Environmental fate, toxicity, characteristics and

Bioresource Technology 108 (2012) 245–251

Contents lists available at SciVerse ScienceDirect

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

Environmental fate, toxicity, characteristics and potential applications of novel bioemulsifiers produced by Variovorax paradoxus 7bCT5 Andrea Franzetti a, Isabella Gandolfi a, Chiara Raimondi a, Giuseppina Bestetti a, Ibrahim M. Banat b,⇑, Thomas J. Smyth b, Maddalena Papacchini c, Massimo Cavallo d, Letizia Fracchia d a

Department of Environmental Sciences, University of Milano-Bicocca, Piazza della Scienza 1, 20126 Milano, Italy School of Biomedical Sciences, University of Ulster, Coleraine BT52 1SA, Northern Ireland, UK ISPESL, Dept. for Production Premises and Interaction with Environment, Via Fontana Candida 1, 00040 Monteporzio Catone (RM), Italy d Department of Chemical, Food, Pharmaceutical and Pharmacological Sciences, University of Eastern Piedmont, Via Bovio 6, 28100 Novara, Italy b c

a r t i c l e

i n f o

Article history: Received 25 November 2011 Received in revised form 30 December 2011 Accepted 3 January 2012 Available online 10 January 2012 Keywords: Biosurfactant Bioemulsifiers Crude oil Environmental sustainability

a b s t r a c t The aims of this work were the characterisation and the evaluation of potential environmental applications of the bioemulsifiers produced by Variovorax paradoxus 7bCT5. V. paradoxus 7bCT5 produces a mixture of high molecular weight polysaccharides. The extracellular bioemulsifiers were able to produce a thick stable oil/water emulsion and maintained the emulsification activity after boiling and at low temperatures. Environmental behavior and impact of bioemulsifiers release were assessed by evaluating biodegradability, toxicity and soil sorption. Respirometric tests showed that moderate biodegradability occurred by soil bacterial inoculum. Furthermore, the produced compounds did not show any toxic properties through different ecotoxicological tests. The Kd values ranged from 1.3 to 7.3 L/kg indicating a high sorption affinity of the bioemulsifier molecules to soil particles. The soil sorption affinity likely affected the bioemulsifier ability to remove hydrocarbons from contaminated soils. In fact, V. paradoxus 7bCT5 bioemulsifiers significantly increased the removal of crude-oil from sandy soil compared to water. Ó 2012 Elsevier Ltd. All rights reserved.

1. Introduction Many microorganisms produce a wide range of amphipathic compounds which exhibit surface activities at interfaces, including the ability to lower surface and interfacial tension of liquids and to form micelles and microemulsions between these different phases (Chen et al., 2010, 2011; Penfold et al., 2011). These surface-active compounds are generally called biosurfactants or bioemulsifiers; the terms are however often used interchangeably. In practice, those that mainly lead to the formation of stable emulsions are referred to as bioemulsifiers or bioemulsans. The former group includes low-molecular-weight compounds, such as lipopeptides and glycolipids while the latter includes high-molecular-weight polymers of polysaccharides, lipopolysaccharides, proteins or lipoproteins which are produced by a number of different bacteria (Rosenberg and Ron, 1999; Smyth et al., 2010a,b). The most commonly studied biopolymer is emulsan, a lipopolysaccharide isolated from Acinetobacter calcoaceticus RAG-1 ATCC 31012, with a molecular weight of around 1000 kDa (Rosenberg et al., 1979). Another well-known high molecular weight bioemulsifier is alasan, a complex consisting of an anionic polysaccharide

⇑ Corresponding author. Tel.: +44 2870123062. E-mail address: [email protected] (I.M. Banat). 0960-8524/$ - see front matter Ó 2012 Elsevier Ltd. All rights reserved. doi:10.1016/j.biortech.2012.01.005

and a protein with a molecular weight of around 1000 kDa isolated from Acinetobacter radioresistens (Navon-Venezia et al., 1995). Yeasts are also known producers of emulsifiers, such as liposan, an extracellular emulsifier produced by Candida lipolytica and mannan protein based emulsifiers produced by Saccharomyces cerevisiae (Banat et al., 2010). The diverse functional properties of bioemulsifiers and biosurfactants, such as emulsification, wetting, foaming, cleansing, phase separation, surface activity and reduction in viscosity of heavy liquids such as crude oil, make them suitable for exploitation in many industrial and domestic applications (Perfumo et al., 2010). Consequently, during the past two decades, they have been under continuous investigation as a potential replacement for synthetic surfactants, since they are expected to have several applications mainly related to detergency, emulsification, dispersion and solubilisation of hydrophobic compounds (Banat et al., 2000). Due to all these properties, in fact, they also have steadily gained increased significance especially in areas such as bioremediation, soil washing, enhanced oil recovery and other general oil processing and related industries (Banat et al., 2010; Perfumo et al., 2010; Thavasi et al., 2011a,b; Urum et al., 2003). The main applications of biosurfactants in environmental remediation involve the enhancement of organic pollutants bioavailability (biosurfactantenhanced bioremediation) and soil washing. The use of chemical surfactants as well as microbial ones has been reported to be

246

A. Franzetti et al. / Bioresource Technology 108 (2012) 245–251

efficient in removing hydrocarbons from soils (Mulligan, 2005). Bioemulsifiers have vast potential for stabilising emulsions between liquid hydrocarbons and water mixtures. However, their application in enhancing bioremediation has provided contrasting results to date (Barkay et al., 1999; Franzetti et al., 2009). The lower toxicity and higher biodegradability of biological surfactants compared to their chemical counterparts is the main reason for their high acceptability. However, these features are often assumed as only direct consequence of their natural origin. For these reasons, the environmental features of novel biosurfactants should be carefully considered and investigated before their release into the environment (Franzetti et al., 2006). In this study, the applicability of novel bioemulsifiers produced by the strain Variovorax paradoxus 7bCT5 isolated from Cambodian soil sample was carried out using a multiphase approach. V. paradoxus is a ubiquitous, aerobic, Gram negative bacterium present in diverse environments and has been associated with a number of interesting biotransformations (Willems et al., 1991; Smith et al., 2005; Nishino and Spain, 2006). Recently, Jamieson et al. (2009) investigated the coordinated surface behaviours in this bacterium and demonstrated the presence of a wetting agent able to reduce the surface tension. However, to our knowledge, no information is available in the literature regarding the potential applications of the biosurfactant compounds produced by this species. The mixture of bioemulsifiers produced by V. paradoxus 7bCT5 was chemically characterised and the effect of releasing of these bioemulsifiers into the environment was assessed. Soil sorption experiments, respirometric biodegradability tests and ecotoxicity tests were carried out to evaluate their possible environmental effects. In addition, to preliminary investigate their potential in environmental remediation, bench-scale washing tests were conducted using three different types of soil contaminated by crude oil. 2. Methods 2.1. Characterization of the bioemulsifier-producing strain The strain 7bCT5 was isolated from a sandy loam Cambodian soil sample. Pure culture was obtained by streaking colonies several times onto LB agar and stored at 80 °C in LB broth (Fluka) supplemented with 30% glycerol. Morphological characteristics were observed on LB agar. Microscopic observation of a Gramstained culture smear and tests for cytochrome oxidase (BioMérieux, France) and glucose oxidation (Oxidation–Fermentation test in OF medium) were carried out. A preliminary identification by Biolog™ GN microplates and MicroLogTM software (Biolog, Hayward, CA, USA) was also carried out. Finally, the genomic DNA of a 24 h culture in LB broth was extracted and sent to BMR Genomics s.r.l. (Padova, Italy) for PCR amplification and sequencing of the 16S rRNA gene.

Surface activity was measured by the oil spreading assay (Satpute et al., 2010) using 20 lL of Motor Oil 10W-40 (Selenia) previously deposited onto the surface of 20 mL of distilled water in a Petri dish (90 mm in diameter) to form a thin film. Twenty microliters of bacterial supernatant were gently added onto the centre of the oil film. Diameter of the displaced circle was measured to determine the presence of biosurfactants (Satpute et al., 2010). Surface tension of V. paradoxus 7bCT5 supernatant was measured by the duNouy method using a ring tensiometer (KRUSS Digital Tensiometer K10). Measurements were carried out in triplicate and results expressed as mN m 1 and compared with sterile LB broth surface tension. 2.3. Preparation of the extracellular crude bioemulsifiers V. paradoxus 7bCT5 was grown on LB agar at 28 °C for 48 h. A loopful of culture was then inoculated in 1 L of LB broth and incubated at 28 °C and 200 rpm for overnight pre-culture. The overnight culture was used to inoculate 40 L of LB broth for fermentation process, carried out at 28 °C for 42 h by Isagro Ricerca S.p.A. (Novara, Italy). After incubation, the culture was centrifuged at 6800g for 20 min and the biomass removed. Ten litres of the cell-free culture were subsequently lyophilized in order to obtain a solid residue. After lyophilisation, the crude bioemulsifiers were suspended in 500 mL of deionised water and precipitated with four volumes of ethanol for 24 h at 20 °C. The solution was then centrifuged at 6800g for 30 min at 4 °C. The supernatant was discarded and the precipitated matter was suspended in water and dialyzed twice in cellulose tube membrane (cut-off: 12.4 kDa) (Sigma–Aldrich, USA) for 24 h against deionised water, then lyophilized to obtain 4 g of a brown coloured powder. The thermal resistance of the obtained product was evaluated using a solution at a concentration of 500 mg/L. The emulsifying activity of the solution was measured after boiling at 100 °C for one hour and cooling to room temperature. The crude extract was also tested for its resistance to very low temperatures by incubating a 500 mg/L solution at 80 °C for 24 h. An emulsification activity test was carried out: 1 mL of heat/cold treated solutions mixed with 0.5 mL hexadecane were placed in an Eppendorf tube and vortexed for 30 s, the height of the emulsion layers was measured and the dimension of the emulsions was qualitatively evaluated. 2.4. Chemical characterization

2.2. Surface activity

2.4.1. Composition of bioemulsifiers The following parameters were measured on a 500 mg/L bioemulsifier solution: (i) protein content by the Coomassie blue method using bovine serum albumin as a standard, (ii), the sugar content by the phenol–sulphuric acid method using glucose as a standard, (iii) lipid content by the sulfo-phospho-vanillin test using triolein as a standard (Izard and Limberger, 2003).

The strain V. paradoxus 7bCT5 was evaluated for emulsion production and stability. After growing in LB broth for 24 h in an orbital shaker at 150 rpm and 28 °C, cells were removed by centrifugation and 5 mL of the cell-free supernatant was mixed with 5 mL hexadecane in a glass test tube. This mixture was vortexed for 2 min and then left to stand. The emulsifying activity of the supernatant was also tested on other hydrophobic substrates such as kerosene, diesel, soybean oil and motor oil. Emulsification index Et (%) was measured after 24 h and at intervals up to 12 months of incubation at room temperature using the following equation: Et (%) – emulsion index at time t = height of emulsion layer/height of total solution 100% (Satpute et al., 2010).

2.4.2. Mass spectrometry MALDI-TOF–MS analysis was carried out using a PerSeptive Biosystems Voyager-DE Mass Spectrometer (Hertfordshire, UK). A 5 mg sample of the isolated bioemulsifiers was dissolved in 1 mL of deionised water. One microlitre aliquots of sample were mixed with 1 lL of sinapic acid (10 mg/mL dissolved in acetonitrile/water (50/50) with 0.1% trifluoroacetic acid (Sigma–Aldrich, UK)). The mixed sample was applied to defined wells on a 100-well sample plate and allowed to air-dry before analysis. The sample was analysed in positive mode detection using a nitrogen laser set at 337 nm, with a pulse duration of 3 ns and a pulse frequency of 3 Hz. Mass spectra were recorded in linear mode while the acceleration voltage (20 kV),


grid voltage (18.8 kV) and the guide wire voltage (10 V) remained constant. The delayed extraction was set at 115 ns. Masses were recorded as mass/charge ratio (m/z) against relative abundance. Various areas of each well were targeted and variable laser intensities were used to achieve optimal spectra. Full mass range was initially analysed followed by narrowing based on the observed masses to 150–200 kDa. 2.5. Ecotoxicity tests A set of four ecotoxicity tests was performed on the bioemulsifier solutions to assess their possible toxic effects on a range of different organisms. MicrotoxÒ assay was carried out in four replicates according to manufacturer’s instructions (Basic Test for Pure Compounds, Azur Environmental, Carlsbad, CA, USA) on a 500 mg/L bioemulsan starting solution and subsequent twofold dilutions, with 5 and 15 min contact times. Toxicity was expressed as a percentage of effect with respect to the control and quantified as a decrease in the bioluminescence of Vibrio fischeri. Acute toxicity test on the crustacean Daphnia magna was carried out in five replicates, according to the Daphtoxkit F™ Magna protocol (Ecotox LDS, Italy), with the exception that only one concentration of the compound was tested (500 mg/L). The number of dead and immobilized crustaceans was recorded at 24 and 48 h, to calculate a percentage effect. Seed germination and root elongation test was performed on Cucumis sativus, Lepidium sativum and Sorghum saccharatum, according to the Italian Technical Guide of Analytical Methods for Contaminated Soil and Sediments RTI CTN_TES 1/ 2004 – APAT, 2004. Filter paper was moistened with a 500 mg/L bioemulsan solution both in the presence and absence of three different soil specimens; the same samples were also used for soil washing experiments. Four replicates of ten seeds each were prepared for each species and each condition; after 72 h, a Germination Index (GI%) was calculated as follows: GI% = 100 (No. germinated sample seeds average root elongation in samples)/ (No. germinated control seeds average root elongation in controls). A contact acute toxicity test was performed on earthworms (Eisenia foetida) using 10 replicates, according to the OECD Guideline for Testing of Chemicals No. 207, as described for the filter paper test. Two bioemulsan concentrations were used: 14.7 and 29.4 lg/cm2, corresponding to 1 mL-spiking of a 500 mg/L bioemulsan solution once and twice, respectively. Mortality of the earthworms was recorded after 48 h. 2.6. Soil characterisation The main parameters affecting the interaction among pollutants, bioemulsifiers and soils were investigated. Organic matter was quantified by the Walkley and Black method; Cationic Exchange Capacity (CEC) was determined by BaCl and triethanolamine method; a granulometric assay was performed according to standard methods (Ministry of Agricultural Policy, 1999). 2.7. Environmental fate 2.7.1. Sorption experiments Sorption experiments were conducted using 100 mL Erlenmeyer flasks filled with 2 g of soil sample and 20 mL of bioemulsifier-water solutions at different concentrations. Preliminary experiments were conducted to determine the initial concentrations that resulted in detectable amounts of bioemulsifier being observed in the water phase after equilibrium. Flasks were shaken at 20 °C and 100 rpm for 30 min to reach sorption equilibrium and then centrifuged at 8700g for 10 min. The residual concentration of

247

bioemulsifiers was determined by anthrone method (Satpute et al., 2010). 2.7.2. Biodegradability test A soil bacteria inoculum was obtained by adding 0.5 g of a soil sample (soil 3 – Table 2) to 50 mL of rich medium LD. Growth was performed at 20 °C overnight and the culture was centrifuged (6800g for 10 min) and resuspended with mineral medium M9; this procedure was repeated to obtain a culture at a desired optical density at 540 nm (OD540). Liquid biodegradability of the bioemulsifiers was assessed by respirometric tests. Experiments were carried out using a Biochemical Oxygen Demand (BOD) apparatus (VELP Scientifica – Italy) and the ratio between actual BOD and Chemical Oxygen Demand (COD) was calculated over time. Bottles were filled with 125 mL of bioemulsifier solution in mineral medium M9, along with 2 mL of inoculum prepared as described above (OD540 = 1). The initial concentrations of bioemulsifiers were 500 and 250 mg/L. A control bottle without bioemulsifiers was assembled to measure the basal activity. Bottles were left at 20 °C for 7 days. BOD was calculated by the difference between oxygen consumption in each bottle and the basal activity. COD was measured using a standard method described elsewhere (APHA-AWWA-WPCF, 1989). 2.8. Washing tests 2.8.1. Soil contamination Aliquots of uncontaminated soils were separately contaminated with crude oil at a concentration of 5% (w/w). The hydrocarboncontaminated soils were obtained by dissolving the crude oil in n-hexane (10% w/w) and adequately mixing the solvents and the soils to uniformly distribute the pollutants. Soil samples were left uncovered at room temperature for 48 h to permit solvent evaporation. 2.8.2. Washing experiments Batch washing experiments were carried out in glass bottles containing 2 g of soil and the suitable washing solution according to the experimental design. Bottles were shaken at 200 rpm for a set period of washing time according to the experimental design. Different factorial experimental designs were used to plan the soil washing experiments. Xanthan gum (polysaccharide of microbial origin commercially available) was used as a positive control. The factors (independent variables) taken into consideration were: (i) the washing solution (as a qualitative variable used to evaluate the removal efficiencies of bioemulsifiers, Xanthan gum and water) (ii) the type of soil, and (iii) the length of washing time. A three level full factorial design (3(3) FFD) (27 experiments) was chosen. Table 1 shows the variables and the levels chosen for this experimentation. At fixed kinetic times, soil samples were subjected to chemical analyses for the determination of residual pollutants. 2.8.3. Chemical analyses The experimental mixtures were centrifuged at 560g for 5 min after the washing procedure and the supernatant solution discarded. The soil was rinsed with water and centrifuged again. After discarding the supernatant solution, the soil was analysed for the residual concentration of pollutants. To determine the residual crude oil, the soil was dried with anhydrous sodium sulphate and extracted four times with 10 mL n-hexane. The extracted crude oil was evaluated by measuring absorbance at 400 nm. The crude oil removal was determined by calculating the difference between the absorbances obtained for the washed and unwashed soils.

248


Table 1 Independent variables and levels of soil washing experiments. Variable

Levels

Washing solution (WS) Time (min) Soil

Water 30 Soil 1

1

3. Results and discussion 3.1. Strain identification, surface activity and emulsion stability Microscopic examination showed the strain 7bCT5 to be a Gram-negative bacillus. It was positive both for the cytochrome oxidase test and for glucose oxidation in the Oxidation–Fermentation test. The strain was preliminarily identified by the conventional method Biolog™ system as V. paradoxus with a percentage of identification of 99%. Subsequently, the genetic identification based on PCR amplification and sequencing of the 16S rRNA gene, confirmed the identification obtained with the phenotypic characterization. The partial sequence of the 16S rRNA gene is available on GenBank (Accession No. JN627864). V. paradoxus, originally classified in either the genus Alcaligenes or Hydrogenomonas, is a ubiquitous, aerobic, Gram-negative bacterium, present in diverse environments (Willems et al., 1991). Strains of V. paradoxus were isolated from various habitats such as contaminated soil and enriched cultures (El-Fantroussi, 2000) and have been associated with a number of biotransformations of chemically-synthesised molecules (Nishino and Spain, 2006; Smith et al., 2005). In spite of its ubiquity and a wealth of interesting metabolic capacities, relatively little has been published on the physiology of V. paradoxus. Recently, Jamieson et al. (2009), studying the coordinated surface behaviors in this bacterium, demonstrated a robust swarming motility, the ability to form biofilms and the presence of a wetting agent able to reduce the surface tension. It is thus plausible to hypothesize that all these properties may be related with the production of surface active compounds such as biosurfactants and/or bioemulsifiers. It is known that these molecules are essential for the motility of the microorganisms and are involved in cell-to-cell interactions such as bacterial pathogenesis, quorum sensing and biofilm formation, maintenance and maturation. The cell-free supernatant of V. paradoxus 7bCT5 was negative for the oil spreading test and positive for the emulsification test. A stable thick emulsion that lasted for several months was produced. In particular, the emulsification activity was of 68% at 24 h (E24h), and of 63% at 48 h and 72 h. After 6 days of incubation at room temperature, the emulsification index reached a value of 59% which remained constant up to 12 months. Moreover, a variety of different hydrophobic substrates, such as kerosene, diesel, soybean oil and motor oil, were efficiently emulsified by V. paradoxus 7bCT5 supernatant. The emulsification activity of cell-free supernatants was also followed during the growth of V. paradoxus 7bCT5 on LB broth. The maximum activity was observed at the end of the exponential phase (after 4 h) and remained constant for 48 h, indicating no further production of the bioemulsifiers during the stationary phase (data not shown). Surface tension values of V. paradoxus 7bCT5 supernatant were similar to that of sterile LB broth (data not shown), confirming that the strain is a bioemulsifier and not a biosurfactant producer. V. paradoxus 7bCT5 crude bioemulsifier extract was tested for its resistance to very high/ low temperatures. The effect of thermal treatment on the emulsifier activity showed that no appreciable changes in emulsification capacity occurred. The extract maintained the emulsification activity even after 1 h boiling and after 24 h incubation at 80 °C; moreover the emulsions remained indefinitely stable. Similarly

0

+1

Xanthan gum (200 mg/L) 50 Soil 2

Bioemulsifiers (500 mg/L) 70 Soil 3

the stability of emulsions at extreme temperature, pH and saline conditions and the ability to emulsify a wide range of hydrophobic mixtures and pure hydrocarbons have been recently described for two Geobacillus pallidus soil strains (Zheng et al., 2011). 3.2. Chemical characterization The total content of sugars, protein and lipids was determined for the bioemulsifiers. The percentages of sugars, lipids and protein were 95%, 4% and 1%, respectively. These values demonstrate that the bioemusifiers produced by V. paradoxus 7bCT5 are polysaccharide based. This also suggested that it was possible to quantitatively determine the concentration of the bioemulsifiers in water using the anthrone method (Satpute et al., 2010). MALDI-TOF–MS analysis revealed the presence of over 30 different mass spectral peaks, showing that the isolated bioemulsifiers are a heterogenic mixture. The presence of this number of structures in the mixture helps to confirm that the bioemulsifiers are polysaccharide based. In fact, it is commonly known that large polysaccharides are present as mixtures, with the isolated emulsan present as a collection of different structures between 165 and 186 kDa. This is due to different lengths of the polysaccharide backbone, based on the number of monosaccharide units present, along with different branching and linkage patterns between polysaccharide chains. Due to the high m/z ions obtained in MALDI-TOF analysis, the mass accuracy could be assumed to be to the nearest kDa, as it is common with high molecular masses within this instrument due to the difficulty in calibration at this mass range. 3.3. Ecotoxicity tests In the ecotoxicity tests the initial bioemulsifier concentration of 500 mg/L for MicrotoxÒ analyses was chosen since it is very close to the maximum water solubility of the compound. Therefore, it represents the highest concentration of bioemulsifiers with which organisms could theoretically come into contact. The results obtained did not permit calculating an EC50 value. In fact, they indicated a percentage of effect of 28 ± 3% after 5 min of contact and of 34 ± 2% after 15 min only for the highest concentration of 500 mg/ L, while the bioluminescence reduction completely disappeared after the first twofold dilution. However, such effect was attributed to the high turbidity of the 500 mg/L solution rather than to an actual toxicity of the solution; this phenomenon is well known and already described in literature (Parvez et al., 2006). On the basis of these results, we chose to limit toxicity tests on D. magna and plants only to the highest bioemulsifier concentration (500 mg/ L). Acute toxicity test on crustaceans did not reveal any effect at all, with a 100% survival of the organisms after 48 h. The Germination Indexes (GI%) calculated from seed germination and root elongation tests did not significantly differ from their controls, both in the presence and in the absence of soil. Therefore, aqueous solutions of the bioemulsifiers did not result in toxicity to any tested organisms (bacteria, crustaceans or plants) even at the water solubility limit, i.e. the highest achievable concentration. The same 500 mg/L solution was then spiked on filter paper for contact tests on earthworms. Also in this case, earthworm survival was 100% after 48 h of contact with two different concentrations of the

249


bioemulsifiers. Thus, this compound can be considered as nontoxic in its solid state as well. Hence, overall results from ecotoxicity tests suggest that the use of a bioemulsifier solution at any concentration, e.g. in soil washing applications, would be environmentally safe since the compound did not cause any appreciable ecotoxicity effects. Furthermore, potential residual traces in the soil, due to incomplete removal or biodegradation of the compound, would not provide any negative effects either. 3.4. Environmental fate Organic carbon content, CEC and granulometric characteristics of soil are among the main factors affecting environmental fate of chemicals in soil system. Analytical results of these parameters for the soils used in this work are shown in Table 2. In Fig. 1, the sorption isotherms of the bioemulsifiers for the three soils are shown. Obtained data were analysed by interpolation with different isotherm models. Sorption of the bioemulsifiers was well modelled by linear isotherms for all the three soils. Soil 2 presented the highest affinity with the biemulsifiers having a Kd of 7.22, while soil 3 and soil 1 had values of Kd equal to 3.99 and 1.39, respectively. Biodegradability of surfactants was preliminarily assessed by liquid respirometric assays. These experiments were assembled using an inoculum cultured from a soil sample. In Fig. 2 the results of respirometric assays are reported. COD values were 234 mgO2/L and 431 mgO2/L when the concentration of bioemulsifiers was 250 mg/L and 500 mg/L, respectively. The maximum oxygen consumption was reached after 6–7 days. At that time the BOD/COD ratios were 0.30 and 0.24 when the concentration of bioemulsifier was 250 mg/L and 500 mg/L, respectively. Amphiphiles are able to alter the physico-chemical conditions at the interfaces affecting the distribution of the chemicals among the phases (Tiehm, 1994). The addition of (bio)surfactants into the

environment leads to numerous and unpredictable partitions. For instance, a hydrocarbon-contaminated soil contains at least six phases: bacteria, soil particles, water, air, immiscible liquid and solid hydrocarbons. The hydrocarbons can be partitioned among different states: solubilised in the water phase, ad/absorbed to soil particles, absorbed to cell surfaces and as a free insoluble phase. Biosurfactants added to this system can interact with both the abiotic particles and the bacterial cells (Banat et al., 2010). The environmental fate properties such as sorption to soil, solubility/ hydrophobicity and biodegradability are strictly correlated to the features linked to the applicability of biosurfactants, such as toxicity and removal capacity. Toxicity is in turn often related to biodegradability since the most degradable molecules are also less toxic (Franzetti et al., 2006). For example, among synthetic surfactants, Tween 80 was easily degraded by mixed bacterial cultures and showed low toxicity towards Vibrio fischeri (Frank et al., 2010; Franzetti et al., 2006), while different classes of biosurfactants showed lower toxicity than sodium dodecyl sulphate (Lima et al., 2011). Bioemulsifiers by V. paradoxus 7bCT5 resembled this behaviour showing very low toxicity to V. fischeri; thus, ecotoxicity data coupled with BOD/COD ratios indicated that these molecules are potentially suitable for environmental releasing. As already mentioned, the estimation of the amount of surfactant lost due to soil sorption is essential for the evaluation of applicability of (bio)surfactants in bioremediation and soil washing/flushing. The sorption behaviour of chemical biosurfactants and of the most studied low molecular weight biosurfactants has been extensively investigated for a wide variety of organic and inorganic particles (Yang et al., 2010). In a previous study on sorption it was reported that soil has a high affinity to Tween 80 and low affinity to Aerosol MA80 (Franzetti et al., 2006), while more recently Ochoa-Loza and colleagues (2007) studied rhamnolipids sorption to different mineral component concentrations relevant for bioremediation and soil washing applications. To the best of our knowledge there is no

Table 2 Characterization of tested soils. Soil

1 2 3

Organic carbon (g/kg)

14 22 24

CEC (cmol(+)/kg)

16.31 17.16 21.24

Granulometric composition