In order to improve the thermostability of β- 1,3-1,4-glucanase, evolutionary molecular engineering was used to evolve the β-1,3-1,4-glucanase from Bacillus subtilis ZJF-1A5. The process involves random mutation by error-prone PCR and DNA shuffling followed by screening on the filter-based assay. Two mutants, EGsl and EGs2, were found to have four and five amino acid substitutions, respectively. These substitutions resulted in an increase in melting temperature from Tm=62.5℃ for the wild-type enzyme to Tm=65.5℃ for the mutant EGsl and 67.5℃ for the mutant EGs2. However, the two mutated enzymes had opposite approaches to produce reducing sugar from lichenin with either much higher （28%） for the former or much lower （21.6%） for the latter in comparison with their parental enzymes. The results demonstrate that directed evolution is an effective approach to improve the thermostability of a mesophilic enzyme.
A derivative ratio spectrophotometric method was used for the simultaneous determination of β-carotene and astaxanthin produced from Phaffia rhodozyma. Absorbencies of a series of the standard carotenoids in the range of 441 nm to 490 nm demonstrated that their absorptive spectra accorded with Beer’s law and that the additivity when the concentrations of β-carotene and astaxanthin and their mixture were within the range of 0 to 5 μg/ml, 0 to 6 μg/ml, and 0 to 6 μg/ml, respectively. When the wavelength interval (?λ) at 2 nm was selected to calculate the first derivative ratio spectra values, the first derivative amplitudes at 461 nm and 466 nm were suitable for quantitatively determining β-carotene and astaxanthin, respectively. Effect of divisor on derivative ratio spectra could be neglected; any concentration used as divisor in range of 1.0 to 4.0 μg/ml is ideal for calculating the derivative ratio spectra values of the two carotenoids. Calibration graphs were established for β-carotene within 0?6.0 μg/ml and for astaxanthin within 0?5.0 μg/ml with their corresponding regressive equations in: y=?0.0082x?0.0002 and y=0.0146x?0.0006, respectively. R-square values in excess of 0.999 indicated the good linearity of the calibration graphs. Sample recovery rates were found satisfactory (>99%) with relative standard deviations (RSD) of less than 5%. This method was suc- cessfully applied to simultaneous determination of β-carotene and astaxanthin in the laboratory-prepared mixtures and the extract from the Phaffia rhodozyma culture.
