Table 3 shows ANOVA for phytase activity units/gd s (Y) indicating
the F value of 2.671 which implied the model to be significant.
Model terms having values of Prob > F(0.149) are
considered highly significant. ANOVA indicated the R2 value
of 0.828 for response Y, which indicated that 82.8 % of data
variability could be explained by the model. A 91% correlation between observed and predicted results for phytase activity re-
flected the accuracy and applicability of the design for process
optimization (see Table 4).
Figs. 9–11 show the response surface plots of phytase production.
Obviously, phytase yield varied significantly upon
changing the glucose, peptone or tween 80 concentrations. It
indicated that the optimum values of each variable were identified
based on the hump in the three dimensional or from the
central point of the corresponding contour plot.
The statistical optimization resulted in about 9.8-fold
compared to the non-optimized medium. The statistical
between observed and predicted results for phytase activity re- flected the accuracy and applicability of the design for process optimization (see Table 4). Figs. 9–11 show the response surface plots of phytase production. Obviously, phytase yield varied significantly upon changing the glucose, peptone or tween 80 concentrations. It indicated that the optimum values of each variable were identified based on the hump in the three dimensional or from the central point of the corresponding contour plot. The statistical optimization resulted in about 9.8-fold compared to the non-optimized medium. The statistical optimization revealed an increase of phytase production by 5-fold by Bacillus subtilis US417 using wheat bran as a supporting material for solid state fermentation (Kammoun et al. 2012) 1.3-fold for Rhizomucor pusillus (Chadha et al. 2004), 1.8-fold for Mucor racemosus (Bogar et al., 2003a), 1.7-fold for Aspergillus ficuum (Bogar et al., 2003b), 1.75-fold for Pichia anomala in synthetic medium (Vohra and Satyanarayana 2002) and 5 fold for cane molasses medium (Kaur and Satyanarayana 2005). These observations clearly suggested that the nutritional and physical requirements of the microbes differ from one another, and therefore, need to be optimized for each strain.
4. Conclusion Waste materials cause a problem all over the world; they always need a new strategy to get rid of it with a benefit target. Millions tons of corn cob and corn bran are accumulated every year as byproducts of industrial work without considerable benefits. In our study, we open a new arena of research in the production of P. purpurogenum GE1 phytase by using corn cob and corn bran as substrates. The enzyme was produced under solid state fermentation (SSF) and the conditions for enzyme production were optimized by using Box–Behnken design. We suggest that our work will have great benefit in solving corn cob and corn bran waste problem. Also, our research introduced a low cost medium and very simple technique in phytase production which is considered as one of the most important enzymes. Acknowledgements This work was supported by the National Research Center, Chemistry of natural and microbial products department (Egypt).