INTRODUCTION

Top Abstract Introduction Materials and methods Results Discussion References

Industrial chemicals, such as hydrocarbons, have been released into the soil environment as a result of mechanical failure, incineration practices, corrosion, leakage, accidental spillage, and improper disposal practices (5, 24, 37, 44). Biodegradation is an attractive method for remediating contaminated sites because of its economic viability and environmental soundness. One limitation of biodegradation, however, is that many hydrocarbons are poorly accessible to bacteria. Heavily contaminated soils contain a separate non-aqueous-phase liquid (NAPL), which may be present as droplets or films on soil surfaces. Biodegradation takes place more readily when the target contaminants are dissolved in an aqueous solution (14, 27, 31, 40, 48, 49), but many hydrocarbons are virtually insoluble in water and remain partitioned in the NAPL. Thus, there have been efforts to improve the bioavailability of hydrocarbons through the use of surfactants.

Both nonionic and anionic surfactants increase the solubility of hydrocarbons by forming micelles (10, 15-17, 35). The surfactants begin to assemble into micelles at the critical micelle concentration (CMC), and the interiors of the micelles provide a hydrophobic environment to solubilize nonpolar compounds, such as hydrocarbons. No enhancement of solubility is observed at concentrations below the CMC. While some research groups have found that the presence of surfactants enhances biodegradation (2-4, 7, 11, 20, 41, 45, 46), others have found that the presence of surfactants inhibits biodegradation (1, 7, 9, 11, 14, 18, 19, 41, 47). This discrepancy within the literature indicates that there is a need to understand the mechanism of biodegradation in the presence of surfactants.

Diffusion of hydrocarbons to bacteria for use as growth substrates can occur by several pathways. Hydrocarbons can dissolve from the NAPL into the aqueous phase. As the bacterial population increases, however, the rate of uptake increases while the rate of dissolution remains constant (40). If surfactant is present at a concentration above the CMC, hydrocarbons can dissolve from the micelles into aqueous solution. Direct interactions between cells and micelles can also occur. Both of these dissolution pathways rely on mixing and diffusion in the aqueous phase to bring the hydrophobic compounds to the bacteria. A more direct pathway is adhesion of the bacteria to the interface between the NAPL and the aqueous phase. A number of species of bacteria are able to degrade liquid hydrocarbons after adhering to the surfaces of droplets (8, 11, 23, 26, 31, 32). This direct contact between a bacterial cell and a target hydrocarbon can significantly increase the rate of diffusion into the cell, thereby enhancing growth and increasing the apparent rate of dissolution of the hydrocarbon. While few bacteria that grow on viscous tars or solid hydrocarbons have been identified, growth of a Mycobacterium species (previously described as a Rhodococcus species) as a biofilm on solid anthracene has been observed (43), suggesting that immediate proximity of a solid carbon source to a cell has the same benefit as it does in the case of liquid hydrocarbons.

The overall impact of addition of a surfactant on biodegradation depends on how the basic diffusion pathways are altered and whether the surfactant itself affects the cells. If the surfactant is neither toxic nor a growth substrate, it can either increase the rate of biodegradation by carrying hydrocarbons in relatively accessible micelles, or it can decrease the rate by inhibiting the adhesion of cells to the NAPL-water interface. The overall impact depends on the importance of each pathway. The presence of surfactants at concentrations above the CMC does inhibit adhesion of bacteria to the surfaces of droplets of liquid hydrocarbons and thus inhibits biodegradation (11, 26, 30). The sorption of surfactants to bacteria and to interfaces can either enhance or inhibit adhesion, depending on the nature of the surfaces and the surfactant itself (25). The alterations of surfaces depend only on the concentration of free surfactant; therefore, they are significant at concentrations below the CMC.

The purpose of this study was to study the adhesion of hydrocarbon-degrading bacteria to the surfaces of NAPL materials that occur in contaminated soils, with and without added surfactants. These NAPL materials are much more viscous than hydrocarbon liquids, such as hexadecane, which have been studied in the past. Our hypothesis was that sorption of surfactants to cells or to NAPL surfaces would inhibit bacterial adhesion to the NAPL surfaces. The adhesion of two polycyclic aromatic hydrocarbon (PAH)-degrading organisms to highly viscous NAPLs was measured in the presence of two surfactants, one nonionic (Triton X-100) and one anionic (Dowfax 8390). Because the sorption of a surfactant to surfaces depends on the free concentration of the surfactant in solution, modifications of surface properties are significant at concentrations below the CMC. So that we could concentrate on surface adhesion and avoid solubilizing the NAPL components in micelles, each surfactant was used at one-half its CMC. Growth of the bacteria on anthracene was used to determine whether surfactants had any influence on bacterial growth on a solid carbon source.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and methods Results Discussion References

NAPL extracts from contaminated soils. NAPL extracts from four different contaminated soils were used in the adhesion experiments. The soils originated from industrial sites in Edmonton, Alberta, Canada; Prince Albert, Saskatchewan, Canada; Devon, Alberta, Canada; and Montreal, Quebec, Canada. The Edmonton (EDM) and Prince Albert (PAA) soils were primarily contaminated with creosote, while the Devon (DEV) and Montreal (MTL) soils were primarily contaminated with petroleum hydrocarbons. The MTL soil had been partially bioremediated before the investigation was begun. The organic material was extracted from approximately 15 g of each soil with 200 ml of methylene chloride by using a Soxhlet extraction apparatus over a period of 8 to 10 h. The viscosity of each NAPL was determined at 23 and 60°C by using a Carri-Med CLS controlled stress rheometer. At 23°C, the stress was applied linearly from 0 to 4,000 dynes/cm² for the DEV and EDM NAPLs and from 0 to 1,000 dynes/cm² for the PAA NAPL. The viscosity of the MTL NAPL could not be determined at this temperature. At 60°C, the stress was applied linearly from 0 to 2,000 dynes/cm² for the MTL NAPL and from 0 to 100 dynes/cm² for the other NAPLs.

Bacteria. In this study we used two hydrocarbon-degrading bacterial strains isolated from a creosote-contaminated soil, a gram-positive Mycobacterium strain and a gram-negative Pseudomonas strain (13). The Mycobacterium strain was originally described as Rhodococcus sp. strain S1 based on chemotaxonomic data (43) but was recently reclassified as a Mycobacterium strain on the basis of 16S rRNA analysis (6). This analysis showed that this strain is closely related to, but not identical to, Mycobacterium fortuitum. The 16S rRNA sequence data has been deposited in the EMBO database under accession no. Y15709 (Mycobacterium sp. strain S1). The Pseudomonas strain was not characterized in as much detail as the Mycobacterium strain was.

Chemicals. Both a nonionic surfactant and an anionic surfactant were used in this investigation. The nonionic surfactant was Triton X-100, obtained from Rohm and Haas Company of Canada Limited, West Hill, Ontario, Canada. The anionic surfactant was Dowfax 8390, obtained from Dow Chemical Company, Midland, Mich. The CMC of each surfactant was determined by using a tensiometer (model 70545; Central Scientific Company, Chicago, Ill.). Anthracene was obtained from Sigma Chemical Company, St. Louis, Mo., and was reported to be 99% pure. Anhydrous D-glucose was obtained from BDH Inc., Toronto, Ontario, Canada. Methylene chloride was obtained from EM Science, Gibbstown, N.J., and was determined to be more than 99% pure (42).

Adhesion experiments. To determine whether surfactants affect the adhesion of bacteria to NAPLs, each organic extract was dissolved in methylene chloride to a concentration of 1 g/liter. Each test tube was prepared by adding 5 ml of this solution. The test tubes were rotated in a roller test tube rack at 12 rpm, which allowed the methylene chloride to evaporate while the inside surface of each test tube became coated with the NAPL. The bacteria were harvested from the growth medium after 14 days by centrifugation at 16,200 × g for 10 min. Each pellet was resuspended three times in 50 mM potassium phosphate buffer (pH 7.2). The resulting suspension, which had an optical density at 600 nm (OD₆₀₀) of approximately 0.6, was then divided into three parts. Triton X-100 was added to one part of the bacterial suspension to a concentration of 0.12 mM, and Dowfax 8390 was added to a second part of the bacterial suspension to a concentration of 0.4 mM. Surfactant was not added to the third part of the bacterial suspension so that it could be used as a control. A 50 mM potassium phosphate buffer solution was also divided into three parts, which were used as blanks; one part contained 0.12 mM Triton X-100, one part contained 0.4 mM Dowfax 8390, and one part did not contain any surfactant. These blanks were used to account for any NAPL that came off the surfaces of the test tubes, which altered the measured OD₆₀₀. The OD₆₀₀ of each of the six solutions was recorded. Each of the six solutions was transferred to test tubes containing one of the four NAPLs and to test tubes that did not contain any NAPL (5 ml per tube). Test tubes for each set of conditions were prepared in triplicate, and the total number of test tubes was 90.

Bacterial growth on solid anthracene. To determine whether the presence of surfactant had any effect on the ability of the bacteria to grow on a solid carbon source, sterile flasks containing 50 mg of anthracene, 90 ml of sterile growth medium, and 10 ml of a bacterial suspension were prepared aseptically. Flasks containing either Triton X-100, Dowfax 8390, or no surfactant (control) were prepared in quintuplicate. A 1-ml sample was taken aseptically from each flask at zero time and on days 1, 2, 3, 5, 7, 10, and 14, and the OD₆₀₀ of each sample was recorded as a measure of bacterial growth. At the end of the experiment, the Mycobacterium strain was streaked onto plate count agar and the Pseudomonas strain was streaked onto Trypticase soy agar to ensure that the cultures remained pure and uncontaminated. This experiment was repeated with the Pseudomonas strain and 50 mg of glucose instead of anthracene in order to determine the effect of surfactants on bacterial growth in the presence of a soluble carbon source.

	RESULTS

Top Abstract Introduction Materials and methods Results Discussion References

Viscosity of the NAPLs. The viscosity of each NAPL was determined at 23 and 60°C, as shown in Table 1. Each NAPL was much more viscous than hexadecane (28), which is a typical liquid hydrocarbon that has been used in previous bacterial adhesion studies (30-33). The MTL NAPL was essentially solid at room temperature.

Determination of the surfactant CMCs. The CMC of Triton X-100 was determined to be 0.24 mM, a value which has been reported elsewhere (18). The CMC of Dowfax 8390 was determined to be 0.8 mM, which is 1 order of magnitude lower than the value which has been reported elsewhere (34). Throughout this study, the surfactants were used at one-half their CMCs (0.12 mM for Triton X-100 and 0.4 mM for Dowfax 8390).

Initial adhesion of bacteria to the NAPL interface. The percentages of cells that adhered to both the coated test tubes and the clean test tubes in the presence and absence of surfactant are shown in Table 2. Approximately 40% of the Mycobacterium cells adhered to the four NAPLs in the absence of surfactant, while approximately 50% of the Mycobacterium cells adhered to the surfaces of the clean glass test tubes in the absence of surfactant. Fewer cells adhered to the surfaces if either surfactant was present. In the presence of Triton X-100, only 5 to 17% of the cells adhered to the NAPLs, while in the presence of Dowfax 8390, only 11 to 36% of the cells adhered to the NAPLs. Similarly, in the absence of any NAPL, only 5 and 19% of the cells adhered to the glass in the presence of Triton X-100 and Dowfax 8390, respectively. Clearly, the Mycobacterium strain was able to adhere to both the glass test tube surfaces and the NAPLs, but the presence of either surfactant inhibited adhesion. Triton X-100 at a concentration that was one-half its CMC, however, inhibited adhesion more effectively than Dowfax 8390 at a concentration that was one-half its CMC. No correlation between adhesion and viscosity of the NAPLs was apparent. The OD₆₀₀ was linearly proportional to the concentration of bacteria, with an OD₆₀₀ of 0.1 corresponding to approximately 10⁸ CFU/ml (43). Examination of the suspension with the microscope revealed that no aggregation or flocculation of the Mycobacterium cells occurred during this experiment.

Removal of bacteria from the NAPL interface. If low levels of a surfactant inhibit the adhesion of cells to surfaces, they may also result in removal of previously adhered cells by sorbing to cell or NAPL surfaces (25). Table 3 shows the effects of surfactants on adhesion of both strains. Between 31 and 52% of the Mycobacterium cells remained adhered to the surfaces if no surfactant was added to the test tubes, but fewer cells remained if either surfactant was added. After Triton X-100 was added, the percentages of cells that remained adhered to the surfaces ranged from 4 to 30%, while after of Dowfax 8390 was added, the percentages of cells that remained adhered to the surfaces ranged from 3 to 26%. Clearly, addition of either surfactant resulted in removal of Mycobacterium cells from the NAPLs.

Growth of bacteria on solid anthracene. Data for the growth of the Mycobacterium strain on anthracene in the presence and absence of surfactants are shown in Fig. 1. In the absence of surfactant, growth occurred until approximately day 7, at which time the OD₆₀₀ was 1.16. In the presence of Triton X-100, the cell concentration increased more slowly, reaching an OD₆₀₀ of 0.82 on day 14. In the presence of Dowfax 8390, the OD₆₀₀ reached a maximum value of 0.32 after day 10. In the absence of either surfactant, the liquid medium turned yellow as growth proceeded, but the medium remained clear when either surfactant was present. The yellow pigmentation did not interfere with OD₆₀₀ measurements.

View larger version (16K): [in this window] [in a new window]   FIG. 1.   Growth of the Mycobacterium strain on anthracene in the presence and absence of surfactants at concentrations that were one-half the CMCs. The error bars indicate standard deviations based on four or five replicates; where there are no error bars, the standard deviations were smaller than the symbols.

View larger version (17K): [in this window] [in a new window]   FIG. 2.   Growth of the Pseudomonas strain on anthracene in the presence and absence of surfactants at concentrations that were one-half the CMCs. The error bars indicate standard deviations based on four or five replicates; where there are no error bars, the standard deviations were smaller than the symbols.

View larger version (15K): [in this window] [in a new window]   FIG. 3.   Growth of the Pseudomonas strain on glucose in the presence and absence of surfactants at concentrations that were one-half the CMCs. The three sets of data are statistically equivalent (error bars are not shown).

	DISCUSSION

Top Abstract Introduction Materials and methods Results Discussion References

In order for biodegradation to occur, the bacteria must have access to the target compounds, either by dissolution of the target compounds in the aqueous phase or by adhesion of the bacteria directly to the NAPL-water interface. Dissolution has been emphasized in the degradation of solid PAHs (3, 22, 46), while adhesion has been stressed in the case of liquid hydrocarbons, such as hexadecane (31, 32). The results of the present study show that bacteria were able to adhere to more viscous NAPL and solid surfaces; however, adhesion was inhibited in the presence of surfactants at concentrations below the CMCs. Addition of surfactants at low concentrations also resulted in removal of adhered bacteria from the NAPL-water interface. Furthermore, the presence of a surfactant had a deleterious effect on the growth of both the Mycobacterium strain and the Pseudomonas strain when solid anthracene was used as a carbon source.

Similar phenomena have been reported for adhesion of cells to droplets of low-viscosity liquid hydrocarbons. Addition of Triton X-100 at a concentration greater than its CMC inhibited adhesion of an Arthrobacter species to a heptamethylnonane-water interface, which in turn prevented degradation of both hexadecane and naphthalene in the heptamethylnonane phase (11, 26). Addition of surfactant also resulted in removal of previously adhered bacteria from the liquid substrates. In a similar study, addition of a nonionic surfactant caused yeast cells to detach from an n-alkane-water interface, leading to a decline in the growth rate of the culture (1). In another study, the presence of a surfactant had little effect on the growth rate of yeast cells on a soluble carbon source, although Triton X-100 at a concentration greater than its CMC did delay the onset of the exponential growth phase (21). The present study extended these findings to more viscous substrates and solid PAHs by showing that surfactants inhibit bacterial adhesion to the surfaces of these compounds and that surfactants inhibit bacterial growth on such carbon sources.

Although the presence of surfactant had a negative impact on the adhesion of both strains to the NAPLs, adhesion of the Mycobacterium strain was inhibited more in the presence of Triton X-100 than in the presence of Dowfax 8390. The reverse was true for the Pseudomonas strain. This result suggests that the relationship between surfactant charge and cell surface characteristics is more important in determining bacterial adhesion than the properties of the NAPL are. The cell walls of acid-fast bacteria, such as corynebacteria, mycobacteria, and Nocardia strains, contain mycolic acids, nocardols, and nocardones. These components do not carry a charge, but they increase the hydrophobicity and polarity of the cell surfaces, making the cells more likely to interact with surfactants that do not carry a charge (39). Conversely, the surfaces of gram-negative cells, such as Pseudomonas cells, carry a charge, indicating that these cells are more likely than gram-positive cells to interact with surfactants that carry a charge. Some evidence suggests that the positive charges on the surface of a bacterial cell play an important role in adhesion (38).

In the test tube experiments, the two strains had similar affinities for both the NAPLs and the glass surfaces. This observation was unexpected, because the NAPLs are hydrophobic and glass is hydrophilic. Hydrophobic interactions regulate the adhesion of bacteria to hydrocarbons (33). As summarized by Neu (25), surfactants can alter adhesion by adsorbing to the cell surface, to the hydrocarbon surface, or to both. If the hydrophobic ends of the surfactant molecules adsorb to hydrophobic surfaces, such as the NAPLs used in this study, then the hydrophilic ends remain in the aqueous phase. This adsorbed layer should make the surface more hydrophilic. Adsorption of surfactant to hydrophobic domains on the cell surface should give the same result. Either surface modification should decrease the hydrophobic interactions between the cells and the NAPL and reduce adhesion. Unfortunately, this mechanism cannot explain why the nonionic surfactant inhibited bacterial adhesion to glass.

An alternative explanation is that adsorption of the surfactants to the cells may have had a dispersive effect by increasing the steric hindrance to cell-cell and cell-surface interactions (36). This mechanism would account for the inhibition of adhesion of the bacteria to both hydrophobic and hydrophilic surfaces. This dispersion mechanism is important in some industrial applications (for example in the pulp and paper industry, where dispersants are used to reduce biofilm deposits) (29).

Regardless of which mechanism is responsible for the role of surfactants in reducing adhesion, the observations made in this study suggest why the previously published data on the benefits of surfactants in biodegradation of hydrophobic compounds are so contradictory. In the absence of toxicity, the net effect of addition of a surfactant to a contaminated soil depends on the benefits that result from enhanced solubility of target compounds versus the reduction in direct adhesion of bacteria to the NAPL. Consequently, the impact of surfactant addition on biodegradation depends on which mechanism is responsible for uptake of compounds in the highly viscous NAPL or solid phase.

The results of the bacterial growth experiment could be explained by two possible mechanisms. First, the surfactants may inhibit growth. Second, the interactions between the cells and the substrate may be altered in the presence of a surfactant. Since further investigation showed that neither surfactant affected the growth of the Pseudomonas strain on glucose, we concluded that the surfactants did not inhibit growth of this organism and did not serve as carbon sources. The adhesion studies clearly showed that surfactants reduced the adhesion of cells to NAPLs, independent of the viscosity of the hydrocarbon. Figures 1 and 2 clearly show that both bacterial strains were able to grow when solid anthracene was used as the sole carbon source and that growth was suppressed in the presence of a surfactant. If transient adhesion of cells to anthracene were a significant factor in the rate of uptake of anthracene by cells, then surfactant would inhibit both adhesion of the cells and cell growth on the solid PAH. Consequently, biodegradation would be inhibited. Transient adhesion would also be consistent with previous observations that cells do not easily adhere to PAHs to form biofilms (43, 48). Anthracene did not adhere well to glass; therefore, adhesion in the presence of surfactant could not be studied directly by using the techniques used for the NAPL materials. Further study is needed to determine whether transient adhesion is necessary for rapid uptake and growth on contaminants that are present in highly viscous or solid phases.