Isolation, identification and characterization of a novel Rhodococcus sp. strain in biodegradation of tetrahydrofuran and its medium optimization using sequential statistics-based experimental designs

Bioresource Technology 100 (2009) 2762–2769

Contents lists available at ScienceDirect

Bioresource Technology journal homepage: www.elsevier.com/locate/biortech

Isolation, identification and characterization of a novel Rhodococcus sp. strain in biodegradation of tetrahydrofuran and its medium optimization using sequential statistics-based experimental designs Yanlai Yao, Zhenmei Lv, Hang Min *, Zhenhua Lv, Huipeng Jiao Institute of Microbiology, College of Life Science, Zhejiang University, 338 Yuhangtang Road, Hangzhou, Zhejiang 310058, China

a r t i c l e

i n f o

Article history: Received 4 November 2008 Received in revised form 31 December 2008 Accepted 11 January 2009 Available online 18 February 2009 Keywords: Tetrahydrofuran degradation Rhodococcus sp. YYL Plackett–Burman design Response surface methodology Box–Behnken design

a b s t r a c t Statistics-based experimental designs were applied to optimize the culture conditions for tetrahydrofuran (THF) degradation by a newly isolated Rhodococcus sp. YYL that tolerates high THF concentrations. Single factor experiments were undertaken for determining the optimum range of each of four factors (initial pH and concentrations of K2HPO4  3H2O, NH4Cl and yeast extract) and these factors were subsequently optimized using the response surface methodology. The Plackett–Burman design was used to identify three trace elements (Mg2+, Zn2+and Fe2+) that significantly increased the THF degradation rate. The optimum conditions were found to be: 1.80 g/L NH4Cl, 0.81 g/L K2HPO4  3H2O, 0.06 g/L yeast extract, 0.40 g/L MgSO4  7H2O, 0.006 g/L ZnSO4  7H2O, 0.024 g/L FeSO4  7H2O, and an initial pH of 8.26. Under these optimized conditions, the maximum THF degradation rate increased to 137.60 mg THF h1 g dry weight in Rhodococcus sp. YYL, which was nearly five times of that by the previously described THF degrading Rhodococcus strain. Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction Tetrahydrofuran (THF) is one of the most polar ethers and a widely used solvent for polar reagents. A large amount of THF is released into the environment and it has been detected in the groundwater, where it causes contamination problems, health problems and even explosions since it is very soluble in water with any ratio and has a relatively low boiling point (Draper et al., 1997; Isaacson et al., 2006; Moody, 1991; van Ravenzwaay et al., 2003). In addition, its carcinogenic activity has been demonstrated in carcinogenesis studies conducted by the National Toxicology Program (NTP) (Chhabra et al., 1998; Hermida et al., 2006). Acute toxicity tests have also illustrated that exposure to THF can cause central nervous system irritation, narcosis, edema and clonic muscle spasms in animals (Chhabra et al., 1990; Katahira et al., 1982; Malley et al., 2001). THF is classified as relatively non-biodegradable over time (Painter and King, 1985). Until now, only a few microorganisms have been reported that can utilize THF as the sole carbon source (Bernhardt and Diekmann, 1991; Kohlweyer et al., 2000; Mahendra and Alvarez-Cohen, 2005; Nakamiya et al., 2005; Parales et al., 1994). In addition, the media used for THF degradation by these microorganisms have usually been inherited from other studies and it has to take a relatively long time to completely remove * Corresponding author. Tel.: +86 0571 88206279. E-mail address: [email protected] (H. Min). 0960-8524/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved. doi:10.1016/j.biortech.2009.01.006

THF contamination (Bernhardt and Diekmann, 1991; Kohlweyer et al., 2000). Therefore, by optimizing the culture conditions to eliminate the potential danger as soon as possible, it may be possible to accelerate THF removal. The optimization of culture conditions is of importance for the development of any biotreatment processes for removing THF contamination. The use of statistics-based experimental design in the optimization of biotreatment processes has been well documented (Annadurai et al., 2008; Yus Azila et al., 2008). The Plackett–Burman design (PBD) has been frequently used to screen variables that have a significant impact on the process (Imandi et al., 2008). The response surface methodology with Box–Behnken design is regularly used to statistically evaluate the main and interactive effects of variables and to optimize the parameters of biotreatment processes (Mohana et al., 2008). In this paper, a pure Rhodococcus strain YYL capable of using THF as a sole carbon source and tolerating high THF concentrations was isolated and identified. Sequential statistics-based experimental designs were applied to optimize the media components and initial pH for obtaining the maximum THF degradation rate by strain YYL. The PBD was used for screening trace elements that significantly increase THF degradation. Single factor experiments were used to determine the optimum range of several parameters. The response surface methodology with Box–Behnken design was subsequently applied to determine the effects of significant parameters, and their interactions, in the removal of THF and to identify the optimum values. Finally, the optimum conditions were

2763

Y. Yao et al. / Bioresource Technology 100 (2009) 2762–2769

Table 1 Assigned concentrations of eight trace elements at two levels in Plackett–Burman design and ANOVA of the results for THF degradation rate by Rhodococcus sp. YYL using model fitting with JMP Software. Trace element

MgSO4  7H2O FeSO4  7H2O ZnSO4  7H2O MnSO4  H2O CoCl2  H2O CuCl2  H2O BaCl2  2H2O CaCl2

Level (g/L) Low

High

0 0 0 0 0 0 0 0

0.400 0.024 0.006 0.006 0.002 0.010 0.015 0.002

Mean square

Coefficient estimate

Standard error

F-value

P-value

0.118 0.059 0.176 0.003 0.098 0.663 0.006 0.074

0.055 0.039 0.067 0.009 0.050 0.131 0.012 0.042

0.017 0.017 0.017 0.017 0.017 0.017 0.017 0.016

10.403 5.175 15.542 0.264 8.663 58.447 0.489 6.497

0.003 0.029 <0.001 0.611 0.006 <0.001 0.489 0.016

experimentally validated by directly comparing the rate of THF degradation under the opitmized culture conditions with the initial, unoptimized conditions. 2. Methods 2.1. Isolation of THF-degrading organisms Four activated sludge samples were obtained from a wastewater treatment plant, a garbage treatment plant, chemical plants and pharmaceutical factory in Hangzhou, Zhejiang, China. The samples were incubated with the addition of THF once a week in concentrations from 0.1% to 1.0% (v/v) for a month to enrich for potential THF-degrading bacteria. The spread-plate method was used for isolation of pure strains. Single colonies were picked up and transferred to additional new plates until homogenous colonies were observed. Finally, the organisms were inoculated into mineral media with THF as the sole carbon source for proving their THF degrading ability.

assessed by testing the ability of the strains to migrate from the point of inoculation through semisolid (0.3%) agar plates. The utilization of some substrates as carbon sources and nitrogen sources by the isolate was performed based on Bergey’s Manual of Determinative Bacteriology (Buchanan et al., 1984). Sequence analysis of the 16S rDNA was performed by amplifying the 16S rDNA of the isolate with PCR (Lü et al., 2003). The obtained 16S rDNA sequence was aligned to sequences in GenBank using the BLAST program. The aligned 16S rDNA sequences of the related species were retrieved from the NCBI nucleotide database. The program Clustal X (version 1.8) with default parameters was run for multiple sequence alignment. Phylogenetic and distance analysis of the aligned sequences was performed by the program MEGA (version 2.1). The resulting unrooted tree topologies were evaluated by bootstrap analysis of the neighbor-joining method based on 1000 resamplings. The accession number of the 16S rDNA in GenBank is EF396959. 2.4. Single factor experiments for determining the optimal range of four factors

2.2. Media and growth conditions Modified base mineral medium (BMM) was initially used with 5 mM THF added as a sole carbon source for isolating and culturing THF-degrading organisms (Bernhardt and Diekmann, 1991). One liter of BMM contained 0.50 g K2HPO4  3H2O, 0.50 g NH4Cl, 0.20 g MgSO4  7H2O, and 0.10 g yeast extract instead of a vitamin solution to provide the necessary growth factors. The initial pH value of media was 7.2. All of the enrichments and cultures were carried out in airtight flasks shaken at 130 rpm and 30 °C. To provide sufficient oxygen for cell growth and THF degradation, each flask contained only 40% of its maximum volume of BMM, with 5 mM THF, during the optimization processes. THF degradation rate was taken as the optimization target. It was calculated as the following formula: THF degradation rate = C/C0, where C was the THF concentration in the solution after certain hours of culture and C0 was the THF concentration at 0 h. 2.3. Identification of strain YYL The colony morphology of the isolated strain was observed on agar plates after 4 days of culturing at 30 °C. The cell morphology of the isolated strain was examined after being cultured in liquid BMM for 24, 48, 72, 96 and 108 h by microscopy (1600 magnification). The cell morphology was also observed using a JEM-120 transmission electron microscope (JEOL, Japan) after 72 h of culture. The physiological and biochemical characteristics of the THFdegrading organism were examined using standard procedures (Dong and Cai, 2001). Gram staining, catalase- and oxidase-activities and other characteristics were investigated. Motility was

The effects of varying the initial pH value (from 5.0 to 10.0) and the concentrations of three other factors (K2HPO4  3H2O, NH4Cl and yeast extract) on THF degradation were initially studied by single factor experiments. Four levels was arranged for K2HPO4  3H2O (0.1, 0.5, 1.0, and 5.0 g/L), NH4Cl (1.0, 2.0, 3.0, and 4.0 g/L)and yeast extract (0, 0.1, 0.5, 1.0 g/L). In each experiment, one factor was changed, with the other factors remaining constant. The effects of these factors were evaluated by measuring the THF degradation rate after 72 h of culture, where each experiment was triplicated. 2.5. Plackett–Burman design for trace element screening Plackett–Burman designs are very efficient screening designs when only main effects, rather than interactive effects, are of interest. Here, it was used for screening the trace elements that would accelerate THF degradation. Eight trace elements (Mg2+, Zn2+, Cu2+, Fe2+, Co2+, Ba2+, Ca2+ and Mn2+) were chosen as factors based on previously described studies (Bernhardt and Diekmann, 1991;

Table 2 Levels of the Box–Behnken experimental design. Independent variables

Initial pH NH4Cl K2HPO4  3H2O Yeast extract

Code

X1 X2 X3 X4

None code

A B C D

Code levels 1

0

1

7.0 1.0 g/L 0.25 g/L 0.025 g/L

8.0 2.0 g/L 0.625 g/L 0.0625 g/L

9.0 3.0 g/L 1 g/L 0.1 g/L

The relations between the code values and none code values were: X1 = (A8.0)/1.0, X2 = (B2.0)/1.0, X3 = (C0.625)/0.375, X4 = (D0.0625)/0.0375.

2764

Y. Yao et al. / Bioresource Technology 100 (2009) 2762–2769

Fig. 1. Unrooted tree based on the polygenetic analysis of 16S rDNA sequences showing the position of strain YYL. The tree was evaluated by bootstrap analysis of the neighbor-joining method based on 1000 resamplings.

Kohlweyer et al., 2000). THF degradation rates were evaluated after 72 h of culture in either BMM containing a specific concentration (Table 1) of each trace element (individually) or BMM alone. The experimental design and analysis was performed using JMP Software (version 7.0.1, SAS Institute Inc., Cary, NC, USA). Each experiment was performed in duplicate and a total of 42 trials were carried out.

Shanghai Tianmei Scientific Instrument Corporation, China) equipped with a PEG6000 column (Dalian Zhonghuida Scientific Instrument. Co. Ltd., China) and a flame ionization detector. The column was operated at 120 °C, the injector at 160 °C and the detection room at 180 °C. Nitrogen gas was used as carrier at a speed of 16 ml/min. 3. Results and discussion

2.6. Response surface methodology for optimizing the medium components and initial pH The optimal ranges of the four factors were identified with single factor experiments for the initial pH, and the concentrations of K2HPO43H2O, NH4Cl and yeast extract. Response surface methodology with a Box–Behnken design was used for analyzing the main and interactive effects of these parameters on THF degradation using JMP Software. The optimum ranges of the four factors obtained in the single factor experiments were used as reference levels. The independent variables along with their levels are shown in Table 2. THF degradation rates were measured after 48 h of culture. Experimental design and analysis were performed with JMP Software. 2.7. Analysis of THF concentration Cultures were centrifuged at 10,000 g for 10 min and 2 ll of the supernatant was analyzed for THF concentration by GC (7890,

3.1. Isolation of THF-degrading strains As THF has been shown to be relatively non-biodegradable, four sludge samples were collected for isolating potentially THFdegrading strains. However, none of the organisms derived from each sample could grow on THF as the sole carbon source, even though they were enriched by adding THF for a long time. Many other media formulations with different vitamin solutions and/or trace elements were tested, but each attempt was unsuccessful in isolating a THF-degrading strain. However, by chance, after four samples were mixed together and incubated with THF, evidence of THF degradation was seen and three purified organisms were isolated from the mixture. One strain, designated as YYL, could utilize THF as the sole carbon source while the remaining two could not. Strain YYL was found to be capable of tolerating high concentrations of THF (up to 200 mM), although the growing lag phase more than 96 h was observed when the concentration was higher than 100 mM (data not shown). Until now, the previously described

2765

Y. Yao et al. / Bioresource Technology 100 (2009) 2762–2769

5.0

0.6

4.0

0.5 0.4

3.0

0.3

2.0

0.2 0.1 0

24

48

72

0.8

6.0

0.7 5.0 0.6 0.5

4.0

0.4

3.0

0.3 2.0 0.2

1.0

0.1

0.0

0

1.0 0.0 0

96

24

0.9

5

0.8

4.5

0.7

4 3.5

OD 560

0.6

3

0.5

2.5 0.4

2

z

0.3

1.5

0.2

1

0.1

0.5

0

0 0

24

48

96

72

96

Time (h)

D 1.2

7.0

1

6.0 5.0

0.8

OD560

B

72

Time (h)

THF concentratin (mM)

Time (h)

48

4.0 0.6 3.0 0.4

2.0

0.2

THF concentratin (mM)

OD 560

0.7

7.0

THF concentratin (mM)

6.0

0.8

0

C 0.9

7.0

0.9

OD560

1

THF concentration (mM)

A

1.0

0

0.0 0

24

48

72

96

Time (h)

Fig. 2. Single factor experiments demonstrating the effect of four factors (the initial medium pH and the concentration of K2HPO4  3H2O, NH4Cl and yeast extract) on THF degradation. One factor was varied with the three remaining constant. The solid and broken lines represent cell growth and THF degradation, respectively. A: Effect of different initial medium pH on THF degradation and cell growth. The symbols indicate initial pH of 5.0(j), 6.0(), 7.0(N), 8.0(d), 9.0(h), 10.0(e). B: Effect of different concentrations of K2HPO4  3H2O on THF degradation and cell growth. The symbols indicate concentrations of 0.1(j), 0.5(), 1.0(N), and 5.0(d) g/L. C: Effect of different concentrations of NH4Cl on THF degradation and cell growth. The symbols indicate concentrations of 1.0(j), 2.0(), 3.0(N), and 4.0(d) g/L. D: Effect of different concentrations of yeast concentration on THF degradation and cell growth. The symbols indicate concentrations of 0(j), 0.1(), 0.5(N), and 1(d) g/L.

THF degrading strains could tolerate a THF concentration of only about 60 mM (Kohlweyer et al., 2000). The tolerance of this isolate to THF was far higher than that of the previously described strains. 3.2. Identification of strain YYL The THF-degrading organism was determined to be pure after the visual observation of colonies on solid medium plates that had been transferred for several generations and the microscopic observation of cells (magnification, 1600) were demonstrated colonies and cells uniform in morphology. Strain YYL was found to be a Gram-positive, rod-shaped bacterium, aerobic and non-motile. The cell morphological cycle of strain YYL began as a short rod, followed by a long rod on the second day and, by fragmentation of the long rod, ending with a short rod again after 3 days. Strain YYL predominantly formed smoothlooking orange colonies after 2 to 3 days of incubation, however some rough-looking orange colonies could be observed. Substrate-utilization experiments were performed demonstrating that strain YYL was capable of using a variety of substrates, including most alcohols, carboxylic acids, aromatic compounds and all tested inorganic nitrogen sources, except for methanol and formic acid, as the sole carbon and nitrogen sources.

An approximately 1.5 kb 16S rDNA fragment was amplified from the total DNA of strain YYL and partially sequenced. After alignment with other 16S rDNA sequences in GenBank, it had a high degree of similarity (96%) to other members of genus Rhodococcus. The 16S rDNA sequence was further aligned with the corresponding sequences from additional strains of species of Rhodococcus in Bergey’s Manual of Determinative Bacteriology, as well as representatives of other constituent taxa of mycolic acid-containing bacteria belonging to Nocardiaceae family that were retrieved from GenBank. A phylogenetic tree based on all known representatives of validly described Rhodococcus sp. and other related species is shown in Fig. 1. Phylogenetic analysis revealed that strain YYL was clustered closely with Rhodococcus rubber, which was first reported to be a pure strain capable of degrading THF. 3.3. Single factor experiments to approach the optimum range of four factors 3.3.1. Effect of the initial pH on THF degradation The initial pH of the culture medium plays an important role in microbial growth and enzyme activity and is one of the most important parameters taken into consideration in the development

2766

Y. Yao et al. / Bioresource Technology 100 (2009) 2762–2769

Table 3 The Box–Behnken experimental design along with the corresponding response in the terms of actual and predicted THF degradation rate by Rhodococcus sp. YYL. Run No.

Model

Initial pH

NH4Cl(g/L)

K2HPO4  3H2O(g/L)

Yeast extract (g/L)

1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27

0+0 +00 +00 00+ ++00 0000 00 0000 +00 0++0 00+ +00 00 0000 0+0 0+0+ 00+ +0+0 00 00 0+0 00+ +00+ 00 0+0 00++ 00

8.0 9.0 7.0 8.0 9.0 8.0 8.0 8.0 9.0 8.0 7.0 9.0 8.0 8.0 8.0 8.0 8.0 9.0 7.0 7.0 8.0 8.0 9.0 7.0 7.0 8.0 8.0

3.0 1.0 3.0 2.0 3.0 2.0 1.0 2.0 2.0 3.0 2.0 2.0 1.0 2.0 3.0 3.0 2.0 2.0 2.0 2.0 1.0 1.0 2.0 1.0 2.0 2.0 2.0

0.250 0.625 0.625 0.250 0.625 0.625 0.250 0.625 0.250 1.000 0.625 0.625 0.625 0.625 0.625 0.625 1.000 1.000 0.250 0.625 1.000 0.625 0.625 0.625 1.000 1.000 0.250

0.0625 0.0625 0.0625 0.1000 0.0625 0.0625 0.0625 0.0625 0.0625 0.0625 0.1000 0.0250 0.0250 0.0625 0.0250 0.1000 0.0250 0.0625 0.0625 0.0250 0.0625 0.1000 0.1000 0.0625 0.0625 0.1000 0.0250

of biotreatment processes (El Hajjouji et al., 2008; Hung et al., 2008; Yus Azila et al., 2008). The effects of the initial pH (from 5.0 to 10.0) on the THF degradation rate and cell growth were investigated. As shown in Fig. 2A, the initial pH affected cell growth and THF degradation greatly. When the initial pH in medium was lower than 7.0, the growth and THF degradation rate significantly decreased and THF could not ultimately be completely utilized within 96 h. The maximum degradation rate was observed when the initial pH was between 7.0 and 10.0. Therefore, the optimum range of the initial pH was thought to be between 7.0 and 10.0. 3.3.2. Effects of the concentration of three media components on THF degradation The components of the growth media provide the necessary nutrition and energy for cell growth and have important roles in the biotreatment process. A proper proportion of components in the media results in the best performance of cells. The concentration of three major components (K2HPO4  3H2O, NH4Cl and yeast extract) were evaluated in single factor experiments. The effects of four concentrations of K2HPO4  3H2O on THF degradation rate were tested, keeping the concentrations of the other factors unchanged. As shown in Fig. 2B, the THF degradation and growth rates increased with the increasing of K2HPO4  3H2O concentration. However, when the concentration was greater than 1.0 g/L, no significant difference in the THF degradation rate was observed. Similarly, four concentrations of NH4Cl were tested As shown in Fig. 2C, the THF degradation rate increased with the increasing of nitrogen concentration. Meanwhile, no significant differences were observed at concentrations above 2.0 g/L. Therefore, the maximum THF degradation rate was achieved at concentrations between 1.0 and 3.0 g/L. Yeast extract is often used for providing necessary growth factors. However, too high concentrations of yeast extract would weaken the use of other carbon sources, such as THF. As seen in Fig. 2D, while higher concentrations of yeast extract improve cell

Response (THF degradation rate) Actual value

Predicted value

0.950 0.983 0.991 1.000 0.665 0.965 0.990 1.000 1.000 1.000 1.000 1.000 0.915 1.000 1.000 1.000 1.000 1.000 0.353 0.715 0.959 1.000 1.000 0.478 0.960 1.000 1.000

0.889 1.000 0.940 1.000 0.716 0.988 0.883 0.988 1.000 1.000 0.898 1.000 0.915 0.988 1.000 1.000 1.047 0.836 0.540 0.694 0.946 1.000 0.948 0.478 0.948 1.000 0.942

growth, it provided no notable benefit to the degradation rate of THF. It was obvious that the lack of yeast extract reduced the THF degradation rate (Fig. 2D). The optimum concentration was found to be between 0 and 0.1 g/L. 3.4. Plackett–Burman design for trace element screening Trace elements are important for both cell growth and the activity of most enzymes (Ferreyra et al., 2002; Wei et al., 2007). To explore the effect of trace elements on THF degradation, eight trace elements were chosen and a Plackett–Burman design was used to select the most significant ones. The assigned concentrations of

Table 4 Analysis of variance (ANOVA) for the regression model. Error source

DF

SS

MS

F-value

P-value

pH(7,9) NH4Cl (1,3) K2HPO4  3H2O (0.25,1) Yeast extract (0.025,0.1) pH* NH4Cl pH* K2HPO4  3H2O NH4Cl* K2HPO4  3H2O pH* yeast extract NH4Cl* yeast extract K2HPO4  3H2O* yeast extract pH*pH NH4Cl* NH4Cl K2HPO4  3H2O* K2HPO4  3H2O Yeast extract* yeast extract Regression Error Total R2 = 0.826

1 1 1 1 1 1 1 1 1 1 1 1 1 1 14 12 26

0.110 0.007 0.033 0.011 0.172 0.092 0.002 0.020 0.002 0.000 0.106 0.010 0.000 0.011 0.611 0.129 0.740

0.110 0.007 0.033 0.011 0.172 0.092 0.002 0.020 0.002 0.000 0.106 0.010 0.000 0.011 0.044 0.011

10.258 0.613 3.035 1.058 16.043 8.573 0.153 1.883 0.167 0.000 9.868 0.903 0.030 0.986 4.056

0.0076 0.4487 0.1070 0.3240 0.0017 0.0127 0.7028 0.1950 0.6900 0.9996 0.0085 0.3607 0.8654 0.3403 0.0100

DF, degree of freedom; SS, sum of squares; MS, mean square. * Significant (P < 0.05). ** Very significant (P < 0.01).

**

** *

**

**

2767

Y. Yao et al. / Bioresource Technology 100 (2009) 2762–2769

7

1 0.9

Growth (OD560)

5

0.7 0.6

4

0.5 3

0.4 0.3

2

0.2

THF concentration (mM)

6 0.8

1 0.1 0

0 0

4

8

12

16

20

24

28

32

36

40

Time (h) Fig. 4. Experimental validation of the optimum conditions identified from the Plackett–Burman design and RSM compared with the original conditions. The real and broken lines represent cell growth and THF degradation, respectively. The symbols indicate the optimum (j) and original (N) values.

Six of the eight elements had a significant effect on THF degradation rate (p < 0.05) and five of these six elements had very significant (p < 0.01) effects. These results demonstrate that trace elements have an important influence on microbe activity: Mg2+, Zn2+ and Fe2+ increased THF degradation, while Cu2+, Co2+ and Ca2+ decreased THF degradation. Cu2+ was the most toxic trace element and the strain could not even grow in the media containing copper at the concentration used in the test. Further single factor experiments showed that the effect of different concentrations of Mg2+, Zn2+ and Fe2+ on THF degradation had no significant difference (data not shown). This was not surprising given the low concentrations needed and the concentrations tested in this research were therefore determined to be optimal. 3.5. Response surface methodology (RSM) The optimum ranges of four critical factors (initial medium pH and concentrations of NH4Cl, K2HPO43H2O and yeast extract) affecting the THF degradation rate were identified by signal factor experiments and selected for further optimization by response surface methodology. A Box–Behnken experimental design was applied to analyze the main and interactive effects of these variables and to find the optimum values within a coded range of 1–+1 in relation to THF degradation rate (Table 2). The experimental design and responses are given in Table 3. The experimental results were modeled with a second-order polynomial equation to explain the dependence of THF degradation rate on the different factors. The mathematical regression model using the coded factors is given as:

Y ¼ 0:988 þ 0:096X 1 þ 0:023X 2 þ 0:052X 3 þ 0:031X 4  0:208X 1 X 2  0:152X 1 X 3 þ 0:020X 2 X 3  0:071X 1 X 4  0:021X 2 X 4  0:00002X 3 X 4  0:141X 1 X 1  0:043X 2 X 2  0:008X 3 X 3 þ 0:045X 4 X 4 ; Fig. 3. Three-dimensional curved surfaces of the effect of four variables on the THF degradation rate. When the effect of two variables was plotted, the other two variables were set at central levels.

the eight trace elements at two levels, along with Analysis of variance (ANOVA) results, are shown in Table 1.

ð1Þ

where Y is the predicted response of THF degradation rate and the initial pH (X1) and concentrations of NH4Cl  (X2), K2HPO4  3H2O (X3) and yeast extract (X4) are the coded variables. The regression coefficients and the interaction between each independent factor were considered statistically significant for p-values below 0.05 and very significant for p-values below 0.01. As measured by ANOVA (Table 4), the quadratic regression model

2768

Y. Yao et al. / Bioresource Technology 100 (2009) 2762–2769

has a correlation coefficient (R2) of 0.826 for THF degradation and is very significant, as indicated by a low probability value ((Pmodel > F) = 0.01) by the Fisher0 s exact statistical test. This demonstrates that the model-predicted values are in perfect agreement with the experimental values. The result presented linear and quadratic coefficients of the initial pH and the interaction between the initial pH and the concentration of NH4Cl as very significant and the interaction between the initial pH and the concentration of K2HPO4  3H2O as significant. Using RSM, the effects of the independent factors (initial medium pH, concentrations of NH4Cl, K2HPO4  H2O and yeast extract) and their interactions on the THF degradation are represented, the response can be predicted and the optimum values of THF degradation can be determined. The three-dimensional response surfaces were generated to directly study the interactions among the four factors tested and to visualize the combined effects of factors on the THF degradation rate (Fig. 3). The response surface plots showed THF degradation rate by Rhodococcus sp. YYL as function of two factors, while the other variables set at their zero level. The initial medium pH played an important role in THF degradation. Plots 3A, 3B and 3C clearly shows that the THF degradation rate was sensitive to alterations of in the initial pH and that it was elevated from 60% to 100% when the initial medium pH value changed from 7.0 to 8.26. When the initial pH value was higher than 8.26, the degradation rate decreased due to the precipitation of trace elements and the inactivation of enzymes. The degradation rate also reached the maximum value when the concentrations of NH4Cl, K2HPO4  3H2O and yeast extract increased to a certain extent. However, the effects of these variables and their interaction were not notable, with the concentration of yeast extract only slightly influencing THF degradation (data not shown). Plots 3A and 3B show the interaction between initial pH and the concentrations of NH4Cl and K2HPO4  3H2O, respectively. These interactions had a significant effect on THF degradation. The degradation rate increased quickly with a small increase of both the initial pH and the concentrations of NH4Cl and K2HPO4  3H2O. The interactions between these variables were most likely due to the buffering capacity of NH4Cl and K2HPO4  3H2O in the solution, as supplementation of the media with relatively high concentrations of K2HPO4  3H2O kept the media at a high pH, which was more suitable for cell growth and THF degradation. Overall, considering the main and interactive effects of the four factors, the optimal conditions were determined to be an initial pH of 8.26, 0.81 g/L K2HPO4  3H2O and 1.80 g/L NH4Cl, respectively, using JMP software by solving the regression equation. 3.6. Model validation Based on the results of the PBD and RSM, the optimum conditions for THF degradation were an initial pH of 8.26, 1.80 g/L NH4Cl, 0.81 g/L K2HPO4  3H2O, 0.06 g/L yeast extract, 0.40 g/L MgSO4  7H2O, 0.006 g/L ZnSO4  7H2O and 0.024 g/L FeSO4  7H2O. These conditions were tested for THF degradation and compared to the original media formulation. THF degradation experiments were conducted in triplicate and the cultures grown in the original formulation were set as blank controls. As seen in Fig. 4, when culture in the optimized media strain YYL completely degraded 6 mM THF within 32 h. Under the initial culture conditions, however, strain YLL could only eliminate 72.76% of the THF in the same period, and the degradation rate slowed after 32 h. Compared with the first reported THF-degrading Rhodococcus species, the maximum growth rate (l) and maximum THF degradation rate increased to 0.0892 h1 and 137.60 mg THF h1?g dry weight from 0.019 h1 and 28.5 mg THF h1?g dry weight, respectively, corresponding

to a yield (Yx/s) of 0.64 (mg dry cell weight/mg THF)(Bernhardt and Diekmann, 1991). 4. Conclusions A pure bacterial strain capable of degrading THF from mixture of four biological samples was isolated and identified as Rhodococcus sp. YYL. Sequential statistics-based experimental designs were adopted to optimize the culturing conditions for maximal THF degradation. By using a Plackett–Burman design, three important trace elements (Mg2+, Zn2+ and Fe2+) were identified that have significant positive effects on the THF degradation. The optimum values of four variables (initial pH and concentrations of NH4Cl, K2HPO4  3H2O and yeast extract) were identified through RSM to be 8.26, 1.80 g/L, 0.81 g/L and 0.06 g/L, respectively. These optimized conditions were further experimentally validated and significantly improved THF degradation. Even though it has been demonstrated that THF is not easily biodegradable, these results demonstrate that THF degradation can be accelerated by the optimization of culture conditions. Acknowledgements This work was financially supported by Research Fund of Science and Technology Bureau of Zhejiang Province (No. 2008C23088) and the National Key Technologies Research and Development Program of China during the 11th Five-Year Plan Period (No. 2006BAJ08B01), 863 High Technology Program (2007AA10Z409). References Annadurai, G., Ling, L.Y., Lee, J.F., 2008. Statistical optimization of medium components and growth conditions by response surface methodology to enhance phenol degradation by Pseudomonas putida. J. Hazard. Mater. 151, 171–178. Bernhardt, D., Diekmann, H., 1991. Degradation of dioxane, tetrahydrofuran and other cyclic ethers by an environmental Rhodococcus strain. Appl. Microbiol. Biotechnol. 36, 120–123. Buchanan, P., Schubert, H.W., Holt, J.K., et al., 1984, 9th[M]. Bergey’s Manual of Determinative Bacteriology The Williams and Wilklins Company, Beltimore. Chhabra, R.S., Elwell, M.R., Chou, B., Miller, R.A., Renne, R.A., 1990. Subchronic toxicity of tetrahydrofuran vapors in rats and mice. Fund. Appl. Toxicol. 14, 338–345. Chhabra, R.S., Herbert, R.A., Roycroft, J.H., Chou, B., Miller, R.A., Renne, R.A., 1998. Carcinogenesis studies of tetrahydrofuran vapors in rats and mice. Toxicol. Sci. 41, 183–188. Dong, X., Cai, M., 2001. Manual of Bacteria Identify. Science Press, Beijing. Draper, A.J., Madan, A., Parkinson, A., 1997. Inhibition of coumarin 7-hydroxylase activity in human liver microsomes. Arch. Biochem. Biophys. 341, 47–61. El Hajjouji, H., Ait Baddi, G., Yaacoubi, A., Hamdi, H., Winterton, P., Revel, J.C., Hafidi, M., 2008. Optimisation of biodegradation conditions for the treatment of olive mill wastewater. Bioresour. Technol. 99, 5505–5510. Ferreyra, O.A., Cavalitto, S.F., Hours, R.A., Ertola, R.J., 2002. Influence of trace elements on enzyme production: protopectinase expression by a Geotrichum klebahnii strain. Enzyme Microb. Technol. 31, 498–504. Hermida, S.A., Possari, E.P., Souza, D.B., a Arruda Campos, I.P., Gomes, O.F., Di Mascio, P., Medeiros, M.H., Loureiro, A.P., 2006. 20 -deoxyguanosine, 20 -deoxycytidine, and 20 -deoxyadenosine adducts resulting from the reaction of tetrahydrofuran with DNA bases. Chem. Res. Toxicol. 19, 927–936. Hung, Y.J., Peng, C.C., Tzen, J.T., Chen, M.J., Liu, J.R., 2008. Immobilization of Neocallimastix patriciarum xylanase on artificial oil bodies and statistical optimization of enzyme activity. Bioresour. Technol. 99, 8662–8666. Imandi, S.B., Bandaru, V.V., Somalanka, S.R., Bandaru, S.R., Garapati, H.R., 2008. Application of statistical experimental designs for the optimization of medium constituents for the production of citric acid from pineapple waste. Bioresour. Technol. 99, 4445–4450. Isaacson, C., Mohr, T.K., Field, J.A., 2006. Quantitative determination of 1, 4-dioxane and tetrahydrofuran in groundwater by solid phase extraction GC/MS/MS. Environ. Sci. Technol. 40, 7305–7311. Katahira, T., Teramoto, K., Horiguchi, S., 1982. Experimental studies on the acute toxicity of tetrahydrofuran in animals. Sangyo Igaku. 24, 373–378. Kohlweyer, U., Thiemer, B., Schrader, T., Andreesen, J.R., 2000. Tetrahydrofuran degradation by a newly isolated culture of Pseudonocardia sp. strain K1. FEMS Microbiol. Lett. 186, 301–306.

Y. Yao et al. / Bioresource Technology 100 (2009) 2762–2769 Lü, Z., Min, H., Wu, S.W., Ruan, A.D., 2003. Phylogenetic and degradation characterization of Burkholderia cepacia WZ1 degrading herbicide quinclorac. J. Environ. Sci. Heath, Part (B) 38, 771–782. Mahendra, S., Alvarez-Cohen, L., 2005. Pseudonocardia dioxanivorans sp. nov., a novel actinomycete that grows on 1,4-dioxane. Int. J. Syst. Evol. Microbiol. 55, 593–598. Malley, L.A., Christoph, G.R., Stadler, J.C., Hansen, J.F., Biesemeier, J.A., Jasti, S.L., 2001. Acute and subchronic neurotoxicological evaluation of tetrahydrofuran by inhalation in rats. Drug Chem. Toxicol. 24, 201–219. Mohana, S., Shah, A., Divecha, J., Madamwar, D., 2008. Xylanase production by Burkholderia sp. DMAX strain under solid state fermentation using distillery spent wash. Bioresour. Technol. 99, 7553–7564. Moody, D., 1991. The effect of tetrahydrofuran on biological systems: does a hepatotoxic potential exist? Drug Chem. Toxicol. 14, 319–342. Nakamiya, K., Hashimoto, S., Ito, H., Edmonds, J.S., Morita, M., 2005. Degradation of 1,4-dioxane and cyclic ethers by an isolated fungus. Appl. Environ. Microbiol. 71, 1254–1258.

2769

Painter, H.A., King, E.F., 1985. Ring test program 1983–84. Assessment of biodegradability of chemicals in water by manometric respirometry. Comm. Eur. Commun. (Rep.). Parales, R.E., Adamus, J.E., White, N., May, H.D., 1994. Degradation of 1,4-Dioxne by an actinomycete in pure culture. Appl. Environ. Microbiol. 60, 4527– 4530. van Ravenzwaay, B., Gamer, A.O., Leibold, E., Kaufmann, W., 2003. Effect of cytochrome P-450 inhibition on tetrahydrofuran-induced hepatocellular proliferation in female mice. Arch. Toxicol. 77, 459–464. Wei, Y.H., Lai, C.C., Chang, J.S., 2007. Using Taguchi experimental design methods to optimize trace element composition for enhanced surfactin production by Bacillus subtilis ATCC 21332. Process Biochem. 42, 40–45. Yus Azila, Y., Mashitah, M.D., Bhatia, S., 2008. Process optimization studies of lead (Pb(II)) biosorption onto immobilized cells of Pycnoporus sanguineus using response surface methodology. Bioresour. Technol. 99, 8549–8552.