Explosives contamination at Department of Defense (DoD) sites has arisen as a result of the manufacture, deployment, and decommissioning of munitions. The contaminants of particular concern are nitroaromatic and nitroamine explosives typified by TNT and RDX, respectively. RDX is a significant issue at a number of DoD sites because of its high mobility through soils and subsequent contamination of groundwater. Bacteria have been isolated that can degrade and metabolize RDX by selective enrichment studies, but to date these bacteria have not been found at DoD field sites. Currently, it is not clear which pathways and enzymes are responsible for RDX degradation in situ. A greater understanding of nitroamine biodegradation in the environment would permit an informed evaluation of natural attenuation.

The objectives of this project were to:

  1. Conduct molecular characterization of microbial communities in soils with RDX degradation
  2. Determine RDX metabolites and metabolic pathways
  3. Determine the diversity of xplA in the environment and possible evolutionary origin
  4. Determine the mechanism of horizontal gene transfer of xplA
  5. Investigate the regulation and transcription of xplA/B
  6. Discover new bacteria capable of degrading RDX and determine the enzymes involved.

Technical Approach

A two-pronged approach was used to characterize RDX degradation in soils: (a) based on known RDX-degrading isolates and (b) a culture independent molecular approach to characterize the microbial communities of soils contaminated to varying degrees with RDX. Sequenced clone libraries combined with stable isotope probing were used to characterize in situ microbial populations of munitions contaminated soils. The occurrence of Rhodococcus and the diversity of xplA/B genes for RDX degradation in the environment were studied. The relative importance of anaerobic and aerobic RDX degrading pathways was determined. The evolutionary origin of known RDX degradation pathways was investigated and the significance of horizontal gene transfer of xplA was studied. Genetic control of RDX degradation was studied and discovery of new RDX degrading bacteria and enzymes was attempted.


Military range soils were screened for RDX degradation activity. Efforts to develop a functional screen for genes from the soil metagenome were unsuccessful. Efficient methods of packaging large fragments of partially digested DNA in vectors were developed; however, clones from the partially digested DNA of any of the known RDX degraders were unable to be enriched. There may be a toxic effect of the expression of the DHP reductase or that expression of xplBA in the vector was inhibited by the upstream DNA.

Degradation of MNX was studied, a key RDX initial product and frequently found as a contaminant with RDX. In the presence of 18O2, Rhodococcus DN22 degraded RDX and produced NO2ˉ that subsequently oxidized to NO3ˉ containing one 18O atom, but in the presence of H218O, NO3ˉ without 18O was detected. A control containing NO2ˉ, DN22, and 18O2 gave NO3ˉ with one 18O, confirming biotic oxidation of NO2ˉ to NO3ˉ. Treatment of MNX with DN22 and 18O2 produced NO3ˉ incorporating two 18O atoms and another incorporating only one 18O. In the presence of H218O, NO2ˉ with two different masses was detected, one representing NO2ˉ and another representing NO2ˉ with the inclusion of one 18O, suggesting auto-oxidation of NO to NO2ˉ.

Horizontal gene transfer of the plasmid containing xplBA using transconjugants that expressed xplAB in the recipient was unable to be demonstrated. Nevertheless, the 454 genome draft sequencing showed the presence in Rhodococcus sp. type strain 11Y of genes potentially involved in plasmid transfer. Overexpression and deletion of the permease aroP showed no change in RDX removal. In resting cells assays, the rate of RDX removal by the marR knock-out was decreased, suggesting that MarR may be involved in the detection of high nitrogen. Activity towards RDX was present in resting cell incubations of R. rhodochrous 11Y , although in the absence of RDX, activity increased when cells were grown with RDX or low concentrations of NH4Cl as the N source, suggesting that xplA expression is a response to low N levels rather than RDX.

The regulation of RDX degrading genes in Rhodococcus strain 11Y has parallels with the regulation of genes encoding atrazine degradation in Pseudomonas sp. strain ADP. Transcription of atzR, a regulator that activates the atrazine gene cluster, is switched on by a metabolite of atrazine degradation, cyanuric acid, and by nitrogen limitation, and this control is thought to link to central nitrogen metabolism in Pseudomonas sp.

Analysis of the R. rhodochrous 11Y sequence identified two putative promoter regions, one upstream of aroP and one upstream of xplB. For the putative promoter region upstream of xplB, primer extension analysis and RACE PCR did not amplify any products. Furthermore, the sequence is only eleven bases from the ATG start codon of xplB, a distance that does not leave much space for RNA polymerase binding prior to translation. Primer extension analysis and RACE PCR did confirm the BPROM-predicted transcription start site upstream of aroP.

2D-Gel electrophoresis of 11Y grown on either RDX or NH4Cl as sole N sources identified some proteins upregulated after RDX treatment involved in nitrogen metabolism; however, no RDX-specific proteins as XplA were found, suggesting that the chosen treatment, as shown for the resting cell assays, results in a similar response and expression of xplA and related proteins arguing that XplA expression is linked to the central nitrogen metabolism and not a specific response to RDX.

Enrichment studies identified a number of RDX degrading bacteria isolated from contaminated soil from England. Belgium and Ukraine. All isolates were gram positive bacteria and belong to the Rhodococcus clade with the exception of Microbacterium isolated from Belgium soil. The different enrichment conditions have not allowed the isolation of gram negative bacteria, or bacteria degrading RDX with an enzyme other than xplA. Our isolates of RDX degrading bacteria are identical in 16S rDNA, which suggests they form a monoclonal population. They constitute a closely related but different genotype in comparison to the other RDX degrader Rhodococcus isolated in the UK. These results indicated that Rhodococcus so far remains the main genus found to biodegrade RDX.

The evolutionary origin of XplA is still unclear. So far, XplA is the only enzyme that actively degrades RDX. Both xplA and xplB are part of a high GC content gene cluster that is nearly identical among Rhodococcus spp, isolated from United Kingdom, Ukraine, United States, Belgium and Israel. The region is also highly conserved in two known genera of actinomycetes: Microbacterium sp. MA1 and Williamsia sp. EG1. The machinery mechanism of transferring the gene cluster around the world is not yet understood; however, the finding that the XplA genomics island is limited to actinomycetes suggests that there is a specific transfer system, possibly via phages.


This project contributed to our understanding of the biodegradation of RDX in situ at DoD field sites. Our study helped identify the microbial communities, pathways and enzymes responsible for the dissimilation of RDX in soils.