Revealing the Potential of Lipid and β-Glucans Coproduction in Basidiomycetes Yeast

Beta (β)–glucans are polysaccharides composed of D-glucose monomers. Nowadays, β-glucans are gaining attention due to their attractive immunomodulatory biological activities, which can be utilized in pharmaceutical or food supplementation industries. Some carotenogenic Basidiomycetes yeasts, previously explored for lipid and carotenoid coproduction, could potentially coproduce a significant amount of β–glucans. In the present study, we screened eleven Basidiomycetes for the coproduction of lipids and β–glucans. We examined the effect of four different C/N ratios and eight different osmolarity conditions on the coproduction of lipids and β–glucans. A high-throughput screening approach employing microcultivation in microtiter plates, Fourier Transform Infrared (FTIR) spectroscopy and reference analysis was utilized in the study. Yeast strains C. infirmominiatum CCY 17-18-4 and R. kratochvilovae CCY 20-2-26 were identified as the best coproducers of lipids and β-glucans. In addition, C. infirmominiatum CCY 17-18-4, R. kratochvilovae CCY 20-2-26 and P. rhodozyma CCY 77-1-1 were identified as the best alternative producers of β-glucans. Increased C/N ratio led to increased biomass, lipid and β-glucans production for several yeast strains. Increased osmolarity had a negative effect on biomass and lipid production while the β-glucan production was positively affected.


Introduction
Recently, coproduction strategies, when two or more valuable products are simultaneously produced in a single fermentation process, have been proposed as a way to reach sustainable production of microbial biomass for a wide range of applications [1]. When developing a coproduction process, it is important to identify suitable microbial cell factories able to perform concomitant production of several metabolites and optimize cultivation conditions for the most optimal and balanced biosynthesis of all the desired products.
Carotenogenic Basidiomycetes could be one of the promising microorganisms used for developing coproduction processes. Carotenogenic Basidiomycetes can be enriched in lipids (up to 70% w/w) [2], carotenoids [3], ergosterol [4] and glucans [5,6]. Currently, there is one main coproduction strategy for Basidiomycetes presented in the literature, and it is related to the coproduction of lipids and carotenoids [3,7]. This strategy has several limitations related to the separation of lipophilic components in the case of their separate use [8]. The strategy of coproduction of lipids and β-glucans Table 1. List of the yeast strains used in this study.

Number
Yeast Strain Phylum Strain Collection Number

Media and Growth Conditions
The cultivation of yeast was performed first on YPD agar medium (Merck, Darmstadt, Germany) to recover frozen cultures and then in YPD broth medium (Merck, Darmstadt, Germany) without nutrient limitation to prepare inoculum. For the screening study, the production broth media (Merck, Darmstadt, Germany) with different C/N ratios (10:1, 40:1, 70:1 and 100:1) were inoculated.
For the YPD agar cultivation, yeast cells from the frozen cryopreserved stock cultures were transferred onto Petri dishes with YPD agar (yeast extract, 10.0 g/L; peptone, 20.0 g/L; glucose 20.0 g/L; agar, 20.0 g/L) (Merck, Darmstadt, Germany) and cultivated for 72 h at 25 • C. Inoculum was prepared by transferring 1µL of yeasts cells from YPD agar into 50 mL of sterile YPD broth medium (yeast extract, 10.0 g/L; peptone, 20.0 g/L; glucose 20.0 g/L) (Merck, Darmstadt, Germany) in Erlenmeyer flask (250 mL) and cultivated for 24 h at 25 • C under shaking regime (100 rpm, 50 mm). To remove the residual medium after the cultivation, the inoculum cells were washed with the sterile 0.1% (w/v) NaCl and resuspended to the original volume. The inoculum was added in the volume ratio of 1:5 to the YPD production broth media with different C/N ratios, which was composed of (g/L): yeast extract, 2; KH 2 PO 4 , 4; MgSO 4 · 7H 2 O, 0.7 and glucose monohydrate. The following C/N ratios were examined, 10:1, 40:1, 70:1 and 100:1. The carbon content of 40% in glucose and nitrogen content of 10.5% in yeast extract was used for the calculation of the C/N ratios. The cultivation in production media was performed in a Duetz Microtiter Plate System [15,22,23] (Enzyscreen, Heemstede, Netherlands) which consists of 24-well extra deep microtiter plates (MTPs) with low-evaporation sandwich covers and clamp system for mounting MTPs onto the incubator shaking platform. Cultivation in the control YPD broth production media was done for 96 h at 25 • C according to Szotkowski et al. (2019) under the shaking regime (100 rpm, 50 mm) [6].

Experimental Design
Three biological replicates were prepared for each strain and cultivation medium by inoculating strains from frozen cryopreserved stocks onto agar plates. Each biological replicate was prepared in a separate microtiter plate. All microtiter plates (MTPs) were prepared by inoculating three wells for each strain and cultivation medium. At the end of cultivation, cell suspensions of the three wells were merged and the final biomass was used for total biochemical profiling by FTIR spectroscopy, lipid analysis by gas chromatography and glucan analysis by Yeast and Mushroom β-glucan Assay Procedure (486 samples in total). Morphology of some yeast strains was additionally examined by optical microscope Leica DM6 B (Leica Microsystems, Wetzlar, Germany).

Preparation of Yeast Biomass for the Glucan and Lipid Analysis
After cultivation, the yeast biomass was centrifuged at 4500 rpm for 5 min at 4 • C and the biomass pellet was then washed three times using 0.1% NaCl solution. Then the yeast biomass was freeze-dried for 48 h and subsequently stored at −20 • C until use.

Analysis of Glucans by Yeast and Mushroom β-Glucan Assay
The total glucan content, and the content of βand α-glucans were determined according to the Yeast and Mushroom β-glucan Assay Procedure K-YBGL (Megazyme Int., Warszawa, Poland) [24][25][26]. To estimate the total glucan content, freeze-dried yeast biomass was hydrolysed with ice-cold 12 M sulphuric acid for 2 h and then incubated for 2 h at 100 • C. Further, acidic hydrolysate was neutralized with 200 mM sodium acetate buffer (pH 5) and 10 M KOH, followed by the effect of enzymes exo-β-(1→3)-D-glucanase and β-glucosidase in acetate buffer (pH 4.5) for 1 h at 40 • C. The α-glucan content was determined after enzymatic hydrolysis with amyloglucosidase and invertase. The β-glucan content was calculated from the assay kit procedure as the difference between total glucan and α-glucan content. The absorbance values indicating the total glucans and α-glucan content were obtained spectrophotometrically at 510 nm after adding glucose oxidase/peroxidase reagent.

Total Lipid Content and Analysis of Fatty Acid Profile
Lipid extraction was performed by a modified Folch method [22,23]. A sample of 15 ± 3 mg of freeze-dried yeast biomass was added into 2 mL polypropylene (PP) tubes together with 250 ± 20 mg acid-washed glass beads (710-1180 µm diameter, Millipore Sigma, St. Louis, Missouri, USA) and 600 µL methanol. For the disruption of yeast cells, the Precellys evolution homogenizer (Bertin Instruments, Germany) was used three times with shaking cycles of 5500 rpm (2 × 20 s). The content of the PP tube was transferred into a glass reaction tube by washing it with a 2.4 mL solvent mixture of methanol:chloroform:hydrochloric acid (7.6:1:1 v/v). A 1 mg dose of tridecanoid acid (C13:0) in the form of triacylglycerol (TAG) was used as an internal standard and added to the reaction mixture. The glass tube was vortexed for 10 s and incubated for 1 h at 90 • C. After cooling to room temperature, 1 mL of distilled water and 2 mL hexane:chloroform (4:1 v/v) mixture were added. The separated upper hexane phase with extracted lipids was evaporated under nitrogen at 30 • C followed by the addition of sodium sulfate and dissolving the fatty acid methyl esters (FAMEs) in 1.5 mL hexane containing 0.01% butylated hydroxytoluene (BHT, Millipore Sigma, St. Louis, Missouri, USA). The hexane containing the extracted lipids was transferred into glass vials for GC analysis. Total lipid content (wt.% of total FAMEs of the dry weight) and the fatty acid profile were estimated by 7820A gas chromatograph, Agilent Technologies equipped with an Agilent J&W GC column (20.0 m × 180 µm × 0.20 µm) and flame ionization detector (FID). The total time of the analysis was 36 min with an initial temperature of 70 • C, which was kept for 2 min, and then increased by 10 • C/min to 150 • C, and finally by 6 • C/min to 230 • C. A dose of 1 µL of the sample was injected in split mode (30:1 split ratio) to an inlet tempered to 280 • C. FAME standard mixture (C4-C24; Millipore Sigma, St. Louis, MO, USA) dissolved in hexane was used for the identification of the FAMEs. Quantification was based on the C13:0 internal standard.

Total Lipid Content and Analysis of Fatty Acid Profile
The biomass of the yeasts grown on different C/N ratios was subjected to the biochemical profiling by FTIR spectroscopy. Washed yeast suspension (4 µL) was transferred on 384-well ZnSi microplate (Bruker Optik, Ettlingen, Germany) in triplicate. Samples were dried at room temperature for 30 min before analysis. FTIR spectra were recorded in transmission mode using the High Throughput screening eXTension unit (HTS-XT) coupled to the Vertex 70 FTIR spectrophotometer (Bruker Optik, Ettlingen, Germany). The spectra were collected in the spectral range from 4000 to 500 cm −1 (spectral resolution of 6 cm −1 , and aperture 5.0 mm), with 64 scans as average for each sample.

Data Analysis
Prior to principle component analysis (PCA), the FTIR spectra were preprocessed. The preprocessing was performed by transforming spectra to the second-derivative using the Savitzky−Golay algorithm with a polynomial of degree 2 and a window size of 11. The second-derivative spectra were preprocessed by extended multiplicative signal correction (EMSC) [27][28][29]. Technical replicates (543 spectra in total) were averaged in order to remove technical variability of the measurements, resulting into 181 spectra. PCA was performed for three spectral regions, lipid (3050-2800 cm −1 combined with 1800-1700 cm −1 ), protein (1700-1500 cm −1 ) and polysaccharide (1200-700 cm −1 ).
The following software packages were used for the data analysis: Unscrambler X version 10.5.1 (CAMO Analytics, Norway), Orange data mining toolbox version 3.24 (University of Ljubljana, Slovenia) [30,31].

Growth, Total Glucan and β-Glucan Content in Basidiomycetes Yeast
The revealing of potential coproducers of lipids and β-glucans and identifying of new promising yeast sources of β-glucans were based on estimating the following parameters: biomass yield, total glucans, βand α-glucans, lipid yield and lipid profile (Tables 2-7, Figures 1-4). Cultivation of yeasts in the media with low (10:1), moderate (40:1) and high (70:1 and 100:1) C/N ratios showed a continuous increase in a biomass yield with the increase of C/N ratio (Table 2). Thus, for most of the studied yeast strains the highest biomass yield was in a range from 2.67 ± 0.26 to 15.33 ± 1.16 g/L and observed in a medium with a C/N ratio of 100. Yeast strains R. kratochvilovae CCY 20-2-26, R. mucilaginosa CCY 19-4-6 and CCY 20-9-7, R. toruloides CCY 62-2-4 and S. pararoseus CCY 19-9-6 showed the highest biomass production in a medium with a C/N ratio of 70:1 (Table 2). Generally, Ascomycetes yeasts showed lower biomass production in comparison to carotenogenic Basidiomycetes yeast ( Table 2). The highest biomass yield 15.3 g/L was obtained for the strain C. macerans CCY 10-1-2 at a C/N ratio of 100. The biomass yield was nearly three times higher than for the reference β-glucan producing strain Saccharomyces cerevisiae CCY 21-4-102. For all Rhodotorula strains and S. salmonicolor CCY 19-4-25, a significant increase in biomass production occurred in the C/N ratio range from 10:1 to 40:1, and from 70:1 to 100:1, while little change was observed in the C/N ratio range from 40:1 to 70:1.

Coproduction of Lipids and β-Glucans in Basidiomycetes Yeast
Basidiomycetes carotenogenic red yeasts showed oleaginous properties and were able to accumulate lipids from 30 to over 47% (w/w) while Ascomycetes noncarotenogenic yeast did not accumulate lipids more than 10% (w/w) (Figure 1). In the present study the noncarotenogenic Metschnikowia did not accumulate more than 10% (w/w). Generally, Ascomycetes yeasts were not affected by the variation in the amount of glucose in the media, and the biomass and lipid yield were unchanged.
toruloides CCY 62-2-4 and S. metaroseus CCY 19-6-20 showed low β-glucan content (10-14% w/w) and can be considered as mainly lipid producers. While some yeast strains as for example C. infirmominiatum CCY 17-18-4 at 100:1 C/N and R. kratochvilovae CCY 20-2-26 at 70:1 C/N were able to accumulate a relatively high content of both lipids (38.21 and 37.92% of w/w) and β-glucans (20.73 and 21.43% of w/w) accompanied by a high biomass yield (15.19 and 12.05 g/L) and therefore could be utilised for developing a strategy of lipid and β-glucan coproduction (in our experiment up to 3.15 g/L of β-glucans for C. infirmominiatum .  Table 1.

Coproduction of Lipids and β-Glucans in Basidiomycetes Yeast
Lipids accumulated in yeasts are represented mainly by triacyl glycerides (TAGs). A detailed fatty acid profile was identified by gas chromatography, where fatty acids present in amounts higher that 2% were taken into consideration. In all the yeast strains the major saturated fatty acids were palmitic (C16:0) and stearic acid (C18:0); monosaturated, palmitoleic (C16:1) and oleic acid (C18:1n9c). Polyunsaturated fatty acids were represented by linoleic (C18:2n6c), γ-linolenic (C18:3n3) and α-linolenic (C18:3n3) fatty acids (Figure 3). The most abundant fatty acid present in all the yeast strains was oleic acid, with a production of over 40% in all strains except for S. cerevisiae CCY 21-4-102, where the major fatty acid was palmitoleic fatty acid. Depending on the C/N ratio and yeast strain, the content of monounsaturated (MUFA), polyunsaturated (PUFA) and saturated (SAT) fatty acids changed significantly (Figure 3). The content of MUFA, specifically oleic fatty acid, increased with the increase in C/N ratio in all yeast strains (Figure 3). The highest amount of MUFA (49% w/w) was detected for the strain S. metaroseus CCY 19-6-20, which showed also one of the highest total lipid and biomass yields (Table 2, Figure 1). The biggest difference in fatty acid profile can be observed for C/N ratios between 10:1 and 100:1, where the highest SAT content in all yeasts was detected at a C/N ratio of 10:1 (Figure 3) while higher C/N ratios favoured oleic acid biosynthesis and suppressed linoleic fatty acid production ( Figure 3). Interestingly, the fatty acid profile of strains from genus Metschnikowia, belonging to the Ascomycetes phylum, was more similar to the fatty acid profile of carotenogenic Basidiomycetes yeast. It differed only by a slightly higher production of C16:1 and lower production of C16:0 ( Figure 3). Among the studied Basidiomycetes yeasts, Rhodotorula yeasts showed a high production of palmitic fatty acid ( Figure 3). The strain S. cerevisiae CCY 19-6-4 showed a very consistent fatty acid profile which was not affected by the different C/N ratios, which had the lowest production of PUFAs and the highest amount of palmitoleic fatty acid (C16:1) at all studied C/N ratios ( Figure 3). The highest content of PUFA, with linolenic fatty acid dominating, was detected in the strain S. salmonicolor CCY 19-6-4 with a yield of up to 40% w/w at a C/N ratio of 40:1 (Figure 3). was detected for the strain S. metaroseus CCY 19-6-20, which showed also one of the highest total lipid and biomass yields (Table 2, Figure 1). The biggest difference in fatty acid profile can be observed for C/N ratios between 10:1 and 100:1, where the highest SAT content in all yeasts was detected at a C/N ratio of 10:1 (Figure 3) while higher C/N ratios favoured oleic acid biosynthesis and suppressed linoleic fatty acid production ( Figure 3). Interestingly, the fatty acid profile of strains from genus Metschnikowia, belonging to the Ascomycetes phylum, was more similar to the fatty acid profile of carotenogenic Basidiomycetes yeast. It differed only by a slightly higher production of C16:1 and lower production of C16:0 ( Figure 3). Among the studied Basidiomycetes yeasts, Rhodotorula yeasts showed a high production of palmitic fatty acid ( Figure 3). The strain S. cerevisiae CCY 19-6-4 showed a very consistent fatty acid profile which was not affected by the different C/N ratios, which had the lowest production of PUFAs and the highest amount of palmitoleic fatty acid (C16:1) at all studied C/N ratios ( Figure 3). The highest content of PUFA, with linolenic fatty acid dominating, was detected in the strain S. salmonicolor CCY 19-6-4 with a yield of up to 40% w/w at a C/N ratio of 40:1 (Figure 3).

Total Cellular Biochemical Profiling by FTIR Spectroscopy
The obtained FTIR spectra at different spectral regions provided information on all the main biochemical building blocks of the yeast cells. Lipids were described mainly by the two spectral regions 3010-2800 cm −1 and 1800-1700 cm −1 and some single peaks related to -CH2 and -CH3 scissoring in a region 1400-1300 cm −1 (Table 4, Figure 4). When analyzing cellular lipid profiles based on FTIR

Total Cellular Biochemical Profiling by FTIR Spectroscopy
The obtained FTIR spectra at different spectral regions provided information on all the main biochemical building blocks of the yeast cells. Lipids were described mainly by the two spectral regions 3010-2800 cm −1 and 1800-1700 cm −1 and some single peaks related to -CH 2 and -CH 3 scissoring in a region 1400-1300 cm −1 (Table 4, Figure 4). When analyzing cellular lipid profiles based on FTIR spectra, the most important lipid associated peaks usually taken into consideration are the following: (1) the peaks 2947 cm −1 , 2925 cm −1 , 2855 cm −1 , 1465 cm −1 and 1377 cm −1 are related to -CH 3 and -CH 2 stretching and indicating mainly the chain length of the carbon skeleton in lipid molecules; (2) the peak 1745 cm −1 is related to the ester bond stretching and indicates the total lipid content in the cell; (3) the peak 1710 cm −1 is related to the carboxyl bond vibrations in free fatty acids, and (4) the peak 3010 cm −1 is related to = C-H stretching in lipids and indicates the unsaturation level of cellular lipids. Proteins were observed in the spectral region 1700-1500 cm −1 with the main peaks for amide I (1650 cm −1 ) and amide II (1540 cm −1 ) bonds and polysaccharides were observed in the region 1200-900 cm −1 which was mainly related to the sugar ring vibrations (Table 4, Figure 4). cm −1 ) and amide II (1540 cm −1 ) bonds and polysaccharides were observed in the region 1200-900 cm −1 which was mainly related to the sugar ring vibrations (Table 4, Figure 4).  Peak numbers correspond to the numbers given in Table 4.
For identifying the biochemical profile of yeasts grown on different C/N ratios based on the obtained FTIR spectra, a principal component analysis (PCA) was applied. The PCA scatter plot shows that the C/N ratio was strongly affecting the lipid and protein profile in all yeasts, where the FTIR spectra of yeasts grown on the medium with a C/N ratio of 10:1 were grouped distinctly from the spectra of yeasts grown on other C/N ratios ( Figure 5). This indicated that both total lipid content  Table 4.
For identifying the biochemical profile of yeasts grown on different C/N ratios based on the obtained FTIR spectra, a principal component analysis (PCA) was applied. The PCA scatter plot shows that the C/N ratio was strongly affecting the lipid and protein profile in all yeasts, where the FTIR spectra of yeasts grown on the medium with a C/N ratio of 10:1 were grouped distinctly from the spectra of yeasts grown on other C/N ratios ( Figure 5). This indicated that both total lipid content and fatty acid profile of accumulated TAGs in yeasts cells obtained from a C/N ratio of 10:1 medium was significantly different from yeasts grown on other ratios ( Figure 5A,C). Further, some strain-and species-specific differences in lipid and protein profiles were observed from FTIR spectra. Thus, lipid and protein FTIR profiles for strains of genus Metchnikowia obtained from all types of media, and strains of species S. salmonicolor and S. cererevisiae grown on C/N ratios of 40:1, 70:1 and 100:1 were clearly different from the others (Figure 5A,C). These results correlate well with the obtained reference fatty acid profile data, where strains of Metchnikowiaa and S. salmonicolor and S. cerereviseae differed from the others in the content of palmitic, palmitoleic, and polyunsaturated fatty acids (Figure 3). The polysaccharide profile was not affected by the C/N ratio in Ascomycetes yeasts, while significant differences were observed for Basidiomycetes yeasts ( Figure 5B). The polysaccharide profile of strains from genus Metchnikowia and species S. salmonicolor differed significantly from other strains ( Figure 5B). ( Figure 3). The polysaccharide profile was not affected by the C/N ratio in Ascomycetes yeasts, while significant differences were observed for Basidiomycetes yeasts ( Figure 5B). The polysaccharide profile of strains from genus Metchnikowia and species S. salmonicolor differed significantly from other strains ( Figure 5B).

Impact of Extracellular Osmolarity on Lipid and β-Glucans Coproduction
In this study, we used sodium chloride (NaCl) at six different concentrations (0.2; 0.5; 2; 5; 8 and 11%) to investigate the influence of different extracellular osmolarity levels on the production of βglucans and coproduction of lipids and β-glucans. The three yeast strains C. infirmominiatum CCY 17-18-4, P. rhodozyma CCY 77-1-1 and R. kratochvilovae  showing the highest β-glucans and a high lipid production in this study, were selected for the study. The S. cerevisiae CCY 19-6-4 strain was used as a control strain for β-glucan production. The growth media with C/N ratios of 40:1 and 70:1 supplemented by NaCl at all studied concentrations led to a decrease in the biomass yield for all yeast strains except for S. cerevisiae (Table 5, 0% NaCl). The biomass production for S. cerevisiae under low osmolarity conditions (0.2 and 0.5% of NaCl) in C/N ratio growth media with C/N ratios of 40:1, 70:1 and 100:1 was higher or very similar to the standard conditions (Table 5, 0% NaCl), while high osmolarity resulted in a significant decrease of biomass (Table 5). Low extracellular osmolarity (0.2%

Impact of Extracellular Osmolarity on Lipid and β-Glucans Coproduction
In this study, we used sodium chloride (NaCl) at six different concentrations (0.2; 0.5; 2; 5; 8 and 11%) to investigate the influence of different extracellular osmolarity levels on the production of β-glucans and coproduction of lipids and β-glucans. The three yeast strains C. infirmominiatum CCY 17-18-4, P. rhodozyma CCY 77-1-1 and R. kratochvilovae CCY 20-2-26, showing the highest β-glucans and a high lipid production in this study, were selected for the study. The S. cerevisiae CCY 19-6-4 strain was used as a control strain for β-glucan production. The growth media with C/N ratios of 40:1 and 70:1 supplemented by NaCl at all studied concentrations led to a decrease in the biomass yield for all yeast strains except for S. cerevisiae (Table 5, 0% NaCl). The biomass production for S. cerevisiae under low osmolarity conditions (0.2 and 0.5% of NaCl) in C/N ratio growth media with C/N ratios of 40:1, 70:1 and 100:1 was higher or very similar to the standard conditions (Table 5, 0% NaCl), while high osmolarity resulted in a significant decrease of biomass (Table 5). Low extracellular osmolarity (0.2% of NaCl) combined with a high C/N ratio resulted in an increase or no change in the biomass yield in comparison to the standard conditions (Table 5, 0% NaCl). The biomass yield was gradually decreasing with the increased amount of NaCl in the media. A high concentration of NaCl combined with a high C/N ratio resulted in the lowest biomass production which was possibly due to the elevated levels of osmolarity caused by both high NaCl and glucose content in the medium ( Table 5). The strain P. rhodozyma CCY 77-1-1 showed the highest sensitivity to the applied osmotic stress, and its growth was highly inhibited in the presence of NaCl 2% and higher (Table 5). The effect of extracellular osmolarity on the total glucan, βand α-glucan content in yeast cells was strain-specific and differed depending on the level of osmolarity and the C/N ratio used (Table 6). Thus, low (0.2 and 0.5% of NaCl) and in some cases moderate (2 and 5%) levels of osmolarity combined with C/N ratios of 40:1, 70:1 and 100:1 led to an increase in the total glucan and β-glucan content in comparison to the reference conditions of 0% NaCl (Table 6). For example, the addition of 0.2% NaCl caused an increase in β-glucan production up to 32.15 ± 0.81 (w/w) in C. infirmominiatum CCY 17-18-4 (Table 6), that is about 21% more than in the standard conditions (Table 6, 0% NaCl) and higher than for the control strain S.cerevisiae CCY 21-4-102 (Table 6). In the case of strain P. rhodozyma CCY 77-1-1, β-glucan content was increased at all osmolarity levels ( Table 6) when combined with C/N ratios of 40:1, 70:1 and 100:1, with the high yield obtained of 2.80 g/L of β-glucans at C/N 100:1 and 0.2% NaCl (w/w), while it showed an inhibiting effect on production of β-glucans in other carotenogenic yeast strains ( Table 6). The β-glucan content was 1.98 g/L for S. cerevisiae CCY 21-4-102 at C/N 100:1 and 0.2% NaCl (w/w), which is higher than at standard conditions without salt.

Discussion
The simultaneous accumulation of multiple products is attractive for economically sustainable microbial biotechnology. Over the decades, the Saccharomyces cerevisiae is the only yeast strain used in industry for β-glucan production. Although the presence of β-1,3-glucan synthase has been identified in all fungal phyla (with the exception of Microsporidia), the carotenogenic Basidiomycetes yeast represents unexplored potential in biotechnological β-glucan production along with the coproduction of lipids [36,37].
In the present study, the screening of eleven carotenogenic Basidiomycetes yeasts for simultaneous β-glucan and lipid production and evaluating them as potential alternative sources of β-glucans was performed in a high-throughput set-up by utilizing high-throughput micro-cultivation in Duetz-MTPS [15,22,23]. Due to the fact that glucan production is usually performed in a low C/N ratio media [5], while lipid accumulation in yeasts is triggered by a high C/N ratio and depletion of a nitrogen source [38], media with both low and high C/N ratios were used for the screening. Further, since β-glucans are important structural components of the cell wall, applying osmotic stress could potentially have a significant effect on their production. In addition, it is well known that any variations in extracellular osmolarity influence the cell volume, and therefore, the concentration of intracellular macromolecules [39,40]. Thus, differences in osmolarity could potentially affect lipid accumulation in yeast cells. Recently, several studies reported the influence of extracellular osmolarity on β-glucans' biosynthesis in Saccharomyces cerevisiae where a decrease in β-glucan content at higher osmolarity was detected [41,42]. Therefore, in our study we used different amounts of NaCl to investigate the effect of osmolarity on the coproduction of lipids and β-glucans in Basidiomycetes yeast. To the authors' knowledge this is one of the first studies reporting evaluation of coproduction lipids and β-glucans in various carotenogenic yeasts.
As a cell wall polysaccharide component, β-glucan content can account for about 10% up to 30% of the DCW and the optimization of culture conditions to achieve high cell mass plays the crucial role in biotechnological β-glucan production [43]. Several researchers reported the screening of yeast strains, metabolic engineering, and optimization of different cultivation parameters: harvesting time, media composition, osmotic stress, for achieving high cell mass, and consequently high β-glucan yield [41][42][43][44][45][46][47][48][49][50][51]. In the literature it was reported quite different yields for β-glucan and biomass production. According to the authors' knowledge, one of the highest reported β-glucan yields for small-scale screening of the S. cerevisiae strain was 2.076 g/L [42] which is higher than our results for S. cerevisiae CCY 21-4-102 (1.98 g/L of β-glucan, C/N 100, 0.2% NaCl w/w), while it is much lower when compared to the results for some Basidiomycetous yeast (C. infirmominiatum CCY 17-18-4, R. kratochvilovae CCY 20-2-26 and P. rhodozyma CCY 77-1-1 at high C/N ratio) reported in this study, where the β-glucan yield was in a range from 2.43 to 3.15 g/L. Some authors report β-glucan yield from 0.35 to 15% w/w for S. cerevisiae [52,53] and 25% w/w for R. mucilaginosa [5], that is lower compared to our results for S. cerevisiae CCY 20-4-102, and C. infirmominiatum CCY 17-18-4. Results on β-glucan yield reported by Valasquez et al. (2015) are higher than for R. mucilaginosa strains evaluated in our study which could indicate strain-specific β-glucan biosynthesis. Another alternative yeast producer of β-glucans, reported in the literature, is the yeast Candida utilis [54,55], which was able to produce 5.3 g/L of β-glucans when cultured on waste potato juice water supplemented with glycerol in a 5L bioreactor and 3.5 g/L when cultured in the shake flasks [55], that is comparable to our results for C. infirmominiatum CCY 17-18-4, which accumulated 3.15 g/L of β-glucans along with a high amount of lipids (38.21% DCW).
It has been previously shown that the C/N ratio is the key feature significantly affecting the accumulation of carbon-based intracellular metabolites (carbohydrates, lipids, pigments, etc.) [3,[56][57][58]. Although there are many reports published on how C/N ratio influenced accumulation of lipids or carotenoids in Basidiomycetes yeast [3,7], the effect on β-glucan biosynthesis has not yet been investigated. In our study we observed that C/N ratio increased biomass production, with the highest yield observed for C.infirmominiatum CCY 17-18-4, which also showed the highest β-glucan yield at the C/N ratio 100:1 (3.15 g/L of β-glucan). Our study is reporting for the first time that Cystofilobasidium yeast can be considered as a potential alternative β-glucan source.
The decrease in β-glucan production for the Cystofilobasidium strains, as well as strains R. toruloides CCY 62-2-4, R, kratochvilovae CCY 20-2-26, and P. rhodozyma CCY 77-1-1 occurred under high C/N ratios and it can be explained by the increased accumulation of lipids which increase from 23 to 38% DCW. While under the condition of low C/N ratio, yeast proliferation is triggered and subsequently formation of the cell wall and carbon is transformed into protein and cell wall components, while lipogenesis is supressed. For some yeast strains (R. mucilaginosa CCY 19-4-6 and CCY 20-9-7, S. pararoseus CCY 19-9-6 and S. salmonicolor , the high amount of glucose (100:1 C/N) led to a decrease in biomass production and, thus, the β-glucan yield. This could be due to the osmotic pressure stress under high glucose concentration that negatively influenced the growth of these strains. Therefore, C/N ratio should be considered as an important optimization parameter when developing coproduction of lipids and β-glucans.
In the last part of the study, eight different osmolarity conditions were applied, that resulted in a decrease of both biomass (12.35 ± 1.21 g/L) and lipid yield (35.49 ± 3.35% of DCW), while it was observed an increase of β-glucan yield, but only at low NaCl addition (22.65 ± 1.19% of DCW; 2.80 g/L). The biomass and β-glucans increased also at S. cerevisiae CCY 20-4-102 (6.32 ± 0.40 g/L of biomass, and 31.32 ± 0.93% w/w of β-glucans at 0.2% NaCl; C/N 100:1). On the other hand, NaCl content higher than 5% (w/w) negatively affected the accumulation of β-glucans. The obtained results are in accordance with those previously published by Varelas et al. (2017), where the 6% NaCl (w/w) resulted in a decrease of the β-glucans.
It is well-known that due to the high activity of AMP-dependent isocitrate dehydrogenase, some carotenogenic Basidiomycetes yeasts are able to produce lipids up to 70% lipids/dry weight [38,59,60] after optimized fed-batch fermentation under controlled conditions. In our screening the highest lipid content of 47.27 ± 2.36% (w/w) was observed for strain C. macerans CCY 10-1-2. Strains C. informiniatum CCY 17-18-4 and R. kratochvilovae CCY 20-2-26 also showed a high lipid content of 38.21 and 37.72% w/w along with the high β-glucan yield, which also was the highest among the studied Basidiomyces yeast. Thus, it could be concluded that these two strains have great potential for the future coproduction of lipids and β-glucan in a scaled-up strategy.
The highest amount of MUFA (49% from all fatty acids) was detected for the strain S. metaroseus CCY 19-6-20, which showed also one of the highest total lipid (45.13 ± 2.56% DCW) and biomass yields (15.11 ± 0.93 g/L) together with the β-glucan production up to 15% of DCW (2.27 g/L). Thus, S. metaroseus CCY 19-6-20 could be considered as a promising candidate for biofuel production, since MUFA is a desired oil component for biodiesel production providing low temperature fluidity and oxidative stability [61] with the concomitant β-glucan coproduction. Noncarotenogenic Metschnikowia phylum did not accumulate more than 10% of lipids, while it has been reported that it can produce up to 40% w/w [14], that could be due to the higher temperature and shorter cultivation time used in our study in comparison to the previously reported 15 • C and 14 days [14].
By achieving high biomass production, we can obtain a high β-glucan yield. As it was previously reported, the high cell density (biomass yield over 50 g/L) can be achieved with an optimized fed-batch strategy in fermenters with fully controlled conditions [43]. As for Saccharomyces cerevisiae, the highest biomass yield of 187.63 g/L was reported by Vu and Kim (2009), while the same biomass yield (185 g/L) was also obtained for carotenogenic yeast strain enriched with high lipid production (40% of DCW) reported by Pan et al. 1986. According to the literature, the biomass, lipid, and β-glucan yield can be optimized by the same strategy. Therefore, choosing Basidiomycetes yeast C. infirmominiatum CCY 17-18-4, R. kratochvilovae CCY 20-2-26, P. rhodozyma CCY 77-1-1, or S. metaroseus CCY 19-6-20 as promising candidates for the concomitant production of lipids and β-glucans, would achieve a sustainable coproduction process. Taking into consideration the increasing demand for β-glucan and lipids, carotenogenic Basidiomycetes yeast can be considered as an innovative source of these nutrients.
Further, several yeast strains showed high coproduction potential, for example strain C. infirmominiatum CCY 17-18-4 was able to accumulate 20.73% (w/w) of β-glucans and 38.21% (w/w) of lipids, the strain R. kratochvilovae CCY 20-2-26 was able to accumulate 38.21% (w/w) of lipids and 20.73% of β-glucans, accompanied by a high biomass yield (15.19 g/L). It was observed that C/N ratio and extracellular osmolarity could influence biomass yield and production of lipids and β-glucans. An increase in the C/N ratio led to an increase in biomass, lipid and β-glucan production for several yeast strains, while osmolarity had a negative effect on the biomass and lipid production but positively affected β-glucan production.