Introduction

Pseudomonas aeruginosa causes a variety of infections to the urinary tract, respiratory tract, skin and soft tissue, bones and joints, and gastrointestinal tract (Tielen et al. 2013; van ‘t Wout et al. 2015). P. aeruginosa infection is a serious problem in patients hospitalized with cancer, cystic fibrosis (CF), and burns with case fatality rate of 50% (Boyd and Chakrabarty 1995). In particular, patients with CF are susceptible to chronic P. aeruginosa infections of the airways, causing heightened inflammation, bronchial asthma, tissue destruction, and pulmonary dysfunction which can lead to death (Bonfield et al. 1999; Wang et al. 2015; Yang et al. 2011, 2008). Initially, patients are colonized with a classic P. aeruginosa morphotype, and subsequently, phase variation or mutation occurs resulting in adapted mucoid strains that predominate during the evolution of the disease (Purevdorj-Gage et al. 2005; Ramsey and Wozniak 2005; Römling et al. 1997; Schmidt et al. 1996; Yang et al. 2011, 2008). Recently, some studies suggested nebulized magnesium sulfate as a clinically significant bronchodilator treatment option in patients resistant to standard therapy of acute asthma (Sarhan et al. 2016). However, the effect of magnesium sulfate on the virulence of P. aeruginosa is lacking.

Virulence of P. aeruginosa is multifactorial, involving both secreted and cell-associated bacterial products. Secreted products include elastase, alkaline protease, protease IV, exotoxins, rhamnolipids, and siderophores while cell-associated bacterial products involve alginate production and O-antigen (Alionte et al. 2001; Baumann et al. 1993; Bever and Iglewski 1988; Bothwell et al. 2003; Döring and Høiby 1983; Finck-Barbançon et al. 1997; Hauser et al. 1998; Hinsa et al. 2003; Hobden 2002; Tielen et al. 2013, 2010; Xiao et al. 2006; Yahr et al. 1996; Yahr and Wolfgang 2006). P. aeruginosa utilizes two major mechanisms to evade the host defense system. The first mechanism is the production of extracellular products, such as proteases, toxins, and lipases (Tielen et al. 2010). Alkaline protease and elastase inhibit the function of the cells of the immune system (phagocytes, NK-cells, T cells), inactivate several cytokines (IL-1, IL-2, IFN-r, TNF), cleave immunoglobulins, and inactivate complement (Høiby et al. 2011; Horvat and Parmely 1988; Klausen et al. 2003; Matz and Girskov 2007; McIver et al. 1995; Otterlei et al. 1991). Moreover, P. aeruginosa uses a type III secretion system (TTSS) as a molecular syringe to deliver bacterial toxins directly into the cytosol of eukaryotic cells and to inhibit host immune response (Franklin et al. 2011; Phillips et al. 2003). The second mechanism is the alginate-dependent biofilm mode of growth in chronic infections (Kharazmi 1991). Mucoid strains overproduce the EPS alginate and are associated with CF (Franklin et al. 2011; Martin et al. 1993). Alginate has a crucial role in the process of pathogenicity; it protects bacterial cells from phagocytosis (Ma et al. 2012; Meluleni et al. 1995; Schlictman et al. 1994). Alginate affects leukocyte functions and plays an immunomodulatory role via suppression of lymphocyte transformation (Mathee et al. 1999). Furthermore, alginate is believed to prevent penetration of antibiotics into the infecting cells and to aid bacterial adherence to epithelial cells of the respiratory tract (Schlictman et al. 1994). Sessile bacteria can be up to 1000 times more resistant to antibiotics than their planktonic counterparts (Gilbert et al. 1997; Høiby 1994; Yang et al. 2011; 2008).

Serotyping of P. aeruginosa is based on the O-specific antigen. P. aeruginosa strain PAO1, a non-mucoid wild-type strain, belongs to the serotype O5 (Wang et al. 2015). P. aeruginosa FRD1, a sputum isolate from a cystic fibrosis patient, is an alginate-producing strain that possesses a distinctive mucoid colony morphology. Expression of virulence factors by P. aeruginosa is controlled by signal molecule-dependent cell-cell communication systems in response to the surrounding environment (Erickson et al. 2002; Singh et al. 2015). The aim of this work is to explore the effect of magnesium sulfate and other environmental variables on the production of alginate and extracellular proteins by P. aeruginosa FRD1 and PAO1 clinical isolates.

The one-dimensional search with successive changes on variables is still employed, even though it is well accepted that it is practically impossible for the one-dimensional search to accomplish an appropriate optimum combination in a finite number of experiments. Single-variable investigation methods are not only tedious, but can also lead to misinterpretation of results, especially taking into account that the interaction between different factors is overlooked. Statistical methods, through fractional factorial design (FFD) and response surface methodology (RSM), offer simultaneous study of many factors as well as the interactive effects of these factors together (Lotfy et al. 2006). To the best of our knowledge, concrete experimental studies of magnesium sulfate and other environmental factors affecting the performance of P. aeruginosa by statistically designed experiments are not employed.

Materials and methods

Microorganisms

Two strains of the species P. aeruginosa were used in this study namely, FRD1 and PAO1. The mucoid FRD1 strain, a sputum isolate from a cystic fibrosis patient, was kindly obtained from Prof. Bernd H. A. Rehm, Institute of Molecular BioSciences, Massey University, Palmerston North, New Zealand. The non-mucoid PAO1 strain (NCCB 2452) was donated from the culture collection NCCB, Netherlands. Stock cultures were maintained on nutrient broth supplemented with 30–40% (v/v) glycerol and kept at − 18 °C. Short-term viable cultures were maintained on nutrient agar plates at 4 °C. Cultures used in the experimental studies were not older than 5 weeks.

Culture media and solutions

For maintaining stock cultures, nutrient broth containing (g/l) beef extract (3.0) and peptone (5.0) was used. The basal culture medium used in the Plackett-Burman experiment contained (g/l) glucose, 20.0; (NH4)2SO4, 0.6; K2HPO4, 2.2; KH2PO4, 0.8; NaCl, 0.2; MgSO4.7H2O, 0.2; CaCl2, 0.05; yeast extract, 0.25; casamino acids, 0.25; FeSO4.7H2O, 0.6 × 10−3; and trace elements (mg/l): ZnSO4.7H2O, 0.2; CuSO4.5H2O, 0.2; MnSO4.H2O, 0.2; and CoCl2.6H2O, 0.2 (Mian et al. 1978). Under some specified experimental conditions, yeast extract (0.5 g/l), casamino acids (0.5 g/l), and/or hydrogen peroxide (0.2 ml/l) were added to the medium.

All media were sterilized by autoclaving at 121 °C for 20 min. Solutions of K2HPO4, KH2PO4, MgSO4, CaCl2, and trace elements were autoclaved separately and added to the rest of the sterilized medium after cooling to room temperature. Ferrous sulfate solution was sterilized by filtration using a Millipore cellulose nitrate filter with a 0.2-μm pore size.

Bacterial growth

Shake flask experiments were carried out in 250-ml baffled Erlenmeyer flasks containing 50 ml medium and kept on a rotary shaker at 200 rpm at 30 °C. The pH was initially adjusted to 7.3 which decreased during the fermentation time due to alginate production. The inoculum was introduced as 10% (v/v) of the culture volume. For each trial in the experimental design matrices, growth parameters such as medium components and culture volume were adjusted as indicated in the “Results” section.

Determination of proteins and alginate

After 48-h incubation on a rotary shaker at 200 rpm at 30 °C, an aliquot of 10 ml culture broth was centrifuged at 5000 rpm at 20 °C for 30 min to precipitate the cells. Total protein was determined according to Lowry et al. (1951) to reflect the concentration of secreted virulence protein. Alginate dry weight was determined gravimetrically by the method described by Clementi et al. (1999). Briefly, the supernatant was cooled and alginate was then precipitated by the addition of three volumes of ice cold acetone which was then recovered by centrifugation at 5000 rpm for 30 min. The precipitate was dissolved in distilled water, precipitated again, centrifuged, and then finally dried at 80 °C for 24 h. For each determination, at least two samples were used.

Experimental designs

To investigate the production of EPS and exo-proteins as responses to environmental variations, two sequential multifactorial experimental designs were applied. The Plackett-Burman design which is a FFD (Lotfy et al. 2017; Plackett and Burman 1946; Yu et al. 1997) was used to reflect the relative importance of various environmental factors on the production of alginate and extracellular proteins by the strains FRD1 and PAO1, respectively, in liquid cultures. In this experiment, 11 self-governing variables were screened in 12 combinations organized according to the Plackett-Burman design matrix described in the “Results” section. These independent variables included culture volume, the components of the basal medium, in addition to yeast extract, casamino acids, and hydrogen peroxide. Phosphates were mixed and treated as a single variable. Similarly, trace elements were prepared as a stock solution to be examined as one factor. Each variable was examined at two levels, a high (+1) and a low (−1) level. All trials were performed in triplicates and the averages of alginate and extracellular proteins production observations were treated as the responses. The main effect of each variable on the response was calculated according to the following equation:

$$ \mathrm{Main}\kern0.17em \mathrm{effect}=\frac{2\left[\sum R(H)-\sum R(L)\right]}{N} $$

where R(H) is the response parameter of an assembly in the screening design that contains the higher quantity of a given component, R(L) is the response parameter of an assembly in the screening design that contains the lower quantity of a given component, and N is the number of assemblies. The factor that had no effect would give a value of zero if no interactions existed. A main effect figure with a positive sign indicates that the high level of this variable is nearer to optimum and a negative sign indicates that the low level of this variable is nearer to optimum. Using Statistica 10 software, statistical t values and p values were calculated for determination of variable significance.

The Box-Behnken design which is a RSM (Box and Behnken 1960; Lotfy et al. 2006) was applied. In this experiment, the most significant variables resolved from the Plackett-Burman design for each strain were treated as independent variables. Each factor was examined at three different levels, low (−), high (+), and basal (0). Sixteen combinations and their observations were fitted to the following second-order polynomial mode:

$$ \mathrm{Y}={\mathrm{b}}_0+{\mathrm{b}}_1{\mathrm{X}}_1+{\mathrm{b}}_2{\mathrm{X}}_2+{\mathrm{b}}_3{\mathrm{X}}_3+{\mathrm{b}}_{12}{\mathrm{X}}_1{\mathrm{X}}_2+{\mathrm{b}}_{13}{\mathrm{X}}_1{\mathrm{X}}_3+{\mathrm{b}}_{23}{\mathrm{X}}_2{\mathrm{X}}_3+{\mathrm{b}}_{11}{\mathrm{X}}_1^2+{\mathrm{b}}_{22}{\mathrm{X}}_2^2+{\mathrm{b}}_{33}{\mathrm{X}}_3^2 $$

where, Y is the dependent variable (alginate or extracellular proteins in mucoid and non-mucoid strains, respectively); X1, X2, and X3 are the independent variables; b0 is the regression coefficient at center point; b1, b2, and b3 are linear coefficients; b12, b13, and b23 are second-order interaction coefficients; and b11, b22, and b33 are quadratic coefficients. Based on the obtained equation, the values of the coefficients and levels that maximize dependent variables were predicted using Statistica 10 software. The coefficient of determination, R2, and a plot showing the distribution of experimental versus predicted values were used to examine the quality of the model.

Results

Elucidation of the factors affecting the expression of EPS and extracellular proteins

The capability of the non-mucoid P. aeruginosa strain to produce virulence factors could be modulated by modifying environmental and culture factors, such as aeration and nutrient amount, or even the extra addition of hydrogen peroxide, a condition that prevailed in vitro (Mathee et al. 1999). In present work, the Plackett-Burman design was applied to screen for the most important environmental factors responsible for controlling EPS expression and the release of the proteinaceous virulence factors by the mucoid P. aeruginosa strain and protein production by the non-mucoid P. aeruginosa strain. The factors chosen for the statistical study were 11 and included the nutritional components together with aeration of the culture and the presence of hydrogen peroxide. The design was applied with 12 different fermentation conditions. The chosen levels of the culture components and the observed results are given in Table 1. With respect to EPS and excreted proteins, the calculated main effect values for each examined variable and statistical analysis results are shown in Table 2. The calculated statistical parameters revealed that the examined variables did not significantly affect the production of extracellular proteins by FRD1 (data not shown). However, the statistical analysis of the response results and the calculated main effects (Table 2) indicated the presence of marked variations with respect to the production of EPS by FRD1 and expression of exo-proteins by PAO1.

Table 1 The Plackett-Burman design applied on the synthesis of EPS by FRD1 and production of exo-proteins by FRD1 and PAO1
Full size table
Table 2 Statistical analysis of the Plackett-Burman experimental results
Full size table

In general, variables with positive main effect signs are predicted to result in high response results at their examined high settings. The main effect results represented in Fig. 1a and Table 2 demonstrate that increasing the level of ammonium sulfate, trace elements, phosphates, or aeration intensity is advantageous for the EPS overall production (g/l) by the mucoid FRD1 strain. On the contrary, the high level of NaCl and the presence of hydrogen peroxide negatively affected alginate production by the same strain. On the basis of the calculated t values, the results of this experiment suggest also that ammonium sulfate, trace elements, and phosphate were the most substantial variables that significantly affect alginate production by the FRD1 strain.

Fig. 1
figure 1

Pareto plot showing the effect of each of the 11 different variables on (a) overall EPS production by FRD1 strain and (b) overall protein production by PAO1 strain. Abbreviations: P, phosphates; S, sodium chloride; M, magnesium sulfate; C, calcium chloride; A, ammonium sulfate; F, ferrous sulfate; T, trace elements; V, culture volume; H, hydrogen peroxide; Y, yeast extract; CA, casamino acid

Full size image

Table 2 and Fig. 1b show the magnitude of effect for each of the 11 different variables on overall production of proteins (g/l) by the PAO1 strain under shake flask culture conditions. Because of the proteinaceous nature of casamino acid and yeast extract, their main effects were not considered in this part of the work. The results suggest that high levels of phosphates, magnesium sulfate, ammonium sulfate, aeration, and the addition of trace elements to the minimal medium were advantageous for production of extracellular proteins (g/l). On the other hand, secretion of proteins was negatively affected by the presence of high levels of FeSO4 and NaCl in the medium. According to the statistical analysis performed (Table 2), phosphates, MgSO4, and FeSO4 were treated as significant variables with respect to extracellular protein production by the PAO1 strain.

The main objective of the applied screening factorial experiment was to get a general picture of how the response is affected by changes in different environmental variables. Thus, a validation experiment was carried out in triplicates to evaluate the accuracy of the Plackett-Burman experiment. The predicted near optimum levels of independent variables were examined and compared to their contrary and their basal settings. As shown in Fig. 2, the near optimized medium formulation resulted in a 1.4-fold increase of alginate production by FRD1 strain when compared to the basal culture condition. On the other hand, total extracellular protein production showed a fourfold increase under the predicted near optimum condition compared to the basal levels (Fig. 2).

Fig. 2
figure 2

Alginate and protein production by FRD1 and PAO1 strains, respectively, grown on basal, near optimum, and reverse optimized conditions

Full size image

Application of the Box-Behnken statistical design

Based on the results of the Plackett-Burman experiment, the levels of the significant variables affecting the production of extracellular biopolymers and proteins were further investigated using the Box-Behnken design. For the FRD1 strain, the interactions of ammonium sulfate with the concentration of phosphates and the presence of trace elements were examined. Whereas the production of extracellular proteins by the PAO1 strain was further studied using concentrations of the most effective key factors, including phosphates, magnesium sulfate, and ferrous sulfate as independent variables. In both cases, each key factor was examined at three different levels in 12 trials as described in Tables 3 and 4, while the other factors were fixed at their basal levels. Trials 13–16 consist of the basal medium of this experiment and represent the control. The basal condition for alginate production by FRD1 strain was as follows: glucose, 20 g/l; phosphates, 3 g/l; sodium chloride, 0.02 g/l; magnesium sulfate, 0.02 g/l; calcium chloride, 0.01 g/l; ammonium sulfate, 2 g/l; ferrous sulfate, 0.3 mg/l; trace elements, 0.6 mg/l; culture volume, 75 ml; hydrogen peroxide, 0.2 ml; yeast extract, 0.5 g/l; and casamino acid, 0.1 g/l. The basal condition for protein production by the PAO1 strain was the same except for the concentration of magnesium sulfate, 1 g/l and casamino acid, 0.5 g/l.

Table 3 Concentrations of the key variables examined in the Box-Behnken experiment and observed alginate production responses of the FRD1 strain
Full size table
Table 4 Concentrations of the key variables examined in the Box-Behnken experiment and observed protein production responses of the PAO1 strain
Full size table

The data shown in Table 3 represent different combination levels and the corresponding biopolymer production results observed after 2 days of incubation. According to the obtained results, the interactions of the three examined variables are presented in the form of three-dimensional surface plots in Fig. 3a–c. From these figures, it is clear that ammonium sulfate had the most considerable effect on alginate production. On the contrary, trace elements and concentration of phosphates within the examined ranges and under the present experimental conditions were very much less effective variables. Furthermore, the tested levels of phosphates and ammonium sulfate did not show an optimum peak in the graph, suggesting that further increases in their concentrations may result in more biopolymer yield.

Fig. 3
figure 3

Three-dimensional surface plots showing the interaction between K2H/KH2PO4 and NH4SO4 (a), NH4SO4 and trace elements (b), and K2H/KH2PO4 and trace elements (c) with respect to alginate production by FRD1 strain based on the Box-Behnken experimental results

Full size image

The statistical analysis and the values of coefficients were estimated by using Statistica 10 software (Table 5). For predicting the optimal point, the following second-order polynomial function was fitted to alginate concentration results of the applied Box-Behnken experiment:

$$ \mathrm{Y}=1.26+0.73{\mathrm{X}}_1-0.17{\mathrm{X}}_2+0.78{\mathrm{X}}_3-0.1{\mathrm{X}}_1{\mathrm{X}}_2+0.9{\mathrm{X}}_1{\mathrm{X}}_3-0.1{\mathrm{X}}_2{\mathrm{X}}_3-0.01{\mathrm{X}}_1^2+0.17{\mathrm{X}}_2^2+0.09{\mathrm{X}}_3^2 $$

where Y is the alginate production in g/l and X1, X2, and X3 are the concentrations of ammonium sulfate (g/l), trace elements (mg/l), and phosphate (g/l), respectively.

Table 5 Parameter coefficients for Box-Behnken design experiments
Full size table

The quality of the polynomial model equation was expressed by the coefficient of determination, R2, which was 0.97. The closer the value of R2 to 1.0, the better the correlation between actual and predicted values. Moreover, a parity plot showing the distribution of experimental versus predicted values by the mathematical model of alginate concentration (Fig. 4) reflects a linear relationship with r = 0.98 indicating a reasonable fitting model. According to this equation, the predicted optimum condition for alginate production contains ammonium sulfate, trace elements, and phosphates with concentrations of 4 g/l, 0.6 mg/l, and 6 g/l, respectively.

Fig. 4
figure 4

Distribution of experimental values of alginate concentration versus predicted values (r = 0.98)

Full size image

A confirmatory experiment was subsequently performed under the best-predicted nutritional conditions for alginate production in the flask culture. The pre-optimized Plackett-Burman culture medium was used as a control against a predicted Box-Behnken optimized medium (Fig. 5) and alginate concentrations of 2.2 and 5.1 g/l, respectively, were observed.

Fig. 5
figure 5

Desirability charts of variables for maximum alginate production by FRD1 strain based on the Box-Behnken experimental results

Full size image

The data shown in Table 4 represent different combinations of independent variables and corresponding concentrations of the exo-proteins virulence factors resulted within 2 days of incubation. The interactions of the three examined variables are shown in Fig. 6a–c. From these figures, it can be concluded that phosphate mixture is the most important variable that affected the production of extracellular proteins.

Fig. 6
figure 6

Response surface plot showing the interaction of FeSO4 with MgSO4 levels (a), FeSO4 with K2H/KH2PO4 levels (b), and MgSO4 with K2H/KH2PO4 levels (c) with respect to proteins production by PAO1 strain based on the Box-Behnken experimental results

Full size image

The results of this Box-Behnken application were fitted to a second-order polynomial model. The coefficient of determination R2, which was found to be 0.996, indicating that 99.6% of the variability in the response can be explained by the model. This revealed that the equation is a suitable model to describe the response of the experiment. As shown in Fig. 7, the parity plot of the experimental versus predicted values indicated a good fitting model (r = 0.984). The following second-order polynomial function was fitted to protein production results of the applied Box-Behnken experiment:

$$ \mathrm{Y}=5.94+5.68{\mathrm{X}}_1-0.58{\mathrm{X}}_2-1.04{\mathrm{X}}_3+1.0{\mathrm{X}}_1{\mathrm{X}}_2-0.28{\mathrm{X}}_1{\mathrm{X}}_3+0.69{\mathrm{X}}_2{\mathrm{X}}_3-1.89{\mathrm{X}}_1^2-0.13{\mathrm{X}}_2^2+0.58{\mathrm{X}}_3^2 $$

where Y is the protein production in g/l and X1, X2, and X3 are the concentrations of phosphates (g/l), magnesium sulfate (g/l), and ferrous sulfate (mg/l), respectively.

Fig. 7
figure 7

Distribution of experimental values of protein concentration versus predicted values (r = 0.984)

Full size image

According to this equation, the predicted optimum condition for protein production contains phosphates, magnesium sulfate, and ferrous sulfate with concentrations of 6 g/l, 0.02 g/l, and 0.3 mg/l, respectively. Finally, for verification, the Box-Behnken optimized medium (Fig. 8) was practically examined for extracellular protein production by PAO1 using the pre-optimized Plackett-Burman culture medium as a control. Extracellular protein concentrations of 10.5 and 6.3 g/l were recorded under the optimized and control conditions, respectively confirming the experimental predictions.

Fig. 8
figure 8

Desirability charts of variables for maximum protein secretion by PAO1 strain based on the Box-Behnken experimental results

Full size image

Discussion

As an opportunistic human pathogen, P. aeruginosa can cause CF and infections of the middle ear, eyes, wounds, and urinary tract (Mesaros et al. 2007). A hallmark of P. aeruginosa is its ability to produce EPS and extracellular virulence proteins to establish biofilm-based infections that are difficult to eradicate. Toyofuku et al. (2012) demonstrated that approximately 30% of the identified matrix proteins of P. aeruginosa were outer membrane proteins, while some proteins were derived from secreted proteins and lysed cells (Toyofuku et al. 2012). Virulence of P. aeruginosa is dependent on a number of factors, involving both secreted proteins and alginate production (Alionte et al. 2001; Baumann et al. 1993; Bever and Iglewski 1988; Bothwell et al. 2003; Döring and Høiby 1983; Finck-Barbançon et al. 1997; Hauser et al. 1998; Hinsa et al. 2003; Hobden 2002; Tielen et al. 2013, 2010; Xiao et al. 2006; Yahr et al. 1996; Yahr and Wolfgang 2006). The main objective of the present study was to investigate various factors that control the production of alginate and extracellular proteins by P. aeruginosa clinical isolates.

Based on the results of the applied FFD and RSM, we should emphasize that the variables affected alginate formation by the mucoid strain were not exactly the same affecting extracellular protein production by the non-mucoid strain. It is, however, noted that the presence of excess phosphates in the medium increases alginate synthesis and secretion of proteins by the FRD1 and PAO1 strains, respectively. This observation is mainly attributed to the effect of phosphate functional groups on medium pH as well as metabolic activities including protein synthesis and cell wall composition. Phosphate is required for the synthesis of virulence proteins such as alkaline phosphatase and phospholipases by P. aeruginosa (Barker et al. 2004; Filloux et al. 1988). Similarly, inorganic polyphosphate has been reported to be essential for long-term survival and expression of virulence factors in Shigella and Salmonella spp. (Kim et al. 2002). Previous results indicated also that acidic metabolites such as free uronic acid and the biopolymer alginic acid are produced by mucoid and non-mucoid strains (Kim et al. 2002). Therefore, in addition to its metabolic roles, the effects of phosphates observed in this work may support the importance of the buffering capacity of the surrounding medium for pathogenic bacteria.

Moreover, it has been previously reported that inorganic polyphosphate (poly-P, a chain of tens or many hundreds of phosphate residues linked by high-energy phosphoanhydride bonds) and the enzyme synthesizing it have been found in all pathogenic bacteria (Kornberg et al. 1999). With respect to alginate synthesis in P. aeruginosa, the sugar subunits are activated with guanosine diphosphate (GDP) (Hay et al. 2009). Depending on the cell type and condition, poly-P performs varied functions (Fraley et al. 2007). For instance, mutants of PAO1 lacking poly-P kinase 1 and the enzyme responsible for most poly-P synthesis in Escherichia coli and other bacteria are defective in motility, quorum sensing, biofilm formation, and the secretion of virulence factors.

Our data suggest also ammonium sulfate and trace elements as substantial variables that significantly affect alginate production by the FRD1 strain. We found that the presence of a relatively high ammonium sulfate level (4.0 g/l) increased the ability of bacterial cells to produce alginate. This is in accordance with previous investigations concerned with the functional characterization of several enzymes and proteins involved in alginate synthesis (Filloux 2011; Keiski et al. 2010; Whitney et al. 2011). Our observations demonstrated also that the presence of trace elements in the surrounding environment supported alginate biosynthesis by the FRD1 strain. A possible explanation for this observation is the prominence of metals to activate enzymes involved in the production of alginate precursor (Regni et al. 2004, 2002).

The results of the applied FFD and RSM experiments demonstrated also that extracellular protein production by the PAO1 strain is significantly affected by the levels of FeSO4 and MgSO4 in the environment. Parallel to our findings, the importance of iron and magnesium availability as major limiting factors with respect to the growth of parasitic bacteria has been previously reported (Meyer 2000).

The results showed that, under our experimental conditions, a concentration of 0.3 mg/l ferrous sulfate supports high protein secretion by the PAO1 strain. Earlier findings proved the presence of a complicated regulation network that involves iron signaling, quorum sensing, and chemotaxis that coordinate the macrocolony formation in non-mucoid P. aeruginosa biofilms (Harmsen et al. 2010; Yang et al. 2011). It has been also demonstrated that P. aeruginosa switch on the expression of iron-regulated outer membrane proteins that assist iron acquisition (Anwar et al. 1991).

Recent investigations have reported that nebulized MgSO4 has significant bronchodilator effect in acute asthma and leads to clinical improvement (Sarhan et al. 2016). However, these studies have focused on host defense rather than parasite virulence. On the other hand, the present work demonstrated that the concentration of magnesium sulfate in the surrounding environment affects biofilm formation as well as the expression of extracellular proteins by P. aeruginosa. The results of the applied FFD indicated that the presence of a relatively high concentration of magnesium sulfate (1 g/l) downregulated the biosynthesis of alginate by the FRD1 strain. Moreover, analyzing the results of the RSM experiment suggested a medium that contains magnesium sulfate in a concentration of 0.02 g/l as an optimized condition for the expression of extracellular proteins by the PAO1 strain. Analogous to this result is the previous finding showing that pathogenesis of PAO1 strain is highly associated with the production of T2SS-dependent exo-proteins such as metalloproteinase whose catalytic mechanisms rely on metals (Bleves et al. 2010; Braun et al. 1998; Voulhoux et al. 2001; Voulhoux et al. 2003). Moreover, the growth of P. aeruginosa in low magnesium resulted in altered subcellular compartmentalization of large enzyme complexes such as ribosomes (Guina et al. 2003).