Submitted in fulfilment for the requirements for the degree of Doctor of Technology: Biotechnology, Durban University of Technology, 2010. / South African edible oil processing plants produce approximately 3 x 105 tonnes of oil annually with up to 3 tonnes of water for every tonne of oil produced. Wastewater that contains oil extracts varies in organic loading from 30,000 to 60,000 mg.l-1 COD. This wastewater can be used to grow oleophilic fungi to produce valuable industrial products. The global vitamin B market is approximately R25.5 billion with 4500 metric tonnes being produced. A large proportion of this is produced using the fungus Eremothecium gossypii using oil substrates. The aim of this study was to to develop a novel method to produce riboflavin with the aid of fungi, using edible oil effluent (EOE) as substrate, and to optimize the production thereof by statistical experimental design. Four fungi were surveyed for their growth potential on EOE and two, E. gossypii (CBS109.51) and C. famata (ATCC 208.50) were found to produce sufficient riboflavin for further study. Mutation of these organisms using ethylmethane sulphonate (EMS) increased riboflavin production from 3.52 mg.l-1 to 38.98 mg.l-1, an 11-fold increase. An enzyme pathway responsible for this was found to involve isocitrate lyase and comparison of this enzyme’s activity in the mutant against the wild-type using Michaelis-Menten kinetics showed a higher reaction velocity (Vmax) with a reduced substrate affinity (Km) indicating that the mutation was associated with this enzyme. Biomass comparisons were fitted to the sigmoid Gompertz model which was used to compare the wild-type to the mutant and increased specific growth rates and doubling times were observed in mutated cultures of E. gossypi. A strategy of statistical experimental design was pursued to optimize media components and iterative fractional factorial experiments culminating in a central composite optimization experiment were conducted. Statistically verified mathematical models were developed at each stage to identify important media components, predict media interactions, show directions for improvement and finally, predict maximum riboflavin production. An eight-factor resolution IV fractional factorial increased riboflavin production to 112 mg.l-1 followed by a four-factor resolution V experimental design which increased riboflavin production to 123 mg.l-1. A two-factor (yeast extract and NaCl) central composite experimental design predicted a maximum riboflavin production of 136 mg.l-1 which was a 3.5-fold increase from the mutant, and 38.6-fold higher than the E. gossypii wild-type. The optimized value was achieved within predicted confidence intervals in confirmatory experiments. Cost implications for production of riboflavin on EOE were calculated and a 10% technology uptake by the edible oil industry could yield a riboflavin industry with a 63.65 million rand turnover and a potential 24.96 million rand gross profit margin. / National Research Fund.

Submitted in fulfillment of the requirements of the Degree of Master of Technology: Biotechnology, Durban University of Technology, 2011. / Riboflavin (vitamin B2), an essential water-soluble vitamin is commercially produced because it cannot be synthesized by vertebrates. Although this vitamin is produced chemically, bioproduction is a better option since it is more economical, requires less energy, produces less waste and can use renewable sources. In this study we investigated spent oil from the food and motor industries as alternative cheap carbon sources for the bioproduction of this vitamin. Commercial fungal strains namely; Eremothecium gossypii ATCC 10895, Eremothecium gossypii CBS 109.51, Eremothecium ashbyi CBS 206.58 and the yeast, Candida famata ATCC 20850, as well as a laboratory mutated Eremothecium gossypii EMS 30/1 strain were used. Statistical experimental design using a series of fractional factorial experimental designs was used to optimize the effect of yeast extract, peptone, malt extract, K2HPO4 and MgSO4.7H2O to supplement the used oils for optimum riboflavin production. Response surface methodology based on central composite experimental designs was then applied and together with the point predictions made, production media for both substrates were further optimized. The optimized conditions were then tested with laboratory experiments. Results showed that mutant E. gossypii EMS 30/1 produced the most riboflavin in spent motor oil (20.45 mg.l-1) while Candida famata ATCC 20850 produced the highest concentration (16.99 mg.l-1) in spent vegetable oil. With these strains and using the experimental designs from the fractional factorial experiments, supplemented spent motor and spent vegetable oils produced 66.27 mg.l-1 and 72.50 mg.l-1 riboflavin, respectively. The central composite optimization results showed that 0.18 g.l-1 and 0.45 g.l-1 K2HPO4 and 12.5 g.l-1 malt extract increased the production to 91.88 mg.l-1 and 78.68 mg.l-1 in spent vegetable oil and motor oil respectively. A point prediction from the response surface methodology was used to validate these and it was found that 103.59 mg.l-1 riboflavin was produced by mutant E. gossypii EMS 30/1 using 2.5 g.l-1 yeast extract, 0.5 g.l-1 peptone, 12.5 g.l-1 malt extract, 0.18 g.l-1 K2HPO4 and 0.3 g.l-1 MgSO4.7H2O. After optimizing K2HPO4 in a one-factor-at-a-time experiment, 82.75 mg.l-1 riboflavin was produced by C. famataon v SVO using 6.5 g.l-1 peptone, 12.5 g.l-1 malt extract 0.15 g.l-1 K2HPO4 and 1.75 g.l-1 MgSO4.7H2O. This is a 5.08 and 4.87 fold increase respectively when compared to spent oil prior to optimization. This shows that spent motor oil and mutant E. gossypii produces 103.59 mg.l-1 riboflavin while spent vegetable oil and C. famata produces 82.75 mg.l-1 riboflavin. Hence, E. gossypii can be used to generate riboflavin using spent motor oil and C. famata, using spent vegetable oil.

Submitted in fulfilment of the requirements of the Degree of Master of Technology: Biotechnology, Durban University of Technology, 2011. / Eremothecium gossypii has the ability to utilize vegetable oils as a carbon source to produce riboflavin. This organism has been known to produce as much as 40 000 times more riboflavin than it requires after genetic modification on simple sugars. Adaptation of this organism to various oil substrates for riboflavin production has been poorly investigated. The aim of this research was thus to investigate the production of riboflavin by Eremothecium gossypii, on various oils and to improve production by mutating the organism and optimising media components using Design of Experiments (DOE). Nine overproducing mutants were obtained after mutating with various concentrations of ethylmethane sulphonate (EMS), n-methyl-n‟-nitro-n-nitrosoguanidine (MNNG) and Ultraviolet light. Riboflavin overproducing mutants were screened on an itaconate-containing medium; the colonies appeared yellow instead of white in the case of the wild-type. The itaconate screening medium isolated mutants with an isocitrate lyase that was insensitive to feedback inhibition. Mutations performed using EMS increased the ability of E. gossypii to produce riboflavin by 611% (7-fold) compared to the wild-type. This was achieved with soybean oil as a carbon source and was better than the other five oils used. Using DOE, fractional factorial experiments were carried out to optimise media components for riboflavin production on soybean oil. The total riboflavin produced by E. gossypii mutant EMS30/1 increased from 59.30 mg l-1 on a standard O&K medium using soybean oil as a carbon source to 100.03 mg l-1 on a DOE improved O&K medium, a 69% increase. The final optimised growth medium was determined from a central composite design using response surface plots together with a mathematical point-prediction tool and consisted of 5.0 g l-1 peptone, 5.0 g l-1 malt extract, 5.1 g l-1 yeast extract, 0.64 g l-1 K2HPO4, 0.6 g l-1 MgSO4 and 20 g l-1 soybean oil. Fractional factorial and central composite media optimization designs increased riboflavin production by several fold over their iterations. There was an overall increase of 1099% (12-fold) in riboflavin production by the mutant grown in an optimized medium compared to the initial riboflavin produced by the wild-type.