INTRODUCTION

Top Abstract Introduction Materials & Methods Results Discussion References

In the United States, toluene is ranked 27th among the top 50 chemical products by production volume, with 931 million gallons (3.5 × 10⁹ liters) of toluene manufactured in 1994 (22). Industry uses toluene in refining gasoline; chemical manufacturing; the manufacture of paints, lacquers, and adhesives; and some printing and leather-tanning processes. Toluene is usually disposed of as a used solvent. Toluene is listed as a priority pollutant (46), because contamination of drinking water can pose a potential health hazard. In a 1988 study, toluene was detected in groundwater, surface water, or soil at 29% of the hazardous waste sites surveyed; the average amounts detected were 0.2 µM in groundwater, 0.1 µM in surface water, and 77 µg/kg in soil (46).

Since toluene is ubiquitous in the environment, it is not surprising that microorganisms capable of degrading toluene have been isolated from a variety of environments, including polluted topsoils and oil tanker ballast waters (4, 5, 10, 29, 38, 48, 50). However, little is known about the role these isolates may play in the conversion of toluene in the environment. Therefore, studies that address the response of stream bacteria to effects of anthropogenic chemical impact are important in enhancing our understanding of the ecology of contaminant degradation.

This study describes the isolation and characterization of two closely related toluene-degrading strains of a novel Mycobacterium species from rock surface biofilms in a toluene-contaminated reach of a small freshwater stream. Ecological data are also presented to assess the role of these Mycobacterium strains in degrading toluene in the rock surface biofilms and to determine the impact of toluene contamination on the stream's microbial community. Earlier studies showed biodegradation to be the most significant sink for toluene in this stream (20), and microorganisms attached to stream sediments and rock surfaces accounted for most of the biodegradative activity (7).

	MATERIALS AND METHODS

Top Abstract Introduction Materials & Methods Results Discussion References

Study site. The East Drainage Ditch is a small stream in an industrial area of Wilmington and Woburn, Mass. (7, 20), and forms part of the Aberjona Watershed, a 90-km² area 20 km north of Boston. Toluene in the stream arises from subsurface contamination below an 80-m-long culvert (11) located 1,600 m upstream from the confluence of the East Drainage Ditch with Halls Brook. Toluene levels typically range from 0.6 to 4.2 µM in the streamwater near the source (20). Sampling stations are upstream 50 m (U50) and 100 m (U100) and downstream 5 m (D5) and 50 m (D50).

Isolation of bacterial strains. Rocks were collected in July 1992 from the streambed at station D5. Rock biomass was scraped with sterile spatulas, serially diluted, and plated onto a mineral salts (MS) agar medium (35) supplemented with trace elements (49). The plates were incubated in a desiccator with a beaker of water (15 ml) previously equilibrated with toluene. The water was replaced every 2 to 3 days. Toluene was supplied by diffusion from the beaker in the desiccator that resulted in theoretical concentrations of approximately 110 µM in the agar based on Henry's law (36), although the actual concentrations might be affected by diffusion kinetics and hydrophobic interactions between toluene and organic constituents of the agar. Plates were monitored over 7 weeks, during which colonies were picked and restreaked on fresh plates. Several colonies grew in the presence of toluene, and four strains, numbered sequentially starting from T101, were selected for further characterization. Strains T103 and T104 possessed traits typical of mycobacteria and are described in this paper.

Phenotypic characterization. Tests for Gram stain, oxidase activity, catalase activity, carbon source utilization, nitrate reduction, and acid fastness (Ziehl-Neelsen method) were performed as previously described (27, 37). Tests for growth on aromatic compounds were conducted in the same manner as with toluene. Substrates were supplied by diffusion from a reservoir in the desiccator resulting in theoretical concentrations in the agar of 10 mg of benzene, o-xylene, m-xylene, p-xylene, phenol, or chlorobenzene per liter. Plates were monitored for 4 weeks, and visible colonies were picked and restreaked on fresh plates that were further incubated in either the presence or absence of the relevant substrate.

Toluene biodegradation kinetics. Rates of toluene degradation were determined by a headspace gas chromatography method (7). Cells were grown in 1,100-ml Teflon-stoppered glass bottles with 600 ml of MS medium initially containing 110 µM toluene. Toluene levels were monitored daily and replenished when depleted. After every five toluene feedings, the headspace was flushed with air to replenish the oxygen. Cells were harvested by centrifugation during the exponential phase and resuspended in 100 ml of fresh medium. Kinetic experiments were performed with 60-ml Teflon-stoppered serum bottles (The West Co., Phoenixville, Pa.) with 19 ml of medium and 0.5 ml of concentrated cell suspension (230 µg of protein/ml for strain T103 and 185 µg of protein/ml for strain T104). Toluene was injected at approximately 1.1, 2.2, 5.4, and 10.9 µM (aqueous concentration). Control experiments were performed with autoclaved cell suspensions. Bottles were shaken in the dark at 20°C on a rotary shaker at 150 rpm and assayed hourly for toluene. Initial rates of toluene disappearance were determined to establish Michaelis-Menten kinetic parameters. Protein was measured with the bicinchoninic acid protein assay (Pierce, Rockford, Ill.).

Fatty acid analyses. Cells were grown on Middlebrook 7H10 agar (9) for 7 days at 28°C, harvested, and saponified to prepare fatty acid methyl esters (33). The analysis was performed at Microbial ID, Inc. (MIDI, Newark, Del.), by using the MIDI Microbial Identification System software for identification of fatty acids. Strains T103 and T104, together with the M. komossense type strain (ATCC 33013), were analyzed in duplicate and compared with profiles from the Microbial Identification System (MIS) library (34).

16S rDNA sequencing. Strains T103 and T104 were cultured in nutrient broth (Difco Laboratories, Detroit, Mich.) in Teflon-stoppered serum bottles by shaking at 20°C on a rotary shaker at 150 rpm. Cells were harvested by centrifugation, and genomic DNA was extracted with a miniprep (1). The nearly full-length 16S rRNA gene was amplified from genomic DNA by PCR with forward primer Eubac27F and reverse primer Universal 1492R (23). All reactions were run in triplicate under PCR conditions as described elsewhere (8), and PCR products were purified with the Wizard PCR Prep (Promega Corp., Madison, Wis.) for automated dye-dideoxy terminator sequencing at the Michigan State University Sequencing Facility with a 373A DNA sequencing system (Applied Biosystems, Foster City, Calif.). Sequences for oligonucleotides complementary to the conserved regions of the eubacterial 16S rRNA gene were kindly provided by Debra J. Lonergan (United States Geological Survey, Reston, Va.); these oligonucleotides were chosen to prime the sequencing reactions. Sequencing reaction mixes consisted of 12 pmol of sequencing primer and 50 to 250 ng of PCR template in a total volume of 20 µl of sterile H₂O. The sequences of approximately 1,425 nucleotide bases, corresponding to the Escherichia coli 16S rDNA sequence from nucleotides 55 to 1501, were obtained in both directions for the two strains.

Phylogenetic analyses. The 16S rDNA secondary structures of strains T103 and T104 were constructed manually with templates published in the Ribosomal Database Project (RDP) (25) to aid in the identification of homologous sequence positions. Sequence alignments were performed manually in the Genetic Data Environment (39). All reference sequences and the basic alignment were obtained from the RDP. Only homologous sites at which the 16S rDNA sequences of strains T103 and T104 could be aligned unambiguously with the reference sequences were included in a final data set of 1,165 nucleotides for further analyses. Distance, parsimony, and maximum likelihood analyses were performed with PHYLIP 3.5 (13), PAUP 3.1 (42), and fastDNAml (12, 28), respectively, as described elsewhere (31).

Bacterial counts. Rock samples were collected from stations U100, U50, D5, and D50 along the East Drainage Ditch on 7 September 1993. Biomass was scraped with sterile spatulas and suspended in 10 ml of MS medium, and the resulting cell suspension was successively diluted to obtain dilutions of 10² to 10⁷ g of biomass/ml. Petri dishes containing 1% PTYG agar (3) were inoculated in duplicate with 100 µl of each dilution and incubated at 20°C for 3 weeks. Heterotrophic colonies were counted on the plates every 3 days, until no new colonies were observed. Since counts were based on the highest-dilution plates, colony overgrowth on the plates was not a problem.

Toluene levels in the stream. Duplicate water samples were collected at stations U100, U50, D5, and D50 on 19 September 1993 with 40-ml vials provided with hollow screw caps and Teflon-coated silicone septa. Mercuric chloride was added to the samples to a final concentration of 15 mg/liter. Toluene was measured as described previously (7).

Nucleotide sequence accession number. The sequences for strains T103 and T104 have been deposited in the GenBank database under accession no. U62889 and U62890. The GenBank accession numbers of the other sequences used in the analyses are as follows: Mycobacterium aurum, X55595; Mycobacterium chelonae subsp. abscessus L948, M29559; Mycobacterium chitae, X55603; Mycobacterium chlorophenolicus PCP-I, X79094; Mycobacterium diernhoferi SN 1418, X55593; Mycobacterium fortuitum subsp. fortuitum, X52933; Mycobacterium gilvum, X55599; Mycobacterium komossense Ko2, X55591; Mycobacterium neoaurum, M29564; Mycobacterium sphagni Sph29, X55590; Mycobacterium thermoresistible, X55602; Mycobacterium vaccae, X55601; Mycobacterium asiaticum N61H, X55604; Mycobacterium avium serovar 1, M29573; Mycobacterium haemophilum, L24800; Mycobacterium tuberculosis H37/Rv, X52917; Mycobacterium xenopi, X52929; Corynebacterium xerosis, M59058; Gordona terrae, X79286; Nocardia otitidiscaviarum, M59056; and Rhodococcus equi, M29574. Primary literature references for these sequences are available from the RDP (25).

	RESULTS

Top Abstract Introduction Materials & Methods Results Discussion References

Enrichment and isolation. Two mycobacterial toluene-degrading strains, T103 and T104, were independently isolated from two yellow colonies, 1 to 2 mm in diameter, that appeared between 3 and 8 days on toluene-incubated MS plates inoculated with 10⁶ g of biomass, indicating that these strains were present in the rock surface biomass at a density of 10⁶ cells/g of biomass. Biomass scrapings averaged 0.018 g (fresh weight)/cm² of rock surface. Both strains grew slowly (relative to other nonmycobacterial toluene-degrading strains similarly isolated) on solid media when incubated with toluene. Mycobacterial colonies were not detected on plates inoculated with higher dilutions of biomass.

Morphological and phenotypic characteristics. Strains T103 and T104 had morphologically similar rod-shaped cells when grown on solid media. Cells were nonmotile, and flagella were not observed under scanning electron microscopy (data not shown). The strains were aerobic, gram positive, and acid-alcohol fast (Table 1), which is characteristic of the mycobacteria (47). Strain T104 differed physiologically from strain T103 in that T104 could grow on xylenes (Table 1).

Toluene biodegradation kinetics. Strains T103 and T104 had maximal toluene consumption rates (V_max) of 1.0 ± 0.1 and 6.0 ± 1.3 µmol of toluene/mg of protein per h, respectively; their half-saturation constants (K_s) were 0.6 ± 0.4 and 3.8 ± 1.9 µM, respectively.

Fatty acid analyses. Strains T103 and T104 were made up of mainly straight-chain saturated and monounsaturated fatty acids, as well as substantial amounts of tuberculostearic acid (Table 2) that are typical of mycobacteria (16). Fatty acid analyses (34) indicated that T103 and T104 were different strains of a novel Mycobacterium sp. that was most closely related to M. aurum (similarity indices of 0.31 and 0.24 with strains T103 and T104, respectively).

16S rDNA sequence analyses. Strains T103 and T104 possessed identical 16S rDNA sequences. The secondary structures of the sequences of T103 and T104 and other fast-growing mycobacteria were identical, and they contained the shortened stem structure bounded by positions 455 to 477 (E. coli numbering) that typically distinguishes the fast-growing from the slow-growing mycobacteria (41).

View larger version (31K): [in this window] [in a new window]   FIG. 1.   Unrooted evolutionary distance tree based on the 16S rDNA sequences of strains T103 and T104, representative members of the genus Mycobacterium, and other high-G+C gram-positive bacteria. Bootstrap values greater than 50% are shown at the nodes. Bar = 0.01 nucleotide difference per sequence position.

Bacterial counts and toluene levels. No toluene was detected at station U100, although very low levels of toluene (0.04 µM) were found at station U50. Downstream stations D5 and D50 had high toluene concentrations of 2.6 and 1.7 µM, respectively, in September 1993 (Fig. 2). The total heterotrophic bacterial plate counts at these stations ranged from 2 × 10⁸ to 7.7 × 10⁸ cells/g of biomass. Toluene-degrading bacteria were not detected for station U100, where no toluene was present. The counts of toluene-degrading bacteria for the other three stations increased with increasing toluene concentration and comprised 0.03, 1.09, and 0.35% of the total heterotrophic plate counts at stations U50, D5, and D50, respectively.

View larger version (27K): [in this window] [in a new window]   FIG. 2.   Bacterial counts (sampled on 7 September 1993) and toluene levels (sampled on 19 September 1993) along the East Drainage Ditch. Error bars represent standard deviations of duplicate samples. , viable count of heterotrophic bacteria on 1% PTYG plates incubated at 20°C for up to 4 weeks in the dark. , plate count of toluene-degrading bacteria on minimal salts agar with 110 µM toluene. Incubations were performed at 20°C for up to 4 weeks in the dark. , toluene concentration.

	DISCUSSION

Top Abstract Introduction Materials & Methods Results Discussion References

Taxonomy and phylogeny. Although strains T103 and T104 both degrade toluene and have major characteristics of the genus Mycobacterium, they exhibit some differences in physiology. Unlike strain T103, strain T104 is able to grow on the xylenes. Also, their fatty acid profiles are sufficiently different for them to be considered different strains of a novel species of fast-growing mycobacteria. Compared to the slow-growing mycobacteria, many of which are human and animal pathogens, most of the fast-growing mycobacteria are common saprophytes in natural habitats (18). They have been isolated from a diverse array of habitats, are able to survive and multiply under a wide range of environmental conditions, and can biotransform a variety of xenobiotic compounds and pollutants, including polycyclic aromatic hydrocarbons (15) and groundwater-pollutant mixtures (4). There is a long history of isolation of mycobacteria that have the capacity to degrade the aromatic fraction of the oil in contaminated soils (44, 45).

Biodegradation kinetics. Toluene contamination of the East Drainage Ditch appears to selectively enrich for toluene-degrading bacteria within the epilithic microbial community. Maximal velocities are somewhat higher than those reported for other aerobic toluene-degrading bacteria; strain T104 also has a larger K_s (3.8 µM) than those reported for other toluene-degrading bacterial strains. Pseudomonas sp. strain T2 had a maximal velocity of 0.30 µmol of toluene/mg of protein per h (assuming that 50% of a typical cell's dry weight is protein) and a K_s of 0.48 µM (32). Corresponding values for a terrestrial strain of Pseudomonas putida, PpF1, were 0.43 µmol of toluene/mg of protein per h and 0.68 µM, respectively (32). That K_s values for strains T103 and T104 lie within the range of toluene concentrations observed in the stream suggests that these strains are adapted to the ambient level of toluene contamination in the stream. Noncarbon nutrients are unlikely to be limiting in this case, because these are present in the stream at high levels (43).

Microbial ecology. Toluene levels in the stream do not appear to have a significant effect on total heterotrophic bacterial counts. On the other hand, counts of toluene-degrading bacteria in samples from the contaminated stations are about an order of magnitude higher than those from pristine stations, which suggests that toluene in the contaminated reaches of the stream has caused the epilithic bacterial communities to adapt by selectively enriching for toluene-degrading bacteria. Other studies of freshwater environments (30, 40) also reported viable heterotrophic counts to be unaffected by the presence of a xenobiotic contaminant, while the numbers of organisms capable of degrading the contaminant of interest increased by between 2 and 3 orders of magnitude. The increased counts in contaminated reaches versus uncontaminated reaches of the East Drainage Ditch are not as high as observed in these other studies, possibly because of greatly reduced microbial activity at lower temperatures (30). The East Drainage Ditch study was conducted in the fall when stream temperatures averaged 11°C. In the other studies, temperatures averaged 18°C (40) or ranged from 19 to 30°C (30).