Raw Glycerol and Parboiled Rice Ef fl uent for Carotenoid Production : Effect of the Composition of Culture Medium and Initial pH

Carotenoids are some of the most important compounds found in food. These natural colorants are responsible for yellow, orange and red colors in many kinds of food and play a decisive role in their acceptability (1). Besides, they have several other benefi ts to health since they protect humans against certain forms of cancer, cardiovascular disease, cataracts and macular degeneration (2–4). Rise in consumers’ health awareness has made them seek functional foods, a fact that may trigger an increase in the demand for carotenoids in the food industry (5). The global market demand for carotenoids has grown 2.9 % per year (6) and is expected to reach almost 10 million tons by 2017 (7). However, most of the commercial carotenoids derive from chemical synthesis and cannot meet consumers’ demands for natural carotenoids (5). Due to concerns over the use of chemical additives in food, there has been an increasing interest in carotenoids that are naturally obtained by biotechnological processes. ISSN 1330-9862 scientifi c note


Introduction
Carotenoids are some of the most important compounds found in food.These natural colorants are responsible for yellow, orange and red colors in many kinds of food and play a decisive role in their acceptability (1).Besides, they have several other benefi ts to health since they protect humans against certain forms of cancer, cardiovascular disease, cataracts and macular degeneration (2)(3)(4).
Rise in consumers' health awareness has made them seek functional foods, a fact that may trigger an increase in the demand for carotenoids in the food industry (5).The global market demand for carotenoids has grown 2.9 % per year (6) and is expected to reach almost 10 million tons by 2017 (7).However, most of the commercial carotenoids derive from chemical synthesis and cannot meet consumers' demands for natural carotenoids (5).Due to concerns over the use of chemical additives in food, there has been an increasing interest in carotenoids that are naturally obtained by biotechnological processes.
A large variety of substrates and microorganisms has been tested with special emphasis on the ever-increasing interest in the production of microbial carotenoids.Therefore, the yeast Phaffi a rhodozyma (also referred to as Xanthophyllomonas dendrorhous) stands out as a natural source of carotenoids (8)(9)(10)(11)(12) due to its ability to assimilate diff erent carbon sources as substrates, heterotrophic metabolism, relatively fast growth rate, ability to reach high cell density and nutrient quality and safety as a food additive.Besides, it has been certifi ed as a GRAS (Generally Recognized as Safe) microorganism.
The production of bioproducts with alternative substrates, such as a source of carbon and/or nitrogen, is very important.Therefore, Brazil has the advantage of having various industrial by-products that can be used to add value, decrease the risk of having them discarded into the environment and reduce costs of medium production.
The eff ect of the experimental design methodology on the conditions of the process that involves the production of carotenoids has been studied (6,13,14).In a previous study carried out by our research group, the production of carotenoids by the wild yeast Sporidiobolus pararoseus with the use of a complex medium containing by-products such as raw glycerol, corn steep liquor and sugar cane molasses was investigated (15).The production of carotenoids from Phaffi a rhodozyma NRRL Y-17268 was investigated in a culture medium containing pure glycerol and raw glycerol as additional carbon sources (8) and parboiled rice effl uent as nitrogen and phosphorus sources (16).Moreover, carotenoids produced by this yeast strain had antioxidant activity (expressed as Trolox equivalent) of 4.00 and 3.66 mM/μg in ABTS radical cation discoloration and ferric reducing antioxidant power assays, respectively (17).However, the combined use of both by-products as substrates in the culture medium to maximize the produc tion of carotenoids by an experimental design has not been studied yet.
Therefore, the eff ect of the concentration of culture medium containing two agro-industrial by-products, raw glycerol and parboiled rice effl uent, as sources of carbon and nitrogen, respectively, on carotenoid bioproduction by Xanthophyllomonas dendrorhous was studied to add value to these by-products, thus possibly reducing the cost of the culture medium.

Alternative substrates
Raw glycerol and parboiled rice effl uent, which were used in this study as alternative sources of carbon and nitrogen, were supplied by industries located in the south of Brazil and previously characterized (8).

Culture conditions and production of carotenoids
The yeast Xanthophyllomonas dendrorhous NRRL Y-17268 used in this research was kept in 10 mL of yeast malt (YM) agar (18) with the addition of 10 % (98 % by mass) analytical grade glycerol (Synth, Diadema, Brazil) and stored at -18 °C.
The inoculum was prepared by transferring 10 mL of the previously thawed microorganism suspension kept in analytical grade glycerol (-18 °C) to 90 mL of the modifi ed YM medium (in g/L: glucose 10, yeast extract 3, peptone 5, malt extract 3 and KNO 3 0.2) (19).It was incubated at 25 °C and 150 rpm for 48 h or for the time required for the cell count to be at least 10 8 cells/mL, counted by a Neubauer chamber (Optik Labor, Lancing, UK) (9).
Bioproduction tests were conducted on an orbital shaker (model TE 425; Tecnal, Piracicaba, Brazil) with controlled stirring rate and temperature.Erlenmeyer fl asks (500 mL) with 153 mL of culture medium were previously sterilized in an autoclave (model 103; Fabbe, São Paulo, Brazil) at 121 °C for 15 min.Each fl ask was inoculated with 17 mL of the previously prepared yeast suspension, which corresponds to 10 7 cells/mL (20), and incubated at 25 °C and 150 rpm for 168 h.Every 24 h 10-mL samples were collected and centrifuged at 1745×g for 10 min for the analysis of the supernatant (pH) and of the pellet (biomass and carotenoids).

Experimental design
In the experimental design, the eff ects of the concentration of the medium components and initial pH on carotenoid production were evaluated using the Plackett --Burman design, assuming that there are no interactions among the diff erent variables (x i ) in the studied range.A linear approach is considered to be suffi cient for screening, according to the following equation: /1/ where y is the estimated target function, β i are the regression coeffi cients and n is number of variables.The Plackett -Burman design is a fractional factorial design whose main eff ects may be simply calculated as the diff erence between the average value of the measurements made at the low level (-1) of the factor and the average value of the measurements at the high level (+1) (21).
The independent variables (factors) under study were raw glycerol, parboiled rice effl uent, glucose, yeast extract, malt extract, peptone and initial pH (Table 1).In this study, seven variables were screened in 12 assays with three central points (Table 2).The responses or dependent variables during 168 h of study were total carotenoids (in μg/mL or μg/g) and biomass (in g/L).

pH determination
The pH of the sample was determined by a potentiometer (model MB-10; Marte, São Paulo, Brazil), in accordance with the AOAC method no.972.44 (22).

Recovery and determination of total carotenoids
Recovery of total carotenoids was carried out in agreement with the methodology adapted from Fonseca et al. (10).Cells were centrifuged at 1745×g for 10 min, dried at 35 °C for 48 h in an oven (Eletrolab, São Paulo, Brazil) and frozen (at -18 °C for 48 h) in a refrigerator (Consul, Joinville, Brazil).Aft erwards, at room temperature, 2 mL of dimethylsulphoxide (DMSO; Synth, Diadema, Brazil) were added (previously heated at 55 °C for 1 h, in a water bath (model 102; Fanem, São Paulo, Brazil) to 0.05 g of dried and frozen cells and shaken on a vortex agitator (model QL-901; Biomixer, Ningbo, PR China) and left to rest for 30 min.
Aft er cell disruption, carotenoids were extracted by adding 10 mL of acetone (Synth).Then, centrifugation was conducted at 1745×g for 10 min.The supernatant was separated and successive extractions were carried out until the solvent and cells were colorless.A volume of 20 mL of NaCl solution (20 % by mass per volume) and 20 mL of petroleum ether (Synth) were added.Aft er agitation and phase separation, excess water was removed with anhydrous sodium sulfate (Synth) and total carotenoids were determined in petroleum ether as astaxanthin by spectrophotometric reading (model SP-220; Biospectro, Zhejiang, PR China) at λ=474 nm (23) with the molar absorption coeffi cient of 2100 L/(mol•cm) (24,25), according to the following equation: where SPC is the specifi c production of total carotenoids (μg/g), A is the absorbance, V is the fi ltrate volume (mL), and m is the dry cell mass (g).

Determination of biomass
Cell count was estimated by absorbance readings at λ=620 nm (26).The conversion of absorbance into biomass concentration was performed using a previously determined standard curve.

Statistical analysis
A statistical analysis of the estimated eff ects of each variable was performed.All analyses were done at 90 % confi dence level (p<0.1) by the STATISTICA v. 5.0 (StatSoft Inc., Tulsa, OK, USA) soft ware.

Carotenoid production kinetics
Total nitrogen (15.2 mg/L) and phosphorus (30.7 mg/L) concentrations in parboiled rice effl uent are in agreement with data found in the literature regarding this agro-industrial by-product (27).The carbon mass fraction (44.4 %) in raw glycerol used in this study was higher than the one (39 %) used by Chatzifragkou et al. (28) in the study of the ability of eukaryotic microorganisms to assimilate raw glycerol derived from biodiesel and convert it into products with high added value.Therefore, the concentrations of nitrogen and phosphorus essential for the cell growth of the parboiled rice effl uent and the carbon mass fraction in raw glycerol infl uence the transformation of these materials into potential sources of such nutrients in the culture medium used in this study.
The determination of the C:N ratio is a way to explain the cell growth and carotenoid production.Results in Table 2 show that reduction of the nitrogen content stimulated the production of carotenoids (and increased the value of the C:N ratio).Vustin et al. (29) found that a C:N ratio above 5 improves the biosynthesis of carotenoids.However, this ratio, by itself, cannot explain cell growth and carotenoid production.Other nutrients, such as vitamins, amino acids and minerals, and the conditions of the process (temperature, pH, oxygen content) exert strong infl uence on the metabolism of yeasts (29,30).On the other hand, the trials with C:N ratio under 5 had the worst performance (trials 7, 9 and 11).Similar results were observed by Rios et al. (16) with the same microorganism in the medium containing alternative carbon and nitrogen sources, but without raw glycerol.
In trial 8 (in g/L: raw glycerol 10, glucose 10, yeast extract 10, malt extract 10, peptone 1, initial pH=6), SPC was the highest, equal to 326.8 μg/g (Fig. 1c), whereas VPC was 4.1 μg/mL (Fig. 1e).The exponential growth phase occurred up to 120 h, followed by the stationary phase up to 168 h (Fig. 1a), when maximum production of carotenoids was achieved, an evidence of the fact that carotenoid synthesis is not directly associated with cell growth of this yeast.Likewise, Chávez-Cabrera et al. (11) found that the concentration of total proteins and pigments shows opposing trends, implying that protein synthesis restriction, necessary for cell growth, stimulates carotenoid synthesis by X. dendrorhous.Liu and Wu (30) also observed that carotenoids are synthesized as secondary metabolites 474 nm 100 SPC 21 by X. dendrorhous, especially when cells are under stress (nutrient limitation) and when cell growth (primary metabolism) is suppressed.
VPC (Figs. 1e and f) trend was similar to biomass production (Figs.1a and b), since it presents a relationship between SPC and cell count.
Good reproducibility of the carotenoid bioproduction can be verifi ed in trials 13, 14 and 15 (central points) of the Plackett -Burman design (Table 2 and Fig. 1).
The pH curves are shown in Figs.1g and h.Depending on the composition of the medium culture and process conditions during its growth, the yeast X. dendrorhous can excrete carbonic intermediaries, such as acetic acid, an alcohol or an intermediate of the citric acid cycle that is subsequently reabsorbed, thus stimulating carotenogenesis (31) and infl uencing pH decrease in the fi rst 24 h of cultivation.This behavior was bett er represented in trial 8 (Fig. 1g), according to the SPC peak observed (Fig. 1c) at 168 h.

Eff ect of the culture medium composition on carotenoid production
Fig. 2 shows the analysis of the main eff ects of the variables on the biomass concentration, SPC and VPC during 168 h of cultivation using the Plackett -Burman design, which was performed at a 90 % confi dence level (p<0.1).Increasing the raw glycerol concentration from 10 to 40 g/L had signifi cant eff ect on the biomass concentration and VPC, leading to an average increase of 6.7 g/L and 0.9 μg/mL, respectively.Glucose had negative eff ect when the medium concentration was increased from 1 to 10 g/L.It decreased biomass concentration to 1.5 g/L, but increased SPC to 44.7 μg/g.
It is likely that these eff ects happened because glucose and raw glycerol are metabolized quickly.Glycerol passes through the cell membrane by facilitated diff usion (32).Inside the cell and under the conditions without nitrogen, it causes stress to the red yeast, thus stimulating carotenoid and other secondary (as well as primary) metabolite production (33,34).Therefore, the energy for efficient acetyl-CoA and NADPH synthesis required for increased carotenoid production may derive from the catabolism of glycerol via glycolysis (34).
Bellou et al. (35) found that the use of glycerol by Mortierella ramanniana did not alter the amounts of biomass in comparison with other strains of Mortierella cultivated on multiple carbon sources.However, these fungi accumulated more interesting compounds (lipids) whose production was associated with primary metabolism during mycelial growth.In the yeast Yarrowia lipolytica, higher assimilation rates of glycerol than of glucose were also observed (36).However, they were not signifi cantly affected by the increase in dissolved oxygen concentration (37).
Under aerobic conditions and easily metabolizable high levels of sugars, Deken (38) found that the Crabtree eff ect occurs and results not only in the decrease of biomass production but also in ethanol and acetic acid formation.Rivaldi et al. (39) showed that the route of glycerol assimilation is similar to the glycolysis, since the oxidation of glycerol by yeast consists in the phosphorylation of glycerol by the glycerol kinase enzyme to form glycerol-3-phosphate, which is reduced to dihydroxyacetone phosphate and is considered an important intermediate to a gluconeogenesis molecule (hexose synthesis), and to obtain a number of compounds by oxidative pathways.
Reynders et al. (40) also observed the Crabtree eff ect in fed-batch fermentation by Phaffi a rhodozyma.In fed--batch culture at feed concentrations of 27 and 55 g/L of glucose, there was a good agreement between the result-  X 1 =raw glycerol (g/L), X 2 =effl uent (g/L), X 3 =glucose (g/L), X 4 =yeast extract (g/L), X 5 =malt extract (g/L), X 6 =peptone (g/L), X 7 =initial pH, Y 1 =biomass (g/L), Y 2 =SPC (μg/g) and Y 3 =VPC (μg/mL) ing biomass formation and the one expected by a mass balance model.At 125 g/L of glucose in the feed, the biomass formation was lower than the expected one and fermentation products, such as ethanol and acetic acid, accumulated in the culture medium.
With the use of 50 g/L of parboiled rice effl uent, there was a signifi cant increase of biomass concentration of 1.1 g/L and decrease of SPC of 49.2 μg/g (Table 2).Therefore, phosphorus and nitrogen found in this substrate were probably used for cell growth.This result also shows that carotenogenesis can be stimulated by nitrogen limitation.Vustin et al. (29) suggested that nitrogen in excess in the medium culture can increase cell growth, but suppress the activity of enzymes in carotenoid conversion.
Likewise, increase in the malt extract concentration from 1 to 10 g/L resulted in the increase in the eff ects of all responses, namely biomass concentration, SPC and VPC of 5.4 g/L, 24.0 μg/g and 1.2 μg/mL, respectively.However, opposite eff ects were observed with the increase of peptone concentration, a variable that was studied at the same concentration range as the previous variable: there was a signifi cant decrease in biomass concentration (3.4 g/L), SPC (32.5 μg/g) and VPC (0.8 μg/mL) responses.Furthermore, no signifi cant eff ect on any response under study was observed when the concentration of the yeast extract increased from the -1 level (1 g/L) to the +1 level (10 g/L), as well as with the increase in the initial pH of the culture medium from 4 to 6.
Rios et al. (16) observed positive eff ects of the increase in malt extract concentration on carotenoid production but not of the yeast extract concentration when using P. rhodozyma in similar culture media without the use of raw glycerol.The yeast extract and other nutrients added to the culture medium did not infl uence the production of carotenoids by Xanthophyllomyces dendrorhous ENM 5 at 20 °C and 250 rpm.However, contrary results to the ones found in this study were obtained, with an increase in the initial pH of the culture medium from 4.5 to 6.5 (30).Studies show that the initial pH of the culture medium for cell growth and production of carotenoids depend on the microbial strain under study (41).
The production of carotenoids using agro-industrial by-products in this study was bett er than that obtained by Zeni et al. (42) using isolated microorganisms capable of producing carotenoids from soil, leaves, fl owers, fruits, agro-industrial waste and decomposed processed products.The isolated microorganisms were cultured on an orbital shaker with 10 % (by mass per volume) inoculum (25 °C and 180 rpm for 48 h) for 120 h in the dark in YM medium.A total of 116 microorganisms were isolated (16 yeasts, 65 bacteria and 35 molds).Three yeasts showed great potential to produce red carotenoids in a concentration up to 707-818 μg/L (99-263 μg/g).One mold and two bacteria produced yellow pigments in a concentration ranging from 1063 to 2563 μg/L (239-2310 μg/g).
The Plackett -Burman design was adequate for screening medium culture variables in the production of carotenoids by X. dendrorhous NRRL Y-17268.Increases in raw glycerol and malt extract concentrations were the most important variables for the concomitant increase in cell growth and production of carotenoids.Increase in the peptone concentration infl uenced cell growth and production of carotenoids negatively.Glucose and parboiled rice effl uent showed the opposite eff ect.The initial pH of the culture medium and the concentration of yeast extract were not signifi cant (p<0.1) in the responses under investigation.

Conclusions
The use of the by-products raw glycerol and parboiled rice effl uent in the culture medium for carotenoid bioproduction by Xanthophyllomonas dendrorhous NRRL Y-17268 under diff erent conditions was observed.The effects of the concentration of culture medium components were verifi ed in shake fl asks by Plackett -Burman design.The maximum value was found at 168 h with volumetric production of carotenoids of 4.1 μg/mL (326.8 μg/g) in the culture medium containing (in g/L): raw glycerol 10, glucose 10, yeast extract 10, malt extract 10 and peptone 1, at initial pH=6.

1 Fig. 1 .
Fig. 1.Carotenoid production kinetics determined by the Plackett -Burman design at 25 °C and 150 rpm: a and b) biomass concentration, c and d) specifi c production of carotenoids (μg/g), e and f) volumetric production of carotenoids (μg/mL), and g and h) pH

Table 1 .
Levels of factors, concentrations of variables and pH values studied by Plackett -Burman design

Table 2 .
Plackett -Burman design with real values, C:N ratio and responses for biomass concentration, specifi c production of carotenoids (SPC) and volumetric production of carotenoids (VPC) in 168 h