بهینه سازی میزان نیکوتین آمید و ریبوفلاوین در سولفور زدایی از دی بنزوتیوفن به روش آماری سطح پاسخ

Introduction

Sulfur is the third most plentiful element of crude oil after carbon and hydrogen (1). Consumption of sulfur-rich fossil fuels leads to releasing sulfur oxides and therefore creates severe air pollution and acid rain. The major part of petroleum sulfur compounds are organic compounds and one of the main problems of oil refining is its refinement from crude oil (2). Although the physical-chemical process can remove inorganic sulfur, organic sulfur compounds such as dibenzothiophene (DBT) remain in oil after this process (3) and these recalcitrant compounds can be removed by biodesulfurization (BDS). Due to specificity and cost-effectiveness, BDS can be an alternative or complementary process for hydrodesulfurization (HDS) technology (4). DBT is often used as a model of heterocyclic sulfuric compounds in biodesulfurization studies. Some of bacterial strains can remove the sulfur of DBT via 4S-pathway and because selective attacking C-S bounds, does not affect the combustion value of petroleum (5- 7). Four enzymes are involved in 4S-pathway: first DBT monooxygenase (Dsz C) oxidizes DBT to DBT-sulfone (DBTO2). In the second step, flavomonooxygenase (Dsz A) catalyze transformation of DBTO2 to 2-2'-hydroxyphenyl benzene sulfinate (HPBS) and in the final step, HPBS is desulfinated by Dsz B to produce 2-hydroxybiphenyl (2-HBP) as a final product (8).Three catabolic genes, dszA, B, C, are clustered in dsz operon. Dsz D, a flavinreductase, prepares FMNH₂ (reduced form of flavin mononucleotide) required for oxygenase reactions via oxidation of NADH (reduced form of nicotinamide-adenine dinucleotide) and is a chromosomal gene (9). Previous study showed that the addition of riboflavin and nicotinamide as precursor of FMNH₂ and NADH in culture medium, led to increasing the BDS activity of Rhodococcus erythropolis USTB-03 (10). In this study, the optimization of riboflavin and nicotinamde using Response Surface Methodology (RSM) was performed and their effects on BDS and growth of Rhodococcus erythropolis R1 were investigated. In ordinary optimization procedures, a parameter is changed and others are invariant. This procedure is time consuming and also the impact of a parameter on other parameters is not clear. Statistical optimization is an alternative approach that rapidly screens a set of factors at different levels and reflects the role of each factor and interaction between them. RSM is a mathematical-statistical technique for condition optimization based on the multivariate non-linear model and is useful to evaluate the affectingfactors (11).

Materials and methods

Chemicals

DBT was purchased from Merck. Riboflavin and nicotinamide were purchased from Sigma. 2-HBP was prepared from FlukaChemika Co. All other chemicals were analytical grade and commercially available.

Bacterial strain and growth condition

Rhodococcus erythropolis R1 (GU570564), a capable strain in the desulfurizing DBT to 2-HBP was previously isolated from oil-contaminated soil (12) and was cultured in basal salt medium (BSM) supplemented with 0.3 mM DBT as the sole sulfur source.

Cell cultivation was carried out in a 250 mL flask containing 50 mL of BSM medium at 30°C on a rotary shaker operated at 180 rpm (n-biotech, inc). The BSM contain following composition: Na₂HPO₄.7H₂O 8 g/l, KH₂PO₄ 4 g/l, NH₄Cl 2 g/l, MgCl₂ 0.2 g/l, FeCl₃ 0.001 g/l, CaCl₂ 0.001 g/l and glucose 15 g/l as carbon source.

The stock solution of riboflavin (10 mM) was prepared by adding 0.376 g of riboflavin in 100 mL of distilled water and for preparing stock solution of nicotinamide (1 M), 12 g of nicotinamide was dissolved in 100 mL of distilled water. Different amount of these stock solutions were filtered and inoculated to sterile medium culture.

Analytical methods

Cell density was measured by the absorbance at 600 nm (OD₆₀₀) (spectronic21D Milton Roy). Desulfurization activity was monitored using the Gibb’s reagent (2,6-dichloroquinone-4-chloroimide). Gibb’s reagent reacts with aromatic hydroxyl groups such as 2-HBP at a pH of 8.0 and forms a blue-colored complex that can be monitored by spectrophotometer at 610 nm. The Gibb’s assay was performed as follows: Microbial culture was centrifuged at 7000 rpm for 10 min and 1 mL of supernatant was placed in a clean test tube and its pH was adjusted to 8.0 by adding 200 µl NaHCO₃ 1M and then 20 µl of Gibb’s reagent (10 mM) was added. The solution was allowed to incubate for 30 min at room temperature to produce full-color development and then absorbance of the solution was measured at 610 nm (spectronic21D Milton Roy). The BDS activity (X_BDS) of the cells was determined as the percentage of desulfurization according to the following equation:

Where  is the initial concentration of DBT (mM) and  is the 2-HBP concentration (mM) after a certain time.

Results

Statistical analysis

The variable interaction can be simultaneously investigated by response surface model. Coded values of experimental variables are shown in Table 1.

Table 1- Coded values of experimental variables for growth rate and

A quadratic polynomial equation was established to recognize the relationship between 2-HBP production of growing cells and variables based on the experimental results of CCD (Table 2).

Table 2- Response surface Central Composite Design (CCD)

The model of coded units is calculated using:

Where  is the predicted response,  is the variable  is constant,  is the linear effect,  is the quadratic effect, and  is the interaction effect.

Biodesulfurization activity

For , the model of coded units after removing non significant parameters can be expressed as:

 = 2.4 – 0.12 X₁– 0.11 X₂ – 0.25 X₁² – 0.067 X₂²

Where  is the response value (%), X₁ is nicotinamide (mM) and X₂ is riboflavin (µM). Positive and negative sign before terms indicates synergistic and antagonistic effect respectively (13). The equation indicates a quadratic linear relationship between variables and . As can be seen in ANOVA Table (Table 3), the model F-value is 62.35. The greate model F value implies that the model is significant. There is only a 0.01% chance that a model F-value could occur due to noise. In addition, the value of Prob> F less than 0.05 indicates model terms are significant. In this case X₁, X₂, X₁² and X₂² are significant model terms. There is a 32.49% chance that a Lack of Fit F-value this large could occur due to noise. The Pred R-Squared of 0.8995 is in reasonable agreement with the Adj R-Squared of 0.9624. Adeq Precision measures the signal to noise ratio. The ratio greater than 4 is desirable. Our ratio of 22.082 indicates an adequate signal. Thus, this model could be used to navigate the design space satisfactorily. According to the present model, both factors are significant model terms.

Table 3- Analysis of Variance (ANOVA)

Growth: R² = 0.9569      R (adjusted) = 0.9262

: R² = 0.9780           R (adjusted) = 0.9624

Growth rate

In growth rate, the model of coded units after removing non significant parameters can be stated as:

 = 2.55 – 0.16 X₁– 0.083 X₂ – 0.061 X₁²

Where  is the response value (OD₆₀₀), X₁ is nicotinamide (mM) and X₂ is riboflavin (µM). The ANOVA for growth rate by resting cells is shown in Table 3. The Model F-value is 31.11 and the Prob> F is less than 0.05, both imply the model and model terms are significant (Table 3). In this case, X₁, X₂ and X₁² are significant model terms. The Lack of Fit F-value of 8.46 indicates that the Lack of Fit is significant. There is 3.32% chance that Lack of Fit F-value occurs due to noise. The Pred R-Squared of 0.7263 is in reasonable agreement the Adj R-Squared of 0.9262. The ratio greater than 4 is desirable. Our ratio of 18.769 indicates an adequate signal. Thus, this model could be used to navigate the design space satisfactorily. According to the present model, both factors are significant model terms.

Discussion and conclusion

The response surface and its contour plot at the base can represent the regression model developed to investigate the interaction between factors and specify the optimum level of each factor. The interaction of two independent factors can be shown by each response surface and a contour plot. The fitted surface and contour plots between nicotinamide and riboflavin are presented in Fig.1 for  and Fig. 2 for growth rate.

Fig. 1- Response surface (a) and contour plot (b) of biodesulfurization of DBT after the addition of cofactor precursors (C7 is the X_BDS).

Fig. 2- Response surface (a) and contour plot (b) of bacterial growth rate after the addition of cofactor precursors (C8 is the OD₆₀₀).