Optimization and Modeling of Process Parameters for Lipase Production by Bacillus brevis

Statistical evaluation of fermentation conditions and nutritional factors by Plackett–Burman two-level factorial design followed by optimization of significant parameters using response surface methodology for lipase production by Bacillus brevis was performed in submerged batch fermentation. Temperature, glucose, and olive oil were found to be the significant factors affecting lipase production. Maximum lipase activity of 5.1 U ml⁻¹ and cell mass of 1.82 g l⁻¹ at 32 h were obtained at the optimized conditions of temperature, 33.7 °C; initial pH, 8; and speed of agitation, 100 rpm, with the medium components: olive oil, 13.73 ml l⁻¹; glucose, 13.98 g l⁻¹; peptone, 2 g l⁻¹; Tween 80, 5 ml l⁻¹; NaCl, 5 g l⁻¹; CH₃COONa, 5 g l⁻¹; KCl, 2 g l⁻¹; CaCl₂·2H₂O, 1 g l⁻¹; MnSO₄·H₂O, 0.5 g l⁻¹; FeSO₄·7H₂O, 0.1 g l⁻¹; and MgSO₄·7H₂O, 0.01 g l⁻¹. The lipase productivity and specific lipase activity were found to be 0.106 U (ml h)⁻¹ and 2.55 U mg⁻¹, respectively. Unstructured kinetic models and artificial neural network models were used to describe the lipase fermentation. The kinetic analysis of the lipase fermentation by B. brevis shows that lipase is a growth-associated product.     

Aqueous medium enzymatic preparation of <Emphasis Type="SmallCaps">l</Emphasis>-alpha glycerylphosphorylcholine optimized by response surface methodology

The ability of Rhizopus chinensis lipase (Thermo 4S-3) to catalyze the deacylation of l,2-diacyl-sn-glycero-3-phosphocholines (sn-1,2-PC) for l-alpha glycerylphosphorylcholine (l-α-GPC) preparation was investigated. Response surface methodology (RSM), based on a modified central composite rotatable design, was employed to examine the effects of substrate concentrations, temperature, enzyme loading, and dosage of CaCl₂ on the l-α-GPC yield. RSM analysis indicated good correlation between experimental and predicated values. The optimal condition was confirmed as follows: reaction time, 3.5 h; temperature, 43 °C; enzyme loading, 28.2 U mL⁻¹; substrate concentration, 51.5 mg mL⁻¹; and dosage of CaCl₂, 1.9 mg mL⁻¹. Under these conditions, the l-α-GPC yield increased by 96.8%, which was close to the amount predicted by the model.     

Recovery of Native Potato Protein Comparing Expanded Bed Adsorption and Ultrafiltration

Obtaining native protein from potato fruit water (PFW) acceptable for food consumption was attempted by comparing expanded bed adsorption (EBA) and ultrafiltration (UF).The methods were assessed on their process performance and the product quality. Extractable tuber proteins were recovered from lab-prepared PFW either by adsorption to an EBA column using a mixed mode resin (0.31 L) or by batch concentration in an UF (10 kDa MWCO, 0.093 m²) unit and then freeze dried. The yields on protein and esterase activity were higher (p < 0.05 and p < 0.01; Mann–Whitney U-test) in UF (3.2 g l⁻¹ PFW and 3.17 kU l⁻¹ PFW) than in EBA (1.8 and 1.21). The performance difference was also reflected in process productivity for esterase activity which was fivefold better (p < 0.01) in UF (4.30 kU h⁻¹) than with EBA (0.84) due to the higher enzyme retention; protein productivities were the same. The content of solanidine glycoalkaloids was depleted to moderate levels but came out unaffected by the processing method: EBA 286 ppm, UF 213 ppm. The low levels of chlorogenic acid in all EBA preparations were negatively correlated to high brightness score (L* = 73.8%), a favorable attribute in food-quality proteins. Both methodologies produced native preparations of comparable protein content (75%). EBA processing, however, increased the fraction of the patatin protein which may offer advantages in food applications.     

Co-production of Lactic Acid and <Emphasis Type="Italic">Lactobacillus rhamnosus</Emphasis> Cells from Whey Permeate with Nutrient Supplements

Lactic acid and cell production from whey permeate by Lactobacillus rhamnosus with different nutrient supplements were investigated in this study. Yeast extract was identified as the most effective nutrient affecting lactic acid production. Increase in inoculum size from 0.05% to 1% (v/v) resulted in a substantial increase in lactic acid productivity from 0.66 to 0.83 g L⁻¹ h⁻¹ (P < 0.001). The optimal temperature for lactic acid production was 37 °C, while the highest cell production was obtained at 42 °C. When whey permeate and yeast extract concentrations were 6.8% (w/v) and 3 g L⁻¹, respectively, lactic acid productivity reached 0.85 g L⁻¹ h⁻¹ after 48-h cultivation, which is 3.40 times of those without nutrient supplements.     

High nisin-Z production during repeated-cycle batch cultures in supplemented whey permeate using immobilized Lactococcus lactis UL719

Nisin-Z production was studied during repeated-cycle pH-controlled batch (RCB) cultures using Lactococcus lactis subsp. lactis biovar. diacetylactis UL719 immobilized in κ-carrageenan/locust bean gum gel beads in supplemented whey permeate. After an initial colonization of gel beads during the first two cycles, nisin-Z production in bulk medium and gel beads was very similar for 1-h and 2-h cycle RCB cultures. A very high nisin-Z production (8200 IU mL⁻¹) was measured in the broth after the 1-h cycles, with a corresponding volumetric productivity of 5730 IU mL⁻¹ h⁻¹. This productivity is much higher than maximum nisin productivities reported in literature or maximum productivities obtained previously for free-cell batch cultures (850 IU mL⁻¹ h⁻¹), and free-cell (460 IU mL⁻¹ h⁻¹) or immobilized-cell (1760 IU mL⁻¹ h⁻¹) continuous cultures, using the same strain and fermentation conditions. The stability of RCB cultures was demonstrated for 24 and 36 1-h cycles carried out over 3 and 6-day periods, respectively. Changing environmental conditions during batch cultures resulted high nisin production.     

Preparation of Alcalase-catalyzed casein plasteins in the presence of proline addition and the ACE-inhibitory activity of the plasteins in vitro

Casein hydrolysates were prepared by hydrolysis of casein with alkaline protease Alcalase for 6 h and showed the highest ACE-inhibitory activity in vitro with an IC₅₀ value of 47.1 μg mL⁻¹. Casein hydrolysates prepared were subjected to Alcalase-catalyzed plastein reaction in the presence or absence of proline addition to prepare casein plasteins. Some optimal reaction conditions of plastein reaction in the presence of proline addition were studied using response surface methodology with the decrease in free amino groups in the casein plasteins as response. When the concentration of casein hydrolysates was fixed at 35% (w w⁻¹) and reaction time at 6 h, the optimal conditions were reaction temperature 48 °C, addition level of proline 0.54 mol/mol free amino groups of casein hydrolysates and addition level of Alcalase 9.5 kU g⁻¹ proteins. With these conditions, the maximal decrease in free amino groups in casein plasteins was 195.7 μmol g⁻¹ proteins. The ACE-inhibitory activities of twelve casein plasteins in vitro, prepared in the presence or absence of proline addition with different reaction extents, were evaluated and compared. The results showed that the ACE-inhibitory activity of the casein plasteins prepared in the presence of proline addition changed irregularly, different to that of the casein plasteins prepared in the absence of proline addition, and might relate to the different linking of proline to the peptides in casein hydrolysates during plastein reaction. When the casein plasteins prepared in the presence of proline addition had a decrease in free amino groups 195.7 μmol g⁻¹ proteins, the IC₅₀ value of the casein plasteins was lowered to 0.2 μg mL⁻¹.     

Analytical Methods for the Evaluation of α-<Emphasis Type="SmallCaps">l</Emphasis>-Fucosidase Activity in Sheep Milk

The lack of analytical methods for measuring the activity of highly thermolable endogenous enzymes in sheep milk is a factor that hampers the protection of typical Protected Designation of Origin (PDO) dairy products made from raw milk. In order to provide a solution, this study assesses, tests, and fully validates analytical procedures for the determination of α-l-fucosidase activity in sheep milk. While the UV–VIS method has been optimized for this matrix in order to solve clarification problems before the spectrophotometric reading, a reversed phase high-performance liquid chromatography literature method proposed for bovine milk has been successfully applied also to sheep milk. Both methods have been fully validated in terms of sensitivity, linearity, precision, and accuracy, displaying low detection and quantification limits, excellent linearity over a wide enzymatic activity interval, very good repeatability and reproducibility, and the lack of any bias. The analysis of a number of real samples of whole sheep milk has allowed the evaluation of an average value (46.78 ± 5.49 U mL⁻¹) and range (from 29.27 ± 2.60 to 72.64 ± 1.17 U mL⁻¹) of α-l-fucosidase activity. Such activity does not seem to differ substantially from those measured for bovine milk. Finally, marked seasonal variability has been observed in this preliminary dataset.     

Properties of <Emphasis Type="Italic">Aspergillus subolivaceus</Emphasis> free and immobilized dextranase

Aspergillus subolivaceus dextranase is immobilized on several carriers by entrapment and covalent binding with cross-linking. Dextranase immobilized on BSA with a cross-linking agent shows the highest activity and considerable immobilization yield (66.7%). The optimum pH of the immobilized enzyme is shifted to pH 6.0 as compared with the free enzyme (pH 5.5). The optimum temperature of the reaction is resulted at 60 °C for both free and immobilized enzyme. Thermal and pH stability are significantly improved by the immobilization process. The calculated K _m of the immobilized dextranase (14.24 mg mL⁻¹) is higher than that of the free dextranase (11.47 mg mL⁻¹), while V _max of the immobilized enzyme (2.80 U μg protein⁻¹) is lower than that of the free dextranase (11.75 U μg protein⁻¹). The immobilized enzyme was able to retain 76% of the initial catalytic activity after 5.0 cycles.     

Large-scale production and application of leucine aminopeptidase produced by <Emphasis Type="Italic">Aspergillus oryzae</Emphasis> LL1 for hydrolysis of chicken breast meat

Leucine aminopeptidase (LAP) production by Aspergillus oryzae LL1 (LAP LL1) was scaled up by varying fermenting conditions. The maximal LAP activity for 5, 20, 250, and 2,000 L fermenters were 0.22, 0.16, 0.19, and 0.12 U mL⁻¹, respectively. LAP LL1 production coincides with increasing pH for all scale cases. A pilot plant scale hydrolysis experiment using LAP LL1 was prepared by removing the cells from 1,000 L of A. oryzae LL1 culture and concentrating the supernatant to ∼60 L, with a LAP LL1 activity yield of ∼20% and a specific LAP activity of 0.054 U mg⁻¹, comparable to Flavourzyme. Using a casein substrate, LAP LL1 specific protease activity (50.9 U mg⁻¹) is 43% higher than Flavourzyme (35.6 U mg⁻¹). We believe LAP LL1 has potential use as an industrial enzyme for food processing. Comparing the hydrolysis effects on chopped chicken breast meat (100 g), the degree of hydrolysis (D_H) of LAP LL1 (50.6%) was about twice as effective as Flavourzyme (25.6%). On a pilot-plant scale hydrolysis experiment of chopped chicken breast meat (54 kg), the D_H of LAP LL1 was 45.6%. Adding 30 μmol L⁻¹ zinc increased D_H by 33.3% for LAPs LL1 but zinc did not enhance Flavourzyme hydrolysis. Together, this study provides possibility that LAP LL1 has significant potential for application in the food industry. Shie-Jea Lin and Yi-Hong Chen have contributed equally to this work.     

Effect of dynamic high-pressure microfluidization at different temperatures on the antigenic response of bovine β-lactoglobulin

The antigenic response of β-lactoglobulin (β-Lg), treated by dynamic high-pressure microfluidization (DHPM) at different temperatures, was determined by an indirect competitive enzyme-linked immunosorbent assay using polyclonal antibodies from rabbit serum. DHPM treatment causes changes in the protein structure and may influence the antigenicity of β-Lg. DHPM treatment of β-Lg at 90 °C showed significant effects with the antigenic response of 5.2 μg mL⁻¹ (untreated), 45 μg mL⁻¹ (40 MPa), 79 μg mL⁻¹ (80 MPa), 132 μg mL⁻¹ (120 MPa), and 158 μg mL⁻¹ (160 MPa). In combination with temperature treatment (70–90 °C), the antigenic response enhanced as the temperature increased at 160 MPa. The β-Lg antigenicities were about 14, 108, and 158 μg mL⁻¹ at 70, 80, and 90 °C, respectively. However, the influence of DHPM pressures on the antigenic response of β-Lg standards was different. DHPM modified β-Lg standards showed a remarkable increase in antigenicity when treated to 80 MPa. Above 80 MPa, the antigenic response decreased.     

Inulinase Production by Agro-Industrial Residues: Optimization of Pretreatment of Substrates and Production Medium

In this work, production of inulinase was studied. Media formulation was optimized by experimental design and response surface techniques, as well as the pretreatment of the agro-industry residues used in the formulation of fermentation medium. Two agro-industry residues were investigated: sugarcane molasses (SCM) and corn steep liquor (CSL). Pretreatment with sulfuric acid was the most effective for clarification of SCM (pH 5.0, 24 h of resting time and final pH 4.0). Clarification of CSL was accomplished with phosphoric acid (pH 3.0, 24 h of resting time and final pH 5.5). A color reduction of approximately 70% was achieved for both substrates. The highest production of inulinase was obtained in a medium containing 100 g l⁻¹ of pretreated SCM, 100 g l⁻¹ of pretreated CSL and 6 g l⁻¹ of Prodex Lac (yeast hydrolysate), yielding 1,139 U ml⁻¹.     

An approach to improve ACE-inhibitory activity of casein hydrolysates with plastein reaction catalyzed by Alcalase

The preparation method of casein hydrolysates with high ACE-inhibitory activity was studied by Alcalase-catalyzed hydrolysis coupled with plastein reaction. Casein hydrolysates with an IC₅₀ value of about 47 μg mL⁻¹ were first prepared by hydrolysis of casein with Alcalase and then modified with plastein reaction catalyzed by the same enzyme. The impacts of four reaction conditions on plastein reaction of casein hydrolysates were studied, and then optimal conditions were determined using response surface methodology with the decrease of free amino groups in the reaction mixture as response. When the concentration of casein hydrolysates was fixed at 35% by weight, the maximum decrease of free amino groups in the reaction mixture of 181.8 μmol g⁻¹ proteins was obtained. The optimum conditions for the above decrease were found to be an E/S ratio of 7.7 kU g⁻¹ proteins, reaction temperature of 42.7 °C and reaction time of 6 h. Analysis results showed that ACE-inhibitory activity of casein hydrolysates prepared could be improved significantly by plastein reaction. When casein hydrolysates were modified by plastein reaction, with a decrease of free amino groups in the mixture of about 154.7 μmol g⁻¹ proteins and 181.8 μmol g⁻¹ proteins, their IC₅₀ values could be decreased to 0.6 and 0.5 μg mL⁻¹.     

Occurrence of aflatoxin M1 in raw, pasteurized and UHT milk commercialized in Esfahan and Shahr-e Kord, Iran

Between September 2006 and September 2007, 236 samples of raw (n = 140), pasteurized (n = 48) and UHT (n = 48) milk were collected from supermarkets and from bulk milk tanks of eight dairy plants in the cities of Esfahan and Shahr-e Kord, Iran. All samples were analyzed for aflatoxin M₁ (AFM₁) contamination by ELISA and 213 (90.3%) were positive with mean concentrations 65 ng.l⁻¹. These concentrations are lower than the standards of Codex Alimentarius and FDA (500 ng.l⁻¹), but 119 samples (55.9%) had higher concentrations than the maximum tolerance accepted by some European countries (50 ng.l⁻¹). Mean concentrations of AFM₁ in raw, pasteurized and UHT milk were 68, 56, and 65 ng.l⁻¹, respectively. Mean concentrations of AFM₁ in autumn and winter samples were significantly higher (P < 0.05) than those of spring and summer but differences between AFM₁ concentrations of spring and summer samples were not significantly different. Concentrations of AFM₁ in milk from Shahr-e Kord were significantly lower (P ≤ 0.05) than those from Esfahan.     

Properties of thermostable extracellular FOS-producing fructofuranosidase from <Emphasis Type="Italic">Cryptococcus</Emphasis> sp.

The present work was devoted to investigations concerning the fructooligosaccharide producing activity of Cryptococcus sp. LEB-V2 (Laboratory of Bioprocess Engineering, Unicamp, Brazil) and its extracellular fructofuranosidase. After cell separation, the enzyme was purified by ethanol precipitation and anion exchange chromatography. The enzyme showed both fructofuranosidase (FA) and fructosyl transferase (FTA) activity. With sucrose as substrate, the data failed to fit the Michaelis–Menten behaviour, showing a substrate inhibitory model. The K _m, K _i and v _max values were shown to be 64 mM, 3 M and 159.6 μmol mL⁻¹ min⁻¹ for FA and 131 mM, 1.6 M and 377.8 μmol mL⁻¹ min⁻¹ for FTA, respectively. The optimum pH and temperature were found to be around 4.0 and 65 °C, while the best stability was achieved at pH 4.5 and temperatures below 60 °C, for both the FA and FTA. Despite the strong FA activity, the high transfructosylating activity allowed for good FOS production from sucrose (35% yield).     

Comparison Between Systems for Synthesis of Fructooligosaccharides from Sucrose Using Free Inulinase from Kluyveromyces marxianus NRRL Y-7571

This work is focused on the synthesis of the fructooligosaccharides (FOS) from sucrose using free inulinase from Kluyveromyces marxianus NRRL Y-7571 in aqueous and aqueous–organic systems. The most significant variables for the aqueous–organic system were identified using a fractional factorial design. The evaluated variables were the temperature, pH, sucrose concentration, inulinase activity, aqueous/organic ratio, and the polyethylene glycol concentration. The use of sequential experimental design methodology was shown to be very useful in the optimization of the FOS synthesis by inulinase either in aqueous or aqueous–organic systems. For the aqueous–organic system, the maximum Y _FOS reached was 16.7 ± 1.1 wt.% with the following operational conditions: temperature of 40 °C, enzyme activity of 4 U mL⁻¹, organic solvent/total system ratio of 25/100, pH of 6.0, and sucrose concentration of 55%. In the aqueous system, the maximum conversion obtained was 12.8 ± 1.0 wt.% under the following conditions: 40 °C, pH 5.0, 55% sucrose, and inulinase activity 4 U mL⁻¹.     

Substrate specificity of a recombinant d-lyxose isomerase from Providencia stuartii for monosaccharides

The specific activity and catalytic efficiency (k_cat/K_m) of the recombinant putative protein from Providencia stuartii was the highest for d-lyxose among the aldose substrates, indicating that it is a d-lyxose isomerase. Gel filtration analysis suggested that the native enzyme is a dimer with a molecular mass of 44 kDa. The maximal activity for d-lyxose isomerization was observed at pH 7.5 and 45 °C in the presence of 1 mM Mn²⁺. The enzyme exhibited high isomerization activity for aldose substrates with the C2 and C3 hydroxyl groups in the left-hand configuration, such as d-lyxose, d-mannose, l-ribose, d-talose, and l-allose (listed in decreasing order of activity). The enzyme exhibited the highest activity for d-xylulose among all pentoses and hexoses. Thus, d-lyxose was produced at 288 g/l from 500 g/l d-xylulose by d-lyxose isomerase at pH 7.5 and 45 °C for 2 h, with a conversion yield of 58 % and a volumetric productivity of 144 g l^− 1 h^− 1. The observed k_cat/K_m (920 mM^− 1 s^− 1) of P. stuartiid-lyxose isomerase for d-xylulose is higher than any of the k_cat/K_m values previously reported for sugar and sugar phosphate isomerases with monosaccharide substrates. These results suggest that the enzyme will be useful as an industrial producer of d-lyxose.     

Partial Purification and Characterization of Extracellular Fructofuranosidase with Transfructosylating Activity from <Emphasis Type="Italic">Candida</Emphasis> sp.

The present work was carried out with the aim to investigate some properties of an extracellular fructofuranosidase enzyme, with high transfructosylating activity, from Candida sp. LEB-I3 (Laboratory of Bioprocess Engineering, Unicamp, Brazil). The enzyme was produced through fermentation, and after cell separation from the fermented medium, the enzyme was concentrated by ethanol precipitation and than purified by anion exchange chromatography. The enzyme exhibited both fructofuranosidase (FA) and fructosyltransferase (FTA) activities on a low and high sucrose concentration. With sucrose as the substrate, the data fitted the Michaellis–Menten model for FA, showing rather a substrate inhibitory shape for fructosyltransferase activity. The K _m and v _max values were shown to be 13.4 g L⁻¹ and 21.0 μmol mL⁻¹ min⁻¹ and 25.5 g L⁻¹ and 52.5 μmol mL⁻¹ min⁻¹ for FA and FTA activities, respectively. FTA presented an inhibitory factor K _i of 729.8 g L⁻¹. The optimum conditions for FA activity were found to be pH 3.25–3.5 and temperatures around 69 °C, while for FTA, the optimum condition were 65 °C (±2 °C) and pH 4.00 (±0.25). Both activities were very stable at temperatures below 60 °C, while for FA, the best stability occurred at pH 5.0 and for FTA at pH  4.5–5.0. Despite the strong fructofuranosidase activity, causing hydrolysis of the fructooligosaccharides (FOS), the high transfructosilating activity allows a high FOS production from sucrose (44%).     

Production of Cell Membrane-Bound α- and β-Glucosidase by Lactobacillus acidophilus

Hydrolytic enzymes, viz. α- and β-glucosidase, were produced from indigenous isolate, Lactobacillus acidophilus, isolated from fermented Eleusine coracana. Production of these enzymes was enhanced by optimizing media using one factor at a time followed by response surface methodology. The optimized media resulted in a 2.5- and 2.1-fold increase in α- and β-glucosidase production compared with their production in basal MRS medium. Localization studies indicated 80% of the total activity to be present in the cell membrane-bound fraction. Lack of sufficient release of these enzymes using various physical, chemical, and enzymatic methods confirmed their unique characteristic of being tightly cell membrane bound. Enzyme characterization revealed that both α- and β-glucosidase exhibited optimum catalytic activity at 50 °C and pH 6.0 and 5.0, respectively. K _m and V _max of α-glucosidase were 4.31 mM and 149 μmol min⁻¹ mL⁻¹ for p-nitrophenyl-α-d-glucopyranoside as substrate and 3.8 mM and 120 μmol min⁻¹ mL⁻¹ for β-glucosidase using p-nitrophenyl-β-d-glucopyranoside as the substrate.     

Protein and Amino Acid Solubilization using Bacillus cereus, Bacillus velesensis, and Chryseobacterium sp. from Chemical Extraction Protein Residue

The exploitation of natural resources and increased environmental pollution have stressed the need for more valued use of residues generated by the fish processing plants, and species with low commercial value. Protein hydrolysis processes—whether chemical or enzymatic—generate insoluble proteins from bones, scales, and skin, which are not recovered and are often used as animal feed or disposed off into the environment. As an alternative, insoluble proteins could be converted in useful biomass protein concentrates or amino acids, by using microbial proteases. This work examines the solubilization of insoluble proteins discarded in the process of pH change in fish residues from Whitemouth croaker (Micropogonias furnieri), through the use of bacterial proteases. Temperature and pH conditions in the fermentations were adjusted for each microorganism and time was set at 96 h. Two substrates (acid and alkaline), three microorganism strains, and the substrate concentration used were examined. Among the three strains, Bacillus velesensis reached the higher proteolytic activity (47.56 U mL⁻¹), followed by Chryseobacterium sp. with 23.46 U mL⁻¹. Bacillus cereus (3.13 U mL⁻¹) showed low proteolytic activity. B. velesensis was the bacterium that presented better results with the analyzed substrates, achieving larger amount of soluble protein and free amino acids. The findings showed that these bacteria could be used to solubilize proteins from fish byproducts, which may be particularly useful to increase the yield of hydrolysis process or food formulations.     

Optimization of the Production of Total Carotenoids by Sporidiobolus salmonicolor (CBS 2636) Using Response Surface Technique

This work aimed at evaluating the conditions of growth and the recovery of total carotenoids produced by Sporidiobolus salmonicolor (CBS 2636). Optimization of carotenoid production was achieved by experimental design technique. A Plackett–Burman design was used, followed by a complete second-order design, to optimize the concentration of total carotenoids in a conventional medium. Maximum concentration of 1,019 μg l⁻¹ of total carotenoids was obtained in a medium containing 40 g l⁻¹ glucose, 10 g l⁻¹ malt extract, and 14 g l⁻¹ peptone, at 180 rpm, 25 °C, and initial pH of 4.0. So far, no previous systematic study using microorganisms of the genus Sporobolomyces (formerly Sporidiobolus) for production of carotenoids has been reported. In this study, very good yields of carotenoids (1 mg l⁻¹) could be obtained after optimization of fermentation medium and operation conditions.