Polycyclic aromatic hydrocarbon-degrading Mycobacterium isolates: their association with plant roots
- First Online:
- Cite this article as:
- Child, R., Miller, C.D., Liang, Y. et al. Appl Microbiol Biotechnol (2007) 75: 655. doi:10.1007/s00253-007-0840-0
- 213 Downloads
Five environmental mycobacterium isolates that degrade polycyclic aromatic hydrocarbons (PAHs) were associated with barley root surfaces after growth of the seedlings from inoculated seed. Mycobacterium cells were detected along the total root length for four of these isolates. These PAH-degrading mycobacterium strains had hydrophilic cell surfaces, whereas one strain, MCS, that was hydrophobic had reduced association along the root length with no cells being detected from the root tips. The root-tip-competent strain, KMS, was competitive for its root association in the presence of the root-colonizing pseudomonad, Pseudomonas putida KT2440. All mycobacterium strains utilized simple sugars (fructose, glucose) and the trisaccharide 6-kestose, present in barley root washes, for planktonic growth, but they differed in their potential for biofilm formation under in vitro conditions. Mineralization of pyrene by the KMS strain occurred when the components in the barley root wash were amended with labeled pyrene suggesting to us that mineralization could occur in plant rhizospheres containing such mycobacterium strains.
KeywordsPolycyclic aromatic hydrocarbonsRoot colonizationMycobacteriumRemediation
Polycyclic aromatic hydrocarbons (PAHs) consist of two or more fused aromatic rings and are formed during the incomplete combustion of coal, oil, gas, wood, and other organic substances (Harvey 1991). They are produced throughout the world as industrial by-products of fossil fuel combustion, asphalt production, wood preservation, and coal-processing. PAHs are persistent because of recalcitrance to microbial degradation because of the stabilization imparted by multiple rings, high hydrophobicity, and ability to sorb strongly to soil particles. The partial breakdown products of PAH degradation have toxic, mutagenic, and carcinogenic properties (Menzie et al. 1992; Patnaik 1992), and thus remediation is desired.
Plants enhance remediation of PAH-contaminated soils, and several mechanisms are proposed to account for these observations (Chen et al. 2003; Kim et al. 2004; Liste and Alexander 2000; Nedunuri et al. 2000). As the root systems acquire water from the soil, they may promote translocation of dissolved pollutants into the rhizosphere (Ferro et al. 1994). Surfactant-active compounds in the root exudates may increase solubility of PAHs and thereby enhance their bioavailability (Kim et al. 2004; Read et al. 2003). Plant roots may improve oxygen diffusion by providing channels for air flow (Tsao 2003). Oxygen availability is important because PAH degradation by certain soil bacteria, including the mycobacteria, requires transformation through dioxygenases (Khan et al. 2001).
Although plant metabolisms may be important in pollutant transformation (Cunningham and Ow 1996), the exudates from plant roots may maintain active populations of bacteria with remediation activities in the root zone (Cerniglia 1993). Plants release up to 35% of assimilated carbon in the form of root exudates (Lynch and Whipps 1990), and these exuded material support microbial growth. Paul and Clark (1989) showed that microbial populations and activities associated with roots are as much as 100 times that of the bulk soil. Anokhina et al. (2004) demonstrated that the efficiency of PAH degradation is increased significantly with plant roots colonized by Pseudomonas with PAH-degrading capabilities.
Several soil mycobacterium isolates mineralize PAHs (Cheung and Kinkle 2001; Dean-Ross and Cerniglia 1996; Khan et al. 2001; Leys et al. 2005; Miller et al. 2004), including the more complex PAHs, such as benzo[A] pyrene (Derz et al. 2004; Miller et al. 2004; Moody et al. 2004). PAH degradation occurs at low nutrient levels and is little affected by imbalanced C/N/P soil ratios (Leys et al. 2005). Consequently, these organisms are attractive for use in bioremediation systems.
We report in this paper studies designed to initiate our understanding of interactions between cells of the PAH-degrading mycobacterium and plant roots. Such associations are little reported in the literature. Mycobacterium presence on roots of another cereal, wheat, was detected by polymerase chain reaction (PCR) techniques (Conn and Franco 2004). PAH-degrading isolates including mycobacterium were detected in the rhizosphere of marsh plants (Daane et al. 2001). Additionally, the pathogenic mycobacterium M. ulcerans has been found colonizing aquatic roots (Marsollier et al. 2004).
Five PAH-degrading mycobacterium strains were used to explore their relationship with plant roots. Three of these strains came from soils from a wood-preservative site (Miller et al. 2004) and two others from PAH-contaminated sludges within the USA (Dean-Ross and Cerniglia 1996; Heitkamp et al. 1988). We determined the extent to which these bacteria when inoculated onto a germinating seed became associated with the developing root surface. At the root’s surface, as well as in the soil matrix, there is competition between microbes for survival. Consequently, we examined competitiveness to associate with the developing root between the mycobacterium and Pseudomonas putida KT2440, an isolate from contaminated soils that is a documented root colonizer (Molina et al. 2005). We observed whether materials washed from plant roots supported mycobacterium planktonic and biofilm growth in a model well plate system. Biofilm formation is important in root colonization by other root-associated bacteria (Bais et al. 2001; Bianciotto et al. 2001; Morris and Monier 2003) and is a known trait of clinical mycobacterial species (Carter et al. 2003; Hall-Stoodley and Lappin-Scott 1998; Rose et al. 2004). Marsollier et al. (2004) showed that aquatic plants stimulate growth and biofilm formation by a pathogenic M. ulcerans isolate. Additionally, a PAH-degrading mycobacterium forms a biofilm in the presence of anthracene as a specific response to increase bioavailability of the PAH (Wick et al. 2002). To understand the nutrition afforded to the mycobacterium in the rhizosphere, we compared the composition of the low-molecular-weight carbohydrates in materials washed from plant roots before and after mycobacterium growth. To determine whether mineralization of PAH could occur in the rhizosphere, we examined whether pyrene mineralization of C14-labeled pyrene would occur in the presence of the root exudate components. Barley was chosen as the model plant for these studies because it grows well on PAH-contaminated soils even under conditions of high salinity (Anderson et al., unpublished).
Materials and methods
Strain origin and maintenance
P. putida KT2440 (Molina et al. 2005), mycobacterium strains from the wood preservative site, JLS, KMS and MCS (Miller et al. 2004), M. flavescens (Dean-Ross and Cerniglia 1996) and M vanbaalenii isolate PYR-1 (Heitkamp et al. 1988), obtained from Dr Carl Cerniglia, were stored at −70°C in 15% glycerol. M. smegmatis was obtained from Dr. Jon Takemoto of Utah State University. Cultures were grown on a Luria–Bertani (LB) medium or on an amended Middlebrook 7H9 medium (Difco, Franklin Lakes, NJ), containing the ADC commercial enrichment (Difco) and 0.25% Tween 80, an addition that prevented cell clumping in liquid conditions.
Barley seed inoculation and growth
Seeds were processed to remove surface microbes by immersion in 30% hydrogen peroxide for 5 min, followed by three washes with sterile water each for 3 min. These seeds were suspended in sterile water and heat treated twice each for 30 min at 50°C to remove endophytes and transferred to LB agar for 48 h at 22°C to ensure microbial sterility upon germination. These barley seeds were inoculated by immersion in a suspension of 108 cells/ml of Mycobacterium strain KMS for 30 s and planted into sterile moist vermiculite. Cells for inoculum were grown for 6 days in amended Middlebrook 7H9 to late log-phase growth, washed twice in sterile water, and suspended in sterile water. To determine the number of mycobacterium cells adhering to the seeds, a single inoculated seed was submersed in 1 ml sterile water and vortexed for 30 s. Serial dilutions of the water fractions showed 2 × 106 cells adhering per seed.
The inoculated or noninoculated control seeds were transferred within a sterile transfer hood into sterilized vermiculite for seedling growth. This growth matrix was prepared by adding 125 ml sterile water to 325 ml vermiculite in Magenta boxes and sterilizing at 121°C for 40 min. The boxes were stored at room temperature for 24 h to allow fungal and bacterial spore germination before sterilization at 121°C for 40 min. No microbial colonies were detected from plating water washes of the vermiculite from the boxes onto LB medium after 5 days of incubation at 26°C. Three seeds were planted per container. The plants were grown at 26°C. The roots of 7-day-old seedlings were removed and gently blotted onto the LB plate medium. The plates were incubated and observed visually to determine the location of the bacterial colonies with respect to seedling root length. The identity of the colonies was confirmed by a PCR assay to detect the presence of a nidB gene, generating a 510-bp product, in their genome. Based on the available gene sequences, as described by Hall et al. (2005) and updated by the Basic Local Alignment Search Tool searches at the time of submission of this paper in December 2006, only PAH-degrading mycobacterium strains possess genes that would generate PCR products from nidB. Additionally, the colonies growing on the medium had the colony morphology, yellow color, and slow growth time characteristic of the mycobacterium strains (Miller et al. 2006).
Roots that were not used in direct plating were harvested and dissected into 2-cm sections and vortexed in 1 ml sterile water for 30 s. Serial dilutions were made from the water onto the LB plate medium, and the number of mycobacterium colonies was ascertained for the different root sections. Colonies were examined using PCR with primers for the nidB gene (Hall et al. 2005) to ensure that the bacteria were mycobacteria. Serial dilutions from root sections of uninoculated sterile control seedlings were performed, and no microbial contamination was observed.
Contact angle measurement
Contact angles were measured as described by Anderson et al. (2005) using a goniometer to determine whether mycobacterium cell surfaces were hydrophilic or hydrophobic. Ethylene glycol was applied as a 1-μl drop to the surfaces of confluent cultures of mycobacterium cells grown for 12 days on LB medium plates. Data are from three independent studies, each with five measurements of contact angle for each isolate.
Competition study for root colonization
To test the competitiveness of the association of soil mycobacteria on roots with other bacteria, sterilized barley seeds were inoculated with a kanamycin (Kn)-resistant strain of isolate KMS and P. putida KT2440. To generate the Kn-resistant KMS strain, cells were grown in liquid cultures in the amended Middlebrook medium. Cells were centrifuged, suspended at a high concentration (optical density OD600 nm > 3) in 10% glycerol, and 100-μl samples were plated onto a LB plate medium containing 25 μg/ml Kn. Colonies that grew were picked and purified by single colony isolation onto the LB–Kn (25 μg/ml) medium. One colony was selected for use that had a generation time equivalent to that of the parental strain on the LB and amended Middlebrook medium that has retained its Kn-resistance during studies for more than 1 year.
P. putida KT2440 was chosen as a coinoculant because it also colonizes plant roots (Espinosa-Urgel et al. 2000; Molina et al. 2005) and is highly sensitive to Kn to facilitate its differentiation from the Kn-resistant mycobacterium used in the study. Barley seeds were sterilized as outlined above and inoculated with suspensions each at 1 × 108 cells/ml of KT2440, mycobacterium strain KMS-KnR, or a mix of these strains. Bacteria used in inoculation were harvested from the LB cultures at an early stationary phase, washed twice, and resuspended in sterile deionized water. Seeds lacking bacterial inoculation were planted as controls.
Seedlings were harvested after 7 days growth, and serial dilutions from 2-cm root sections were made as described above. Serial dilutions were plated onto the LB medium for growth of P. putida KT 2440 and the LB–Kn medium for growth of mycobacterium strain KMS. This study was repeated three times with 20 plants used for each treatment.
Root wash growth studies
Root washes used in our study were obtained from barley seedlings grown from disinfested seeds in sterilized vermiculite and watered with sterile water for 7 days. Roots were immersed into sterile, deionized water for 10–15 min using 50 ml water/100 roots. The roots were removed and then wash filtered through two micron filters (Nalgene, Rochester, NY). The neutral sugar concentration was determined by the method of Dubois et al. (1956), with glucose as the standard, and protein was assessed by the bicinchoninic acid method as described by Pierce Chemical, Rockford, IL.
Cultures of Mycobacterium strains MCS, JLS, KMS, PYR-1, and M. flavescens were grown in the amended Middlebrook medium. The cells were harvested by centrifugation while still in logarithmic-phase growth. The cells were washed, and 2 × 105 cells were transferred into 1 ml sterilized root wash solution in 24-well polystyrene tissue culture plates treated by vacuum gas plasma (Becton Dickinson, Franklin Lakes, NJ). After 10 days growth at 23°C, serial dilutions of the liquid cultures were performed to determine planktonic cell density. Cell counts were taken from six separate wells for each strain and growth medium used, and averages and standard deviations were calculated. Results were compared to cell densities of cultures in the amended Middlebrook medium.
Biofilm formation was quantified by a modification of a crystal violet assay described in Djordjevic et al. (2002). The planktonic cells and media were removed from the wells by washing twice carefully with 1 ml/well water. A crystal violet solution, 1 ml/well (0.25%), was added. After 5 min gentle shaking, excess dye was removed by decanting and washing twice with water. The biofilms were visible as purple deposits on the bottom and side of the wells. The crystal violet was extracted from the stained cells by adding 1 ml 80:20 (v/v) ethanol/ethyl acetate per well and shaking for 10 min. The absorbance of the ethanolic extracts was measured at 570 nm. The absorbance values from control wells lacking the bacterial inocula were subtracted from all test values. This biofilm assay was performed six times, each assay with three wells, for all Mycobacterium strains, and averages and standard deviations were calculated.
Identification of simple sugars in the root wash medium
The carbohydrate composition of the barley root wash was analyzed by separating the components by anion exchange chromatography and pulsed amperometric detection on a Dionex DX-500 BioLC system, with a CarboPac PA-100 column (4 × 250 mm). Root wash was diluted 1:50 with sterile double-deionized water, and 25 μl was injected onto the column. Carbohydrates were eluted from the column in 150 mM NaOH with a sodium acetate gradient of 25 mM (0–1 min), 25–100 mM (1–12 min), and 25 mM (12.1–16 min). The elution times of sugars from the root wash samples were compared with the elution profiles of standard solutions of authentic sugars. Areas under the peak were used to calculate the amount of each sugar. Each sample was assayed twice, and two independent studies were performed. This evaluation was repeated with a cell-free culture filtrate obtained after growth of the mycobacterium isolates for 7 days.
Pyrene mineralization in the root wash
The potential to mineralize pyrene in liquid medium was tested for the Mycobacterium isolate KMS. Barley root wash was used at the prepared strength and when diluted tenfold in water. LB was used at full strength or at a tenfold dilution. Briefly, pyrene (300 μg) dissolved in methanol was mixed with radiolabeled 14C pyrene and coated onto 50-ml flasks. Total decomposable plant material in each flask was 211,071 ± 2,166. After the methanol was evaporated, 20 ml of growth media was added. Mycobacterium cells (approximately 1 × 107 cells) were added to three flasks of each media. Other flasks were not inoculated as controls. Each flask was placed into a 500 ml I-Chem Jar (A Nalgene, New Castle, DE) together with a 7-ml scintillation vial that contained 1 ml 0.1 M KOH as a 14CO2 trap. All jars were shaken at 120 rpm at 30°C in the dark. The vials containing the KOH were replaced every 2–7 days, and 14CO2 was counted in a liquid scintillation counter (Beckman Coulter, Fullerton, CA). Assays were run in triplicate and averages, and standard deviations were counted.
Growth on roots under competitive conditions
Growth on root components
Planktonic growth and biofilm formation for strains MCS, KMS, JLS, PYR-1, M. flavescens, and M. smegmatis in barley root wash and Middlebrook medium
Strain of Mycobacterium
Planktonic growth: Middlebrook (Cfu × 108/ml)
Biofilm formation: Middlebrook (Cfu × 108/ml)
Planktonic growth: root wash (Cfu × 108/ml)
Biofilm formation: root wash (Cfu × 108/ml)
7.7 ± 0.8
0.2 ± 0.1
10.3 ± 0.1
1.2 ± 0.1
8.5 ± 0.0
0.2 ± 0.1
10.3 ± 0.5
1.2 ± 0.3
8.6 ± 0.7
0.1 ± 0 .0
8.9 ± 1.3
0.4 ± 0.1
10.0 ± 0.4
0.4 ± 0.2
10.6 ± 0.6
1.1 ± 0.2
9.1 ± 0.7
0.4 ± 0.2
10.5 ± 0.1
1.8 ± 0.1
8.3 ± 0.0
0.4 ± 0.1
9.0 ± 0.0
0.4 ± 0.2
Biofilm formation in the root wash was greater than that in the amended Middlebrook medium for all strains (Table 1) with the exception of M. smegmatis and isolate MCS, where biofilm formation was lower and similar for both media. The presence of surfactant Tween 80 in the amended Middlebrook medium had no effect on biofilm formation (data not shown) and was used because it prevented cell aggregate formation.
Composition of barley root wash before and after growth of mycobacterial strains for 7 days
Uninoculated root wash
1278.1 ± 6.8
708 ± 3
132.3 ± 1.2
74.5 ± 0.2
77 ± 2
168.8 ± 2
192 ± 2
326.8 ± 0.1
330.1 ± 0.4
214 ± 1
81 ± 0
58 ± 0
Mineralization of pyrene in the presence of root wash materials
We observed that mycobacterium cells associated with the root surfaces as roots developed from inoculated barley seeds. Four of the five PAH-degrading mycobacterium isolates tested effectively followed the root tip down into the root matrix from a seed inoculum. Strong root tip association is regarded as a beneficial trait for microbes with remediation potential (Lugtenberg et al. 2001). The plant roots may serve as a bioinjector to augment the spread of the PAH-degrading organisms into soil layers where they could have wider access to deeper pockets of contamination. The finding that one isolate was less able to associate with the plant root tip suggests that not all PAH-degrading mycobacterial strains would have the same potential if used as a seed inoculant to remediate contaminated soils. Our studies add to the few other reports of mycobacterium association with plant roots (Conn and Franco 2004; Daane et al. 2001; Marsollier et al. 2004).
Lack of association with the root tip as the barley roots developed correlated with a high degree of cell surface hydrophobicity for the PAH-degrading MCS strain and another hydrophobic cell line, M. smegmatis. The role of microbial cell hydrophobicity in root surface interactions is not resolved. The first step in root colonization adherence has been studied with mixed results. James et al. (1985) and Macrae et al. (1988) found no correlation between bacterial cell hydrophobicity and their root adherence. However, Van Loosdrecht et al. (1987) and Jana et al. (2000) report that greater cell wall hydrophobicity correlated with nonspecific attachment/adherence. Whether other surface features of MCS and M. smegmatis contributed to the poor root tip association is under further study. Bacterial root surface association may involve products of att genes involved in attachment (Matthysse and McMahan 1998) and/or genes that are later expressed governing biofilm formation, such as those required for the production of cellulose fibrils (Jackson et al. 2005; Matthysse and McMahan 1998; Walker et al. 2004). Interestingly, open reading frames encoding proteins implicated in cellulose breakdown are present in several mycobacterial genomes (Varrot et al. 2005), but whether they are involved in regulation of biofilm formation is unknown. Using a model well-plate-based system, we observed differences in biofilm formation between strains; the more hydrophobic strains produced biofilms with fewest cells.
Medium composition also influenced biofilm formation. The greater extent of biofilm formation on the barley root components compared to the amended Middlebrook medium may be due to stimulatory factors. Autoinducer-like compounds that stimulate upregulation of gene expression of quorum-sensing genes are present in some root exudates (Teplitski et al. 2000), and quorum sensing is postulated to have a role in the development of mature biofilms (Davies et al. 1998). Barley root exudates may contain such compounds that influence gene expression and biofilm formation.
In coinoculation studies with both mycobacterium isolate KMS and the pseudomonad strain KT2440, the root supported higher total microbial load. These two types of bacteria may have utilized discrete root exudate components, allowing for reduced competition for nutrients. Using Biolog test plates with more than 90 different substrates, the mycobacterium cells were observed to use far fewer of the test compounds than the pseudomonad (Miller et al. 2006). P. putida KT2400 is well known for its metabolic diversity (Molina et al. 2005). The relatively slow doubling time (7–8 h; data not shown) of the mycobacterium isolates in liquid medium compared to the doubling time of P. putida KT2440 (37–40 min; data not shown) in liquid medium did not reduce the mycobacterium cell numbers under the competitive conditions used for this study. Future studies will determine mycobacterium population densities and survival in a more microbially complex rhizosphere matrix.
We do not know all of the root components that are utilized for growth by the mycobacterium cells. Low-molecular-weight carbohydrates were detected in the barley root washes. Each of the mycobacterium isolates used glucose and fructose and the oligomers kestose and nestose during growth. Kestoses are common in wheat and barley (Chatterton and Hardson 2003). Additionally, many genera of bacteria both produce and degrade fructans (Henry and Wallace 1993). The nutrients provided from the barley root wash permitted mineralization of pyrene by mycobacterium isolate KMS to occur at a rate similar to that observed in the LB medium, although the latter had a higher protein-to-carbohydrate ratio. The finding that mineralization was greater in 1/10 diluted root wash suggested that mineralization could occur with mycobacterium associated with the rhizosphere. The more limited rate and extent of mineralization in the concentrated media were consistent with catabolic repression as has been observed in other plant–bacterial PAH-degradation studies (Keuth and Rehm 1991; Rentz et al. 2004).
The ability of PAH-degrading mycobacteria to associate along the length of developing roots from a seed inoculum and to survive in the rhizosphere may be useful traits in bioremediation strategies. Our findings suggest that these bacteria would be distributed through polluted soils as the roots grow. Utilization of root exudate components as the bacteria colonize the rhizosphere additionally would provide into the contaminated soils metabolically active bacteria able to mineralize PAHs.