- Top of page
- Materials and methods
- Results and Discussion
Hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX) is a serious environmental pollutant on military land. This compound is the most widely used explosive and pollution has arisen primarily as the result of military training, along with munition manufacturing and disassembly processes. This toxic explosive is recalcitrant to degradation in the environment and leaches rapidly into groundwater, where accumulation in aquifers is threatening drinking water supplies (Clausen, et al., 2004). While plants have only limited degradative activity towards RDX, microorganisms, including Rhodococcus rhodochrous 11Y, have been isolated from contaminated land. Despite the presence of microbial RDX-metabolising activity in contaminated soils, the persistence of RDX in leachate from contaminated soil indicates that this activity or biomass is insufficient, limiting its use to remediate polluted soils. Bacterial activity in the rhizosphere is of magnitudes greater than in the surrounding soil, and the roots of grass species on training ranges in the United States are known to penetrate deeply into the soil, producing a compact root system and providing an ideal environment to support the capture of RDX by microorganisms in the rhizosphere. Here, we have investigated the ability of the root-colonising bacterium Pseudomonas fluorescens, engineered to express XplA, to degrade RDX in the rhizosphere.
- Top of page
- Materials and methods
- Results and Discussion
The nitramine hexahydro-1,3,5-trinitro-1,3,5-triazine (RDX) makes up a major component of military explosives. This compound is both toxic; it is classed as a possible carcinogen by the U.S. Environmental Protection Agency and recalcitrant to degradation in the soil. Following extensive manufacturing, use and decommissioning, there is now significant environmental contamination with RDX on military land (Rylott et al., 2011).
A specific problem is the high mobility of RDX through the soil column, resulting in the subsequent contamination of groundwater on military training ranges, which is now threatening drinking water supplies. Current decontamination methods such as incineration or composting are expensive and, with 24.6 million acres of active military ranges in the United States alone, unsuitable for the scale of the problem. In addition, decontamination using these approaches would require closure of active training ranges. To contain and remediate existing and future contamination, alternative in situ strategies are required. A number of RDX-degrading microorganisms have been isolated from RDX-contaminated soils using enrichment studies whereby RDX is supplied as the sole nitrogen source for growth. However, despite the presence of RDX-degrading soil microorganisms, RDX continues to be a pollutant, suggesting that other factors are limiting these bacteria from utilising RDX in the soil. Bioaugmentation, a process to supplement the growth of RDX-degrading bacteria, has so far yielded only limited, short-term effects (Pennington et al., 2001). The rhizosphere, a zone surrounding the roots, is of great microbial activity resulting from the extensive exudation of compounds including organic acids, sugars and amino acids from plant roots (Lugtenberg & Dekkers, 1999), and rhizosphere soil can be populated by 10–1000 times more bacteria than adjacent bulk soil (Lugtenberg & Kamilova, 2009). The bacterial community within the rhizosphere is extremely diverse and dependent on a variety of factors such as soil properties and plant species (Kent & Triplett, 2002; Berg, 2009). The high levels of bacterial activity and biomass, along with predicted increases in other soil microorganisms, are advantageous for the degradation of toxic compounds. In line with this, rhizoremediation systems for pollutants such as cadmium (Wu et al., 2006), polychlorinated biphenyl (Villacieros et al., 2005) and 2,4-dinitrotoluene, a manufacturing by-product of the explosive compound 2,4,6-trinitrotoluene (Monti et al., 2005), have been shown to decrease levels of these contaminants.
A study has demonstrated that growth of Arabidopsis in 2,4-dinitrotoluene-contaminated soil could be enhanced by adding a Pseudomonas strain modified to express genes from a Burkholderia sp., which confer the ability to degrade this toxic pollutant (Monti et al., 2005). In complement with this, experiments have demonstrated that maize inoculated with P. putida JLR11, a bacterium able to use 2,4,6-trinitrotoluene as the sole source of nitrogen, enhanced the removal of 2,4,6-trinitrotoluene from a hydroponic environment (Van Dillewijn et al., 2007). We have shown that plants expressing XplA, efficiently remove RDX from soil leachate, utilising the resultant nitrogen for growth (Rylott et al., 2006, 2011; Jackson et al., 2007). However, no studies have yet investigated the degradation of RDX by bacteria expressing XplA in the rhizosphere.
A number of bacterial species able to transform RDX have now been identified (Crocker et al., 2006), and xplA, which was first isolated from Rhodococcus rhodochrous 11Y (Seth-Smith et al., 2002, 2008; Rylott et al., 2006; Jackson et al., 2007), has now also been identified in several of these species. None of the RDX-degrading species isolated so far are known root colonisers. Here, we have tested the ability of R. rhodochrous 11Y to degrade RDX in the rhizosphere. Alongside this, we have engineered a species known to colonise alfalfa roots efficiently: Pseudomonas fluorescens F113 (Villacieros et al., 2005) to express the RDX-metabolising enzyme XplA.
Results and Discussion
- Top of page
- Materials and methods
- Results and Discussion
Our experiments used the gnotobiotic system described above to investigate the ability of the RDX-degrading bacterium R. rhodochrous 11Y to colonise Arabidopsis roots and to degrade RDX in the rhizosphere. Results showed that there was no significant RDX degradation or removal of RDX from soil by R. rhodochrous 11Y (data not shown). Rhodococcus rhodochrous 11Y is not a known root coloniser of Arabidopsis, and thus, XplA activity was introduced into Pseudomonas fluorescens F113. To test the activity of XplA in the P. fluorescens genetic background, resting cell assays were performed. During earlier studies (Jackson et al., 2007), we had found α-aminolevulinic acid, a well-characterised tetrapyrrole precursor (Avissar & Beale, 1989), to be required for XplA expression in Escherichia coli. Thus, we used resting cell assays to test the requirement of α-aminolevulinic acid for P. fluorescens to synthesis active XplA. The results, in Fig. 1, showed that P. fluorescens containing XplA removed 49% of the RDX over the time course of the experiment. When α-aminolevulinic acid was present, significantly more, 72%, RDX was removed indicating that supply of haem precursors for the assembly of XplA is a limiting factor in XplA expression.
Figure 1. RDX-metabolising activity of Pseudomonas fluorescens strain F113 expressing xplA. (a) Resting cell assays showing removal of RDX. Pseudomonas fluorescens transformed with xplA were dosed with 160 μM RDX and uptake monitored at 30 °C. Results are mean ± SD of three biological replica. (b) Levels of RDX in alfalfa aerial tissue four weeks following inoculation with Pseudomonas fluorescens containing the xplA-expressing plasmid pME6031- TAC-xplA or the empty vector control pME6031. Seedlings were grown in RDX-contaminated sand in a gnotobiotic system and, where stated, supplemented with ALA. Results are mean ± SE of eight biological replica.
Download figure to PowerPoint
We conducted studies to determine whether α-aminolevulinic acid would also be limiting production of active XplA and thus RDX removal by P. fluorescens in the rhizosphere. Alfalfa plants were inoculated with P. fluorescens and grown in the gnotobiotic system described above. Four weeks after inoculation, the levels of RDX remaining in the bulk soil and rhizosphere were measured, and the results are presented in Table 1. In both the rhizosphere and bulk samples, significantly higher levels of RDX were removed from the samples containing plants inoculated with pME6031-TAC-xplA, when compared to those containing plants inoculated with the pME6031 empty vector control. This result was in agreement with the resting cell assays and indicates that the supply of haem precursors for XplA production by P. fluorescens is also limited in the rhizosphere. Following these observations, α-aminolevulinic acid was included in subsequent experiments designed to test individually the contribution of alfalfa and P. fluorescens to the removal of RDX and the ability of XplA-expressing P. fluorescens to populate the alfalfa rhizosphere. The mass balance data are presented in Table 2. When treatment 1 with neither plants nor bacteria was compared with treatment 2, the results showed that the presence of untransformed P. fluorescens did not significantly affect the levels of RDX in the bulk sand. However, in treatment 3, inoculated with P. fluorescens expressing XplA, significantly more RDX was removed than by the untransformed P. fluorescens in treatment 2, indicating that the XplA-expressing bacteria are metabolising some RDX in the absence of a plant root environment. In treatment 4, the presence of alfalfa, in addition to untransformed P. fluorescens, enhanced the removal of RDX when compared to untransformed P. fluorescens in treatment 2. Plants are known to have, albeit low, abilities to degrade RDX (Best et al., 1999; Bhadra et al., 2001; Just & Schnoor, 2004).
Table 1. Effect of ALA amendment on RDX removal from bulk and rhizosphere sand samples. Results are mean ± SE of eight biological replica
|Sample||% RDX remaining in soil containing plants inoculated with pME6031- TAC-xplA, relative to plants inoculated with pME6031 empty vector control|
|No ALA||+ ALA|
|Rhizosphere||75.3 ± 5.5||27.6 ± 3.1|
|Bulk||96.4 ± 7.3||59.4 ± 3.9|
Table 2. Mass balance showing percentage RDX recovered 4 weeks postinoculation. Results are mean of 16 biological replica ± SE
|Treatment||Bulk||Rhizosphere||Aerial plant tissue||Total|
|1||No bacteria, no plant||70.7 ± 0.9||–||–||70.7 ± 0.9|
|2||Bacteria with pME6031||69.6 ± 0.8||–||–||69.6 ± 0.8|
|3||Bacteria with pME6031- TAC-xplA||51.2 ± 0.8||–||–||51.2 ± 0.8|
|4||Bacteria with pME6031 and plant||29.7 ± 1.3||7.6 ± 3.0||16.6 ± 2.6||53.9 ± 0.12|
|5||Bacteria with pME6031- TAC-xplA and plant||21.0 ± 1.5||6.4 ± 0.4||10.7 ± 0.6||38.1 ± 0.08|
The mass balance data show that when alfalfa plants were inoculated with XplA-expressing P. fluorescens (treatment 5), the total remaining RDX in the system was significantly lower than in all other treatments. Furthermore, the aerial alfalfa tissues had 50% less RDX than alfalfa inoculated with untransformed P. fluorescens. RDX is known to preferentially accumulate in the aerial tissue (Vila et al., 2007), and this result indicates that the expression of XplA in the root-colonising P. fluorescens effectively reduced the level of RDX in the rhizosphere available for uptake by the plant. We have previously observed significant reductions in the accumulation of RDX in plants engineered to express XplA (Rylott et al., 2011) and reducing levels would reduce exposure of herbivores to this toxic compound.
To investigate the stability of the xplA gene within P. fluorescens, samples from the sand were grown on agar amended with tetracycline, the substrate for the selection gene product encoded in the empty and xplA-containing plasmids. Colony counts are shown in Table 3, and PCR and 16S sequencing techniques were used to identify P. fluorescens F113 and colonies containing xplA. The results demonstrate that both untransformed and xplA-expressing P. fluorescens successfully colonised the sand to similar levels, populating predominantly the rhizosphere over the bulk soil. Additionally, the percentage of bacteria retaining xplA decreased significantly. These observations indicate that the ability to metabolise RDX, conferred by xplA, did not give a selective advantage under these experimental conditions and that, possibly as a result of this, the xplA-containing plasmid was not stable over the experimental time.
Table 3. Levels of Pseudomonas fluorescens and xplA in bulk and rhizosphere sand 4 weeks postinoculation. Results are mean of 16 biological replica ± SE, cfu = colony-forming units
|Time of inoculation t = 0 days||Population cfu g−1|
|P. fluorescens expressing xplA||1.3 ± 0.3 × 105||1.7 ± 0.3 × 105|
|P. fluorescens and empty vector||1.9 ± 0.5 × 105||2.6 ± 0.4 × 105|
|Time of inoculation t = 24 days||Population cfu g−1|
|+ tetracycline||− tetracycline|
|Rhizosphere||Bulk soil||Rhizosphere||Bulk soil|
|P. fluorescens expressing xplA||0.05 ± 0.007 × 105||0||54 ± 10 × 105||10 ± 0.5 × 105|
|P. fluorescens and empty vector||0.03 ± 0.006 × 105||0||62 ± 10 x105||1.3 ± 0.2 × 105|
To test our experimental system in conditions more similar to those found in the environment, we conducted studies using soil artificially contaminated with RDX. There was no significant difference in the level of RDX in the soil leachate from plants inoculated with bacteria expressing XplA, when compared with plants inoculated with bacteria transformed with the empty vector only (results not shown).
While we demonstrate that RDX removal can be enhanced by rhizosphere bacteria engineered to express the RDX-metabolising gene product XplA, in a sand-based system, our results also show that there are a number of factors limiting the effectiveness of RDX uptake. Transgene stability could be improved by increasing the selective pressure of RDX. In our experiments, non-RDX sources of nitrogen might not have been limiting, reducing the need to metabolise RDX. Nitrogen is likely to be a limiting nutrient on training range soils. Our results also indicate that availability of the haem precursor α-aminolevulinic acid is a factor in XplA activity in P. fluorescens. Alternative microorganisms might have the ability to synthesise precursors more readily than P. fluorescens.
Another limiting factor is the activity of XplA. It has been shown that the expression of the R. rhodochrous 11Y reductase partner XplB alongside XplA in Arabidopsis can enhance RDX uptake by up to 30-fold. In the RDX-degrading species tested so far, the xplA and xplB genes are plasmid encoded (Andeer et al., 2009; Indest et al., 2010) and there is evidence for lateral transfer to related bacteria (Andeer et al., 2009; Jung et al., 2011). Additionally, codon optimisation of xplA for expression in Pseudomonas would be expected to enhance XplA activity. The lack of nutrients in RDX-contaminated military soils is likely to be a significant factor limiting the number of RDX-degrading microorganisms. If the XplA plasmid could be transferred to species that thrive in the relatively nutrient-rich rhizosphere environment, then our studies predict that in situ remediation and containment of RDX on military training land would be significantly enhanced. Furthermore, if this technology was employed in combination with XplA-expressing plants, this could be a powerful tool for the remediation of RDX from polluted soils.