Application of Response Surface Methodology for Optimization of Siderophore Production

Iron is the fourth most abundant element in the earth’s crust, vital for growth of living organisms as it acts as cofactor for enzymes involved in various metabolic processes (Saha et al., 2016). Iron is an important bioactive metal indispensable for the growth and metabolism of bacteria. It is usually abundant in the environment, particularly in clay soils. The major roles of iron in plants and animals include the biosynthesis of chlorophyll, redox reactions in ATP, ribonucleotide synthesis, formation of heme, cell cycle regulation and detoxification (Silva-Stenico et al., 2005). Despite being most abundant element in earth’s crust, the availability of iron is limited due to very low solubility of the dominant ferric iron (Fe³⁺) in soil and become unavailable to plants as a micronutrient (Troeh et al., 2005). Some bacteria have the capability to produce low molecular weight (500-1000Da) metal chelating compound including iron, called as siderophore (Gupta and Gopal, 2008).
        
Siderophores chelate iron from mineral phases by formation of soluble Fe³⁺ complexes that can be taken up by energy dependent membrane transport mechanism and make it available to plants or bacterial cells (Ali and Vidhale, 2013). The term Siderophore was coined by Lankford (1973) which in Greek means “iron carrier” (Ahmed and Holmström, 2014). Siderophores are low molecular weight, non-ribosomal peptides, secreted under low iron stress conditions and capture iron from the environment (Budzikiewicz, 1993). These are compounds with small peptidic molecules having side chains and functional groups which have high-affinity ligand to bind ferric ions and transport them through the cell membrane (Niehus et al., 2017). In nature, more than 500 types of siderophores are studied; of which 270 have been structurally characterized (Boukhalfa et al., 2003). Various species of bacteria belonging to genus Escherichia, Pseudomonas, Azotobacter, Bacillus, Rhizobium, Salmonella, Klebsiella, Vibrio, Aeromonas, Aerobacter, Enterobacter, Yersinia and Mycobacterium are known to produce siderophores. Besides bacteria, several common species of fungi e.g., Penicillium, Mucor, Rhizopus,Saccharomyces, actinomycetes e.g., Nocardia, Streptomyces and algae e.g., Anabaena are also known to produce siderophores (Kannahi and Senbagam, 2014). Though, the primary application of siderophore is to provide soluble iron to microbes for its growth. They also play various roles in fields such as agriculture, bioremediation, biosensor and medicine. Siderophore producers are also play role in induction of systemic resistance to biocontrol (Akhtar and Ali, 2011). Siderophores are as diverse as the microorganisms which produce them. This diversity is thought to be due to evolutionary pressure on bacteria that facilitated the development of siderophores which could not be utilized by other organisms (Lee et al., 2012).
        
In the present study rhizosphere soil samples were screened for isolation of siderophore producing bacteria. A statistical technique of CCD had been selected for optimization of siderophore production. In this study, the effect of pH, incubation temperature, time and succinic acid conc. for maximizing siderophore production had been evaluated.

Collection of soil samples and enrichment
 
In the present study, rhizospheric soil of sugarcane, mango and capsicum were collected from Baramati Pune, Maharashtra, India (18.1792°N, 74.6078°E). The intact plant with root was dug out carefully. Then 250 gm of clumps of soil tightly bound to the roots were collected from all the sites and carefully stored in sterile polyethylene bag and was used for the isolation.
        
Standard serial dilution method was used for isolation. Soil samples were air dried to remove the excess moisture. One gm of each soil sample was then suspended in 9 ml sterile distilled water followed by transfer of one ml solution from each tube sequentially in next tube, a dilution range of 10^-¹ to 10^-6 was prepared. 0.1 mL of sample from 10^-6 dilution was inoculated in sterile minimal and Ashby’s broth which were deprived of carbon source and kept for enrichment for 72 hrs on rotary shaker. The loopfull sample from these broths was streaked on sterile minimal medium and sterile Ashby’s agar plates and incubated for 24 hrs at 37°C. The isolated colonies were characterized based on their micro and macroscopic traits. All the isolates were further screened for siderophore production by qualitative and quantitative assays.
 

 
Siderophore detection by qualitative and quantitative methods
 
The assay is based on ability of siderophore to bind to ferric iron with high affinity. The agar contains CAS dye which, when complexed with Fe³⁺, is blue in color. If the inoculated bacterial culture produces siderophores, ferric iron is stripped from the dye, causing the colour change of the media from blue to orange or yellow. Therefore, colour change at the site of bacterial growth is the indication of the presence of siderophores. Further siderophore production was qualitatively assayed by CAS plate assay (Schwyn and Neilands, 1987). This assay is a sensitive and widely used approach for detecting siderophore production. Freshly grown bacterial isolate was streaked on King’s B agar plate in which CAS solution was added and the plate was incubated at 30±2°C for 24-48 hrs (Neilands, 1987: Kumar et al., 2017). After incubation, siderophore production was confirmed by the presence of orange colour zone around the colony on CAS agar plates.
        
Quantitative detection of siderophore was carried out as per protocol given by Payne (Payne, 1994). In the CAS-shuttle assay, the isolates were grown in succinate medium (Meyer and Abdallah, 1978) and incubated for 24-48 hrs at 30±2°C. After incubation five mL of culture containing medium was centrifuged at 10,000 rpm for 15 mins at 4°C and the cell-free supernatant was tested for the presence of siderophore by using CAS test (Schwyn and Neilands, 1987). In this method, one ml of the culture supernatant was mixed with one ml of CAS reagent. Absorbance was measured at 630 nm against a reference consisting of one ml of uninoculated broth and one ml of CAS reagent. Siderophore content in the aliquot was calculated by using the following formula and expressed as percent siderophore unit (SU), which is defined as the percent (v/v) of siderophore present in the given sample (Shaikh et al., 2016).
   
Where,
Ar is absorbance of the reference at 630 nm (CAS reagent) and as is absorbance of the sample at 630 nm.
 


Characterization of the selected isolate
 
Isolate showing efficient siderophore production was further characterized by morphological, biochemical and molecular methods. Colony characters of the isolate were recorded such as size, shape, color, margin, elevation, opacity and consistency. The Gram nature of the isolate was also determined by using standard Gram staining procedure (Shen and Zhang, 2017). Motility was determined by using Hanging drop technique (Bisen, 2014). The selected isolate was further characterized on the basis of biochemical tests described in Bergey’s manual of determinative bacteriology. On the basis of morphological characters various biochemical tests were conducted like amylase, gelatinase, catalase, citrate utilization, indole, Voges Proskauer, methyl red, nitrate reduction, H₂S production and sugar fermentation test (glucose, mannitol, lactose, sucrose, maltose and xylose). The isolate was further identified upto species level by 16S ribosomal RNA gene sequencing (Panda et al., 2017).
 
Optimization of siderophore production by CCD
 
In the first phase of siderophore optimization, significant variable(s) or factor(s) were determined to optimize the siderophore production. In the next phase, RSM through CCD was used to find the optimum concentration of the selected variable. The effect of selected variables on the responses was analyzed to maximize the siderophore production (Murugappan et al., 2012). CCD was used to study the interaction between the significant components and also to determine their optimum levels. It involved steps such as procedures to find the optimum region, the responses in the optimum region of variables, estimation of the optimal conditions and verification of the data (Tanyildizi et al., 2005).
        
A set of 31 runs for four variables (A to D) at high and low levels were constructed by CCD using MINITAB software (Table 1).
 

        
The concentrations of the four factors i.e., pH (6-8), temperature (22-34°C), incubation time (24-96 hrs.) and succinic acid conc. (0.28-0.52%) were optimized. The frequency of high and low values of each variable was maintained according to the rules of CCD. These parameters were optimized for siderophore production using succinate media. Siderophore produced by the isolate in different experimental setup was quantified after incubation.
        
Statistical analysis of the yield and interpretation of the data was carried out for the RSM experimental design using Minitab statistical software. The analysis of variance (ANOVA) was used to value the effects and determine regression coefficients of model. The response surface plots were adopted to represent the effect of the independent variables on siderophore production. These response curves were then used to predict the optimum level of the factors (Kwak et al., 2006). Minitab software was used throughout the study for designing the experiments of RSM using CCD, regression and graphical analysis.

Collection of soil samples and enrichment of the isolates
 
The different rhizospheric soil samples were successfully collected from the different regions of Baramati. The soil samples collected were found to be the rich source of microbial diversity. Total eight isolates were obtained and they were further screened on the basis of siderophore production. Rhizospheric soil samples are the rich source of microorganisms having the ability to produce variety of biocompounds (Pahari and Misra, 2017).
 
Estimation of siderophore production by qualitative and quantitative assay
 
Yellow coloration around the test colony was observed which indicated positive result for qualitative test.
        
Quantitative estimation of siderophore production was done in terms of percent siderophore units (% SU). Maximum siderophore production was 98%.
        
Pattan et al., (2017) also used qualitative and quantitative methods for the estimation of siderophore production.
        
Qualitative estimation by CAS assay and quantitative by CAS shuttle assay are popularly used to estimate siderophore production. Similar results were reported by Pahari and Mishra, (2017). Raval and Desai, (2015) reported that the use of the CAS assay as a comprehensive, exceptionally responsive and most convenient. Their study showed that 37% of bacteria were positive and mainly belonged to Enterobacter, Pseudomonas and Bacillus, showed presence of orange halo around the colonies. For quantitative estimation, the isolates were grown in iron-deficient succinate medium and CAS assay was employed and OD value was measured at 630 nm. Kumar et al., 2017 assayed siderophore production qualitatively by using CAS assay. Their results showed positive for the strains Bacillus thuringiensis VIT VK5 and Enterobacter soli VIT VK6.
        
Shaikh et al., (2016) also used similar approach for the quantitative estimation of siderophore production.They also used CAS shuttle assay for quantitative estimation of siderophore using succinate medium and found that P. aeruginosa RZS9 produced 63.38% SU. Marathe et al., (2015) used same method for the quantitative estimation of siderophore produced by Pseudomonas which was determined by CAS-shuttle assay. The isolated strain of Pseudomonas sp. showed 71% siderophore production. Mokracka et al., (2004) also reported efficient siderophore production using Enterobacter spp.
 
Characterization of the efficient siderophore producing isolate
 
Further the bacterial strain having the ability to produce siderophore was successfully characterized upto species level by using morphological, biochemical and molecular methods. The colony characters of the selected isolate were determined and were found to be Gram negative, rod shaped, non-motile and non-endospore forming. The isolate showed positive results for Voges-Proskauer, citrate, starch hydrolysis, catalase and nitrate reduction tests. It showed negative methyl red, gelatinase, oxidase and H₂S production and was able to ferment glucose with acid and gas production. On the basis of results obtained of biochemical tests by using Bergey’s manual of determinative bacteriology, the organism was identified upto genus level as Enterobacter sp.The isolate was identified as Enterobacter hormaechei subsp. oharae DSM 16687(T) by using 16S r RNA gene analysis technique. The 16S rRNA FASTA analysis showed 99.12% similarity with the strain Enterobacter hormaechei (Table 2).
 

 

        
The Neighbor-Joining method was used to inferred evolutionary history (Saitou and Nei, 1987). The optimal tree with the sum of branch length 0.46613182 is shown. Associated taxa clustered together in the bootstrap test (1000 replicates) are shown next to the branches (Felsenstein, 1985).The evolutionary distances used to infer the phylogenetic trees, with branch lengths in the same units. The Tajima-Nei method was used to compute the evolutionary distances (Kimura, 1980) and is in the units of the number of base substitutions per site. A gamma distribution (shape parameter = 1) was used to modeled the rate variation among sites. The positions showing upto 95% coverage of site were excluded. MEGA6 was used to conduct evolutionary analysis (Tamura et al., 2013).
 
Optimization of siderophore production and statistical analysis of the CCD (by RSM and ANOVA)
 
The CCD under the RSM was employed in order to illustrate the nature of the response surface in the experimental region and elucidate the optimal conditions of the most significant independent variables. The average siderophore production given in Table 2 was subjected to multiple linear regression analysis. The effect of pH, time, temperature and succinic acid on siderophore production was described in the form of following equation as
 
For siderophore production, 49.55% of variability in the response could be explained by the model (Table 3). The closer the R² value to unity, the better the empirical model fits the actual data. The model explains 50.45% variability in the observed response value. Probably, 50.45% of the total variations would be due to other factors which were excluded in the model. However, the R² (pred) of 29.30% which is less than R² indicates that the model is overfit. In this study, the adjusted R² (41.79%) was close to the R² (49.55%) value. The higher the adjusted R² implies better the model. The relationships between the response and the predictors, pH (P value 0.003) and time (P value 0.002) are significant. The relationship between the response, siderophore production and the predictor, temperature (P value 0.128) and succinic acid (P value 0.867) is not statistically significant because their p-value is higher than the α-level. A commonly used α-level is 0.05.
 

          
The RSM model signifies better fit to the experimental data when the f value was large and the p-value is less than 0.05. Based on the above discussion, the high f and low p values with 6.83 and 0.001 respectively indicates that the regression model found in this study was significant (Table 4). The test for lack of fit was also calculated by Minitab software. Lack of fit explains the variation in the data around the fitted model. Table 4 shows the results of the lack of fit and it was found that the f and p values for the lack of fit were 9.25 and 0.006, respectively. Besides, the absence of any lack of fit (p>0.05) also strengthened the reliability of the models. Thus, it exhibits that the model was fitted well to the experimental data.
 

        
In the normal probability plot the data doesn’t fit a straight line and this verified that the residuals in the data were not normally distributed. In the normal probability plot of the effects, that are farther from 0 are statistically significant (Chantarangsi et al., 2016). The color and shape of the points differ between statistically significant and statistically insignificant effects. The normal probability plot of the effects displays negative effects on the left side of the graph and positive effects on the right side of the graph. On this plot, the main effects for the factors A and C are statistically significant at the 0.05 level because their p-values are less than the of 0.05. These points have a different color and shape from the points for the insignificant effects shown in Fig 4a.


        
A standardized Pareto chart (Fig 4b) consists of bars with a length proportional to the absolute value of the estimated effects, divided by the standard error. The bars are exhibited in the order of the size of the effects, with the largest effect on top. The Pareto chart illustrates the order of significance of the variables affecting siderophore production at p≤0.05. In this Pareto chart, the bars that represent the factors C, A, CC and AA cross the reference line that is at 2.120. The variables namely time and pH were found to be statistically significant.
        
Surface plot is a plot that shows the 3D model of contour plot. A contour plot gives a 2D view of all the points connected to produce contour lines of constant responses.
        
The response surface plot shows effects of time and succinic acid on siderophore production (Fig 5a). At less incubation time and succinic acid concentration, the yield of siderophore was low. Significant improvement in production could be obtained by increasing time and succinic acid concentration when the incubation time was set at high level i.e. 60 hours and succinic acid concentration at 0.40%, the siderophore production reached a maximum.
 

        
The Fig 5b gives the relationship between temperature and succinic acid on final yield. Maximum production was achieved at temperature 28^oC and succinic acid 0.4%.
        
The 3D response surface plot (Fig 6a) represents a rising ridge surface. It shows that higher siderophore yield was obtained with high time (60 hours) and low temperature (28^oC), within the chosen experimental range while Fig 6b gives effect of pH (7) and succinic acid (0.4%) on final yield of siderophore. Similarly the Fig 7a and 7b represents the relationship between pH-time and pH-temperature on siderophore yield respectively. According to Mouafia et al., (2016) the CCD and RSM allowed the creation of a polynomial model for the optimum production of biosurfactant.
 

 

The study shows that collected rhizospheric soil samples are the thriving source of potent microorganisms able to produce variety of ion chelating compounds. Enrichment increases the number of desired organisms from sample upto detectable levels. For selective enrichment Minimal and Ashby’s broth were used. It was found that enrichment method is very effective for isolation of indigenous rhizobacteria.
        
CAS-HDTMA complex has high affinity towards ferric ion resulting in dark blue color. When the siderophore is added, the siderophore binds with the ferric iron and iron-ligand complex is formed and release of the free dye is accompanied by a color change. Instant decolorization of CAS reagent from blue to orange red was observed with three cultures and it was confirmed by qualitative CAS test. All three isolates were Pseudomonas sp. which produced 71%, 72.33%, 33% units of siderophores in succinate medium, respectively.
        
Morphological and biochemical tests can be readily used for the identification of organisms upto genus level by using Bergey’s manual while16S rRNA gene sequencing is used for sp. level identification effectively.
        
The statistical based optimization using CCD and RSM offered an efficient and feasible approach. The optimum conditions for the maximum siderophore production were predicted from the proposed model. Present study showed that maximum yield of siderophore was obtained at pH 7, time 60 hrs, incubation temp. 28^oC and succinic acid conc. of 0.40%. The confirmation experiment was performed successfully within the range of levels predicted from the model whichgives maximum yield siderophore production by the isolate.
              
Report suggests that the proposed model accurately predict the maximum siderophore production by providing optimum values of pH, incubation period, temperature and succinic acid concentration. Authors conclude that these types of models are of great importance for developing industrially viable processes.