INTRODUCTION

Top Abstract Introduction Materials and Methods Results Discussion References

Biodegradation of heterogeneous petroleum waste streams in refinery-based fermentation systems relies on mixed microbial cultures. Physical, metabolic, and community interactions contribute to the dynamic nature of these systems, with different populations employing different methods to access hydrocarbons. These methods include uptake of water-soluble hydrocarbons, direct adherence of microbes to hydrocarbon-water interfaces, and biosurfactant-mediated micellar transfer (6, 12). Consequently, methods that alter hydrocarbon solubility, e.g., addition of a chemical or biological surfactant, will not have uniform effects on different members of the microbial community (8, 32). An understanding of these interactions is necessary when developing treatment techniques for hydrocarbon biodegradation in refinery-based or field-based fermentation units.

Pseudomonas spp. and Rhodococcus spp. often are isolated from hydrocarbon-contaminated sites and hydrocarbon-degrading cultures (6, 17). These two genera have a broad affinity for hydrocarbons and can degrade selected alkanes, alicyclics, thiophenes, and aromatics (1, 5, 11, 14, 20, 21, 25). Strains within each genus also produce a range of biosurfactants (9, 20), e.g., rhamnolipids (26) and trehaloselipids (7, 15, 23, 33). Typically, Rhodococcus spp. are more hydrophobic and have a higher affinity for hydrocarbon-water interfaces than Pseudomonas spp. (28), suggesting that these strains utilize different modes of hydrocarbon accession. Interactions between these genera can be important for the biodegradation of contaminants; e.g., cocultures of Pseudomonas and Rhodococcus spp. can mineralize chloronitrobenzenes (22). Conversely, Ko and Lebeault (13) showed that Rhodococcus equi P1 competitively uses hexadecane in a hydrocarbon mixture, reducing hexadecane-dependent cooxidation of decalin by Pseudomonas aeruginosa K1.

Previously, strains isolated from a mixed bacterial culture growing on crude oil and crude oil fractions (32) were screened for crude oil biodegradation and emulsification abilities. Two strains identified by fatty acid analysis were particularly interesting. Rhodococcus sp. strain F9-D79 forms excellent, though transient, crude oil-water emulsions between 24 and 48 h of incubation. Pseudomonas sp. strain JA5-B45 does not emulsify oil but does efficiently degrade crude oil in the presence of nonylphenol ethoxylate.

We hypothesized that a coculture of the two organisms will enhance total petroleum hydrocarbon (TPH) removal through the combination of superior emulsification and degradation capabilities. In this study, we used simple pure cultures and cocultures to study metabolic and physiological interactions that may occur in a more complex mixed-culture fermentation system. In addition, we determined how exogenous chemical surfactants may affect treatment outcomes, depending on the organisms involved.

	MATERIALS AND METHODS

Top Abstract Introduction Materials and Methods Results Discussion References

Culture medium. The medium used for culture selection and biodegradation of 20 g of Bow River crude oil per liter contained, per liter, 1 g of KH₂PO_4, 1.5 g of Na₂HPO₄, 2 g of urea, 0.2 g of MgSO₄ · 7H₂O, 0.1 g of Na₂CO₃, 50 mg of CaCl₂ · 2H₂O, 5 mg of FeSO₄, 20 mg of MnSO₄, and 3 ml of trace element solution. The trace element solution contained, per liter, 14 mg of ZnCl₂ · 4H₂O, 12 mg of CoCl₂, 12 mg of Na₂MoO₄ · 2H₂O, 1.9 g of CuSO₄ · 5H₂O, 50 mg of H₃BO₄, and 35 ml of 12 N HCl. Yeast extract (Difco Laboratories, Detroit, Mich.) was added at 1 g/liter, and the initial pH was 7.0.

Substrates. We used Bow River crude oil (density, 0.905 g/ml; 22% volatiles, 25% saturated compounds, 42% aromatics, 5.8% resins, and 5.5% asphaltenes) (Imperial Oil, Sarnia, Ontario, Canada). The oil was stored at 4°C in a glass bottle sealed with a Mininert cap (VICI Precision Sampling Inc., Baton Rouge, La.). No changes in oil composition were noted by SARA (saturate, aromatic, resin, and asphaltene) and solid-phase microextraction analysis (30, 31). The nonlyphenol ethoxylate surfactant Igepal CO-630 (hydrophile lipophile balance, 13; critical micellization concentration, 54 mg/liter) (Rhône-Poulenc, Cranbury, N.J.) was used without further purification. Hexadecane (99%) (Sigma, St. Louis, Mo.), D-glucose, sodium acetate (BDH Ltd., Toronto, Ontario, Canada), Trypticase soy broth (TSB) (30 g/liter) (Becton Dickinson, Cockeysville, Md.), and diesel fuel from a local PetroCanada filling station (Waterloo, Ontario, Canada) also were used as growth substrates.

Culture source and maintenance. Isolates were obtained during a study of the community structure (32) of a mixed bacterial culture obtained from petroleum-contaminated soil (30) and identified by MIDI (Newark, Del.) fatty acid analysis (32). Isolates were stored at 80°C using the Microbank bacterial preservation system (Pro-Lab Diagnostics, Richmond Hill, Ontario, Canada). The isolates were identified as P. aeruginosa and Rhodococcus globerulus and are referred to here as Pseudomonas sp. strain JA5-B45 and Rhodococcus sp. strain F9-D79, respectively. Prior to initiation of a degradation experiment, a single frozen Microbank bead was streaked onto a Trypticase soy agar-40-g/liter soybean-casein digest agar (Becton Dickinson) plate and grown for 3 days at 30°C. A single colony was used to inoculate each flask, as Rhodococcus sp. strain F9-D79 could not be resuspended, which precluded optical density measurements.

Removal of TPH from crude oil. Microbial removal of TPH from Bow River crude oil (20 g/liter) with and without 0.625 g of Igepal CO-630 per liter was carried out in triplicate 250-ml Erlenmeyer flasks containing 50 ml of medium and 1 g of yeast extract per ml. Flasks were incubated at 30°C (with 175-rpm orbital shaking) prior to extraction of the entire contents with 50 ml of dichloromethane (DCM). Samples were then centrifuged at 12,000 × g for 10 min to break oil-in-water emulsions, and the oily solvent phase was dehydrated over granular anhydrous sodium sulfate. Extracts were concentrated under vacuum and transferred to preweighed 50-ml beakers. After drying to a constant weight in a fume hood, the total nonvolatile hydrocarbon level was measured gravimetrically and compared to those in both uninoculated and killed (0.1% [vol/vol] perchloric acid) controls (30). Following extraction, supernatant pHs were measured.

SARA analysis and gas chromatography (GC)-flame ionization detection analysis of the saturated fraction. SARA fractionation of crude oil was previously described (30). Prior to extraction of crude oil from flasks, 50 µl of a mixture of squalene (28 µl/ml) and fluoranthene (10 mg/ml) (Sigma) in DCM was added as internal standards for the saturated and aromatic fractions.

Surface tension. Surface tension measurements were made with a CSC-du Noüy ring tensiometer (Central Scientific Company, Fairfax, Va.) equipped with a platinum-iridium ring. Whole broth surface tensions were measured prior to centrifuging cultures at 12,000 × g for 10 min. Cell-free supernatants were discarded, and cell pellets were resuspended in 50 mM phosphate buffer, pH 7.0. Surface tensions of both fractions were measured to determine if biosurfactants were released into the culture medium. We took five readings for each sample and for fresh medium with carbon source.

Staining and microscopy. We used an Optiphot phase-contrast microscope (Nippon Kogaku K. K., Tokyo, Japan) to examine live cultures, India ink as a negative stain for capsule visualization, and Ziehl-Neelsen stain to detect mycolic acids (4). Live cultures were stained as described by the manufacturer with a LIVE BacLight bacterial Gram stain kit (Molecular Probes, Eugene, Oreg.) and visualized with a fluorescence microscope.

Determination of surfactant toxicity by biomass measurement. We evaluated surfactant toxicity to Rhodococcus sp. strain F9-D79 by quantifying biomass production following 5 days of growth at 30°C on TSB supplemented with Igepal CO-630. Cells recovered by centrifugation (10 min, 12,000 × g) were washed three times with deionized water prior to being dried to a constant weight at 100°C for 24 h.

Emulsification assay. The method of Nadarajah (18) for monitoring biological deemulsification of hydrocarbon-water emulsions was adapted as follows. Five milliliters of culture supernatant or resuspended cells (50 mM phosphate buffer, pH 7.0) was vortexed at the highest setting while 5 ml of kerosene was added dropwise over 30 s in 10-ml graduated cylinders. Once the kerosene was added, a Teflon cap was used to seal the cylinder prior to vortexing for an additional 30 s. The volume of the middle emulsion phase was normalized to the total volume to calculate percent emulsification after 1 h. All emulsions were stable for at least 24 h. Cultures were grown on 2% (wt/vol) Bow River crude oil for the assay, and cells were recovered by pipetting from the solvent-water interface of DCM-extracted samples.

	RESULTS

Top Abstract Introduction Materials and Methods Results Discussion References

Coculture effects on crude oil emulsification. We observed the physical state of Bow River crude oil in shake flasks containing Rhodococcus sp. strain F9-D79, Pseudomonas sp. strain JA5-B45, or a coculture of the two. The F9-D79 culture passed through a stage (24 to 48 h) where emulsification occurred, and the coculture produced a similar but more stable emulsion. The lack of turbidity in the F9-D79 culture sharply contrasted with the turbid, nonemulsified JA5-B45 culture, indicating that F9-D79 grew on the oil-water interface.

Microscopic characterization. We observed gram-negative JA5-B45 rods in the aqueous phase of crude oil cultures (Fig. 1a) and clumps of oil-associated gram-positive F9-D79 rods (Fig. 1b). Strands of fibrous material extended between F9-D79 cell clumps, and, in coculture, JA5-B45 cells attached to this material (Fig. 1c). When we used phase-contrast microscopy, we saw opaque JA5-B45 rods, both motile and nonmotile, in the aqueous phase of crude oil-grown cultures (Fig. 1d). Nonmotile F9-D79 cells growing in clumps appeared white under phase contrast and accumulated solely on oil droplets (Fig. 1e). In coculture, strands of JA5-B45 cells remained in the aqueous phase adjacent to adherent F9-D79 cells (Fig. 1f). Live fluorescent Gram stain probe binding was inefficient, although gram-negative and gram-positive rods were observed only in the aqueous and hydrocarbon phases, respectively.

View larger version (90K): [in this window] [in a new window]   FIG. 1.   Photomicrographs of 48-h crude oil cultures. Bars, 20 µm. (a to c) Gram staining of JA5-B45 (a), F9-D79 (b), and coculture (c). (d to f) Phase-contrast microscopy of live cultures of JA5-B45 (d), F9-D79 (e), and coculture (f). (g) Negative staining of F9-D79. (h) Phase-contrast microscopy of crude oil-grown F9-D79 (culture from panel e) treated with Igepal CO-630 at 48 h of incubation.

Alteration of strain F9-D79 morphology by Igepal CO-630. With crude oil as the substrate, adherent F9-D79 cells could be washed from the oil-water interface with Igepal CO-630 treatment (Fig. 1h). When Igepal CO-630 was added at the beginning of a fermentation, F9-D79 cells were dispersed in the aqueous phase on soluble and insoluble substrates. A capsule was not produced, and cells were single, nonmotile cocci 0.5 µm in diameter rather than rods 2.5 to 3.5 µm long (Table 1). The surfactant was not acutely toxic to strain F9-D79, with biomass yields ranging from 70 to 80% of that for a no-surfactant control in TSB broth containing 0.312 to 2.5 g of Igepal CO-630 per liter. No surfactant effect on strain JA5-B45 morphology was evident.

Surface tension and emulsification abilities. Surface tension measurements were confounded by crude oil and the viscous nature of the bioemulsified cultures. No surface tension reductions were observed in cultures of strain F9-D79 or strain JA5-B45 grown on TSB or 0.2% (wt/vol) glucose-acetate. No surface tension reductions were observed in strain JA5-B45 cultures grown on diesel fuel (0.2 or 2%, wt/vol) or hexadecane (0.2%, wt/vol). Conversely, significant reductions in surface tension were associated with the cells of F9-D79 cultures grown on 2% diesel fuel and 0.2% hexadecane (data not shown).

Biological removal of crude oil: coculture, chemical surfactant, and pH. By weight, TPH removal by the pure cultures was minimal (Fig. 2), although strain F9-D79 removed 15 to 20% of the saturated fraction (Fig. 3A), corresponding to 55 to 60% of the GC-resolvable compounds from C₁₆ to C₃₂, including branched compounds such as pristane and phytane (data not shown). Strain JA5-B45 removed <10% of the saturated fraction by weight. Insignificant quantities of the aromatic fraction were removed in both cases (Fig. 3B). A slight enhancement in TPH removal by the coculture was noted. The saturated fraction was removed at the same rate as for the strain F9-D79 culture, and the aromatics were not affected.

View larger version (22K): [in this window] [in a new window]   FIG. 2.   Time course curves for Bow River crude oil removal. Closed symbols, no surfactant; open symbols, 0.625 g of Igepal CO-630 per liter.  and , JA5-B45;  and , F9-D79;  and , coculture; , coculture plus surfactant at day 4. Error bars indicate standard deviations; n = 3.

View larger version (24K): [in this window] [in a new window]   FIG. 3.   Biodegradation of saturated compounds (A) and aromatics (B) by cultures growing on crude oil. Symbols are as described in the legend to Fig. 2. Errors were <10% of the plotted values; n = 3.

View larger version (22K): [in this window] [in a new window]   FIG. 4.   Supernatant pH values from crude oil biodegradation samples. Symbols and errors bars are as described in the legend to Fig. 2.

	DISCUSSION

Top Abstract Introduction Materials and Methods Results Discussion References

We found that a nonylphenol ethoxylate chemical surfactant (Igepal CO-630) can enhance biological transformation of water-insoluble substrates by a coculture through disruption of hydrophobic cell-substrate interactions of one strain. This observation suggests that the various stimulatory and inhibitory effects of nontoxic biological and chemical surfactants reported previously (12, 18, 34) are not contradictory but, rather, describe unique cases based on surfactant properties and the physiology of the organisms involved. Of specific importance is the mode by which bacteria access insoluble hydrophobic substrates. Knowing how bacteria respond to surfactant amendments, either alone or in the presence of other species, is useful when developing fermentation protocols involving mixed cultures.

We used a Rhodococcus strain that produced a mycolic acid capsule and formed transient crude oil-water emulsions while growing solely on the oil-water interface. Direct adherence to hydrocarbons is a common uptake mechanism for Rhodococcus spp. (16). We use the term hydrocarbon accession mode rather than the traditional term hydrocarbon uptake mode to describe this phenomenon, as bioemulsification or adherence may enhance access to a substrate but can be quite distinct from passive or active transport across the cell wall. We observed the capsule extending into the aqueous phase, and hydrocarbon-grown F9-D79 cells had surface-active and emulsification properties. In other Rhodococcus spp., most of the surface-active agents are cell-bound glycolipids (3, 16). For example, Whyte et al. (35) observed flocs of Rhodococcus sp. strain Q15 when it was grown on hydrocarbons. These flocs were connected by an extracellular polymer, which was a complex mixture of glygoconjugates.

Igepal CO-630 washed F9-D79 cells from the oil phase, presumably through disruption of cell-oil hydrophobic interactions upon emulsification. Alternatively, the hydrophobic nonylphenol moiety in the surfactant may have intercalated with the distal moieties of bound and free mycolic acids in the outer layer of the cell envelope. Such an interaction would render the cells more uniformly hydrophilic, but this did not appear to be the case, as the cells removed from the interface clumped together. Natural surface amphiphiles of mycolata might orient themselves with their hydrophobic heads pointing away from the cell in the lipid layer (29). F9-D79 cells from cultures supplemented with Igepal CO-630 at the beginning of a fermentation were dispersed, capsule-free cocci rather than clumped, oil-associated rods. This dispersed nature contrasted with the cell clumping observed in water-soluble media. Cell clumping was not observed by Whyte et al. (35) for strain Q15 when it was grown on glucose-acetate, and Stelmack et al. (27) have described a nonaggregating hydrophobic Mycobacterium strain.

Supernatants from Igepal CO-630-amended F9-D79 cultures had significantly lower pH values than did surfactant-free cultures. F9-D79 does not produce acidic metabolites when grown on Igepal CO-630 alone, which is consistent with a disruption in the accumulation of alkane-derived fatty acids into cell wall mycolic acids. This hypothesis is also supported by a recent study by Rodgers et al. (24), who have shown that Rhodococcus rhodochrous (ATCC 53968) directly incorporates C₁₆ and C₁₈ alkanes into cell wall lipids.

The linear rate of alkane removal by F9-D79 in the presence of Igepal CO-630 may indicate mass transfer limitations. Surfactants can reduce biodegradation by inhibiting attachment of microorganisms to oil-water interfaces (18, 34). For example, Stelmack et al. (27) showed that two nontoxic surfactants inhibit both adhesion to non-aqueous-phase liquid and growth on anthracene for a Mycobacterium strain and a Pseudomonas strain. In addition, Foght et al. (10) concluded that emulsan from Acinetobacter calcoaceticus RAG-1 inhibited degradation of saturated compounds by a mixed culture and several isolates by coating oil droplets in a hydrophilic shell.

In contrast to F9-D79 cultures, strain JA5-B45 did not adhere to hydrocarbons and was limited by hydrocarbon insolubility. When Igepal CO-630 was added, JA5-B45 cultures could use both saturated and aromatic compounds in the crude oil. This was a micellar phenomenon, as the surfactant was added at above the critical micellization concentration. Interestingly, Al-Tahhan et al. (2) recently showed that rhamnolipid addition increased the cell surface hydrophobicity of a P. aeruginosa strain by removing lipopolysaccharide. They hypothesized that this removal may explain why subcritical micellization concentrations of surfactants can enhance biodegradation in some cases.

The coculture of F9-D79 and JA5-B45 cells had a slightly enhanced crude oil removal rate. However, TPH removal by the coculture was highest when the chemical surfactant was added at the beginning of the fermentation, preventing strain F9-D79 from growing on the oil-water interface. Indeed, addition of Igepal CO-630 to coculture preemulsified by strain F9-D79 increased the removal of saturated and aromatic compounds, probably by washing F9-D79 cells from the interface and forming a more dynamic micellar solution. It appears that the emulsion formed primarily by strain F9-D79 does not allow strain JA5-B45 access to aromatic hydrocarbons. Since F9-D79 cells normally coat oil droplets and do not appear to be easily removed, less water-soluble compounds should reside in the organic interior. Conversely, fine micellar solutions of chemical surfactant and oil will be in a constant state of flux, breaking apart and reforming, resulting in the opportunity for increased mass transfer rates.

We suggest a dual role for Igepal CO-630 in this hydrocarbon-degrading coculture system. First, Igepal CO-630 directly increases the availability of both the saturated and aromatic fractions to JA5-B45 cells. Second, the surfactant inhibits attachment to, or removes F9-D79 cells from, the oil-water interface with the formation of a more dynamic emulsion, indirectly increasing crude oil availability for strain JA5-B45. This second role is an important mechanism to help explain the varied effects of nonylphenol ethoxylate, and perhaps other chemical surfactants, in mixed culture systems degrading non-aqueous-phase substrates.