Available online at www.sciencedirect.com

www.elsevier.com/locate/chemosphere

Chemosphere 70 (2008) 982–989

Comparison of the removal of 2,4-dichlorophenol and phenolfrom polluted water, by peroxidases from tomato hairy roots,

and protective effect of polyethylene glycol

Paola Solange Gonzalez, Elizabeth Agostini *, Silvia Raquel Milrad

Departamento de Biologıa Molecular, FCEFQN, Universidad Nacional de Rıo Cuarto, 5800 Rıo Cuarto, Cordoba, Argentina

Received 16 April 2007; received in revised form 3 August 2007; accepted 7 August 2007Available online 27 September 2007

Abstract

Multiple efforts have been directed towards optimized processes in which enzymes, like peroxidases, are used to remove phenolic com-pounds from polluted wastewater. Here we describe the use of peroxidase isoenzymes from tomato hairy roots, which were able to oxi-dise 2,4-dichlorophenol (2,4-DCP) and phenol from aqueous solutions. This could be an interesting alternative for the removal of thesecompounds from contaminated sites. We used different enzyme fractions: total peroxidases (TP), ionically bound to cell wall peroxidases(IBP), basic (BP) and acidic peroxidases (AP). We analyzed the optimum conditions of removal, the effect of Polyethyleneglycol (PEG-3350) on the process and on the enzyme activities, to obtain the maximum efficiency. The optimal H2O2 concentrations for 2,4-DCP andphenol removal were 1 and 0.1 mM, respectively. TP, IBP and BP showed better removal efficiencies than AP, for both contaminants.The addition of different concentrations (10–100 mg l�1) of PEG-3350 to solutions containing 2,4-DCP showed no effect on the removalefficiencies of the isoenzymes. However, PEG (100 mg l�1) increased the removal efficiency of phenol by BP and IBP fractions. On theother hand, peroxidase activities from BP and IBP fractions were 3 and 13 times higher, respectively, than those detected for the samefractions in phenol treated solutions without PEG. The protective effect of PEG, which depends on the contaminant as well as of theenzyme fraction used, would be important to improve the removal efficiency of phenol by some peroxidase isoenzymes.� 2007 Elsevier Ltd. All rights reserved.

Keywords: Bioremediation; Phenolic compounds; Peroxidase isoenzymes; Additives; Hairy roots; PEG

1. Introduction

High concentration of pollutants generated through var-ious industrial activities, have conduced to the contamina-tion of the aquatic environment (Sakurai et al., 2003). Forexample, phenol and several derivatives, like 2,4-dichloro-phenol (2,4-DCP), can be found in aqueous effluents fromvarious industries, such as the manufacturing of resins andplastics, textiles, pulp and paper, coal conversion, etc.Although the toxicity and the environmental impact ofphenolic compounds can vary depending on the number,types and positions of substitute groups on the aromatic

0045-6535/$ - see front matter � 2007 Elsevier Ltd. All rights reserved.

doi:10.1016/j.chemosphere.2007.08.025

* Corresponding author. Tel.: +54 358 4676537; fax: +54 358 4676232.E-mail address: eagostini@exa.unrc.edu.ar (E. Agostini).

ring (s), these chemicals are considered to be toxic for var-ious organisms including humans. Hence phenolic com-pounds removal is of great importance (Entezari andPetrier, 2004).

In recent years, an enzyme catalyzed polymerization andprecipitation process has been explored as a new methodfor the treatment of phenol solutions. Such methods haveseveral advantages due to their specificities. Besides, theyare very efficient in removing targeted compounds andthe enzymes are easy to handle and store (Wilberg et al.,2002). Among these methods, the use of horseradish perox-idase (HRP) to remove phenolic compounds from aqueoussolutions was firstly proposed by Klibanov and Morris(1981). This enzyme is, by far, the most commonly usedperoxidase and catalyze the oxidation of aromatic com-pounds by hydrogen peroxide. The free radical oxidation

P.S. Gonzalez et al. / Chemosphere 70 (2008) 982–989 983

products have limited water solubility and tend to precipi-tate from solution (Aitken, 1993). However, due to highcost of the enzyme, much effort has been devoted to findan alternative source of peroxidases. Treatment costs aresignificant not only due to the price of the purified enzymebut, also, because peroxidase is susceptible to permanentinactivation by various undesirable side reactions of thetreatment process (Ghioureliotis and Nicell, 1999). In orderto minimise such inactivation, some researchers havesuggested the addition of compounds such as polyethyleneglycol (PEG) to decrease the adsorption of polymers ontothe enzyme’s active site (Nakamoto and Machida, 1992).It has been demonstrated that PEG would combine withthe polymerization products formed during the reaction,because it has a higher partition affinity with the polymerproducts than with peroxidases (Wu et al., 1993). The ben-eficial effect of adding PEG to protect soybean peroxidase(SBP) from inactivation has also been demonstrated (Kins-ley and Nicell, 2000).

In a previous work, we established that tomato (Sola-

num lycopersicum L. cv Pera) hairy root cultures andenzymes derived from them, represented a good alternativefor removal phenol treatment (Gonzalez et al., 2006).These cultures produced some peroxidase isoenzymes,which may be useful to remove phenol from aqueous solu-tions. However, the roles of different peroxidase isoen-zymes in the removal of other phenolics like 2,4-DCPand the effect of additives, need to be clarified.

Thus, the objectives of this study were: (1) to comparethe removal efficiencies of different peroxidase extractsand partially purified peroxidase isoenzymes, from tomatohairy root cultures, in the oxidation of phenolic com-pounds like phenol and 2,4-DCP; (2) to determine thekinetic parameters of oxidation of both phenolic com-pounds catalyzed by the different isoenzyme fractions, inorder to establish their relationship with the removal effi-ciencies and (3) to study the effect of PEG addition onthe removal efficiency and peroxidase activity after removalassays.

2. Materials and methods

2.1. Plant material

Hairy root cultures of S. lycopersicum L. cv Pera wereobtained inoculating sterile leaf explants with Agrobacte-

rium rhizogenes strain LBA 9402 as it was previouslydescribed (Talano et al., 2003).

They were sub-cultured every 30 d in Murashige–Skoogliquid medium (Murashige and Skoog, 1962) enriched withvitamins (Agostini et al., 2002) and maintained in an orbi-tal shaker at 100 rpm at 25 ± 2 �C in the darkness. Afterthis period of time, roots were harvested and used for theexperiments described below.

Hydrogen peroxide (30% v/v), buffer substances (Ana-lytical Reagent Grade) and phenol were purchased fromMerck; o-dianisidine was supplied by Sigma. All solutions

were prepared using deionised water. The concentrationof H2O2 was determined spectrophotometrically usinge240nm = 43.6 M�1 cm�1.

2.2. Total and differential enzyme extraction and peroxidase

activity determination

Roots were homogenised in a mortar with 10 mMsodium acetate/acetic acid buffer pH 4, containing 1 MNaCl (1 g fresh weight per 4 ml of buffer) at 4 �C andagitation for 2 h. Homogenates were centrifuged at5000g, for 5 min and the supernatants were considered astotal peroxidase extracts (TP), which contained both solu-ble and ionically bound to cell walls peroxidases (IBP).These extracts were used to determine the total peroxidaseactivity.

In order to perform differential enzyme extraction, itwas necessary to separate soluble peroxidases (SP) fromperoxidases ionically bound to cell wall. It was achievedby root homogenisation with 10 mM sodium acetate/ace-tic acid buffer, pH 4.0 (1 g fresh weight per 2 ml of buffer).The solid residue was recovered by centrifugation (5000g,5 min) and supernatants were considered the soluble pro-tein extracts (SP). The pellet was washed with 10 mMsodium acetate/acetic acid buffer, pH 4.0 until no peroxi-dase activity was detected in supernatants. Then the pel-lets were re-suspended and were incubated with thesame buffer supplemented with NaCl 1 M (2 ml for eachgram of fresh weight) and at 4 �C for 2 h. The homoge-nates were centrifuged at 5000g for 15 min, and the super-natants were considered as IBP and used for enzymeassays.

Peroxidase activity was determined with o-dianisidine assubstrate in a 1 ml reaction mixture containing 0.63 mMo-dianisidine, 0.5 mM H2O2 and 100 mM sodium acetate/acetic acid buffer pH 5.3 with 2 ll of crude extracts or10 ll of the purified fractions.

Activity was measured following the appearance ofo-dianisidine oxidation product (e460nm: 11.3 mM�1 cm�1)due to the increase in A460nm at 37 �C. The enzyme interna-tional unit (IU) was defined as the amount of enzyme form-ing 1 lmol of product in 1 min under the conditionsdescribed.

For the detection of peroxidase activity after 2,4-DCPand phenol removal, an aliquot of the remanent solutionwas used to determine residual peroxidase activity. Theactivity was measured, as described above and the resultswere expressed as percentage, considering 100% of enzymeactivity that was obtained with the enzyme concentrationused at the beginning of the experiment (200 IU ml�1 for2,4-DCP and 2000 IU ml�1 for phenol solutions). Theseunits were used because in previous works, we establishedthat with 200 IU ml�1 for 2,4-DCP and 2000 IU ml�1 forphenol, we obtained similar values of removal efficiencies.Peroxidase activity was determined in the same mannerwhen PEG was added to the reaction medium in differentconcentrations.

984 P.S. Gonzalez et al. / Chemosphere 70 (2008) 982–989

2.3. Enzyme purification

Purification of peroxidase isoenzymes was performed byion-exchange chromatography on DEAE Sephacel col-umns. Total protein extracts were dialysed during 3 h at0–4 �C, with 20 mM sodium phosphate buffer pH 6.5, bychanging the buffer twice or thrice. Samples of 2–4 ml wereloaded on a DEAE Sephacel column (2 cm · 18 cm) previ-ously equilibrated with the same buffer. The column waswashed with 50 ml of 20 mM phosphate buffer pH 6.5,and proteins were eluted using 200 ml of a linear gradient(0–0.3 M NaCl) in the same buffer at a flow rate of1 ml min�1 controlled with a peristaltic pump (BioRad).

2.4. Analysis of isoenzyme patterns

Peroxidase isoenzymes from the crude hairy rootextracts and partially purified fractions were identified byisoelectric focusing (IEF) in a pH range of 3.0–10.0 onpolyacrylamide gels using a BioRad Mini Protean III sys-tem following the manufacturer’s instructions. The sampleswere previously desalted with Sephadex G-25. A BioRad’sIEF Standard, with isoelectric points (pI) ranging from4.45 to 9.60 was used. Gels were stained with 4-chloronaphtol (0.3%) and H2O2 (7.6 M) in Phosphate Buffer Sal-ine 0.2 M, pH 6.0 to detect peroxidase activity. Markerproteins, were stained by immersing the gel for 30 min ina solution, containing Coomasie Brilliant Blue R-250(0.04%) and Crocein Scarlet (0.05%). Then the gel wasrepeatedly washed using a mixture of methanol:water:ace-tic acid (5:5:1).

2.5. 2,4-DCP and phenol removal assays with extracts and

purified peroxidase isoenzymes

The removal reactions were carried out using 10 mg l�1

of 2,4-DCP and phenol solutions. Phenol solutions weretreated with 0.1 mM H2O2 and 2000 IU ml�1 of differentextracts (TP and IBP) and purified basic and acidic perox-idase isoenzymes (BP and AP, respectively) as it was previ-ously described (Section 2.2).

For the treatment of 2,4-DCP solutions, H2O2 wasadded in different concentrations (0.1–1.0 mM) in orderto establish the optimal concentration in the presence of200 IU ml�1 of the above cited enzyme extracts or purifiedisoenzymes.

The reaction mixture was incubated during 1 h at25 ± 2 �C in an orbital shaker at 100 rpm. After the incu-bation, residual 2,4-DCP and phenol in the reaction mix-tures were measured through a colorimetric assay(Klibanov and Morris, 1981). A 5 ml aliquot of the solu-tion after removal reacted with 0.025 ml of 6.0 M ammo-nium hydroxide, 0.025 ml of 4-aminoantipyrine (2% w/vaqueous solution) and 0.05 ml of potassium ferricyanide(8% w/v). After several minutes, the coloured compoundformed was extracted with 2.5 ml of chloroform. Theabsorbance of the extract was determined at 510 nm and

was proportional to the concentration of 2,4-DCP or phe-nol in the range of 0–10�4 M.

The results were expressed as removal efficiency, whichis defined as the percentage of pollutant removed fromsolution under established experimental conditions(Miland et al., 1996). Results are the average of at leastsix determinations, corresponding to three independentexperiments.

2.6. Removal of contaminants by addition of PEG-3350 to

the reaction medium

The removal reactions were carried out following thesame protocols described above, but in these experiments,10 or 100 mg l�1 of PEG-3350 were added to the reactionmedium.

2.7. Determination of kinetic parameters

Kinetic properties of root extracts (TP and IBP) andpurified isoenzymes, were studied by a colorimetric method,with phenol (e510 = 7870 M�1 cm�1) or 2,4-DCP (e510 =6710 M�1 cm�1) (Metelitza et al., 1991). The catalytic effi-ciency was determined using a millimolar absorption coeffi-cient (e403 = 90 mM�1 cm�1) to estimate peroxidaseconcentration (Agostini et al., 2002).

The assays were performed with 2,4-DCP or phenol, 4-aminoantipyrine (4-AAP) and H2O2 as colour-generatingsubstrates. The formation of the reaction product, a redquinoneimine, was followed by monitoring A510nm at afinal time of 1 min. The reaction mixture (1 ml) contained2.4 · 10�3 M 4-AAP, 2,4-DCP or phenol in the concentra-tion range of 0.1–10 mM, and H2O2 in the concentrationrange of 1 · 10�3 to 1 mM. In order to establish the opti-mal pH, a buffer solution composed by 0.2 M acetic acid,0.2 M boric acid, 0.2 M phosphoric acid and 1 M NaOH,was used to provide a pH in the range between 4.0 and8.0. All assays were carried out by triplicate, and the resultsare the mean of three independent experiments, using theprogramme Statistica 6.0.

3. Results and discussion

3.1. Purification of basic and acidic peroxidase isoenzymes

from tomato hairy root cultures

In a previous work (Gonzalez et al., 2006) we describedthat tomato hairy root cultures, represented a good alter-native for phenol removal. During phenol removal experi-ments, a lost of peroxidase activity in the roots wasdetected and changes in the isoenzyme patterns wereobserved. According to these results, we have focused ourresearch in the analysis of those peroxidase isoenzymeswhich seems most likely involved in the process of transfor-mation of phenolic compounds.

Simple extraction methods were applied in order toobtain peroxidase extracts, which were referred as TP

S TP IBP BP AP

pI 9.6

pI 3.9pI 3.6

Fig. 1. IEF in the pH range 3.0–10.0 of peroxidase extracts and partiallypurified isoenzymes. Line S: standard of pI, Line TP: total peroxidases,Line IBP: ionically bound to cell wall peroxidases, Line BP: basicperoxidase, Line AP: acidic peroxidases.

P.S. Gonzalez et al. / Chemosphere 70 (2008) 982–989 985

and IBP. The isoenzyme profile of TP extract, included sev-eral peroxidase isoenzymes, with pI values ranging from9.6 to 3.6, while IBP fractions, included only the groupof basic and neutral peroxidase isoenzymes (Fig. 1).

The main BP, with a pI 9.6 and a group of AP, pI 3.6and 3.9, were separated and partially purified from TPextract by ion exchange chromatography. The IEF isoen-zyme profiles obtained for TP, IBP, BP and AP fractionsafter purification, are shown in Fig. 1.

3.2. Determination of kinetic parameters for the oxidation of

2,4-DCP and phenol

Kinetic studies were performed with different extracts(TP and IBP) and partially purified isoenzymes (BP andAP), using 2,4-DCP and phenol as substrates, in order todetermine the apparent kinetic constants for their oxida-tion. Typical hyperbolic curves were obtained with bothsubstrates in concentrations ranging from 0.1–10 mM andH2O2 in concentrations from 10�3 to 1 mM.

All fractions showed high activity over a wide range ofpH, and the optimal pH was 6.0 except for BP removalof 2,4-DCP, which was 7.0 (Table 1). Similar results weredescribed for other peroxidase isoenzymes (Akhtar andHusain, 2006; Bodalo et al., 2006) and it was recommendedthe use of pH close to neutrality for the treatment ofphenolic compounds (Akhtar and Husain, 2006). This

Table 1Kinetic parameters of different root extracts (TP and IBP) and partiallypurified isoenzymes (BP and AP) from tomato hairy roots obtained with2,4-DCP and phenol as substrates

Enzymefraction

2,4-DCP Phenol

Km

(mM)Catalytic efficiency(mM�1 s�1)

Km

(mM)Catalytic efficiency(mM�1 s�1)

TP 8.6 9.7 6.3 2.1IBP 9.0 12.8 0.6 19.1BP 6.3 13.6 4.5 7.2AP 9.1 10.1 11.4 2.3

characteristic is also of great importance since the enzymescould be used for wastewater treatment without previouspH adjustment.

The ratio kcat/Km, is generally considered a good mea-sure of the catalytic efficiency of an enzyme. In order tocompare the efficiency of different enzyme fractions fromtomato hairy root cultures, the value of catalytic efficiencywas calculated for each extract and for the partially puri-fied isoenzymes. The apparent kinetic constants shownfor TP and IBP fractions represent the relative contribu-tions of all peroxidases present in these extracts. The prep-arations contained multiple forms of the same enzyme, sothe kinetic constants are the average of the parameters ofeach enzyme in the mixture. However, the IPB extract isenriched in BP (pI 9.6), so the kinetic properties are influ-enced by this main isoenzyme.

The results are summarised in Table 1. When 2,4-DCPwas used as substrate, the higher catalytic efficiencies wereobserved for the IBP and BP fractions. In addition, theseenzyme fractions were the samples with the best values forKm and catalytic efficiency, for phenol oxidation (Table 1).

The catalytic efficiency depends on the chemical natureof the reducing substrate and the results recorded in Table1, suggest that 2,4-DCP was a better substrate than phenol.The exception was the IBP fraction, which showed thehighest catalytic efficiency and the lowest Km for phenol.These results are in agreement with the data presented bySantos de Araujo et al. (2004). These authors describedthat peroxidases obtained from D. carota and I. batatas

presented greater catalytic efficiencies for 2,6-DCP as sub-strate than for phenol. This could be explained by the factthat the di-substitution of the aromatic ring with chlorineincreases the electron withdrawal from the phenolic oxy-gen, leading to higher acidity of the attached proton thusincreasing the chances for the formation of compound II(second intermediate in the enzyme catalytic cycle). Onthe other hand, these authors demonstrated that peroxi-dases from other plant species (Solanum aviculare), weremore active with regard to the oxidation of phenol and cat-echol than guaiacol and 2-chlorophenol, similar to theresults obtained with the IBP fraction. These differencesmay be due to slight variations in the isoenzyme composi-tion of the different fractions.

From these results, we suggest that the group of thebasic peroxidase isoenzymes from IBP and BP fractionsof tomato hairy roots could be responsible for 2,4-DCPand phenol removal, while the participation of AP fractionin the oxidation of phenol would be low. However AP frac-tion was able to oxidize 2,4-DCP more efficiently than phe-nol, and with similar catalytic efficiency than the otherfractions.

3.3. Removal of 2,4-DCP and phenol by peroxidase fractions

and the effect of PEG

In order to compare the removal efficiencies, allfractions obtained (TP, IBP, BP and AP), were used for

0

20

40

60

80

100

Phe

nol R

emov

al E

ffici

ency

(%

)

TP IBP BP AP

PHENOLb0

20

40

60

80

100

2,4-

DC

P R

emov

al E

ffici

ency

(%

)

Control

10 mg l-1 PEG

100 mg l-1 PEG

2,4-DCPa

Fig. 3. Comparison of removal efficiencies of different extracts andpartially purified isoenzymes, incubated without PEG (control), and in thepresence of 10 and 100 mg l�1 PEG. (a) Removal of 2,4-DCP (10 mg l�1)with H2O2 1 mM and 200 IU ml�1 of enzyme fractions. (b) Removal ofphenol (10 mg l�1) with H2O2 0.1 mM and 2000 IU ml�1 of enzymefractions.

986 P.S. Gonzalez et al. / Chemosphere 70 (2008) 982–989

2,4-DCP and phenol removal assays, using a solution of10 mg l�1 of the pollutant and H2O2 in optimal concentra-tion, because it would be a limiting factor in the removalprocess. An optimal concentration must be establishedbecause when a peroxide concentration higher than opti-mum was used, and enzyme inactivation could happen. Aconcentration of 0.1 mM H2O2 was optimal for phenolremoval, as it was previously established (Gonzalez et al.,2006). However, this concentration was not suitable toremove 2,4-DCP, as it is shown in Fig. 2. So, a concentra-tion of 1 mM H2O2 was selected when 2,4-DCP was used assubstrate.

Removal assays reactions were carried out with thesame enzyme units for the different fractions to evaluatethe relative contribution of each isoenzyme to the process.The results are shown in Fig. 3a and b. The removal effi-ciencies of 2,4-DCP (control without PEG) were similarfor TP, IBP and BP fractions (approximately 80%). How-ever, when AP fraction was used in the assay, removal effi-ciency decreased to 47% (Fig. 3a). When phenol was usedas pollutant, the highest removal efficiencies were obtainedusing IBP and BP fractions (control without PEG)(Fig. 3b), while AP fraction showed low removal efficiency(lower than 20%). These results are in agreement with thekinetic parameters (Km and catalytic efficiency) obtainedfor IBP and BP fractions. The data suggested a major par-ticipation of basic peroxidases in the transformation ofboth phenolic compounds. The AP showed lower removalefficiency for phenol than for 2,4-DCP (control withoutPEG). This result was in concordance with the high cata-lytic efficiency showed for AP with 2,4-DCP as substrate.

Cheng et al. (2006) have reported that the use of someadditives like PEG, can improve the efficiency of phenolremoval, by forming a protective layer in the vicinity ofthe active centre of horseradish peroxidase (HRP) torestrict the attack of free phenoxy radicals formed in thecatalytic cycle. PEG is of particular interest as a syntheticadditive since it significantly reduces the cost of treatmentand it is a non-toxic compound which has been declared

0

20

40

60

80

APBPIB PTP

Re

mo

valE

ffic

ien

cy(

%)

H2O2 0,1 mMH2O2 1 mM

Fig. 2. Comparison of removal efficiencies of extracts of isoperoxidasesand partially purified isoenzymes incubated with different H2O2 concen-trations, using 2,4-DCP as substrate.

fit for human consumption by the United States Foodand Drug Administration (Kinsley and Nicell, 2000). How-ever, they reported that PEG could be an additive withpotential environmental impact, such as oxygen demand,so its concentration should be minimized. In view of theseresults, we evaluated the effect of PEG concentrationsranging from 10 to 500 mg l�1 using a simple toxicity testemploying Amphibian embryos (Bufo arenarum). Theseorganisms are more susceptible to environmental contami-nants than birds and mammals. We have found that theabove mentioned concentrations did not affect the viabilityof these embryos (data not showed).

So, to analyze if the addition of PEG affects the removalof 2,4-DCP and phenol with the studied enzyme fractions,this compound was added to the reaction medium, in dif-ferent concentrations (10 and 100 mg l�1). The results aresummarized in Fig. 3a and b.

The use of PEG with an average molecular weight of3350 g mol�1, was based on the results of Nakamoto andMachida (1992) who determined that PEG with molecularweights above 1000 g mol�1 were most effective in protect-ing HRP. Furthermore, it has been demonstrated that PEGwith molecular weights of 1000 g mol�1 and below are inef-fective in protecting SBP (Kinsley and Nicell, 2000).

0

10

20

30

40

50

60

70 PHENOL

Pos

t-re

mov

al p

erox

idas

e ac

tivity

(%

)

TP IBP BP AP

b0

10

20

30

40

50 2,4-DCP

Pos

t-re

mov

al p

erox

idas

e ac

tivity

(%

)

Control

10 mg l-1 PEG

100 mg l-1 PEG

a

Fig. 4. Comparison of peroxidase activity after 2,4-DCP (a) and phenol(b) removal, obtained with different root extracts and partially purifiedisoenzymes from tomato hairy roots, without PEG and in the presence of10 and 100 mg l�1 of PEG.

P.S. Gonzalez et al. / Chemosphere 70 (2008) 982–989 987

When 2,4-DCP was used as contaminant, the presenceof PEG showed no effect on the removal efficiency, forall fractions analyzed (Fig. 3a). Kinsley and Nicell (2000),showed that PEGs with molecular weights greater than1500 g mol�1 improved the life of SBP and the degree ofimprovement increased with increasing molecular weight,but they have not investigated the abilities of PEGs to pro-tect other peroxidases. From these results, probably, themolecular weight of PEG used in this study, was verylow to give an adequate protection to peroxidase isoen-zymes from tomato hairy roots against 2,4-DCP. So, theuse of PEGs of greater molecular weight than 3350 g mol�1

will be studied in our laboratory.However, when 100 mg l�1 of PEG was added to solu-

tions containing phenol, the removal efficiency increasedwith IBP and BP fractions (Fig. 3b). Both fractions (IBPand BP) displayed high affinity towards phenol and highervalues of catalytic efficiencies than the other peroxidaseisoenzymes.

Similar results were presented by Bodalo et al. (2006).They concluded that PEG has a positive effect on the phe-nol elimination process whether catalyzed by HRP or SBP,although its effect is much more pronounced in the case ofHRP. PEG showed some protective effect on HRP againstinactivation. This differential PEG effect could be due tothe different purity of the two enzymes (HRP and SBP)or to the isoenzyme composition of each protein fraction.

3.4. Determination of peroxidase activity after 2,4-DCP or

phenol removal

After phenolic compounds removal, residual peroxidaseactivity was measured in the reaction mixtures. It wasobserved that IBP and BP fractions showed the lowestactivity after 2,4-DCP removal in the absence of PEG(Fig. 4a) while the TP extracts showed more activity thanIBP and BP fractions, with both contaminants. Thesecould be attributed to a protective action of some compo-nents of the plant tissues that act as stabilizing agents andavoid peroxidase inactivation. Nevertheless, the activity ofAP fraction after 2,4-DCP removal assays was the lessaffected. On the basis of this observation, and consideringthat AP showed a removal efficiency of approximately50% for 2,4-DCP and retained 45% of its activity, we sug-gest that probably AP could be used in removal processesfor periods longer than an hour. Furthermore, this isoen-zyme would be an interesting alternative for usage in con-tinuous removal processes in which the enzymes are usedrepeatedly.

On the other hand, as it can be seen in Fig. 4b, after phe-nol removal, residual peroxidase activity presented similarresults for IBP, BP and AP fractions (in the absence ofPEG). These fractions showed very low activity, althoughAP fraction is less efficient to remove phenol. From theseresults, as well as in concordance with the removal efficien-cies, we suggest that the group of basic peroxidase isoen-zymes would be the most likely involved in 2,4-DCP and

phenol removal. For this reason, the activity of these per-oxidase isoenzymes would be the most affected after theprocess.

As it is well known, peroxidases are inactivated by freeradicals as well as oligomeric products formed in the reac-tion. These products attach to the enzyme and inactivate it(Buchanan and Nicell, 1997). So, various additives havebeen studied in an attempt to increase the catalytic life ofthe enzyme. Among them, polyethylene glycol (PEG) hasbeen demonstrated to be particularly effective in diminish-ing the inactivation of HRP (Nakamoto and Machida,1992) even though the mechanism of protection is not yetfully understood. It was proposed that PEG can beattached to the polymeric products of the enzyme activity.This may contribute to the reduction of inactivationthrough the mechanism of polymer entrapment as pro-posed by Buchanan and Nicell (1997).

Therefore, as it was previously described, experimentswere conducted to assess the impact of PEG-3350 on thecatalytic life, of the different enzyme fractions analysedand to compare the residual enzyme activity of the frac-tions with and without PEG.

In general, peroxidase activities of all fractions analysedafter 2,4-DCP and phenol removal, were not affected by theaddition of PEG (10 or 100 mg l�1). Exceptions to theseresults were the peroxidase activity of BP and IBP fractionsafter phenol removal. These activities were 3 and 13 times

988 P.S. Gonzalez et al. / Chemosphere 70 (2008) 982–989

higher for BP and IBP, respectively, when PEG 100 mg l�1

was added to the reaction mixture, than those detected forthe same fractions in solutions without PEG (Fig. 4b). Theincrease in peroxidase activity after phenol removal,detected for IBP and BP fractions, was similar to the resultsshown by Entezari and Petrier (2004), who described thepositive effect of PEG on the HRP activity. Moreover,Kinsley and Nicell (2000) indicated that the optimumPEG dose for the treatment of phenol with SBP was100 mg l�1. In addition, Wu et al. (1993), reported thatthe quantity of PEG required, in order to provide optimumprotection of HRP, varied between 10 and 100 mg l�1,depending on the phenolic compound used. These findingssuggest that the protective effect exhibited by PEG uponperoxidase isoenzymes is variable and must be determinedfor each enzyme sample. So, with the results obtained, wecannot generalize the effect of PEG, since this effectdepended on the contaminant, as well as on the enzymefraction used and the concentration of PEG employed.

4. Conclusions

Our results demonstrated the potential of peroxidasesfrom tomato hairy roots for the efficient oxidation of toxicphenols.

We found that TP, BP and IBP could be used to remove2,4-DCP and phenol with high efficiencies (approximatelyfrom 78% to 98%), when 2,4-DCP and phenol were usedas substrates.

AP showed lower removal efficiencies for 2,4-DCP andphenol than the other fractions studied. However, APwas more efficient to remove 2,4-DCP than phenol, whichwas in agreement with its high catalytic efficiency for thissubstrate.

The effect of PEG as an additive was variable. PEGshowed no effect on the removal of 2,4-DCP by all enzymefractions analysed. Nevertheless, PEG increased the phenolremoval efficiency of BP and IBP fractions and protectedthese fractions from inactivation, after phenol removal.So, it could be used to improve phenol removal with theseenzyme fractions.

From these findings, we suggest that the group of thebasic peroxidase isoenzymes from tomato hairy root cul-tures, would be the most likely isoenzymes involved in2,4-DCP and phenol removal. On the other hand, the useof peroxidase extracts could be a good alternative insteadof pure enzymes, because the latter are more expensive toobtain. Moreover, purified isoenzymes could be inactivatedfaster than root extracts during phenols removal, with aconsequent low efficiency. So IBP and TP extracts seemto be more suitable for the removal of both contaminants.These enzyme extracts appear to be quite interesting inview of possible application in waste-water treatment orother environmental applications. Additional research isrequired to explore methods to scale-up this process. How-ever, a simple stirred tank reactor would be used as a con-tinuous removal system, for treatment of great amounts of

polluted effluents. New advances in this sense would beuseful to optimize commercial applications for theseenzyme extracts making their use more competitive.

Acknowledgements

We wish to thanks to SECyT-UNRC, FONCyT andConsejo Nacional de Investigaciones Cientıficas y Tecnicas(CONICET) (Argentina) for financial support. E.A. is amember of the research career from CONICET. Theauthors acknowledge to M. Woelke for technical assistanceand Monica Gonzalez, for language correction of themanuscript.

References

Agostini, E., Hernandez-Ruiz, J., Arnao, M.B., Milrad, S.R., Tigier,H.A., Acosta, M., 2002. A peroxidase isoenzyme secreted by turnip(Brassica napus) hairy-root cultures by hydrogen peroxide andapplication in diagnostic kits. Biotechnol. Appl. Biochem. 35, 1–7.

Aitken, M.D., 1993. Waste treatment applications of enzymes: opportu-nities and obstacles. Chem. Eng. 52, 49–58.

Akhtar, S., Husain, Q., 2006. Potential applications of inmovilized bittergourd (Momordica charantia) peroxidase in the removal of phenolsfrom polluted water. Chemosphere 65, 1228–1235.

Bodalo, A., Gomez, J.L., Gomez, E., Bastida, M.F., Maximo, 2006.Comparison of commercial peroxidases for removing phenol fromwater solutions. Chemosphere 63, 626–632.

Buchanan, I., Nicell, J.A., 1997. Model development for horseradishperoxidase-catalyzed removal of aqueous phenol. Biotechnol. Bioeng.54, 251–261.

Cheng, J., Mings, S., Zuo, P., 2006. Horseradish peroxidase immovilizedon aluminium-pillared interlayered clay for the catalytic oxidation ofphenolic wastewater. Water Res. 40, 283–290.

Entezari, M.H., Petrier, C., 2004. A combination of ultrasound andoxidative enzyme: sono-biodegradation of phenol. Appl. Catal. B:Environ. 53, 257–263.

Ghioureliotis, M., Nicell, J.A., 1999. Assesment of soluble products ofperoxidase-catalyzed polymerization of aqueous phenols. EnzymeMicrob. Technol. 25, 185–193.

Gonzalez, P.S., Capozucca, C., Tigier, H., Milrad, S., Agostini, E., 2006.Phytoremediation of phenol from wastewater, by peroxidases oftomato hairy root cultures. Enzyme Microb. Technol. 39, 647–653.

Kinsley, C., Nicell, J.A., 2000. Treatment of aqueous phenol with soybeanperoxidase in the presence of polyethylene glycol. Bioresource Tech-nol. 73, 139–146.

Klibanov, A.M., Morris, E.D., 1981. Horseradish peroxidase for theremoval of carcinogenic aromatic amines from water. Enzyme Microb.Technol. 3, 119–122.

Metelitza, D.I., Litvinchuk, A.V., Savenkova, M.I., 1991. Peroxidase-catalyzed co-oxidation of halogen-substituted phenols and 4-amino-antipyrine. J. Mol. Catal. 67, 401–411.

Miland, E., Smyth, M.R., Fagain, O.C., 1996. Phenol removal bymodified peroxidases. J. Chem. Technol. Biotechnol. 67, 227–236.

Murashige, T., Skoog, F., 1962. A revised medium for rapid growth andbioassays with tobacco tissue culture. Plant. Physiol. 15, 473–479.

Nakamoto, S., Machida, N., 1992. Phenol removal from aqueoussolutions by peroxidase-catalyzed reaction using additives. WaterRes. 26, 49–54.

Sakurai, A., Masuda, M., Sakakibara, M., 2003. Effect of surfactants onphenol removal by the method of polymerization and precipitationcatalysed by Coprinus cinereus peroxidase. J. Chem. Technol. Biotech-nol. 78, 952–958.

Santos de Araujo, B., Omena de Oliveira, J., Salgueiro Machado, S.,Plestch, M., 2004. Comparative studies of the peroxidases from hairy

P.S. Gonzalez et al. / Chemosphere 70 (2008) 982–989 989

roots of Daucus carota, Ipomea batatas and Solanum aviculare. PlantSci. 167, 1151–1157.

Talano, M.A., Agostini, E., Medina, M.I., Milrad, S.R., Tigier, H.A.,2003. Tomato (Lycopersicon esculentum cv. Pera) hairy roots cultures:characterization and changes in peroxidase activity under NaCltreatment. In Vitro Cell. Develop. Biol. Plant 39, 354–359.

Wilberg, K., Assenhaimer, C., Rubio, J., 2002. Removal of aqueousphenol catalyzed by a low purity soybean peroxidase. J. Chem.Technol. Biotechnol. 77, 851–857.

Wu, J.N., Taylor, K.E., Bewtra, J.K., Biswas, N., 1993. Optimization of thereaction conditions for enzymatic removal of phenol from wastewaterin presence of polyethylene glycol. Water Res. 27, 1701–1706.