A study was undertaken to thoroughly characterize the petroleum hydrocarbon (PHC) degradative ability of a recovered microbial consortium for later use in a managed microbial biodegradation system. Liquid bioreactors at three different temperatures (16, 22, and 30^oC) were used to determine the recovered microbes' petroleum hydrocarbon (PHC) degradative abilities. The PHCs within the bioremediation systems were extracted at set time intervals using organic solvents and liquid-liquid extraction techniques and the total petroleum hydrocarbons (TPHs) were analyzed and quantified by GC-FID. GC-MS was also used on some samples to determine if any preferential degradation of compounds was occurring within the complex mixture. Acute toxicity and mutagenicity of the water soluble fraction (WSF) of the test systems was evaluated using the Microtox and Salmonella/microsome (or Ames) assays, respectively. In the 16^oC and 30^oC systems, biodegradation reduced TPH levels from approximately 950 ppm to below 350 ppm in 6 days. TPH levels in the 22^oC system dropped from approximately 950 ppm to less than 150 ppm after 4 days, achieving a maximum rate of degradation of 10mg/L/hr. The recovered microbes showed an excellent ability to grow exclusively on vapor-phase PHCs. Mutagenicity and toxicity test results substantiated the TPH biodegradation results. The mutagenic potential of the WSF reached background levels at all temperatures after 6 days. After an initially low WSF EC₅₀ of 2.5%, the toxicity of the WSF was reduced after 4 days to an EC₅₀ of greater than 6.0%. The microbial community showed broad PHC degradative abilities, was able to reduce the toxicity and mutagenicity of the remaining PHCs and degradation products, and shows excellent promise for incorporation into a managed microbial biodegradation system.

Various bioremediation technologies have been accepted as cost-effective methods for destruction of petroleum contaminants (1-5). However, more importantly than just the total loss of petroleum products, the changes in the toxicity and/or mutagenicity of the starting material are important for determining the risk associated with residual contaminants and biodegradation byproducts. Often as bioremediation proceeds, the toxicity of the contaminant will increase initially and then decrease with time (2,6-8). Thus, it is important to monitor the toxicity and mutagenicity of the bioremediation system to determine that the biodegradation of the contaminant is actually reducing the adverse effects of the contaminants still present. We chose to evaluate the toxicity and mutagenicity of the water soluble fraction (WSF) in the test systems, as this fraction is what would be most bioavailable in a bioremediation system (9,10). A toxicity build-up in bioremediation systems leads to inefficiency and poor removal rates due to decreased microbial activity. In addition, should the contaminant of interest be a suitable carbon and energy source, the bacterial populations performing the biodegradation should increase in number over time (2). The use of microbial growth to evaluate biodegradation is termed growth-linked biodegradation (2).

The overall goal of this project was to recover a consortium of microorganisms able to degrade a wide range of PHCs while detoxifying the remaining contaminants and exhibiting excellent growth on exclusively PHCs, and to evaluate the consortium's suitability for future use in a managed bioremediation system such as a biofilter or liquid reactor. Many biodegradation systems are often set-up with little knowledge about the microbial communities which are being utilized (11). In order to better understand what is happening in an operational biodegradation system, an understanding of the microbial community is needed (11). A comprehensive study using liquid systems was chosen, as liquid systems are relatively inexpensive, easy to set-up, and contain parameters which can be monitored and/or manipulated more readily than large bioremediation systems. We chose to study a consortium of microorganisms isolated from a site currently undergoing bioremediation to specifically i) assess the consortium's growth on a complex mixture of PHCs, ii) assess the consortium's ability to degrade a complex mixture of PHCs, and iii) determine the effect this degradation has on the toxicity and mutagenicity of the test system.

Sources of microorganisms and chemicals. Petroleum hydrocarbon-contaminated soil was obtained from a site currently undergoing bioremediation in Vancouver, BC. All chemicals were analytical-grade and obtained from Sigma Chemical Co. (St. Louis, MO). P-9 oil, the surrogate PHC, was obtained from the Institute of Wood Research at Michigan Technological University (MTU). P-9 oil was chosen as the surrogate PHC due to its complex nature (i.e., many different classes of compounds found in the mixture) and the fact it is an environmental contaminant at numerous wood preserving sites.

Media and Microbial Enrichment Conditions. Enrichment cultures were prepared by adding 10 g of soil to a 250-ml Erlenmeyer flask containing 100 ml of Bushnell-Haas broth (BHB; Difco Laboratories, Detroit, MI) (designated as Flask 1). P-9 oil was added to each flask at ~1% (v/v) to assure the PHC did not limit growth. The flasks were then placed on a gyratory shaker at approximately 250 rpm at 23±1^oC. Bacterial growth was monitored daily using a 10µl loopful of culture and phase contrast microscopy until >30 cells per field of view were observable at 1000x magnification. After 14 days, 1 ml from Flask 1 was transferred to a second 250-ml Erlenmeyer flask (designated as Flask 2) containing 100-ml of fresh BHB with ~1% (v/v) of P-9 oil. Flask 2 was incubated as for Flask 1 and bacterial growth was again observed daily as for Flask 1. After 10 days, 1 ml from Flask 2 was transferred to a third 250-ml Erlenmeyer flask (designated as Flask 3) containing 100 ml of fresh BHB with ~1% (v/v) of P-9. Flask 3 was incubated and monitored for microbial growth as for Flask 1 and 2. The microbial inoculum for all experiments was prepared from Flask 3 and maintained in liquid culture thereafter by subsequent weekly transfers. The microbial consortium was also frozen using liquid nitrogen for long-term storage at -70^oC.

Characterization of Consortium Members. Bacterial isolates were recovered using a recently developed filter-pad method (9). After 5 days of incubation at 26^oC, colonies were observed for differences in colony morphology. Colonies with different morphologies were subsequently transferred to Tryptic Soy Agar (TSA) plates for Biolog analysis (Biolog, Inc., Hayward, CA). Biolog GN MicroPlates were used for the identification of the two recovered bacterial isolates. The isolates were grown on TSA+5.0% (v/v) defibrinated sheep blood (Remel Laboratories, Lenexa, KS) for 24 hours at 26^oC prior to Biolog testing. All standard BIOLOG manufacturer-specified procedures were followed for the identification of the isolates.

Liquid system set-up and incubation. A total of forty-six (46) 125-ml EPA-grade volatile organic compound (VOC) bottles were set-up at the beginning of this experiment for all testing (PHC degradation, growth, and toxicity/mutagenicity). For each sample day, three VOC bottles with teflon closures (Fisher Scientific, Chicago, IL) were set-up (one control and two samples). To each VOC bottle, 25 mls of BHB were added along with 25µls of P-9 oil (~950 ppm). The controls received no microbial inoculum. To the test bottles, 1 ml of microbial inoculum was added. The inoculum for the liquid degradation systems was prepared by adding 2 mls of a stock microbial consortium, stored at -70^oC, to a 1L Erlenmeyer flask containing 250 mls of BHB and 2.5 mls of P-9 oil (~950 ppm). The culture was allowed to grow for 7 days and then was centrifuged (Beckman, J2-21 Centrifuge, Minneapolis, MN) at 15,300 x g for 10 minutes. The resulting supernatant was then discarded and the pellet resuspended in 250 mls of BHB. This washing procedure was repeated two times and the final resuspension was made in 25 mls of BHB to yield a tenfold concentration of cells. One ml of this inoculum was then added to each of the non-control bottles (approximately 1x 10¹⁰ CFU/ml).

Analytical methods. For each sampling time, the liquid from an entire bottle was extracted using liquid-liquid extraction techniques in separatory funnels with hexane as the organic solvent. Upon extraction, final extract volume was determined and samples were stored in borosilicate glass autosampler vials (Hewlett-Packard Co., Palo Alto, CA) at 4^oC until analysis. Analysis was performed within 48 hours. Extracts were analyzed by a gas chromatograph (GC) (Hewlett-Packard Model 5890; Palo Alto, CA) with a 30-m macrobore (0.25 mm diameter) bonded 5% diphenyl-dimethyl polysiloxane phase. Temperatures for the analysis were as follows: injection port, 150^oC; FID, 250^oC; and oven initially 60^oC and programmed to reach 280^oC at 20^oC min^-1. Analytical standards were prepared at each analysis time using stock P-9 standards. All P-9 quantities were expressed as milligrams of PHC per milliliter of liquid. For determination of specific compound degradation, extracts from the 22^oC systems were also analyzed by a GC (Hewlett-Packard Model 5890 Series II) with a 30-m macrobore (0.25 mm diameter) bonded 5% diphenyl-dimethyl polysiloxane phase coupled with a mass selective detector (Hewlett Packard Model 5874). Temperatures for the analysis were the same as described above. Analytical standards were prepared at each analysis time using stock P-9 standards.

Microbial enumeration. Flasks set-up in an identical manner to those described above were used for microbial enumeration. There were two microbial enumeration flasks per sampling time. At each sampling time, 1 ml of sample was removed for microbial enumeration and spread-plated onto TSA. The inoculated plates were incubated at 26^oC for 48 hours before counting. Direct counts utilized the same 1 ml extract of cells as the plates counts (each counting method used 0.1 ml from the 1 ml extract). Direct counts were done by staining the cells with methylene blue, placing them on a hemocytometer, and counting cells at 400x magnification.

Toxicity and mutagenicity studies. Flasks set-up as described above were used for the toxicity and mutagenicity studies. There were two toxicity and mutagenicity flasks per sampling time. From each flask, 5 mls were withdrawn and placed in a borosilicate glass vial with teflon seal. Toxicity samples were stored at 4^oC until analyses were performed. Mutagenicity samples were stored at -20^oC until analyzed. Toxicity analysis was performed using the Basic Test protocol (Azure Environmental, Hayward, CA) with a Microtox Model 500 analyzer (13). Mutagenicity analysis was performed using a modification of the microsuspension version of the Salmonella/microsome (or Ames) assay (14). Tester strain TA98, with positive and negative controls and without S9 metabolic activation, was used for all assays. The results were reported as revertants per 5 µls of the WSF. Spontaneous revertant levels with dimethyl sulfoxide as a solvent (5 µls) were 39 revertants per plate and spontaneous revertant levels with BHB (5 µls) were also 39 revertants per plate. The positive control, 2-nitrofluorene, showed revertant levels of 837 (SD±10) at a concentration of 400 ng/plate. Revertant levels were compared to P-9 oil solubilized in BH media for one hour, and incorporated in the assay in the same way as the other samples. P-9 saturated BHB showed a revertant level of 89 revertants per 5 µls of sample.

WSF PHC determination. The remaining 20 mls from the toxicity/mutagenicity VOC bottles were extracted for TPH determinations using liquid-liquid extraction techniques. The extract was then analyzed and quantified using GC-FID as described above.

Characterization of Surrogate PHC. P-9 oil is composed of numerous aromatic and aliphatic compounds including numerous substituted benzene and naphthalene compounds, long-chain aliphatics, and polycyclic aromatic hydrocarbons (PAHs). GC-MS analysis determined that 2-methyl naphthalene is the largest single constituent in P-9 oil (Figure 1).

Microbial Characterization and Identification. Two morphologically distinct gram-negative, rod-shaped bacteria were recovered from the enrichment cultures and subsequent bioremediation studies. Both isolates were oxidase and catalase positive and were identified using the Biolog system. Isolate GK-622 was identified as Pseudomonas fluorescens, with a similarity index match (SIM) of 0.908, and isolate GK-104 was identified as Alcaligenes xylosoxydans, with a SIM of 0.863. The SIM matches of the two isolates were very good, as a SIM of >0.500 is required for a positive identification after 24 hours of incubation, using standard Biolog protocol. The identification procedure was repeated with identical results.

Liquid Degradation Results. All test systems showed decreases in P-9 oil from greater than 950 ppm to below 350 ppm in 6 days. The 22^oC culture showed the most biodegradation (to below 150 ppm in 4 days) and the fastest rate of degradation at greater than 10mg/L/hr for the first 96 hours (Figure 2); after the initial 96 hours, only a small amount (~50 ppm) of degradation was evident in the test system. The 22^oC liquid system showed greater than 90% removal of the complex mixture over the 12 days of the study. The 16^oC and 30^oC test systems had rates of degradation of 3.67mg/L/hr and 5.7mg/L/hr, respectively (Figures 3 and 4) over the first 6 days of the study. Both the 16^oC and 30^oC systems showed overall PHC removal of greater than 65% over the course of the study.

While PHC levels in the test systems were greater than 100 ppm, the PHCs present in the WSF of all systems, as determined by the liquid-liquid extraction of the water phase, were between 26 and 59 ppm. The rate of 10mg/L/hr in the 22^oC system led to a relatively even destruction of compounds in the complex mixture and not just the degradation of selective classes or groups of compounds. GS-MS data show the lower molecular weight aromatic compounds (e.g., substituted benzene and naphthalene compounds) are degraded most rapidly in the first 48 hours and the C₁₀-C₂₀ compounds (alkanes) are degraded more slowly from 0-48 hours (Figure 5). Five of the major components found in the complex mixture are listed in Table 1. As can be seen by each compound's percent of the total PHCs present, the smaller substituted benzene and naphthalene components are present in all controls and in the day 0 sample. For example, if the total PHCs were 100 mg/L and naphthalene composed 3.2% of that total (Table 1) then there was approximately 3.2 mg/L naphthalene present in the complex PHC mixture. However, after 48 hours these compounds are either not present or are a much smaller percentage of the total PHCs present. Conversely, the alkanes increased in their share of the total PHCs present. After 48 hours, levels of the individual PHCs are relatively low and further trends are difficult to distinguish due to the masking of the compounds in the complex baseline.

Microbial Growth. There was an inverse relationship between the degradation of PHCs from all the test systems and microbial growth. The initial concentration of microorganisms increased in all systems from approximately 1.6 x 10⁶ to greater than 5.0 x 10⁸ cells ml^-1 in less than 6 days (Figures 2-4) and to greater than 1.0 x 10⁹ cells ml^-1 in the 22^oC test system (as determined by both enumeration methods). Growth rate constants for the 22^oC system were calculated as k=0.10 hr^-1 for the viable counts and k=0.11 hr^-1 for the direct counts. The growth rate constants for the 30^oC system using direct and plates counts were k=0.058 hr^-1 and k=0.061 hr^-1, respectively. The growth rate constants for the 16^oC system using direct and plates counts were k=0.056 hr^-1 and k=0.058 hr^-1, respectively. The microorganisms were also grown on solid media using VOCs as the sole source of carbon and energy (12). Both isolates, as well as a mixture of the two organisms, showed excellent growth on these plates. Control plates with no VOCs were compared to the plates with VOCs to account for any growth which might be attributed to the agar or impurities in the media. In all cases, the inoculated plates with VOCs present showed confluent growth, indicating the recovered microorganisms were indeed able to utilize not only liquid phase PHCs, but also VOCs as their sole source of carbon and energy (12).

Toxicity and Mutagenicity. Toxicity and mutagenicity results corroborated both the PHC removal and microbial growth data of the test systems by confirming that the PHC degradation reduced the toxicity and mutagenicity within the test system. The EC₅₀s from the Microtox system decreased (indicating an increase in toxicity) in all systems in the second day of the study (Figure 6), followed by increases in EC₅₀s (indicating a decrease in toxicity) over the last days of the study. At all temperatures, the toxicity decreased in the systems from an initial EC₅₀ of 2.5% of the WSF to EC₅₀s of greater than about 6.0% of the WSF. Salmonella/microsome testing (tester strain TA98 without S9 mix) revealed a similar trend in that the test systems (Figure 7). Levels of revertants per 5 µls of the WSF in the test systems were initially at the levels of BHB saturated with P-9 oil. As biodegradation proceeded, revertants per 5 µls of WSF decreased to approximately background spontaneous revertant levels by about 6 days in all test systems.

It was not unexpected to find gram-negative microbes, especially the Pseudomonas fluorescens, in the consortium of PHC degraders, as Pseudomonads are the most commonly found PHC-degrading genera in contaminated soil (15). Alcaligenes spp. have also been reported to be involved in various PHC degradation systems (4,5,15). However, often times they are only identified to genus (escpecially in the case of Alcaligenes xylosoxydans) and thus further comparisons between degradative ability and specific species is difficult. Given the fact both the direct and plate count enumeration methods yielded similar results, the plate count technique is presumed to have recovered the majority of the microbes present in the systems. In addition, it can also be assumed the majority of the microorganisms present were viable. If there were many nonviable microbes present, the direct counts levels would be expected to be greater than the plate count levels (which enumerate only viable cells able to grow under the selected conditions).

The growth rate constants and the degradation rates show an inverse relationship at all temperatures. The degradation of the PHCs and microbial growth were most rapid in the 22^oC system. There was also a direct correlation between the slowing of microbial growth and the depletion of PHCs from the test system after approximately day 6. As PHC levels became low, bacterial growth also slowed and entered stationary phase. This comparison shows excellent correlation between two methods of measuring biodegradation (growth-linked measurements and PHC loss measurements) in the tests systems.

While the percentage of the PHCs removed was greater than 90%, there was approximately a TPH level of 100 ppm left in the most efficient system (i.e., 22^oC). Of this, only about 50 ppm was soluble, or bioavailable, in the WSF. This last fraction of the complex mixture likely represents low-solubility compounds (e.g., polyaromatic hydrocarbons), recalcitrant breakdown products, and very slowly biodegraded compounds. Given the complexity of the mixture, it was not apparent what specific compounds were remaining, as all compounds were masked at the baseline of the chromatograms. Also, the soluble amount of PHCs was less that 40 ppm in six days, at all temperatures. It was also noteworthy that the consortium was effective in degrading numerous classes of compounds and not just certain classes of compounds (e.g., alkanes or aromatics).

As has been reported previously, PHC biodegradation systems generally show a small initial increase in toxicity which is then followed by a steady decrease in toxicity (7) . While others have observed these toxicity results of soil extracts from soil systems undergoing bioremediation, we are reporting the same effect in simple liquid bioremediation systems. During other studies a steady decrease in mutagenicity has been observed as biodegradation proceeds (7,10). Our liquid systems showed similar results, with all systems reaching approximately the spontaneous revertant levels after about 6 days. In all, the mutagenicity and toxicity data, PHC degradation data, and microbial growth data are all in excellent agreement in these systems. As PHC degradation proceeded and microbial growth continued, a decrease in both the toxicity and mutagenicity of the contaminants in the test system was observed.

Several other researchers have used similar studies to evaluate biodegradation potential for soil systems (7,8,10). However, none of these studies addressed all of the parameters discussed in this paper. Other studies generally do not consider microbial growth as a test parameter, mutagenicity testing coupled with Microtox testing, and two-phase liquid degradation systems.

This study illustrates an inexpensive model biodegradation study which could be undertaken prior to the inclusion of a microbial consortium in a managed microbial system such as a liquid bioreactor or biofiltration unit. This study provides essential information about the consortium's degradative abilities, toxicity/mutagenicity of biodegradation end-products, and microbial characteristics. These laboratory results may be most applicable to a managed microbial system in which the conditions for contaminant degradation and microbial growth can be controlled for optimum removal. In these types of managed systems, studies of the type described in this paper may be useful in determining the starting conditions or parameters for a biodegradation system.

This work was funded in part by DOE/NCASI contract No. DE-FC07-96D13440. The authors would also like to thank Dr. Gary D. McGinnis and the Institute of Wood Research at MTU for the use of analytical equipment which was invaluable. Thanks to Jay Leversee for his assistance in the microbial enumeration portion of this project. Also, special thanks to Wayne P. St. John for advice on analytical methodologies.

1. Bossert, I. and R. Bartha. (1986) Structure-Biodegradability Relationships of Polycyclic Aromatic Hydrocarbons in Soil. Bull. Env. Cont. Tox.. 37:490-495.
2. Alexander, M. (1994) Biodegradation and Bioremediation. Academic Press. New York, NY.
3. Flathman, P.E., D.E. Jerger, and J.H. Exner (1994) Bioremediation: Field Experiences. CRC Press, Boca Raton, FL.
4. Riser-Roberts, E. (1992) Bioremediation of Petroleum Hydrocarbon Contaminated Sites. CRC Press, Boca Raton, FL.
5. Baker, K., and D. Herson (1994) Bioremediation. CRC Press, Boca Raton, FL.
6. Song, H., X. Wang, and R. Bartha (1990) Bioremediation Potential of Terrestrial Fuel Spills. Applied Env. Micro.. 56:652-656.
7. Wang, X, X. Yu, and R. Bartha (1990) Effect of Bioremediation on Polycyclic Aromatic Hydrocarbon Residues in Soil. Environ. Sci. Technol.. 24:1086-1089.
8. Wang, X. and R. Bartha (1990) Effects of Bioremediation on Residues, Activity and Toxicity in Soil Contaminated by Fuel Spills. Soil. Biol. Biochem. 22:501-505.
9. Diks, R.M.M., and S.P.P. Ottengraf (1991) A Biological Treatment System for the Purification of Waste Gases Containing Xenobiotic Compounds, p. 452-463. In R.E. Hinchee and R.F. Olfenbuttel (eds.), On-Site Bioreclamation: Processess for Xenobiotic and Hydrocarbon Treatment. Butterworth-Heinemann, Stoneham, MA.
10. Dasappa, S.M. and R.C. Loehr (1991) Toxicity Reduction in Contaminated Soil Bioremediation Processes. Water Research. 25:1121-1130.
11. Corsi, R.L., and L. Seed (1995) Biofiltration of BTEX: Media, Substrate, and Loading Effects. Environmental Progress. 14:151-158.
12. Kleinheinz, G.T. and S.T. Bagley (1997) Simple Method for the Recovery and Cultivation of Petroleum Hydrocarbon Utilizing Microorganisms. Journal Microbiological Methods (in Press).
13. American Public Health Association. (1996) Standard Methods for the Examination of Water and Wastewater. 19th Ed. Method 8050. American Public Health Association, Washington, D.C.
14. Bagley, S.T., K. J. Baumgard, L.D. Gratz, J. H. Johnson, and D.G. Leddy (1996) Characterization of Fuel and Aftertreatment Device Effects on Diesel Emissions. Research Report No. 76. Health Effects Institute. Cambridge, MA.
15. Atlas,R.M. and C.E. Cerniglia (1995) Bioremediation of Petroleum Pollutants, Diversity and Environmental Aspects of Hydrocarbon Biodegradation. BioScience. 45:332 338.