Chitinase Expressed as an Inducible Trait in Pseudomonas aeruginosa Schröter P-15 79
ance (ANOVA). According to ANOVA (Table 7 ),
the model was significant with an F-test of a very
low probability [( P > F) < 0.0001]. The goodness
of fit for the model was expressed by coefficient
of determination R^2 and the value was 0.9769.
The value of R^2 indicated that the experimental
values were in agreement with the predicted val-
ues and suggested that the model is suitable and
practicable.
Model run showed that the predicted and
obtained response values were with minimum
alteration as shown in Fig. 1. The 3-D surface
and contour plots illustrate the response over a
region of interesting factor levels, the relation-
ship between the response and experimental lev-
els of each variable and the type of interactions
between the test variables in order to deduce the
optimal composition of the culture medium. But
at higher concentration of crude glycerol, the ef-
fect of different concentrations of peptone was
little on chitinase production. It is evident from
fig. 3 that at low concentration of colloidal chitin,
production of chitinase increased with decrease
in concentration of peptone and on increasing
crude glycerol. When colloidal chitin was high-
er, higher chitinase production was obtained at
lower values of both the variables. There was
maximum production of chitinase at low peptone
value and highest colloidal chitin (Fig. 4 ). We can
determine the optimal concentration of the medi-
um components from the data obtained from the
3-D surface and contour plots and the equations
obtained from the multiple regression analysis.
The model predicts that the chitinase pro-
duction (2.88 Uml−^1 ) is located at the actual
values: crude glycerol—47.70 ml/l, colloidal
chitin—38.75 g/l, and peptone—25.59 g/l. The
predicted values of the response obtained and
their corresponding concentration of medium
components vary accordingly. Thus, graphical
optimization of the overall desirability function
was performed to determine the best possible
combination for each response simultaneously.
To validate optimized media components for
chitinase activity, periodical production curve
with 0.2 % culture inoculation up to 120 h was
carried out and compared with both unoptimized
KBCG and KB in presence of 1.5 % of colloidal
chitin. It was found that optimal media compo-
nents were responsible for premature production
peak (72 h) with 3.75 Uml−1 which was 12 h ear-
lier as well as 46.4, 30.20, and 5 % higher thanFig. 3 Effects of peptone and crude glycerol