Brewers’ spent grain
BSG was kindly provided by Asia Pacific Breweries (Singapore) Pte, Ltd. and was stored in airtight plastic containers at −80 °C till used.
Strains
Yeast strain R. toruloides (CBS 5490, Central Bureau voor Schimmelcultures, Utrecht, The Netherlands) were maintained on agar plates containing (per liter) 20 g of agar (Sigma, St. Louis MO, USA), 10 g of yeast extract (Biobasic Canada Inc.), 20 g of peptone (Biobasic Canada Inc.) and 20 g of d-dextrose (Sigma, St. Louis MO, USA).
Medium and culture conditions
All growth media were prepared with deionized water. The biofermentation and the extraction of nutrient for the preparation of BSG growth media were performed according to a method proposed (Kirana et al. 2012) with minor modification. The unfermented BSG media was prepared using the same method for the nutrient extraction. The extracts were centrifuged (14,500 rpm, 20 min, 4 °C), filtered through a 0.45 μm filter and autoclaved (121 °C, 20 min). These solutions were used as the fermented BSG media and the unfermented BSG media to grow R. toruloides in. YPD fermentation media was prepared with 20 g of peptone, 10 g of yeast extract and 20 g of dextrose (per liter).
Rhodosporidium toruloides was initially grown on YPD agar plates (2% glucose), made of 10 g of yeast extract, 20 g of peptone, 20 g of agar and 20 g of dextrose dissolved in 1000 ml of ultrapure water. Overnight culture of R. toruloides grown in YPD fermentation media (2% w/w glucose) was used to inoculate 50 ml of the fermentation media with an optical density (OD600) of 0.2. Fermentation was carried out at 30 °C in 250 ml Erlenmeyer flasks incubated for 5 days shaken at 200 rpm. Three biological replicates were carried out during the experiments.
Growth analysis
The growth in the yeast cultures was found out using the optical density (OD600) of the culture media. Samples were collected every 24 h for 5 days and OD600 was measured using NanoDrop 2000 UV–Vis spectrophotometer (Thermo Scientific, Waltham, MA, USA). The OD600 values were correlated to dry weight to construct the growth curve for the yeast. Samples with varying concentrations of cells were centrifuged (10,000g, 5 min), washed twice with ultrapure water and analysed using the infrared moisture analyser (Sartorius MA37) to find the dry cell weight. This data was used in constructing a calibration curve for OD600 values and yeast cell dry weight.
Carotenoid measurement
1 ml of the culture was used to measure the intracellular carotenoid production by R. toruloides. The samples were centrifuged for 10 min at 10,000 rpm, 4 °C and washed with Milli-Q water. The cell palettes were broken down using glass beads in the Fast Prep Grinder (MP Biomedicals, Solon, OH, USA). Carotenoids were extracted with 1 ml of acetone until the cell pallets were colourless.
Agilent 1100 high-performance liquid chromatography (HPLC) equipped with a photodiode array detector was used to identify and quantitate the carotenoids extracted. Carotenoids, torularhodin, torulene and β-carotene, were identified using standards from CaroteNature (Ostermundigen, Switzerland). Li-Chrospher 100RP-18 column (250 mm × 4.6 mm id, 5 μm) and a guard column (4 mm × 4 mm id) of the same material (Merck, Rahway, NJ, USA) were used for the separation. Acetone and Milli-Q water were used as the mobile phases with a gradient from 70 to 100% (acetone in Milli-Q water) at a 0.5 ml/min flowrate. Detection was performed at 450 nm, and the UV–Vis absorption spectra were recorded online using the photodiode array detection system (Lee et al. 2014).
Fatty acid measurement
Lipids were extracted from the yeast cells employing the chloroform–methanol 2:1 method adopted by Chen and Chen (2014). 1 ml of the culture medium was extracted from the fermentation flask. The cell pallets were washed three times using Milli-Q water. These washed cells were centrifuged for 10 min at 10,000 rpm, 4 °C and separated. The cell pallets were resuspended in 1 ml of 0.9% NaCl and acidified with 200 μl of acetic acid. 10 μl of 10 mg/ml of heptadecanoic acid (dissolved in ethanol) (Sigma, St. Louis MO, USA) was added as the internal standard (IS) to account for the fatty acid loss during processing. Around 300 μl of glass beads were added and the cells were disrupted in the FastPrep®-24 instrument for 30 s for four times.
3 ml of a chloroform–methanol (2:1) mixture was added to the samples, inverted several times, vortexed vigorously and centrifuged (10,000g, 10 min, 4 °C). The bottom chloroform layer was transferred to another tube. This step was repeated once again. The collected chloroform was dried overnight to dryness on a heatblock at 30 °C.
Fatty acid derivatization was performed according to a method proposed by Horak et al. (2009). The dried lipid residue was redissolved in 500 μl 10% BF3-methanol and incubated in a sealed screw cap tube in a heatblock at 95 °C for 20 min. Then the tubes were cooled down to near room temperature. 300 μl of saturated NaCl in water and 300 μl of n-hexane were added. Samples were centrifuged (14,000 rpm, 10 min) at room temperature. The upper hexane layer containing the extracted Fatty acid methyl esters (FAMEs) were transferred to glass vials for GCMS analysis. FAME mix C8-C24 (Sigma, St. Louis MO, USA) was used as the standard for quantitation.
Chromatography was performed using Agilent Technologies 7890A GC-5975C inert MS system. 1 μl samples were injected into the HP-5MS capillary column by splitless mode using an auto-injector. Helium was used as a carrier gas at a flow rate of 1.1 ml/min. The inlets and MS source temperatures were maintained at 250 and 230 °C respectively. The oven temperature was maintained at 80 °C for 1 min and ramped to 250 °C at a rate of 7 °C/min, then held at 250 °C for 10 min. Data were acquired in full scan from 35 to 600 m/z.
Media metabolite analysis
1.5 ml of the original sample media was spiked with 10 µl internal standard (IS, ribitol, 2 mg/ml dissolved in water) and freeze dried. The lyophilised samples were derivatized for GCMS analysis according to Wang et al. (2010). Methoximation was performed by dissolving the samples in 50 µl of methoxyamine hydrochloride (20 mg/ml in pyridine) (Sigma, St. Louis MO, USA) to protect the carbonyls and incubating at 37 °C for 60 min. Afterwards, silylation was carried out by adding 100 µl of N-methyl-N-(trimethylsilyl)-trifluoroacetamide (MSTFA) with 1% trimethylchlorosilane (TMCS) (Sigma, St. Louis MO, USA) to each sample and incubating at 70 °C for 30 min. Subsequently the samples were shaken for 60 min at room temperature and then analysed in GCMS. All samples were analysed within 24 h in a random order.
Metabolites were analysed using Agilent 7890A GC-5975C inert MSD (with Triple Axis Detector) system (Agilent Technologies, CA, USA) equipped with a HP-5MS, 5% Phenyl-Methyl-Silox capillary column (30 m × 250 μm × 0.25 μm Agilent J&W Scientific, Folsom, CA, USA). 1 μl samples were injected to the system by the auto-sampler in split less mode. The solvent cut off was set to 5 min. Helium was used as a carrier gas at a flow rate of 1.1 ml/min. The inlet and ion source temperatures were maintained at 250 and 230 °C. The oven temperature was maintained at 75 °C for 4 min and increased at 4 °C/min to 280 °C and remained for 2 min. Data was acquired in full scan from 35 to 600 m/z.
Chromatographic peak deconvolution and identification was processed using Agilent MassHunter Qualitative Analysis software (B.06.00). Multivariate data analysis and statistical analysis were performed using Mass Profiler Professional (B.02.01) software. The list of compounds extracted was subjected to alignment, normalization (according to the IS, ribitol) and filtering. Afterwards, principal component analysis (PCA) was performed on the obtained data to eliminate any outliers present in the data. Clustering heatmap was created using K-Means algorithm and Euclidean distance calculation. Statistical analysis of the data were performed by one-way ANOVA followed by the post hoc Tukey’s Honest Significant Difference (HSD) to determine significant metabolic changes; p < 0.05 was considered significant. The multiple testing was corrected by Benjamin Hochberg false discovery rate (FDR). Accurate masses of the compounds were searched against NIST mass spectral library with similarity above 75% for feature identification.