Biodegradation

, Volume 21, Issue 4, pp 501–511

Enhanced decolourization of Direct Red-80 dye by the white rot fungus Phanerochaete chrysosporium employing sequential design of experiments

Authors

  • Sukhwinder Singh
    • Department of BiotechnologyIndian Institute of Technology
    • Department of BiotechnologyIndian Institute of Technology
  • Achlesh Daverey
    • Department of BiotechnologyIndian Institute of Technology
Original Paper

DOI: 10.1007/s10532-009-9319-2

Cite this article as:
Singh, S., Pakshirajan, K. & Daverey, A. Biodegradation (2010) 21: 501. doi:10.1007/s10532-009-9319-2

Abstract

Decolourization of Direct Red 80 (DR-80) by the white rot fungus Phanerochaete chrysosporium MTCC 787 was investigated employing sequential design of experiments. Media components for growing the white rot fungus were first screened using Plackett-Burman design and then optimized using response surface methodology (RSM), which resulted in enhancement in the efficiency of dye removal by the fungus. For determining the effect of media constituents on the dye removal, both percent dye decolourization and specific dye removal due to maximum enzyme activity were chosen as the responses from the experiments, and the media constituents glucose, veratryl alcohol, KH2PO4, CaCl2 and MgSO4 were screened to be the most effective with P values less than 0.05. Central composite design (CCD) followed by RSM in the optimization study revealed the following optimum combinations of the screened media constituents: glucose, 11.9 g l−1; veratryl alcohol, 12.03 mM; KH2PO4, 23.08 g l−1; CaCl2, 2.4 g l−1; MgSO4, 10.47 g l−1. At the optimum settings of the media constituents, complete dye decolourization (100% removal efficiency) and a maximum specific dye removal due to lignin peroxidase enzyme of 0.24 mg U−1 by the white rot fungus were observed.

Keywords

Direct Red-80Phanerochaete chrysosporiumLignin peroxidaseDye decolourizationAzo dyes

Introduction

Azo dyes such as Direct Red-80 (DR-80) are important colorants and constitute the largest class of dyes for application in textile, paper, leather, gasoline, foodstuffs and cosmetics industries (Dias et al. 2003). These dyes, generally characterized by the presence of one or more azo linkages (–N=N–) and aromatic rings (Nigam et al. 1996), are known to be toxic and mutagenic to living organisms (Moller and Wallin. 2000; Stolz. 2001; Mechichi et al. 2006). Discharge of wastewaters from dye manufacturing units and textile processing industries containing such azo dyes therefore results in pollution of the receiving aquatic environment (Keharia and Madamwar 2003). Hence, treatment of such dye containing wastewater is essential to prevent deterioration as well as to sustain the natural flora and fauna of the ecosystem. Currently, the major methods of treating dye containing wastewaters involve physical and/or chemical techniques such as photo catalytic degradation (Korbahti and Rauf 2008), electrocoagulation (Aleboyeh et al. 2008), ozonation (Hassan and Hawkyard 2002), treatment using Fenton’s reagent, electrochemical destruction etc. (Lin and Chen 1997). However, these techniques are either inefficient in complete removal of dyes or are economically non-viable (Chang and Lin 2000). On the other hand, microbial degradation and decolourization of azo dyes is considered more efficient and cost competitive compared to the physico-chemical techniques. The white-rot fungus Phanerochaete chrysosporium, which belongs to a group of lignin-degrading basidiomycetes, has received considerable attention in the past for their very good bioremediation potential (Bumpus and Brock 1988; Cripps et al. 1990; Bakshi et al. 1999) owing to its natural capability to degrade complex lignin by their extracellular non-specific and non-stereoselective enzyme system composed of lignin peroxidases (LiP, EC 1.11.1.14), and manganese peroxidases (MnP, EC 1.11.1.13) that function together with H2O2 producing oxidases and secondary metabolites (Hatakka 1994). The same unique non-specific mechanisms that give these fungi the ability to degrade lignin also facilitate degradation of a wide range of recalcitrant pollutants including polycyclic aromatic hydrocarbons, chlorinated phenols, polychlorinated biphenyls, dioxins, pesticides and explosives (Pointing 2001). The DR-80 is a water-soluble azo dye and is used for dyeing silk, wool, cellulose and cotton (Mahmoodi et al. 2005). Although important to decolorize or degrade this dye in aqueous systems, there is no report as such on its biodegradation. However, decolourization of DR-80 by UV oxidation in the presence of hydrogen peroxide employing TiO2 as a photo catalyst or by electrochemical method using three different materials as anodes, iron, polypyrrole (PPy) and boron doped diamond (BDD), have been reported. As mentioned earlier, all these methods suffered from the main drawback in not being efficient to degrade the dye (Mahmoodi et al. 2005; Aleboyeh et al. 2008). Hence, biodecolorization of DR-80 dye by P. chrysosporium was investigated in the present study. Since media constituents play vital role in growth of P. chrysosporium and in production of the lignin degrading enzyme—lignin peroxidase, which is directly involved in dye decolourization process, proper understanding of the effects of media constituents and optimization of their levels are essential to significantly improve the dye decolourization efficiency by the fungus.

Screening and optimization of media constituents can be done by either varying one variable at a time or non-conventionally by using statistical methods. Conventional screening and optimization techniques involve varying factor and their levels by maintaining the other factors at an unspecified constant level; by doing so, the combined effect of the factors is, however, neglected. In addition, these techniques are time-consuming and require sufficiently large number of experimental runs. Such limitations of a classical method can be eliminated by screening and optimizing all the affecting factors collectively by employing statistical experimental design and empirical model building using regression analysis. The most successfully used technique for optimization based on statistical methods is known as response surface methodology (Chen 1994), which further helps in understanding the effects of individual variables and their interaction on a given response (Montgomery 1991). To reduce the number of factors to be used in an optimization study, screening of factors is normally performed by employing another statistical design such as Plackett-Burman. Although RSM has been effectively used in optimization of various biotechnological and industrial processes (Asgher et al. 2008; Mohana et al. 2008; Reddy et al. 2008; Purama and Goyal 2008; Sheth and Dave 2009; Tir and Moulai-Mostefa 2008), the use of such nonconventional statistical techniques in screening followed by optimization of media constituents for enhancing dye decolourization has not been reported so far in the literature. Therefore, in the present study, decolourization of DR-80 by P. chrysosporium was studied systematically by screening and optimization of media constituents using Plackett-Burman design and RSM, respectively, with an aim to enhance the efficiency of dye removal by the fungus.

Materials and methods

Chemicals and reagents

Azo dye (Direct Red-80) and veratryl alcohol (3, 4-dimethoxybenzyl alcohol, 96% pure) were purchased from Sigma (St. Louis, Mo, USA). All other chemicals and solvents, of Guaranteed Reagent (GR) grade, were purchased from either High Media Mumbai (India), SRL (India) or Merck (India).

Microorganism and its maintenance

The fungus P. chrysosporium MTCC 787 used in this study was procured from IMTECH, Chandigarh, India, and was maintained at 25°C on potato dextrose agar (PDA) slants. For spore production, the slants were incubated at 39°C for 2–5 days in media containing glucose: 10 g l−1, malt extract: 10 g l−1; peptone: 2 g l−1; yeast extract: 2 g l−1; asparagine: 1 g l−1;KH2PO4: 2 g l−1; MgSO4·7H2O: 1 g l−1; thaimin-HCl: 1 mg l−1 and agar: 20 g l−1 (Tien and Kirk 1983). The media used for studying decolourization of DR-80 dye composed of basal medium (KH2PO4, 20 g l−1; MgSO4: 5 g l−1; CaCl2: 1 g l−1); trace elements (MgSO4: 3 g l−1; MnSO4: 0.5 g l−1; NaCl: 1 g l−1; FeSO4·7H2O: 0.1 g l−1; CoCl2: 0.1 g l−1; ZnSO4·7H2O: 0.1 g l−1; CuSO4: 0.1 g l−1; AlK(SO4)2·12H2O:10 mg l−1; H3BO3:10 mg l−1; Na2MoO4·2H2O: 10 mg l−1; nitrilotriacetate: 1.5 g l−1) and other ingredients—glucose: 100 g l−1; 2,2-dimethylsuccinate: 0.1 M (pH 4.2); thiamine:100 mg l−1 (filter sterilized); veratryl alcohol: 4 mM stock (filter sterilized) and NH4Cl: 4.68 g l−1 (Radha et al. 2005).

Plackett-Burman design for screening of media constituents

A Plackett-Burman design of total 12 experimental runs was employed to screen the media constituents for their effects on DR-80 decolourization by P. chrysosporium. All the media constituents, except the trace elements, and inoculum size, which are known to influence both the fungus growth and its lignin peroxidase enzyme activity, were chosen as factors in the screening study. Since the trace elements, temperature, pH etc. are known to play a role in the fungal growth and not necessarily on the activity of its enzyme activity, these were not included in the study. Table 1 presents the experimental combinations of the factors and their levels adopted in the screening study, where −1 and +1 are the coded levels of each factor representing the respective minimum and maximum feasible concentrations of the factors in the study. In all the experimental runs that were performed in duplicates using an initial dye concentration of 50 mg l−1, percentage dye decolourization and specific dye removal were recorded as the responses, and the results analyzed using the statistical software package Minitab®, 15, PA, USA. The response expressed as percentage dye decolourization was simply calculated from equation (1)
$$ D = \left( {{\frac{{{\text{C}}_{ 0} - {\text{C}}}}{{{\text{C}}_{ 0} }}}} \right) \times 100 $$
(1)
where D is percentage dye decolourization, C0 and C are the respective initial and final dye concentrations (mg l−1) in the media. The response on specific DR-80 removal (mg U−1) due to maximum enzyme (LiP) activity was calculated using equation (2)
Table 1

Plackett-Burman design matrix showing the experimental combinations adopted in the screening study along with the observed percentage dye decolourization and specific dye removal

Exp. run no.

Experimental variables

Percentage dye decolourization (%)

Sp. dye removal (mg U−1)

Glucose

(g l−1)

NH4Cl

(g l−1)

KH2PO4

(g l−1)

MgSO4

(g l−1)

CaCl2

(g l−1)

Veratryl alcohol

(mM)

Tween 20 (%)

Inoculum size (OD)

1

15 (+1)

0.68 (−1)

30 (+1)

1 (−1)

0.2 (−1)

1 (−1)

0.2 (+1)

1.2 (+1)

97

0.4529

2

15 (+1)

8.68 (+1)

10 (−1)

10 (+1)

0.2 (−1)

1 (−1)

0.05 (−1)

1.2 (+1)

100

0.5549

3

5 (−1)

8.68 (+1)

30 (+1)

1 (−1)

1.8 (+1)

1 (−1)

0.05 (−1)

0.4 (−1)

38

0.6784

4

15 (+1)

0.68 (−1)

30 (+1)

10 (+1)

0.2 (−1)

7 (+1)

0.05 (−1)

0.4 (−1)

62

0.7643

5

15 (+1)

8.68 (+1)

10 (−1)

10 (+1)

1.8 (+1)

1 (−1)

0.2 (+1)

0.4 (−1)

100

0.5863

6

15 (+1)

8.68 (+1)

30 (+1)

1 (−1)

1.8 (+1)

7 (+1)

0.05 (−1)

1.2 (+1)

47

0.7580

7

5 (−1)

8.68 (+1)

30 (+1)

10 (+1)

0.2 (−1)

7 (+1)

0.2 (+1)

0.4 (−1)

32

0.5130

8

5 (−1)

0.68 (−1)

30 (+1)

10 (+1)

1.8 (+1)

1 (−1)

0.2 (+1)

1.2 (+1)

55

0.5333

9

5 (−1)

0.68 (−1)

10 (−1)

10 (+1)

1.8 (+1)

7 (+1)

0.05 (−1)

1.2 (+1)

83

0.8635

10

15 (+1)

0.68 (−1)

10 (−1)

1 (−1)

1.8 (+1)

7 (+1)

0.2 (+1)

0.4 (−1)

99

0.6104

11

5 (−1)

8.68 (+1)

10 (−1)

1 (−1)

0.2 (−1)

7 (+1)

0.2 (+1)

1.2 (+1)

59

0.5774

12

5 (−1)

0.68 (−1)

10 (−1)

1 (−1)

0.2 (−1)

1 (−1)

0.05 (−1)

0.4 (−1)

76

0.1955

Coded levels of the factors are indicated in parentheses

$$ {\text{SDR}} = \left( {{\frac{{{\text{C}}_{ 0} - {\text{C}}_{\text{t}} }}{{{\text{E}}_{ \max } }}}} \right) $$
(2)
where SDR is specific DR-80 removal due to maximum enzyme activity (mg U−1), Emax is the maximum enzyme activity (U l−1) and Ct is the dye concentration remaining in the media corresponding to the time at Emax.

Optimization of media constituents using response surface methodology

Based on the results obtained from Plackett-Burmann design for screening the media components, central composite design (CCD), containing the five factors glucose, MgSO4·7H2O, KH2PO4, CaCl2 and veratryl alcohol, was employed in optimizing the levels of the media constituents for enhancing both the responses on percentage dye decolourization and specific dye removal due to maximum enzyme activity by P. chrysosporium. Table 2 shows the 25 full factorial CCD and 10 center point replicates along with the observed and predicted responses in the study. All the experiments were carried out in 250 ml Erlenmeyer flasks containing 100 ml of the media containing DR-80 (50 mg l−1) and the media constituents, as per the design shown in Table 2. After inoculation with five day old spores (OD650 nm = 0.5) produced on PDA slants, the flasks were incubated at 39°C and 180 rpm for 6 days.
Table 2

Central composite design matrix showing the factors and their levels in the optimization study along with the measured and predicted responses

Exp. run no.

Experimental variables (mg l−1)

Percentage dye decolourization (%)

Sp. dye removal (mg U−1)

X1

X2

X3

X4

X5

Exp.

Pred.

Exp.

Pred.

1

6.9 (−1)

4.2 (−1)

13.8 (−1)

5.6 (−1)

1.6 (−1)

66

63

1.3689

0.9881

2

16.9 (+1)

4.2 (−1)

13.8 (−1)

5.6 (−1)

1.6 (−1)

75

70

0.8841

0.7829

3

6.9 (−1)

10.2 (+1)

13.8 (−1)

5.6 (−1)

1.6 (−1)

64

79

0.4969

0.4997

4

16.9 (+1)

10.2 (+1)

13.8 (−1)

5.6 (−1)

1.6 (−1)

86

88

0.3370

0.3718

5

6.9 (−1)

4.2 (−1)

33.8 (+1)

5.6 (−1)

1.6 (−1)

65

66

0.6327

0.6262

6

16.9 (+1)

4.2 (−1)

33.8 (+1)

5.6 (−1)

1.6 (−1)

65

69

0.4083

0.4605

7

6.9 (−1)

10.2 (+1)

33.8 (+1)

5.6 (−1)

1.6 (−1)

98

94

0.2881

0.3958

8

16.9 (+1)

10.2 (+1)

33.8 (+1)

5.6 (−1)

1.6 (−1)

100

99

0.3135

0.3074

9

6.9 (−1)

4.2 (−1)

13.8 (−1)

13.6 (+1)

1.6 (−1)

76

74

0.4948

0.6415

10

16.9 (+1)

4.2 (−1)

13.8 (−1)

13.6 (+1)

1.6 (−1)

75

77

0.4376

0.5197

11

6.9 (−1)

10.2 (+1)

13.8 (−1)

13.6 (+1)

1.6 (−1)

94

91

0.3426

0.2915

12

16.9 (+1)

10.2 (+1)

13.8 (−1)

13.6 (+1)

1.6 (−1)

95

95

0.3384

0.2470

13

6.9 (−1)

4.2 (−1)

33.8 (+1)

13.6 (+1)

1.6 (−1)

78

75

0.6929

0.5329

14

16.9 (+1)

4.2 (−1)

33.8 (+1)

13.6 (+1)

1.6 (−1)

77

74

0.6466

0.4506

15

6.9 (−1)

10.2 (+1)

33.8 (+1)

13.6 (+1)

1.6 (−1)

100

104

0.4274

0.4408

16

16.9 (+1)

10.2 (+1)

33.8 (+1)

13.6 (+1)

1.6 (−1)

100

104

0.4124

0.4359

17

6.9 (−1)

4.2 (−1)

13.8 (−1)

5.6 (−1)

3.2 (+1)

69

68

0.5296

0.6170

18

16.9 (+1)

4.2 (−1)

13.8 (−1)

5.6 (−1)

3.2 (+1)

69

71

0.5872

0.5361

19

6.9 (−1)

10.2 (+1)

13.8 (−1)

5.6 (−1)

3.2 (+1)

96

94

0.2688

0.3302

20

16.9 (+1)

10.2 (+1)

13.8 (−1)

5.6 (−1)

3.2 (+1)

95

98

0.2734

0.3266

21

6.9 (−1)

4.2 (−1)

33.8 (+1)

5.6 (−1)

3.2 (+1)

65

65

0.3278

0.2968

22

16.9 (+1)

4.2 (−1)

33.8 (+1)

5.6 (−1)

3.2 (+1)

66

64

0.3232

0.2555

23

6.9 (−1)

10.2 (+1)

33.8 (+1)

5.6 (−1)

3.2 (+1)

98

102

0.3720

0.2680

24

16.9 (+1)

10.2 (+1)

33.8 (+1)

5.6 (−1)

3.2 (+1)

98

103

0.3555

0.3039

25

6.9 (−1)

4.2 (−1)

13.8 (−1)

13.6 (+1)

3.2 (+1)

69

70

0.6354

0.4653

26

16.9 (+1)

4.2 (−1)

13.8 (−1)

13.6 (+1)

3.2 (+1)

70

69

0.6406

0.4678

27

6.9 (−1)

10.2 (+1)

13.8 (−1)

13.6 (+1)

3.2 (+1)

95

97

0.3371

0.3168

28

16.9 (+1)

10.2 (+1)

13.8 (−1)

13.6 (+1)

3.2 (+1)

95

97

0.3488

0.3967

29

6.9 (−1)

4.2 (−1)

33.8 (+1)

13.6 (+1)

3.2 (+1)

66

66

0.4135

0.3984

30

16.9 (+1)

4.2 (−1)

33.8 (+1)

13.6 (+1)

3.2 (+1)

67

60

0.3898

0.4405

31

6.9 (−1)

10.2 (+1)

33.8 (+1)

13.6 (+1)

3.2 (+1)

100

103

0.4501

0.5079

32

16.9 (+1)

10.2 (+1)

33.8 (+1)

13.6 (+1)

3.2 (+1)

100

100

0.4445

0.6273

33

0.008 (−α)

7.2 (0)

23.8 (0)

9.6 (0)

2.4 (0)

100

96

0.3665

0.4786

34

23.79 (+α)

7.2 (0)

23.8 (0)

9.6 (0)

2.4 (0)

100

100

0.3697

0.3764

35

11.9 (0)

0.065 (−α)

23.8 (0)

9.6 (0)

2.4 (0)

14

22

0.2868

0.5972

36

11.9 (0)

14.34 (+α)

23.8 (0)

9.6 (0)

2.4 (0)

100

88

0.4300

0.2385

37

11.9 (0)

7.2 (0)

0.016 (−α)

9.6 (0)

2.4 (0)

75

72

0.5144

0.6522

38

11.9 (0)

7.2 (0)

47.58 (+α)

9.6 (0)

2.4 (0)

80

80

0.5149

0.4960

39

11.9 (0)

7.2 (0)

23.8 (0)

0.086 (−α)

2.4 (0)

88

81

0.4012

0.4877

40

11.9 (0)

7.2 (0)

23.8 (0)

19.11 (+α)

2.4 (0)

88

91

0.4277

0.4601

41

11.9 (0)

7.2 (0)

23.8 (0)

9.6 (0)

0.497 (−α)

91

89

0.4622

0.6030

42

11.9 (0)

7.2 (0)

23.8 (0)

9.6 (0)

4.305 (+α)

93

90

0.4114

0.3894

43

11.9 (0)

7.2 (0)

23.8 (0)

9.6 (0)

2.4 (0)

96

96

0.2671

0.2772

44

11.9 (0)

7.2 (0)

23.8 (0)

9.6 (0)

2.4 (0)

97

96

0.2704

0.2772

45

11.9 (0)

7.2 (0)

23.8 (0)

9.6 (0)

2.4 (0)

98

96

0.2708

0.2772

46

11.9 (0)

7.2 (0)

23.8 (0)

9.6 (0)

2.4 (0)

97

96

0.2646

0.2772

47

11.9 (0)

7.2 (0)

23.8 (0)

9.6 (0)

2.4 (0)

97

96

0.2702

0.2772

48

11.9 (0)

7.2 (0)

23.8 (0)

9.6 (0)

2.4 (0)

97

96

0.2702

0.2772

49

11.9 (0)

7.2 (0)

23.8 (0)

9.6 (0)

2.4 (0)

97

96

0.2702

0.2772

50

11.9 (0)

7.2 (0)

23.8 (0)

9.6 (0)

2.4 (0)

97

96

0.2702

0.2772

51

11.9 (0)

7.2 (0)

23.8 (0)

9.6 (0)

2.4 (0)

97

96

0.2702

0.2772

52

11.9 (0)

7.2 (0)

23.8 (0)

9.6 (0)

2.4 (0)

97

96

0.2702

0.2772

X1 = Glucose (g l−1), X2 = Veratryl alcohol (mM), X3 = KH2PO4 (g l−1), X4 = MgSO4 (g l−1), X5 = CaCl2 (g l−1)

Coded levels of the factors are indicated in parentheses

Analytical methods

Estimation of lignin peroxidase enzyme and DR-80 dye concentration

Samples taken during the experiments were centrifuged at 10,000×g for 10 min at 4°C to remove fungal biomass. Biomass free samples so obtained were divided into two portions: one portion containing lignin peroxidase enzyme was assayed by spectrophotometric method (Kirk et al. 1990), which was based on the oxidation of veratryl alcohol to veratraldehyde. One unit (U) of lignin peroxidase enzyme activity was defined as the amount that converted 1 mol of veratryl alcohol to veratraldehyde per min per ml of the culture filtrate; the enzyme activities were expressed as U l−1 (Linko and Haapala 1993). The other portion was used for determining residual DR-80 concentration by measuring its absorbance (λmax DR- 80) at 528 nm using a UV–visible spectrophotometer (Carry 100, Varian, USA).

Results and discussion

Lignin peroxidase activity and dye decolourization profiles

In the present study both percentage dye decolourization and specific dye removal due to maximum LiP activity by the fungus were simultaneously investigated by screening and optimization of the media constituents with an aim to enhance the dye decolourization efficiency. Typical profiles of both LiP activity and DR-80 decolourization obtained from experimental run 1 of the Plackett-Burman screening design are presented in Fig. 1. It could be seen that the enzyme activity reached a maximum value of 85.14 U l−1 at 48 h of culture, and after which the activity decreased considerably. Similar profiles of lignin peroxidase enzyme by the fungus have been reported in the literature by other authors, as well (Sayadi and Ellouz 1995). Lignin peroxidase is one such important enzyme produced by P. chrysosporium that it is mainly responsible for decolourization of the azo dye. This aspect is clearly revealed in Fig. 1 where the concentration of DR-80 in the media was found to decrease with an increase in the enzyme activity that resulted in a significant decrease (~80%) of the dye within the first day itself (Fig. 1). Similar profiles of lignin peroxidase and decolourization of DR-80 were obtained in all the other experimental runs, and these are depicted in Table 1. However, variations in these two responses were obtained depending upon the factor and combinations thereof in each of the runs strongly suggesting the role of media constituents and their levels on the dye decolourization efficiency by the fungus.
https://static-content.springer.com/image/art%3A10.1007%2Fs10532-009-9319-2/MediaObjects/10532_2009_9319_Fig1_HTML.gif
Fig. 1

LiP activity and DR-80 decolourization profiles obtained in experimental run no. 1 of the Plackett-Burman screening design

Plackett-Burman design for screening of media constituents

The results obtained in the screening study (Table 1) were analyzed in the form of analysis of variance (ANOVA), which is a statistical technique that subdivides the total variation in a data set into component parts associated with specific sources of variation for the purpose of testing null hypotheses on the parameters in a model (Montgomery 1991). Table 3 presents the ANOVA of percentage dye decolourization and specific dye removal at maximum LiP enzyme activity, in which the mean sum of squares (MS) of the model terms was obtained from the ratio of sum of squares (SS) and degrees of freedom (df); the Fisher’s F value was calculated by dividing the MS owing to the model by the MS owing to error. Generally, a large F value and a corresponding low value of P for a term in the model indicate high significance of the term (Montgomery 1991). Therefore, from the ANOVA table, the term for main effects of media constituents in both the models for percentage dye decolourization and specific dye removal were highly significant at P value less than 0.01. Table 3 also shows a term for error in the models, the MS value of which indicates that the amount of variation in the response data that was left unexplained by the models is low. To assess the significance of each individual medium constituent on percentage dye decolourization and specific dye removal, Student’s t-test was performed and the results are presented in Table 4. From this table, glucose was found to have a maximum effect on percentage dye decolourization followed by KH2PO4, NH4Cl and veratryl alcohol; however, only CaCl2 and MgSO4·7H2O were found to significantly effect specific removal of the dye. Hence, out of the seven media constituents, the five media constituents of glucose, KH2PO4, veratryl alcohol, CaCl2 and MgSO4·7H2O were further chosen for optimizing their levels. In a similar study by Mohana et al. (2008), glucose and such media components were studied to be important in decolourization of Direct Black 32 dye by a bacterial consortium. Moreover, the observation that NH4Cl did not show any significant effect on both the responses was found to be in conformity with the literature where it is reported that under nitrogen excess conditions enzyme activity shown by the fungus is low, which therefore results in poor degradation of xenobiotics (Tien and Kirk 1983; Nagarajan and Annadurai 1999).
Table 3

ANOVA of percentage dye decolourization and specific dye removal in the screening study

Source

Percentage dye decolourization (%)

Specific dye removal (mg U−1)

SS

df

MS

F

P

SS

df

MS

F

P

Main effects

11252.7

8

1406.58

18.92

0.000

0.45784

8

0.06789

3.03

0.031

Curvature

751.9

1

751.89

10.11

0.005

0.1070

3

0.0356

1.59

0.233

Residual error

1264.0

17

74.35

  

0.33628

15

0.02241

  

Pure error

1148.2

14

82.01

  

0.04028

14

0.00287

  

Total

13268.5

26

    

26

   

df degrees of freedom, SS sum of squares, MS mean sum of squares

Table 4

Student t test for screening of media constituents by the Plackett-Burman method

Media constituents

Sp. dye removal due to maximum LiP activity (mg U−1)

Percentage dye decolourization (%)

T

P- value

T

P- value

Constant

21.81

0.000

42.35

0.000

Glucose

1.24

0.236*

5.33

0.000

Ammonium chloride

1.25

0.230*

−3.91

0.001

Potassium di hydrogen phosphate

−0.48

0.635*

−9.45

0.000

Calcium chloride

3.01

0.009

−0.40

0.692*

Magnesium sulphate

1.91

0.075*

0.07

0.944*

Tween 20

−1.72

0.106*

1.68

0.111*

Veratryl alcohol

0.26

0.801*

−3.39

0.004

Inoculum Size

0.95

0.357*

2.01

0.060*

* Indicates insignificant values of the factors at P > 0.050)

Optimization using response surface methodology

Table 2, which was earlier referred to present the experimental combinations of the levels of screened media constituents in the optimization study, also presents the results of percentage dye decolourization and specific dye removal. Based on the results obtained, regression model equations were developed for depicting relationship between the various media constituents and the responses on percentage dye decolourization and specific dye removal. These two models are presented in equations 3 and 4 respectively.
$$ \begin{array}{lll} f(X) &= & - 13.9 + 11.56X_{2} + 1.29X_{3} + 4.63X_{4} + 8.61X_{5} - 0.81X_{2}^{2} - 0.03X_{3}^{2} - 0.12X_{4}^{2} \\ && -1.88X_{5}^{2} + 0.097X_{2} X_{3} + 1.01X_{2} X_{5} - 0.01X_{3} X_{4} - 0.68X_{4} X_{5} \\ \end{array} $$
(3)
$$ \begin{array}{lll} f(X) & = & 3.08 - 0.16X_{2} - 0.578X_{3} - 0.11X_{4} - 0.62X_{5} + 0.002X_{2}^{2} + 0.005X_{3}^{2} + 0.002X_{4}^{2} \\ && + 0.056X_{5}^{2} + 0.002X_{2} X_{3} + 0.021X_{2} X_{5} + 0.002X_{3} X_{4} + 0.015X_{4} X_{5} \\ \end{array} $$
(4)
ANOVA results of the models for percentage dye decolourization and specific dye removal (mg l−1) were obtained as before in the screening study, and are given in Table 5. The Fisher’s F values of both the models owing to regression were found to be high, which indicates that most of the variation in the responses can be explained by the regression model equations. The associated P value, which is used to judge whether F is large enough to indicate statistical significance or not, suggest that all the interaction terms in both the regression models were found to be highly significant (P < 0.03). Overall, the regression models for percentage dye decolourization and specific dye removal was found to be highly accurate with P value less than 0.003. Also, from Table 2, both the experimental and predicted values of the two responses were found to match well with each other. These findings indicate that the second-order polynomial models for percentage dye decolourization and specific dye removal were adequate in representing the actual relationship between the media constituents and the responses. To determine significance of the regression coefficient of the factors in these models, the results were further subjected to Student t-test which is presented in Table 6. This particular result reveals that regression coefficients for the linear terms due to veratryl alcohol (X2) and MgSO4 (X4) in the model for percentage dye decolourization were found to be highly significant (P < 0.02); also, the coefficient of KH2PO4 (X3) was significant at P < 0.05 compared to those of other media constituents. However, the term for glucose (X1) was found to be only slightly significant. Compared to the findings on the significance of coefficients of linear terms in the models, all the coefficients of the quadratic terms in the model were found to be highly significant (P < 0.05). On the other hand, most of the regression coefficient terms for two-way interaction between factors, i.e. X1 × X2, X1 × X3, X1 × X4, X1 × X5, X2 × X4 and X3 × X5, were found to be statistically insignificant (P > 0.05) on both the responses.
Table 5

ANOVA of percentage dye decolourization and specific dye removal in the optimization study

Source

Percentage dye decolourization (%)

Specific dye removal (mg U−1)

SS

df

MS

F

P

SS

df

MS

F

P

Regression

13182.7

20

659.13

25.84

0.000

1.22805

20

0.06140

3.14

0.002

Linear

8614.9

5

1722.99

67.55

0.000

0.40173

5

0.08035

4.11

0.006

Square

3772.4

5

754.47

29.58

0.000

0.30597

5

0.06119

3.13

0.021

Interaction

795.4

10

79.54

3.12

0.007

0.5202

10

0.05204

2.66

0.018

Residual error

790.7

31

25.51

  

0.60619

31

0.01955

  

Pure error

 

9

0.12

  

0.00004

9

0.00001

  

Total

13973.4

51

    

51

   

df degrees of freedom, SS: sum of squares, MS mean sum of squares

Table 6

Estimated regression coefficients and their significance in the models for percentage dye decolourization and specific dye removal in the optimization study

Term

Percentage dye decolourization (%)

Specific dye removal (mg U−1)

Coef

SE Coef

T

P

Coef

SE Coef

T

P

Constant (C)

96.38

1.59

60.716

0.000

0.27720

0.0440

6.307

0.000

Glucose (X1)

2.00

1.825

1.099

0.280

−0.0510

−0.0505

1.011

0.320

Veratryl alcohol (X2)

32.95

1.825

18.053

0.000

−0.18

0.0505

−3.549

0.001

KH2PO4 (X3)

3.73

1.825

2.042

0.050

−0.078

0.0505

−1.545

0.132

MgSO4 (X4)

4.60

1.825

2.522

0.017

−0.014

0.0505

−0.273

0.786

CaCl2 (X5)

0.51

1.825

0.281

0.781

−0.107

0.0505

−2.114

0.043

X12

1.85

3.735

0.494

0.625

0.150

0.1034

1.454

0.156

X22

−41.15

3.735

−11.02

0.000

0.141

0.1034

1.360

0.184

X32

−20.30

3.735

−5.436

0.000

0.297

0.1034

2.871

0.007

X42

−10.33

3.735

−2.766

0.009

0.197

0.1034

1.902

0.067

X52

−6.64

3.735

−1.779

0.085

0.220

0.1034

2.118

0.042

X1X2

2.6

5.050

0.515

0.611

0.109

0.1398

0.782

0.440

X1X3

−5.21

5.050

−1.032

0.310

0.056

0.1398

0.400

0.692

X1X4

−5.78

5.050

−1.144

0.261

0.118

0.1398

0.844

0.405

X1X5

−5.60

5.050

−1.109

0.276

0.176

0.1398

1.258

0.218

X2X3

16.45

5.050

3.267

0.003

0.364

0.1398

2.609

0.014

X2X4

0.96

5.050

0.190

0.851

0.195

0.1398

1.399

0.172

X2X5

13.71

5.050

2.715

0.011

0.285

0.1398

2.039

0.050

X3X4

−2.61

5.050

−0.518

0.608

0.358

0.1398

2.562

0.015

X3X5

−8.58

5.050

−1.699

0.099

0.059

0.1398

0.422

0.676

X4X5

−12.45

5.050

−2.465

0.019

0.276

0.1398

1.971

0.058

Coef coefficient, SE Coef coefficient of standard error

In general, 2D response surface contour plots are graphical representation of relationship between response and experimental variables that can be used for determining the optimum conditions (Haider and Pakshirajan 2007; Tanyildizi et al. 2005). These contour plots show relative effects of any two variables when the concentration of the remaining variable is held constant at its center-point value; from the nature of the response surface contours whether elliptical, circular or saddle point, interaction between the variables may be predicted. Figures 2, 3, 4 and 5 depict response surface contour plots with significant interaction effects between the media constituents obtained in the study. The response surface contour plots between the variables CaCl2 and veratryl alcohol and that between KH2PO4 and veratryl alcohol (Figs. 2 and 3, respectively) were found to be elliptical and, hence, indicated significant interactions on percentage dye decolourization. Similarly, contour plot of mutual interaction between the variables KH2PO4 and veratryl alcohol (Fig. 4) on specific dye removal due to maximum enzyme activity indicated good significance. The interaction between glucose and KH2PO4 also showed good importance on the response (Fig. 5) which was, however, not revealed from the Student t test results presented earlier in Table 6. Ravikumar et al. (2005) also demonstrated such interaction effects between independent variables using contour plots on the removal of Astrazone Blue FRR (Basic Blue 69) and Teflon Blue ANL (Acid Blue 125) by a novel adsorbent based on 1:1 ratio of carbon and fly ash.
https://static-content.springer.com/image/art%3A10.1007%2Fs10532-009-9319-2/MediaObjects/10532_2009_9319_Fig2_HTML.gif
Fig. 2

2D contour plot showing interaction between calcium chloride and veratryl alcohol on the dye decolourization efficiency by P. chrysosporium

https://static-content.springer.com/image/art%3A10.1007%2Fs10532-009-9319-2/MediaObjects/10532_2009_9319_Fig3_HTML.gif
Fig. 3

2D contour plot showing interaction between KH2PO4 and veratryl alcohol on the dye decolourization efficiency by P. chrysosporium

https://static-content.springer.com/image/art%3A10.1007%2Fs10532-009-9319-2/MediaObjects/10532_2009_9319_Fig4_HTML.gif
Fig. 4

2D contour plot showing interaction between KH2PO4 and veratryl alcohol on specific dye removal by P. chrysosporium

https://static-content.springer.com/image/art%3A10.1007%2Fs10532-009-9319-2/MediaObjects/10532_2009_9319_Fig5_HTML.gif
Fig. 5

2D contour plot showing interaction between KH2PO4 and glucose on specific dye removal by P. chrysosporium

The model equations for percentage dye decolourization and specific dye removal due to maximum enzyme activity were solved by partially differentiating the equation and then equating it to zero. Corresponding local maxima were further checked by second-order sufficient condition using a Hessian matrix, and the optimized levels of the screened media constituents were thus obtained to be glucose: 11.9 g l−1, veratryl alcohol: 12.03 mM, KH2PO4: 23.08 g l−1, CaCl2: 2.4 g l−1, MgSO4: 10.47 g l−1. The optimum values of the constituents were also in close agreement with those found from the response surface contour plots (Figs. 2, 3, 4 and 5). Further, maximum predicted percentage dye decolourization and specific dye removal were found from the surface confined in the smallest curve of the contour diagrams. Experimental validation of the results were performed at the optimized settings of the media constituents in batch shake-flasks, and the corresponding percentage dye decolourization and specific dye removal were found to be 99.98% and 0.24 mg U−1 respectively, which perfectively matched with the predicted values of 99.99% of dye decolourization and 0.23 mg U−1 of specific dye removal.

Overall, the study not only gave good insight into the effect of various media constituents on DR-80 dye decolourization and LiP activity by the fungus, but also proved helpful in significant enhancement in dye decolourization by the fungus.

Conclusion

The present study clearly demonstrated enhancement in DR-80 decolourization efficiency by P. chrysosporium following screening and optimization of media constituents. Based on the Plackett-Burman design for screening the media constituents, glucose, MgSO4, KH2PO4, veratryl alcohol and CaCl2 were found to be the most influential factors affecting the percentage dye decolourization and specific dye removal due to maximum enzyme activity by the fungus. At the RSM optimized levels of the media constituents, P. chrysosporium showed complete (100%) dye decolourization efficiency with a specific DR-80 removal 0.24 mg U−1 due to its maximum LiP activity.

Acknowledgments

This study was funded by the Council of Scientific and Industrial Research (CSIR), India, under Scheme No. 38(1171)/07/EMR-II.

Copyright information

© Springer Science+Business Media B.V. 2009