Extracellular cholesterol oxidase production by Streptomyces aegyptia, in vitro anticancer activities against rhabdomyosarcoma, breast cancer cell-lines and in vivo apoptosis

In recent years, microbial cholesterol oxidases have gained great attention due to its widespread use in medical applications for serum cholesterol determination. Streptomyces aegyptia strain NEAE-102 exhibited high level of extracellular cholesterol oxidase production using a minimum medium containing cholesterol as the sole source of carbon. Fifteen variables were screened using Plackett–Burman design for the enhanced cholesterol oxidase production. The most significant variables affecting enzyme production were further optimized by using the face-centered central composite design. The statistical optimization resulted in an overall 4.97-fold increase (15.631 UmL−1) in cholesterol oxidase production in the optimized medium as compared with the unoptimized medium before applying Plackett Burman design (3.1 UmL−1). The purified cholesterol oxidase was evaluated for its in vitro anticancer activities against five human cancer cell lines. The selectivity index values on rhabdomyosarcoma and breast cancer cell lines were 3.26 and 2.56; respectively. The in vivo anticancer activity of cholesterol oxidase was evaluated against Ehrlich solid tumor model. Compared with control mice, tumors growth was significantly inhibited in the mice injected with cholesterol oxidase alone, doxorubicin alone and cholesterol oxidase/doxorubicin combination by 60.97%, 72.99% and 97.04%; respectively. These results demonstrated that cholesterol oxidase can be used as a promising natural anticancer drug.


Results and Discussion
Statistical screening of factors influencing cholesterol oxidase production by Streptomyces aegyptia strain NEAE 102 using Plackett-Burman design. Firstly, Plackett-Burman statistical design was used to assess the impacts of carbon sources (glucose, cholesterol, starch), energy source (K 2 HPO 4 ), metals (NaCl, MgSO 4 .7H 2 O, FeSO 4 . 7H 2 O), nitrogen sources (yeast extract, peptone, (NH 4 ) 2 SO 4 ), in addition to the physical parameters (pH, medium volume, time of incubation, temperature, inoculum size) on cholesterol oxidase production by the selected strain. Secondly, a face centered central composite design was used for optimization of significant variables and studying their interactions on cholesterol oxidase production.
Compared with other strategies for designing a growth medium, the Plackett-Burman design is uncomplicated and quick method for screening a large number of variables in one experiment to evaluate the significant variables affecting the cultural requirements and the production of enzyme in fermentation broth 21,22 . It was carried out by performing 20 runs for identifying the key variables that increase cholesterol oxidase production by the selected strain. Table 1 illustrates the experimental design and the fifteen independent variables as well as levels of each variable used in the experimental design. Table 2 represents the effect of fifteen independent factors with coded levels on the cholesterol oxidase production using twenty-trials according to Plackett-Burman design. Results of different trials show a variation in findings from 0.837 to 7.861 UmL −1 . This change reflects the importance of medium optimization to achieve high enzyme production. The maximum cholesterol oxidase production (7.861 UmL −1 ) was achieved in the sixteenth run, while the minimum cholesterol oxidase production (0.837 UmL −1 ) was observed in the third run.
To determine the correlations between the independent factors and production of cholesterol oxidase, a multiple-regression mathematical model was used. Statistical analysis was achieved and summarized in Tables 3, 4. Table 3 and Fig. 1 illustrate the estimated effect of the tested parameters on the enzyme production. Main effect enables the estimation of the influence of each factor on the enzyme production. Both large positive or negative effects indicate that a variable has a large impact on the production, while the factor has little effect or considered as uneffective with value close to zero. From the main effect results, we can find that nine of the fifteen variables named cholesterol, yeast extract, (NH 4 ) 2 SO 4 , K 2 HPO 4 , NaCl, MgSO 4 , FeSO 4 , temperature and the volume of medium have a positive influence on the production of enzyme, where the other six variables namely glucose, peptone, inoculum size, starch, pH and incubation time negatively affect the cholesterol oxidase production. "Variables with positive impacts on the production of cholesterol oxidase have been used at high level, while the variables that have a negative effect are kept at low level for further optimization by face-centered central composite design".
The contribution percentage of each factor are presented in Table 3; pH, cholesterol and incubation time are the most contributing components with 18.64, 11.76, and 11.54%; respectively. As shown in Fig. 2, the Pareto chart offers an easy way to view the results obtained by Plackett-Burman design, it illustrates the significance order of the factors affecting cholesterol oxidase production. "It shows the absolute values of the effects, and draws a reference line on the chart. Any effect that extends past this reference line is potentially important". Pareto chart in design expert version 7.0 reproduce the relation between t-value (effect) vs. ranks. Among the 15 assigned variables, pH was the most significant variable affecting cholesterol oxidase production at 99.98% confidence followed by cholesterol at 99.89% confidence then incubation time at 99.88% confidence.
To validate the obtained results of the experimental design, analysis of variance (ANOVA) for cholesterol oxidase production was performed and confidence level, F-value, sum of square, P-value, t-value (t-Stat) and the mean square are presented in Table 4. The P-value is the probability which serves as a tool for testing the significance of the model and each variable. "A low P-value indicates a real or significant effect. The values of P < 0.05 indicate model terms are significant. The significance of each variable was determined by applying the Student's t-test". The P-value of 0.0018 and the model F-value of 34.43 indicate that this model is significant.
The ANOVA analysis displayed that, pH (p), with t-value of −13.286, P -value of 0.0002 and confidence level of 99.98% was found to be more significant variable, followed by cholesterol (A) (probability value of 0.0011, t-value 8.383, and confidence level 99.89%), then incubation time (M) (t-value −8.214, probability value of 0.0012 and confidence level 99.88%), the lower probability values indicate significant variables influencing the production of cholesterol oxidase. Furthermore, it was obvious that among the three factors, only cholesterol exerted a positive effect, whereas the other factors (pH and incubation time) exerted negative effects on cholesterol oxidase production, which means that the increase in cholesterol concentration and a decrease in pH and incubation time could exert positive impact on cholesterol oxidase production. Screened significant variables (cholesterol, starch, glucose, (NH 4 ) 2 SO 4 , K 2 HPO 4 , NaCl, MgSO 4 .7H 2 O, inoculum size, time of incubation and pH) strongly affects the cholesterol oxidase production, while temperature, medium volume, yeast extract, peptone and FeSO 4 .7H 2 O, of no significant influence on the enzyme production.
A good correlation between the experimental and predicted values was indicated by a higher correlation coefficient (R = 0.9961) value. "The coefficient of determination (R 2 ) value explains a measure of how much variability in the observed response values can be illustrated by the experimental factors. The value of R 2 is always between 0, 1. The closer the R 2 to 1, the stronger is the model and the better predicted response" 23 . The coefficient of determination value (R 2 = 0.9923) showed that 99.23% of the variations in the enzyme production can be demonstrated by the independent factors and only 0.77% of the variations are not demonstrated by these factors. Also, a very high adjusted coefficient of determination value (Adj. R 2 = 0.9635) indicates the highly significance of the model 24 . The "Pred R-Squared" value of 0.8079 is close to the "adj R-Squared" value of 0.9635. This elucidates a good correlation between the predicted and experimental values of cholesterol oxidase production. "Adeq Precision estimates the signal to noise ratio. A ratio greater than 4 is desirable". Ratio of 25.183 indicates an adequate signal. "The coefficient of variation% (C.V.%) is a measure of residual variation of the data relative to the size of the mean. Usually, the higher the value of C.V., the lower is the reliability of experiment". In this study, the C.V.% value (7.7419%) is very low which gives a greater precision of the performed experiments. "The predicted residual sum of squares (PRESS) is a measure of how well the model fits each point in the design. The smaller the PRESS statistics, the better the model fits the data points". The PRESS value is 9.3473. In this experiment, the model gives values of mean and standard deviation of 3.9491 and 0.3057; respectively.
A first order polynomial equation was applied to represent the optimum cholesterol oxidase production as a function of the independent factors. By ignoring the insignificant terms, the following equation of regression in terms of coded factors was obtained:  (Table 3) and the calculated t-values (Table 4), cholesterol concentration (A), incubation time (M) and pH (P) were chosen to further optimize by FCCD, as those variables represented the more significant effect on cholesterol oxidase production. Adequacy of the model. The normal probability plot (NPP) is given in Fig. 3A. "The normal probability plot of the residuals is an important diagnostic tool to reveal and demonstrate the systematic departures from the normality" 25 . NPP of internally studentized residuals illustrates the points near the diagonal line which means 3.798 0.001 Table 2. The effect of fifteen independent factors with coded levels on cholesterol oxidase activity using twentytrials according to "Plackett-Burman experimental design". "Two level design, each variable is tested at a low "−1" and high "+1" value. "  that the residuals are normally distributed and this model was well fitted with the experimental results. Box-Cox plot represents a potential best practice for selecting a better power transformation to further improve the model.     increase in cholesterol oxidase activity of 7.75 UmL −1 in comparison of the obtained activity before employing Plackett-Burman design (3.1 UmL −1 ). An enhanced cholesterol oxidase production was reported by using compounds like yeast extract, cholesterol 15 , potato starch, malt extract and peptone 20 as substrates. Cholesterol is largely metabolized by microorganisms as a source of energy and carbon 26 . Different microorganisms like Brevibacterium, Arthrobacter, Mycobacterium, Nocardia, Streptomyces and Corynebacterium have the capability of degrading cholesterol. The initial step of cholesterol degradation by microorganisms is the oxidation of the 3β-hydroxyl group by cholesterol oxidase. Consequently, cholesterol degrading microbes are commonly considered to be cholesterol oxidases producer 27 . Yehia et al. 28 reported that the decomposition of cholesterol by the tested bacterial isolates was affected by cholesterol concentration in the cultural medium. The highest percentage of cholesterol breakdown (80.2%) by the Enterococcus hirae was displayed when 1 g/L cholesterol was added and the highest cholesterol breakdown by Rhodococcus erythropolis 29 and Streptomyces fradiae 15 was accomplished at 2 g/L cholesterol.
The pH value of the cultural medium is extremely crucial for growth of microorganisms, their metabolic characteristics and also for the biosynthesis of metabolites. The hydrogen ion concentration can have a direct impact on the cell or may cause an indirect effect on the cell by changing the degree of dissociation of the constituents of the medium 30 . Yehia et al. 28 found that the change in the cultural medium pH significantly affect the maximal physiological performance of the cells, the transfer of different nutrients through the membrane of the cell and the breakdown of cholesterol. Yazdi et al. 15 reported that the best pH value for cholesterol breakdown by Streptomyces fradiae was 7.2, while Sojo et al. 29 found that pH 6.75 was the best for cholesterol breakdown by Rhodococcus erythropolis ATCC 25544. In addition, Yehia et al. 28 reported that the highest pH value needed for maximum decomposition of cholesterol and growth of Enterococcus hirae in liquid medium was pH 7.0. A maximal pH 8 was reported for a new alkaline Streptomyces species which isolated from the east African soda lakes 31 . Also, Solingen et al. 31 highlighted the effect of alkaline pH on the growth and adaptation of Streptomyces species.
The incubation period is an important factor that must be considered. Since the production of cholesterol decomposing enzyme could be subjected to alteration with the long incubation period. For example, the maximum decomposition of cholesterol by non-irradiated and Nd-YAG irradiated bacteria was detected at the end of the sixth day. The percentage of cholesterol decomposition, on the sixth day, amounted 82.8 in the case of irradiated Streptomyces fradiae as compared to 66.9% in the case of non-irradiated one. Generally, the decomposition of cholesterol, at any incubation time, by the irradiated Streptomyces fradiae significantly exceeded that recorded by non-irradiated bacterium. Several investigators indicated different incubation periods for different microorganisms. Abo-El-Khair 32 reported that the cholesterol decomposing activity of Mycobacterium fortuitum increased rapidly during the first three days of incubation followed by insignificant variations with extension of incubation up to 10 days. It can be mentioned here that the process of cholesterol decomposition by microorganisms occurs during 6 days for Streptomyces spp. 33 and 43 hr. for the most active strains of actinomycetes 34 . Watanabe et al. 2 reported that most Rhodococcus strains complete the cholesterol degradation during 3-7 days of incubation period. A maximum production obtained by Bacillus cereus strain KAVK4 at 32 hours of incubation period was 1.67 UmL −1 . Niwas et al. 35 reported that the period of incubation has taken an important role in cholesterol oxidase enzyme production by Streptomyces sp. The culture grew exponentially up to 56 h and then entered into a stationary phase that was continued for up to 96 h. In the exponential phase the enzyme production was initiated, at stationary phase of the growth the enzyme production reached to the maximum and then declined as a result of the consumption of carbon source and other nutrients. Streptomyces lavendulae 22 and Streptomyces parvus 17 have been also outlined for their optimal production of cholesterol oxidase enzyme at the stationary phase of growth.

Optimization of fermentation process using face-centered central composite design (FCCD).
Experiment of Plackett -Burman design showed that, cholesterol concentration (X 1 ), time of incubation (X 2 ) and the value of pH (X 3 ) influenced the cholesterol oxidase production effectively. So, FCCD was employed to determine the optimum levels of these variables which give maximum production of cholesterol oxidase and their interactions with each other. "Variables with positive impacts on the production of cholesterol oxidase have been used at high level, while the variables that have a negative effect are kept at low level for further optimization by FCCD". However, among the variables that negatively affect cholesterol oxidase production, only inoculum size was kept in all trials at its low level because it was associated with the composition of the growth medium and can't be deleted. Starch and glucose that have negative influences on the production of cholesterol oxidase were omitted from medium composition. On the other hand, peptone, yeast extract and FeSO 4 .7H 2 O, which had no significant effects on cholesterol oxidase production, also omitted from the subsequent experiments. The effect of cholesterol concentration (X 1 ), time of incubation (X 2 ) and the value of pH (X 3 ) and their interactions on cholesterol oxidase production were studied by FCCD at three levels (−1, 0, 1) using 20 runs and the results are presented in Table 5.
On the basis of experimentaly obtained data; the activity of cholesterol oxidase extended from 6.193 to 15.631 UmL −1 . The maximum enzyme activity was observed in the run number 18 (center point) with a value of 15.631 UmL −1 , where cholesterol concentration 3 g/L, incubation time 5 days and pH 6 were used. Whereas the minimum activity of cholesterol oxidase was obtained in the run number 11 with a value of 6.193 UmL −1 , where cholesterol concentration 2 g/L, incubation time 4 days and pH 5 were used. In Table 5 each experimental value of cholesterol oxidase activity was in good correlation with the model predictable value.
In the current study, the maximum enzyme activity (yield) of cholesterol oxidase of Streptomyces aegyptia NEAE-102 was 15.631 UmL −1 , which was higher than the activity of cholesterol oxidase enzyme of Streptomyces lavendulae (2 UmL −1 ) 36  Multiple regression analysis and ANOVA. The obtained data were subjected to regression analysis and analysis of variance (ANOVA) using the Design Expert 7.0 software, and the generated data are represented in Tables 6-8. The coefficient of determination (R 2 ) checks the goodness of fit of the model. In the present investigation, the obtained coefficient of determination (R 2 ) value (0.9895) ( Table 6) reflected that 98.95% of the variations in the enzyme production can be explained by the independent factors and only 1.05% of the variations are not explained by these factors. According to Chen et al. 43 , "a regression model is considered highly correlated when the R 2 value is higher than 0.9". The adjusted determination coefficient (Adj R 2 ) found to be 0.9801 which confirm the model significance. The values of the "Pred R-Squared and the "Adj R-Squared" were found to be 0.9494 and 0.9801, respectively, which elucidates a very good fit between the experimentally obtained results and the   theoretically predicted values by the model 44 and implied that the model is reliable for cholesterol oxidase production in the present study. Furthermore, Adeq Precision value was very high (29.4015) which indicates that the model can be used to navigate the design space. The lower value of coefficient of variation % (3.810) indicated a better precision and reliability of the experimental performance 45 . The obtained PRESS value is 9.971, the model gives values of mean and standard deviation of 11.920 and 0.454; respectively (Table 6). With respect to the coefficients (Table 6), interactions between two factors could appear as a synergistic effect (positive coefficient) or an antagonistic effect (negative coefficient). The positive coefficients for X 1 , X 2 , X 3 , X 1 X 3 ( Table 6) indicate that linear effect of X 1 , X 2 , X 3 and interaction effects for X 1 , X 3 increase production of cholesterol oxidase, whereas other negative coefficients indicate decrease in the production of cholesterol oxidase. Table 7 illustrates the analysis of variance (ANOVA) values for the quadratic regression model obtained from FCCD employed in the optimization of cholesterol oxidase production. The P-value was calculated to validate the significance of each coefficient and is also necessary for understanding the mutual interactions between the factors. The P-values are listed in Table 7. Values of "Prob > F" (P-values) less than 0.05 indicates the significance of model terms. Values greater than 0.05 indicate the model terms are not significant. The analysis of variance confirms that the model is highly significant as is evident from Fisher's F-test (104.99) and a very low P-value < 0.0001. It can be seen that the linear coefficients of cholesterol concentration, incubation time and pH, the interaction between cholesterol concentration, incubation time, interaction between cholesterol concentration, pH and interaction between incubation time, pH and quadratic effect of cholesterol concentration (X 1 ) are significant while quadratic effect of incubation time (X 2 ) and pH (X 3 ) is not significant (P-value < 0.05) ( Table 7). The probability values of the coefficient indicated that, among the studied three variables pH, the  interactions between incubation time and pH and quadratic effect of cholesterol concentration shows maximum P-value > 0.0001 demonstrating 99.99% of the model affected by these variables. As illustrated in Table 8, the statistics of the quadratic model summary showed the highest adjusted, R-squared and predicted R-squared of 0.9801, 0.9895 and 0.9494; respectively and lower standard deviation of 0.4541. The fit summary confirmed the adequacy and the high significance of the quadratic model with a very low P-value < 0.0001.
In order to determine the relationship between the production of cholesterol oxidase and the independent variables and to determine the optimal concentration of each component involved in cholesterol oxidase production mainly, concentration of cholesterol, time of incubation and value of pH, the equation of second-order polynomial was obtained to define the predicted cholesterol oxidase production (Y) in terms of the independent variables: Where Y is the predicted activity of cholesterol oxidase and X 1 , X 2 and X 3 are concentration of cholesterol, time of incubation and pH; respectively.
Three dimensional plots. The three-dimensional surface graphs and its corresponding contour plots explained the interaction of different studied factors and optimal levels of each factor involved in cholesterol oxidase production. Response plotting curves represented the effect of one fixed variable at its optimum level when the other two factors are varying (Fig. 4A-C). Figure 4A illustrates cholesterol oxidase activity as affected by the concentration of cholesterol (X 1 ) and incubation time (X 2 ) by keeping the initial pH (X 3 ) at the best value. It showed that, when cholesterol concentration increases, production of cholesterol oxidase gradually increases, but increasing cholesterol concentration higher than 3.5 g/L can lead to reduction in cholesterol oxidase production. Also, increasing of incubation time increases the enzyme production and increasing of incubation time above 5 days can lead to reduction in cholesterol oxidase production. Figure 4B showed the effect of cholesterol concentration (X 1 ), value of initial pH (X 3 ) by keeping incubation time (X 2 ) at optimal value on cholesterol oxidase production. In this experiment, moderate values of both cholesterol concentration and initial pH yielded the maximum activity of cholesterol oxidase, so any further increase will result in a gradual decrease in the enzyme production. Furthermore, significant interactions between these variables have greatly helped to increase the activity of cholesterol oxidase. Figure   Cholesterol oxidase has the ability to convert membrane cholesterol to cholestenone and block the formation of lipid rafts 52 . "Cholesterol oxidase differs from other cholesterol-depleting agents, in disrupting lipid rafts by displacing cholesterol with cholestenone. Cholesterol oxidase does not attack all cell types because it's action depend on the microenvironment of membrane cholesterol, like the composition of phospholipid, cholesterol content and ionic strength. So far, despite the fact that cholesterol oxidase has been well known as a destroyer for lipid rafts, few investigations have been conducted to directly determine the effect of cholesterol oxidase on signal transduction in cancer" 13 . Liu et al. 13 reported that cholesterol oxidase from Bordetella spp. made the cells of lung cancer to undergo irreversible apoptosis both in vitro and in vivo. Table 9 shows the selectivity index of cholesterol oxidase or doxorubicin tested against the cancer cell lines. Treatments of cancer cell lines with cholesterol oxidase (23 UmL −1 ) gave selectivity index ranged from 1.26 to 3.26 which is higher than that of the standard anticancer drug doxorubicin which ranged from 0.73 to 1.45 (Table 9), demonstrating superiority of cholesterol oxidase over the clinically used anticancer doxorubicin and that the cholesterol oxidase has therapeutic potential. The selectivity index of the purified cholesterol oxidase on RD cell line reached 3.26. This value means that the cholesterol oxidase is toxic to the cancer cell line by more than thrice in comparison to the normal cell line. "The selectivity index considered interesting if the value is greater than 3 as reported by Bézivin et al. 53 . On the other hand, the selectivity index of the purified cholesterol oxidase on HepG-2, HCT-116, MCF-7 and Hela cell lines reached 1.94, 1.47, 2.56 and 1.26; respectively. Selectivity index value greater than or equal to 2 is an interesting as reported by Suffness and Pezzuto 54 ". These results demonstrated that cholesterol oxidase is a promising anticancer drug. In vivo effect of cholesterol oxidase on solid tumor growth inhibition. The use of natural products to treat or prevent the cancer is currently receiving considerable attention 55 . The traditional application of individual drugs fails to treat cancer adequately. Therefore, synergistic combinations of drugs are a promising means of enhancing effectiveness. However, to assess in vivo the role of cholesterol oxidase as an effective inducer of apoptosis, the effect of cholesterol oxidase and cholesterol oxidase/ doxorubicin (Dox) in synergistic combination treatment on growth and apoptosis of Ehrlich ascites carcinoma (EAC) solid tumor has been analyzed (Figs 6-8, Table 10). The average tumor volume in the EAC control mice increased from 60.2 mm 3 to 962.1 mm 3 after 20 days treatment. Compared with EAC control mice, tumors growth was significantly inhibited in the EAC bearing mice injected with cholesterol oxidase and Dox-treated groups (Fig. 6) by 60.97% and 72.99%; respectively. Moreover, cholesterol oxidase /Dox combination treatment showed a significant tumor growth inhibition (by 97.04%) as compared with both Dox-treated and cholesterol oxidase-treated mice. The quantified weight of tumor lumps in mice injected with cholesterol oxidase, Dox and the combination of cholesterol oxidase/Dox was found to be significantly much smaller compared to EAC control mice. On the other hand, weight of tumor lumps in mice injected with the combination therapy was significantly smaller (0.1 ± 0.01) as compared with the mice injected with Dox (1.7 ± 0.52) or the mice injected with cholesterol oxidase (2.3 ± 0.16) (Fig. 7, Table 10).
Our results showed that the administration of cholesterol oxidase alone significantly reduced the tumor volume and weight compared to control group. The anti-tumor effect of cholesterol oxidase was near to the therapeutic effect caused by treatment with Dox. The results showed that a significant reduction in tumor in case of the combination of cholesterol oxidase/Dox treatments compared to each treatment alone. These results collectively showed that the combination therapy was able to inhibit tumors growth much more effectively than any of the cholesterol oxidase or Dox alone in vivo. Liu et al. 13 reported that "cholesterol oxidase isolated from Bordetella species oxidizes membrane cholesterol to 4-cholesten-3-one with the reduction of O 2 to H 2 O 2 , which results in a decrease of cholesterol content and increase of reactive oxygen species (ROS) levels which leads to cell apoptosis by inactivation of the protein kinase B and extracellular signal-regulated kinase 1/2 pathway as well as activation of caspase-3. Cholesterol oxidase also oxidizes exogenous cholesterol to 4-cholesten-3-one and H 2 O 2 . Therefore, cholesterol addition is not able to prevent the cholesterol oxidase induced apoptosis. These findings suggest that cholesterol oxidase may be a promising therapeutic agent for cancer by targeting membrane cholesterol" 13 .
Histopathological examination of the tumor sections stained with hematoxylin and eosin (Fig. 8) showed the cords of malignant cells, invasion of cellular infiltration and large cancer cells in between muscle fibers in mice that carry untreated tumor (control). Treatment of EAC bearing mice with cholesterol oxidase or doxorubicin showed less tumor invasion, cellular infiltrates, markedly increased apoptotic bodies and moderate mitotic rate. Treatment of EAC bearing mice with both cholesterol oxidase and doxorubicin showed a significant protective synergistic effect on the histopathological architecture.

Materials and Methods
Microorganisms and cultural conditions. The Streptomyces sp. that used in the present study is a local isolate formerly identified and published as novel species; Streptomyces aegyptia NEAE-102 56 . This isolate was preserved on starch-nitrate agar medium slopes 57 . Slopes were incubated at 30 °C for a time of 7 days. The isolate was stored as spore suspension in glycerol (20%, v/v) at −20 °C for the subsequent studies.
Qualitative screening of actinomycetes isolates for cholesterol oxidase production potentialities using colony staining method. The potentiality of actinomycetes for cholesterol oxidase production was screened on agar plate medium which containing the cholesterol substrate as the sole carbon source    according to the method of El-Naggar et al. 46 . This experimental medium contained/L: "2 g cholesterol, 2 g KNO 3 , 1 g K 2 HPO 4 , 0.5 g MgSO 4 .7H 2 O, 0.5 g NaCl, 3 g CaCO 3 , 0.01 g FeSO 4 .7H 2 O, 20 g agar and distilled water up to 1 L". To confirm cholesterol oxidase producing strain, colony staining method was carried out on the grown colonies. Discs of filter papers were dipped in 100 mM potassium phosphate buffer (pH 7.0) containing 0.5% cholesterol; 3000 U/L horseradish peroxidase; 1.7% 4-aminoantipyrine and 6% phenol. "Thereafter, soaked discs were localized on the grown colonies and the plates were incubated at room temperature for 24 h. Cholesterol oxidase activity was assessed by development of pink color surrounding the tested colonies because of the quinoneimine dye formation" 58 .
Inoculum preparation. The  Production of cholesterol oxidase in submerged fermentation. Erlenmeyer flasks (250 mL) containing 100 mL of fermentation medium were inoculated with previously prepared inoculum and incubated at 30-37 °C on a rotatory shaker incubator (150 rpm). After the incubation time, the mycelium was centrifuged at 5000 × g for 15 min (at 4 °C). The obtained cell free supernatant after centrifugation was used as crude enzyme for further investigations.
Cholesterol oxidase activity assessment. Cholesterol oxidase activity was determined spectrophotometry by the modified method of Sasaki et al. 59 which based on generation of hydrogen peroxide during the oxidation reaction of cholesterol. The enzyme reaction was composed of: 0.1 mL crude enzyme, 1 mL cholesterol (3 μM) as the reaction substrate prepared in Triton X-100 (1%), 21 μM of phenol, 1.2 μM 4-aminoantipyrine, 20 U of horseradish peroxidase, 300 μM of potassium phosphate buffer (pH 7) in final reaction volume of 3 mL. The enzyme reaction was incubated with shaking for 10 min at 37 °C. To terminate the reaction, the assay mixture was boiled for 3 min. The developed pink color was measured spectrophotometry at 500 nm. "One unit of the enzymatic activity (U) was defined as the amount of enzyme required to form 1 μmol of H 2 O 2 /min. at 37 °C".

Screening of main factors affecting the production of cholesterol oxidase by Plackett-Burman
design. The Plackett-Burman statistical design was used to screen and evaluate the influence of significant constituents of medium and the environmental conditions with respect to their main effect on cholesterol oxidase production. In the current work, fifteen independent (assigned) factors were screened with four un assigned variables (commonly referred as dummy variables D 1 , D 2 , D 3 and D 4 ). Based on the Plackett-Burman factorial design, each variable was examined in two levels +1 for the high level and −1 for the low level 60 where, Y is the activity of cholesterol oxidase, β 0 is the coefficient of the model, β i is the linear coefficient and X i is the independent factors levels.
Optimization of cholesterol oxidase production by response surface methodology. Face centered central composite design (FCCD) was used in the next step to determine the optimum levels of significant factors for the production of cholesterol oxidase. The face centered central composite design is a statistical experimental design where each factor has three levels (−1, 0, 1) and 6 runs at center point resulting in a total of 20 trials. The second-order polynomial equation was used to calculate the relationship between the independent factors and the response. Considering all the linear terms, square terms and by linear interaction terms, the quadratic regression model can be illustrated as: where "Y is the predicted value of cholesterol oxidase production, β 0 is the regression coefficients, β i is the linear coefficient, β ij is the interaction coefficients, β ii is the quadratic coefficients and X i is independent factors levels". Assessment of in vitro cytotoxicity and anticancer activities using microculture tetrazolium assay (MTT assay). The anticancer activities of Streptomyces aegyptia NEAE-102 purified cholesterol oxidase were measured in vitro on both cancerous (rhabdomyosarcoma (RD), breast cancer (MCF-7), hepatocellular carcinoma (HepG −2), cervical epithelioid carcinoma (Hela), colon carcinoma (HCT-116) and non-cancerous cell line (human lung fibroblast, WI-38). "Which obtained from ATCC via holding company for biological products and vaccines (VACSERA), Cairo, Egypt". The cell lines mentioned above were used to determine the inhibitory effects of cholesterol oxidase on cell growth using standard 3-(4, 5 dimethythiazol-2-yl)-2, 5-diphenyl tetrazolium bromide (MTT) assay 63  Ehrlich solid tumor model. Mice were injected subcutaneously into the right hind limb (thigh) of all animals, each with 5 × 10 5 cells, to form solid tumor. Five days post-tumor inoculation when the solid tumor volumes reached approximately 50-100 mm 3 (day 0), the mice were randomly divided into four groups comprising five animals in each group. Group I served as tumor control injected with EAC; group II represents EAC bearing mice injected in tumor with cholesterol oxidase alone (3 U, q 3 days); group III represents EAC bearing mice injected with the anti-cancer doxorubicin (Dox) alone (2 mg kg 1 day); group IV represents EAC bearing mice injected with cholesterol oxidase in combination with Dox. Tumor volumes were measured from the fifth day of tumor induction and before the beginning of the treatment (day 0) and were carried out every five days for a period of 20 days. "Volume of solid tumor was monitored by vernier caliper and was calculated using the following formula: V = (L × S 2 ) × 0.5 65 , where (L) is the longest diameter of tumor and (S) is its shortest perpendicular diameter. Anti-tumor effectiveness was evaluated by the estimation of the average change of tumor volume in the treatment group (ΔT) and the average change of tumor volume in the control EAC bearing mice (ΔC). The degree of tumor growth was calculated as ΔT/ΔC × 100 which was then subtracted from 100% to estimate the percentage (%) of tumor growth inhibition" 66 . On the day 21 of treatment process (at the end of the experiment), mice were sacrificed and tumors lumps were removed, weighed and preserved for further histopathological analysis in buffered formalin solution. Five-micrometer sections were stained with hematoxylin and eosin. The slides were examined for histopathological changes using light microscopy.

Conclusion
The present study involved the use of statistical experimental designs to optimize nutritional and environmental variables for production of cholesterol oxidase from Streptomyces aegyptia strain NEAE-102. Three variables which are cholesterol concentration, incubation time and pH were identified by the experiment of Plackett-Burman design as significant for cholesterol oxidase production. These variables were further optimized using a face centered central composite design. The maximum level of cholesterol oxidase activity (15.631 UmL −1 ) was produced when the fermentation medium variables were set as follows: cholesterol concentration 3 g/L, incubation time 5 days, pH 6. The methodology as a whole was proved to be adequate for optimization of medium components for obtaining a therapeutically valuable product as cholesterol oxidase. Treatments of cancer cell lines with cholesterol oxidase (23 UmL −1 ) gave selectivity index ranged from 1.26 to 3.26 which is higher than that of the standard anticancer drug doxorubicin which ranged from 0.73 to 1.45, demonstrating superiority of cholesterol oxidase over the clinically used anticancer doxorubicin and that the cholesterol oxidase can be used as potential natural anticancer. In vivo study, the results showed that the combination therapy was able to inhibit tumors growth much more effectively than either cholesterol oxidase or Dox treated groups alone.