X = 21 mg/L/h), for the 80 and 160 mol/mol conditions, respectively. Polysaccharide production was 7.3 g/L (K_cPS/X = 52.3% w/w; P_cPS = 22 mg/L/h) and 5.2 g/L (K_cPS/X = 55.0% w/w; P_cPS = 12 mg/L/h), respectively. Nitrogen-limited conditions favored mannitol production and secretion (Figure S2b,c). Specifically, under a C/N ratio of 80 mol/mol, mannitol production reached 5.3 g/L (Y_MANN/S = 0.09 g/g; P_MANN = 16 mg/L/h), while under more severe nitrogen limitation (C/N = 160 mol/mol), it increased to 9.1 g/L (Y_MANN/S = 0.22 g/g; P_MANN = 20 mg/L/h). The reduction in nitrogen availability also led to a moderate increase in lipid accumulation, reaching 12.9% (w/w) (L = 1.8 g/L) and 15.8% (w/w) (L = 1.5 g/L) for the 80 and 160 mol/mol conditions, respectively, with corresponding productivities of approximately 5 and 3 mg/L/h.

The cultivation of the strain under nitrogen-rich conditions (C/N = 20 mol/mol) in glucose-based media initially resulted in the production of 17.6 g/L of dry biomass and 4.6 g/L of mannitol after 68 h, following complete glucose consumption and partial nitrogen assimilation (FAN_CON = 320 mg/L) (Figure 1a). The microorganism subsequently catabolized the accumulated mannitol, leading to a final dry biomass concentration of 25.5 g/L, with total nitrogen consumption reaching 380 mg/L. The biomass contained 49.0% (w/w) polysaccharides (cPS = 12.5 g/L) and 5.1% (w/w) lipids (L = 1.3 g/L). At 68 h, the productivities of dry biomass and mannitol were 259 and 67 mg/L/h, respectively, while by the end of cultivation, the productivity of biomass reached 131 mg/L/h, and those of polysaccharides and lipids were 64 and 7 mg/L/h, respectively. Under C/N ratios of 80 and 160 mol/mol, the microorganism also fully consumed glucose, at slower rates compared to nitrogen-rich conditions, after 155 and 345 h, respectively. In both cases, the available FAN was depleted within the first 18 h (FAN_CON = 103 and 48 mg/L, respectively) (Figure 1b,c). The resulting dry biomass reached 15.0 g/L (P_X = 97 mg/L/h) and 9.0 g/L (P_X = 26 mg/L/h), for the 80 and 160 mol/mol conditions, respectively. Polysaccharide content was 6.6 g/L (K_cPS/X = 44.0% w/w; P_cPS = 43 mg/L/h) and 3.8 g/L (K_cPS/X = 41.0% w/w; P_cPS = 11 mg/L/h), respectively. Nitrogen-limited conditions, also, favored mannitol production and secretion (Figure 1b,c). Mannitol production reached 7.0 g/L (Y_MANN/S = 0.12 g/g; P_MANN = 45 mg/L/h) under a C/N ratio of 80 mol/mol, while it increased to 13.1 g/L (Y_MANN/S = 0.21 g/g; P_MANN = 38 mg/L/h) under C/N = 160 mol/mol). The lipid accumulation presented a moderate increase as nitrogen availability was reduced, reaching 11.3% (w/w) (L = 1.7 g/L) and 16.7% (w/w) (L = 1.5 g/L) for the 80 and 160 mol/mol conditions, respectively, with corresponding productivities of approximately 11 and 4 mg/L/h.

3.2. Batch-Bioreactor Cultivation on SCW

C. albidus Y-6965 and C. curvatus NRRL YB-775 (previously screened on commercial substrates at various carbon sources and C/N ratios, see [19]) were batch-cultivated on pretreated SCW as substrate (Vw = 1.0 L—50%, v/v) in bioreactor system and the end-point results of the cultures are recorded in

Table 2. Data from C. albidus Y-6965 and C. curvatus YB-775 cultures on SCW (S₀ ≈ 50 g/L and FAN₀ ≈ 50 mg/L).

Strain	Time (h)	SCON (g/L)	X (g/L)	L (g/L)	cPS (g/L)	YX/S (g/g)	YL/S (g/g)	YcPS/S (g/g)	KL/X (g/g)	KcPS/X (g/g)	PX (mg/L/h)	PL (mg/L/h)	PcPS (mg/L/h)	FANCON (mg/L)
Y-6965	32 b	38.6 a ± 0.1	16.7 a ± 0.2	1.0 b ± 0.1	7.8 a ± 0.1	0.44 a ± 0.00	0.03 b ± 0.01	0.20 a ± 0.01	0.063 b ± 0.004	0.464 a ± 0.014	523.4 a ± 4.7	32.7 b ± 1.7	242.8 a ± 4.1	30 a ± 3
YB-775	51 a	37.8 a ± 0.8	15.9 a ± 0.3	2.5 a ± 0.2	6.4 b ± 0.3	0.42 a ± 0.02	0.07 a ± 0.00	0.17 a ± 0.02	0.157 a ± 0.016	0.402 b ± 0.027	311.8 b ± 5.8	48.8 a ± 4.1	125.3 b ± 6.1	32 a ± 5

Different letters (a,b) indicate statistically significant differences according to Tukey’s test (p < 0.05) among the two cases.

. The kinetics of biomass and lipids production, as well as of the lactose and FAN assimilation, for both strains, are presented in Figure 2.

The microorganism C. albidus Y-6965 assimilated approximately 77% of the available lactose (S_CON = 38.6 g/L) after 32 h of cultivation, producing dry biomass of 16.5 g/L (Y_X/S = 0.44 g/g) with a productivity value of 523 mg/L/h. Thereafter, the residual lactose, FAN, and the produced metabolites stabilized (Figure 2a). The biomass contained polysaccharides at a proportion of 46.4% (w/w), corresponding to 7.8 g/L (Y_cPS/S = 0.20 g/g), with a polysaccharide productivity of 243 mg/L/h, and lipids at a proportion of 6.3% (w/w) (L = 1.0 g/L − Y_L/S = 0.03 g/g − P_L = 33 mg/L/h). In the case of C. curvatus YB-775 (Figure 2b), approximately 75% of the available lactose (S_CON = 37.8 g/L) was consumed after 51 h of cultivation, producing 15.9 g/L of dry biomass (Y_X/S = 0.42 g/g) with a productivity value of 312 mg/L/h. The polysaccharides production was 6.4 g/L (K_cPS/X = 40.2%, w/w − Y_cPS/S = 0.17 g/g − P_cPS = 125 mg/L/h) and for lipids 2.5 g/L (K_L/X = 15.7%, w/w − Y_L/S = 0.07 g/g − P_L = 49 mg/L/h). Lactose monomers (glucose and/or galactose) were not HPLC-detected in the culture medium at any stage of the cultivation process, regardless of the microorganism used.

3.3. Yeast-Lipid Profiles

The lipids of C. albidus Y-6965, when cultivated on commercial lactose-, glycerol-, or glucose-based media, under various C/N ratios and on SCW, as well as those of C. curvatus YB-775 when cultivated on SCW were recovered, derivatized to FAMEs, GC-analyzed, and the results are presented in

Table 3. Fatty acid profile of the lipids produced by C. albidus Y-6965 after shake-flask batch-cultivated on semi-defined substrates containing lactose, glycerol, or glucose, under three nitrogen conditions (C/N =20, 80, 160 mol/mol), as well as of those produced by C. albidus Y-6965 and C. curvatus YB-775 when batch-cultivated on SCW.

Strain	Carbon Source	C/N (mol/mol)	Time (h)	g/100 g of Total FA
				C16:0	C18:0	Δ9 C18:1	Δ9,12 C18:2	SFA	UFA
Y-6965	Lactose	20	200	21.6 b ± 0.5	9.7 a ± 0.3	47.9 b ± 2.0	20.8 a ± 0.9	31.3 b ± 0.8	68.7 a ± 2.8
		80	142	28.7 a ± 1.0	9.1 a ± 0.4	50.2 ab ± 1.6	12.0 b ± 0.0	37.8 a ± 1.5	62.2 b ± 1.6
		160	164	30.8 a ± 1.2	9.0 a ± 0.2	52.7 a ± 1.0	7.5 c ± 0.2	39.8 a ± 1.3	60.2 b ± 1.3
	Glycerol	20	160	29.3 ab ± 1.1	15.3 a ± 0.6	38.2 b ± 1.2	17.2 a ± 0.4	44.6 b ± 1.8	55.4 b ± 1.6
		80	237	32.9 a ± 1.7	16.1 a ± 0.5	42.2 b ± 1.3	8.8 b ± 0.2	49.0 a ± 2.2	51.0 c ± 1.5
		160	450	28.3 b ± 0.6	11.5 b ± 0.2	50.9 a ± 1.9	9.3 b ± 0.3	39.8 c ± 0.8	60.2 a ± 2.2
	Glucose	20	68	19.8 b ± 1.6	11.4 a ± 0.6	47.0 c ± 1.1	21.8 a ± 0.4	31.2 ab ± 2.2	68.8 a ± 1.6
		80	195	24.1 a ± 1.2	10.8 a ± 0.4	52.6 b ± 1.4	12.5 b ± 0.5	34.9 a ± 1.5	65.1 a ± 1.9
		160	155	19.9 b ± 0.7	11.6 a ± 0.2	58.5 a ± 1.0	10.0 c ± 0.2	31.5 b ± 0.9	68.5 a ± 1.3
	SCW	29	32	22.0 ii ± 0.8	10.1 i ± 0.4	48.4 i ± 1.6	19.5 i ± 0.4	32.1 ii ± 1.3	69.7 i ± 2.0
YB-775	SCW	29	51	31.3 i ± 1.4	10.3 i ± 0.8	48.4 i ± 2.0	10.3 ii ± 1.1	41.3 i ± 2.0	58.7 i ± 3.2

Different letters (a–c) indicate statistically significant differences according to Tukey’s test (p < 0.05) among the lipids derived from the three different C/N ratios tested, for each carbon source in commercial substrates; Different letters (i,ii) indicate statistically significant differences according to Tukey’s test (p < 0.05) among the lipids derived from the two batch-bioreactor cultures on SCW; The symbol Δ (Delta) refers to the position of a double bond in a fatty acid.

.

Based on the GC results for C. albidus recovered lipids prior to screening cultivations, in all cases, the predominant fatty acid was oleic acid (ranging between 38.2 and 58.5%, w/w), with the lowest content observed in lipids recovered when the microorganism was cultivated on a glycerol-based medium at a C/N = 20 mol/mol. The highest oleic acid content was found in cultures grown on glucose-based substrate at a C/N ratio of 160 mol/mol. Palmitic acid was the next most abundant fatty acid in most cases, with concentrations ranging from 19.8% to 32.9% (w/w). In all cultures, a statistically significant increase in oleic acid content was observed as the C/N ratio increased. Conversely, a significant decrease in palmitic acid content was noted as nitrogen availability decreased, with concentrations approximately halved at C/N = 160 mol/mol compared to those at C/N = 20 mol/mol. Stearic acid concentrations ranged between 9.0% and 11.6% (w/w) in most cases, except in glycerol-based cultures at C/N ratios of 20 and 80 mol/mol, where slightly higher values were recorded (15.3% and 16.1% w/w, respectively). The lipids of the yeasts grown on SCW, were also derivatized and analyzed regarding their fatty acid composition. The profile of those derived from C. albidus Y-6965 presented high content of oleic acid (48.4%, w/w), followed by palmitic acid (22.0%, w/w), linoleic acid (19.5%, w/w), and stearic acid (10.1%, w/w). In the case of C. curvatus YB-775, the lipids consisted primarily of oleic acid (48.4%, w/w) and palmitic (31.3%, w/w), while linoleic and stearic acid were detected in lower content values (10.3%, w/w, each one).

The lipids of these latter two cases were analyzed through Thin Layer Chromatography and the results are depicted in Figure 3. The two produced lipids mainly in the form of triacylglycerols (TGAs), although the quantities of storage lipids were not high (K_L/X values were ≤0.20 g/g). Ergosterol was identified in both samples, along with several other neutral lipid compounds—such as monoacylglycerols and diacylglycerols—that could not be specifically identified based on the available standards. Polar lipids, such as phospholipids, were observed as intense bands at the injection point of all samples, due to their inability to move through the stationary phase.

3.4. β-Glucan Content in Yeast Biomass

The β-glucans contents (g/100 g of dry biomass-X) were indirectly determined after physicochemical and enzymatic treatments of the dry biomasses, using a β-glucan enzyme assay kit. Τhe β-glucans were indirectly calculated after the quantification of total and α-glucans and the results are presented in

Table 4. Presentation of total cPS, total-, α- and β-glucans in the dry biomass (X), derived from the strains C. albidus Y-6965, and C. curvatus YB-775, when bioreactor-cultivated on SCW. The ratios of β-glucans to total cPS were, also, calculated.

Strain	Total cPS (g/100 g X)	Total Glucans (g/100 g X)	α-Glucans (g/100 g X)	β-Glucans (g/100 g X)	β-Glucans/ Total cPS (g/100 g cPS)
Y-6965	46.4 a ± 1.4	32.5 a ± 0.7	0.3 a ± 0.1	32.2 a ± 0.6	70.0 a ± 2.9
YB-775	40.2 b ± 2.7	29.2 b ± 0.5	0.1 a ± 0.0	29.1 b ± 0.4	72.6 a ± 4.8

Different letters (a,b) indicate statistically significant differences according to Tukey’s test (p < 0.05) among the two cases.

.

The dry biomass of C. albidus Y-6965 contained 32.2% (w/w) of β-glucans, and that of C. curvatus YB-775 29.1% (w/w). The ratios of β-glucans to total cPS were calculated and the values performed were 72.6% (w/w) for C. curvatus YB-775 and 70.0% (w/w) in case of C. albidus Y-6965. In both cases, the contents of α-glucans remained at very low levels (<0.5%, w/w).

3.5. Characterization of Polysaccharide-Rich Yeast Extracts (FTIR Characterization, Monomers Determination, Antioxidant Capacity Evaluation)

The cPS-rich isolates from C. albidus Y-6965 and C. curvatus YB-775, namely 775-cPS and 6965-cPS, were isolated after lipid extraction using hexane as solvent, high-pressure homogenization on the dry lipid-free biomass, centrifugation, filtration, deproteinization, and ethanol precipitation processes prior to FTIR analysis.

The recorded spectra (Figure 4) were presented very similar to each other with common peaks at 3300 (O-H stretching in phenolic compounds and carbohydrates; N-H stretching), 2927 (CH₂ asymmetric stretching), 1650 (C=O stretching), 1545 (N-H bending), 1452 (CH₂ deformation), 1245 (C-H deformation), 1150/1080/1030 (C-O stretching in polysaccharides), 927 (asymmetric pyranose ring vibration) and 850 (C-H deformation in α-pyranose compounds) cm⁻¹ based on Socrates [55].

To determine the monomer profile of the cPS isolates, chemical pretreatment followed by HPLC analysis were carried out, and the results are presented in

Table 5. Monomers profiles of cellular (cPS) or extracellular (EPS) polysaccharides derived from the two studied strains C. albidus Y-6965, and C. curvatus YB-775, when bioreactor-cultivated on SCW.

Monomers	Strains
	6965-cPS	775-cPS
	(%, w/w)
Glucose	74.1 a ± 1.1	71.8 a ± 1.0
Galactose	3.8 d ± 0.4	7.6 b ± 0.6
Mannose	6.5 c ± 0.5	7.9 b ± 0.4
Xylose	10.0 b ± 0.7	8.5 b ± 0.2
Glucuronic acid	3.9 d ± 0.2	1.7 c ± 0.0

Different letters (a–d) indicate statistically significant differences according to Tukey’s test (p < 0.05) among the contents of the five monomers.

. In the case of C. albidus Y-6965 cPS, glucose was the predominant monomer, accounting for 74.1% (w/w), followed by xylose (10.0%, w/w) and mannose (6.5%, w/w). Galactose and glucuronic acid were present at levels of 3.8% and 3.9% (w/w), respectively. C. curvatus YB-775 cPS were also rich in glucans (71.8%, w/w), followed by xylans (8.5%, w/w), mannans (7.9% w/w), and galactans (7.6% w/w). Glucuronic acid was detected at 1.7% (w/w).

The putative antioxidant capacity of the cPS-rich isolates was determined by using the FRAP assay and the results are presenting in

Table 6. Presentation of the antioxidant capacity of cPS-rich isolates derived from the strains C. albidus Y-6965, and C. curvatus YB-775, when bioreactor-cultivated on SCW.

Strain	Antioxidant Capacity (μΜ Trolox eq.)
Y-6965	44.1 b ± 3.9
YB-775	68.5 a ± 3.0

Different letters (a,b) indicate statistically significant differences according to Tukey’s test (p < 0.05) among the two cases.

. The antioxidant capacities of isolates from C. curvatus YB-775 and C. albidus Y-6965 were 44.1 and 68.5 μM Trolox eq., respectively.

3.6. Effects on Human Cells Viability

The two cPS-rich isolates—namely 6965-cPS, 775-cPS—were tested regarding their effect on the viability of two cell lines, Caco-2 and HT29-ΜΤΧ. For each sample, three different concentrations were tested—150, 300, and 600 μg/mL—to assess potential dose-dependency. The results of Caco-2 cell line are presented in Figure 5a and those of HT29-MTX cell line in Figure 5b.

According to the results of the statistical analysis, no statistically significant effect was observed on the viability of the Caco-2 cell line by any of the cPS-rich solutions at any of the three tested concentrations. In the case of the HT29-MTX cell line, statistically significant negative effects on the viability of the cells were observed when using the cPS-rich solutions derived from C. albidus Y-6965 (indicated with an asterisk [*] in Figure 5b). More specifically, 6965-cPS significantly reduced cell viability at all three concentrations tested. On the other hand, 775-cPS solutions did not significantly affect the viability of the HT29-MTX cell line at any of the tested concentrations.

4. Discussion

Agro-industrial residues are by-products with a high environmental impact; however, within the framework of microbial biotechnology, they can serve as ideal feedstocks for the production of high-value-added microbial metabolites, thus lowering the high cost of raw materials used in bioprocesses [56]. Cheese whey is one such pollutant, commonly found in regions with intense dairy production. When referring to dairy products made from cow’s milk, whey constitutes a year-round waste stream. In this study, an effort was made to valorize secondary (deproteinized) cheese whey derived from the production of feta cheese and the subsequent manufacture of the whey cheese Mizithra. The production of Mizithra leads to significant deproteinization of whey; however, pretreatment of this stream using microfiltration membranes was necessary to ensure complete recovery of protein aggregates. The clarified and sterilized nutrient medium (SCW), without the addition of any further nutrients or salts, was used for the cultivation of two non-conventional and poorly studied yeast strains, namely C. albidus NRRL Y-6965 and C. curvatus NRRL YB-775. The main objective of this bioprocess was to produce high-value-added microbial metabolites while simultaneously contributing to whey remediation, primarily through the utilization of residual lactose, nitrogenous compounds, and salts.

Preliminary experiments were conducted in semi-synthetic commercial media with varying carbon sources (lactose, glycerol, or glucose) and nitrogen levels to evaluate the metabolic responses of these yeasts. For C. curvatus YB-775, these studies had been previously reported [19]. C. albidus Y-6965 was tested on lactose-, glycerol-, or glucose-based media under three carbon-to-nitrogen (C/N) molar ratios (20, 80, and 160 mol/mol). The yeast was capable to catabolize all three carbon sources, although assimilation rates depended on both the type of carbon source and the C/N ratio. To our knowledge, this study represents the first report of C. albidus cultivation on lactose- or glycerol-based media. In lactose-based cultures, complete sugar consumption occurred only under nitrogen-rich conditions (C/N = 20 mol/mol). Under conditions of lower nitrogen availability, lactose utilization stopped once the free amino nitrogen (FAN) was depleted. No glucose or galactose monomers were detected in the medium, indicating intracellular rather than extracellular lactose hydrolysis, consistent with previous findings [12,19,57]. The highest biomass and polysaccharide yields, as well as the corresponding productivities and polysaccharide content of the biomass, were achieved at the lowest C/N ratio tested (20 mol/mol). Although reduced nitrogen availability did not significantly enhance lipid production, a modest increase in lipid content was observed relative to the C/N = 20 condition. Analysis of the lipid profile revealed that increasing the C/N ratio from 20 to 160 mol/mol resulted in significant increases in oleic and palmitic acids, accompanied by a decrease in the polyunsaturated linoleic acid. This trend, observed in both yeasts and fungi, aligns with the role of polyunsaturated fatty acids in cellular membranes and proliferation rather than in storage lipid accumulation [58,59,60]. Notably, palmitic acid content increased markedly under higher C/N ratios, highlighting its metabolic significance in these conditions.

In glycerol-based cultures, a low consumption rate was observed only under strict nitrogen starvation conditions (C/N = 160 mol/mol), leading to termination of the culture before complete assimilation of glycerol. Similarly to the previous case, higher biomass and polysaccharides production values were noted under nitrogen excess conditions (C/N = 20 mol/mol). However, polysaccharides’ and lipids’ contents in dry biomass increased significantly as the C/N ratio increased. Lipid production was notably higher in these cultures compared to those in lactose, with peak production values found at lower C/N ratios. The conversion coefficient of glycerol to biomass or polysaccharides was higher in cultures with C/N = 20 mol/mol but decreased significantly with increasing C/N molar ratio. This difference is reasonable, as mannitol production was observed in nitrogen-starved cultures, reaching maximum production efficiency at C/N = 160 mol/mol. The production of mannitol is particularly noteworthy, as it is a metabolite that is rarely secreted, and its secretion has only been reported from strains of related species (Cryptococcus curvatus, C. neoformans) according to the international literature [61,62,63]. To the best of our knowledge, this is the first time that the production of this polyol has been reported from a C. albidus strain. The microorganism’s choice to secrete mannitol instead of accumulating lipids suggests a potential role in osmoregulation or thermoprotection. Therefore, it is possible that either the temperature or the carbon source was not at optimal levels.

In cultures using glucose as the main carbon source, the microorganism managed to fully assimilate the carbon source in all three different nitrogen conditions. Glucose consumption rate under nitrogen excess (C/N = 20 mol/mol) was the highest among all experiments carried out. Glucose was fully consumed within the first 68 h to produce biomass and mannitol. Subsequently, the microorganism catabolized the produced mannitol and generated additional biomass, cellular polysaccharides, and lipids. The low assimilation duration of total glucose in this case consequently led to higher production values for the bioproducts. Notably, the microorganism did not produce mannitol during the metabolism of lactose, in contrast to cultures grown on glucose, despite lactose being composed of glucose monomers. In both cases, the lipid profile shifted as available nitrogen decreased (from C/N = 20 to = 160 mol/mol), with an increase in oleic acid content and a reduction in linoleic acid. No such significant differences were observed in palmitic acid content across the three C/N molar ratios tested, unlike what was seen in the lactose-based cultures. Palmitic acid contents were generally lower in lipids derived from the glucose-based experiments, while the highest oleic acid content was observed in this substrate, at C/N = 160 mol/mol ratio.

Glucose-based cultivations of C. albidus strains have been previously studied in the literature, and it has been shown that they are capable of producing and accumulating lipids. In an early study by Pedersen [64] using C. terricolus (strain No. 1) (synonym to C. albidus), a gradual increase in lipid accumulation was observed (from 55.0% to 69.3%, w/w) as the C/N ratio increased from 1.9 to 400.0 g/g. This early research demonstrated that lipid accumulation in this species does not necessarily require nitrogen-limiting conditions. Boulton & Ratledge [25] repeated this experiment under bioreactor conditions and did not observe the high lipid accumulation levels previously reported. The strain C. terricola IFO 1322 was cultivated by Boulton & Ratledge [25] in batch mode on a glucose-based medium under nitrogen-limited conditions, resulting in lipid accumulation of approximately 38.0% (w/w) of dry biomass (X = 16 g/L), with nitrogen remaining available and being assimilated throughout the cultivation. The same authors noted that a subsequent semi-continuous cultivation, in which 90% of the culture medium was replaced with fresh medium, led to lipid accumulation of around 60.0% (w/w).

The cultivation of the strain C. albidus var. albidus CBS 4517 for single-cell oil production has been extensively investigated in batch systems under both nitrogen-limited and nitrogen-rich conditions [32]. By investigating different temperatures under nitrogen-limited conditions, the highest lipid content (44.4% w/w, L = 3.9 g/L) was observed in flask cultures at 20 °C. When the process was scaled up to a bioreactor, lipid content under nitrogen limitation was approximately 32% (w/w) (L = 4.5 g/L), while under nitrogen excess it decreased to 26.0% (w/w) (L = 3.3 g/L), confirming that lipid accumulation in this strain occurs under both conditions but is more pronounced under (partial) nitrogen deficiency. Furthermore, the rapid lipid production under nitrogen limitation resulted in higher productivity (P_L = 129 mg/L/h) compared to conditions with greater nitrogen availability (P_L = 77 mg/L/h). In a semi-continuous cultivation system with continuous glucose feeding under nitrogen-limited conditions, lipid content reached 46.0% (w/w), total lipid production was 12.5 g/L, and productivity was 139 mg/L/h [32]. The same research group also examined lipid production by the same strain in continuous culture systems, under both nitrogen and carbon limitation. Maximum lipid accumulation was recorded at low dilution rates (D = 0.031 h⁻¹) for both cases—41% and 37% (w/w) under nitrogen and carbon limitation, respectively [33]. Fu et al. [26] cultivated C. albidus ATCC 10672 in a continuous bioreactor system, yielding a lipid productivity of 160 mg/L/h (L = 7.9 g/L, 55.9% w/w). When a cell recycling (CR) system was implemented, lipid productivity increased substantially to 630 mg/L/h (L = 8.8 g/L, 35.5% w/w).

In a previous study, Fei et al. [35] batch-cultivated C. albidus ATCC 10672 using volatile fatty acids (VFAs) as the carbon source, while also testing different temperature conditions and nitrogen sources. The optimal cultivation temperature was found to be 25 °C and pH = 6.0. The highest lipid content (27.8%, w/w − L =0.33 g/L) was achieved when ammonium chloride was used as the nitrogen source and the VFA mixture had an acetic/propionic/butyric acid ratio of 8:1:1 (S₀ = 2 g/L). In this case, the lipids were particularly rich in linoleic acid (61.1%, w/w), followed by palmitic acid and oleic acid (16–18%, w/w). Vajpeyi & Chandran [36] also cultivated C. albidus ATCC 10672 using a commercial VFA-based substrate under different nitrogen availability conditions. The highest lipid production and accumulation (L = 0.32 g/L, 28.3% w/w in batch culture and L = 0.30 g/L, 29.9% w/w in continuous culture) were achieved at an approximate C/N ratio of 11 mol/mol (COD/N = 25 mg/mg, COD₀ = 6500 mg/L). Lipid accumulation was slightly lower (14.9% w/w, L = 0.15 g/L) when the VFAs were derived from the fermentation of food waste. Lipids from batch cultures at C/N = 11 mol/mol were rich in oleic acid (50.4% w/w), followed by palmitic acid (28.7% w/w) and linoleic acid (15.3% w/w). In contrast, lipids from continuous cultures or those using VFAs from food waste exhibited lower oleic acid content, ranging between 36.0% and 38.0% w/w. In another study, Sathiyamoorthi et al. [27] used hydrolysates derived from fruit and vegetable waste (nutrient media resulting from the hydrolysis of apple and onion residues) for lipid production via C. albidus KCTC 17541 in batch, semi-continuous, and continuous cultures with cell recycling. The highest lipid accumulation (35.7% w/w, L = 2.0 g/L) was observed in the semi-continuous system, whereas the greatest overall lipid yield (2.7 g/L, 20.0% w/w) was achieved in the cell-recycling continuous system after 168 h. The lipids were interestingly rich in linoleic acid (67–74%, w/w), followed by palmitic acid (18–27%, w/w). All these findings indicate that strains of this yeast species can accumulate lipids regardless of the medium or the type of cultivation system (batch, semi-continuous, or continuous).

In their study, Hansson & Dostálek [34] investigated the effect of various cultivation conditions (such as temperature, and nitrogen or carbon content) on fatty acid profiles and observed that the lipids were rich in oleic acid, followed by linoleic acid and palmitic acid. The composition of individual fatty acids was not significantly affected by the cultivation conditions, and the content of unsaturated fatty acids decreased as fermentation progressed. When comparing all the above results with those obtained in the present study, it becomes apparent that the strain C. albidus var. kuetzingii Y-6965 did not reach such high lipid contents (K_L/X < 17.0%, w/w), indicating intraspecific variation relative to other strains of the species. Furthermore, regarding fatty acid composition of the lipids, oleic acid was the predominantly fatty acid, followed by palmitic acid and then linoleic acid. However, both strains produced significantly higher amounts of dry biomass and, consequently, cellular polysaccharides—particularly in nitrogen-rich cultures—which could exhibit putative properties of interest.

For this reason, cellular polysaccharides were extracted from the biomasses of C. albidus Y-6965 and C. curvatus YB-775, after bioreactor-cultivation on pretreated secondary cheese whey. During their cultivation on pretreated SCW, both strains failed to completely assimilate the main carbon source—similar to what was observed during the screening experiments under nitrogen starvation. This observation contrasts with the results expected based on the nitrogen-rich conditions of the cultivation medium, as indicated by the calculated carbon-to-total Kjeldahl nitrogen (C/N) ratio of approximately 29 mol/mol. It is likely that not all of the TKN was bioavailable, which subsequently limited lactose assimilation and, consequently, microbial biomass and metabolite production. Nevertheless, in the case of C. albidus Y-6965 the highest productivity values were achieved for both dry biomass and polysaccharides, along with the highest content of β-glucans in dry biomass. In the case of C. curvatus YB-775, lactose was also not fully assimilated, and the culture was terminated before a significant decrease in polysaccharide productivity occurred. However, this strain presented the highest lipid production.

According to the TLC analysis results, the lipid fractions of both strains were dominated by triacylglycerols (TAGs), although the microorganisms did not accumulate lipids significantly (K_L/X < 16%, w/w), confirming previous findings of our research group [12]. At the level of dry cellular biomass, both strains contained a significant proportion of β-glucans in their dry matter (>29%, w/w), while C. albidus Y-6965 presented the highest content. However, no statistically significant differences were observed in the β-glucan to total cellular polysaccharide ratio. β-Glucans are molecules with notable biological activity, and therefore, the yeast biomass of C. curvatus YB-775, being the only one rich in β-glucans that did not exhibit any cytotoxic effect, could potentially be used by cosmetics, pharmaceuticals, food, or feed industries. Apart from glucans (>71%, w/w of total cPS), xylose, galactose, mannose, and glucuronic acid, in concentrations of less than 10% (w/w) were also detected. Both samples tested, presented antioxidant capacity (>44 μM Trolox eq) and the one from C. curvatus YB-775 exhibited the highest value.

Various structural features—such as solubility, covalently attached phenolic compounds, protein content, or the presence of charged groups—play a crucial role in determining the properties (such as antioxidant capacity) exhibited by polysaccharide-rich extracts [65]. Based on the ATR-FTIR spectra, the peaks at 1410, 1245 and 1080 cm⁻¹ could be related to the presence of carboxymethylated polysaccharides, while the peaks at 1650 and 1545 cm⁻¹ could be assigned to residual polypeptides [66]. Furthermore, the combined peaks at 2927 and around 1740 cm⁻¹ might indicate the presence of residual fatty acids in both samples [66]. The peak at 1373 cm⁻¹, observed exclusively in the 6965-cPS sample, also combined with the peak at 1650 cm⁻¹, can be associated with N-linked acetyl groups of polysaccharides, such as those present in chitin and chitosan. These are structural components of cell membranes, with the 1373 cm⁻¹ peak corresponding to the symmetric bending vibration of CH₃ and the 1650 cm⁻¹ peak corresponding to the C=O stretching vibration [66]. This assumption is enhanced by the presence of absorptions at 1650 and 1545 cm⁻¹ [66]. Finally, the combined peaks at 1245, 1030 and 850 cm⁻¹ could also indicate the attachment of sulfonyl groups to polysaccharides, corresponding to symmetric, asymmetric stretching of S=O and stretching of C-O-S, respectively [67].

Exopolysaccharides (EPS) produced by encapsulated Cryptococcus species, such as C. neoformans and C. gattii, are known to contribute to pathogenicity due to the presence of a dense polysaccharide capsule. This capsule is mainly composed of acidic glucuronoxylomannan, galactoxylomannan, and/or mannoproteins [31,68]. In the strains analyzed in this study, no detectable secretion of polysaccharides or increase in culture medium viscosity—typically associated with a thick capsule—was observed during growth, in contrast to previous reports [12,19]. The cellular polysaccharide (cPS)-rich isolates were evaluated for cytotoxic effects on two human cell lines, Caco-2 and HT29-MTX, at concentrations of 150, 300, and 600 μg/mL. None of the samples showed cytotoxicity toward Caco-2 cells at any concentration. In contrast, cPS from C. albidus Y-6965 exhibited cytotoxic effects on HT29-MTX cells, even at the lowest tested concentration. The pronounced cytotoxicity against the HT29-MTX cancer cell line suggests that these polysaccharides could be promising candidates for anticancer applications. Their high content of β-glucans, known biological response modifiers, may underlie their capacity to inhibit cancer cell proliferation and metastasis [46,69]. Conversely, polysaccharides from C. curvatus YB-775 appear suitable for food or feed applications, given their high β-glucan content, lack of cytotoxicity, and notable antioxidant properties.

In the international literature, studies investigating the cytotoxicity of polysaccharides derived from yeasts are relatively limited. In their study, Kogan et al. [46] reported an enhancement of the activity of a cytostatic drug against mouse leukemia cells in the presence of sulfoethyl glucan from the cell wall of S. cerevisiae and glucomannan from Candida utilis. Glucan particles from S. cerevisiae yeast did not exhibit significant cytotoxicity toward peripheral blood mononuclear cells or RAW 264.7 macrophages [70]. Microparticles from the S. cerevisiae cell wall, primarily composed of β-1,3-D-glucan, either in their native form or loaded into polymer–lipid hybrid nanoparticles, were evaluated for their effects on the viability of RAW 264.7 cells. In all cases, cell viability remained above 80%, which was considered by the authors to be not statistically significant [71]. More recently, our research group investigated the cellular polysaccharides of two Papiliotrema laurentii strains, which exhibited cytotoxic activity against the HT29-MTX cell line, suggesting a potential biological response-modifying effect against cancer cells, similar to that of the cPS from C. albidus examined in the present study [72].

5. Conclusions

To sum up, the non-conventional yeast strain C. albidus Y-6965 demonstrated the ability to grow on semi-defined commercial substrates rich in lactose, glycerol or glucose and produce high concentrations of microbial biomass, primarily enriched in polysaccharidic compounds. The rate of glucose consumption was significantly higher compared to the other two carbon sources, particularly under nitrogen-rich conditions. Cultivation on glucose and glycerol also promoted the secretion of mannitol into the fermentation medium, which was subsequently utilized as a secondary carbon source once the primary source was depleted. In contrast, cultivation on lactose did not favor mannitol production, and lactose assimilation was markedly reduced upon nitrogen depletion. The incapability of mannitol production during lactose catabolism is of high scientific interest and warrants further investigation. Although the produced lipids were present at lower concentrations than those observed in related oleaginous species, they mainly consisted of triacylglycerols rich in oleic acid. Lipid production increased under nitrogen-limited conditions but did not reach levels that were sufficient to classify these strains as “oleaginous”, as observed in other previously discussed species. Both C. albidus Y-6965 and C. curvatus YB-775 were grown on SCW, partially assimilated lactose and produced cPS-rich biomass, mainly consisting of β-glucans, molecules of high functional properties, which also exhibited antioxidant capacity. The cPS isolates of C. albidus Y-6965 demonstrated cytotoxic activity, suggesting their putative role as biological response modifiers against cancer cells, while the cPS from C. curvatus YB-775 could be studied further for potential applications in the food, feed, biomedicine or cosmetic industries.