Prof. Christof Holliger, Dr. Pierre Rossi, Christiane Devenoges
6th Framework Programm of the EU
September 2004 – December 2007
Dr. D. Pieper, GBF, Braunschweig, Germany ; Victor de Lorenzo, Centro Nacional de Biotecnología CSIC, Spain; Stefan Trapp, Technical University of Denmark, Copenhagen, Denmark; Vladimir Brenner, Czech Academy of Sciences, Institute of Microbiology, Prague, Czech Republic; Ulrich Karlson, NERI, Denmark; Hermann Heipieper, Centre for Environmental Research Leipzig-Halle GmbH, Germany; Jan Jurak, KAP Ltd, Czech Republic; Juan Rodriguez, BIONOSTRA S.L., Spain.
The objective of BIOTOOL was the assessment, evaluation and prediction of natural attenuation processes to implement natural attenuation as the accepted key groundwater and soil remediation strategy in Europe. This requires benchmarked monitoring tools for diagnosing biological status and predicting evolution of contaminated soil and groundwater, which have to be rooted in biological processes. The generation and validation of such novel instruments will be materialized through the application of a suite of state-of-the-art genomic, proteomic and analytical technologies to environmental samples and sites themselves. The translocation of indicator chemicals from below ground into above-ground vegetation was assessed as a cheap and rapid monitoring tool for subsurface contamination. Diagnosis of the biological status and evolution models for polluted environments will be achieved through [i] the design and utilization of DNA and specifically DNAarray technology for examining the catabolic potential of any given particulate sample and [ii] the identification of protein biomarkers as descriptors of soil and groundwater conditions and biological attenuation.
The progress in microbial community functional genomics and proteomics was employed to gain a mechanistic understanding of microbial responses to chemical insults, plant/microbe interactions and microbial community adaptations that determine microbial-driven soil and groundwater attenuation processes. Such mechanistic understanding will add a considerable predictive power to the genomic and proteomic approaches. Determining the links between environmental factors and expression of degradation abilities will be crucial for strategies aiming at an optimal expression of the catalytic power of the indigenous microbial community. The robustness of diagnostic instruments for future normative applications was validated in microcosms and used for assessment of contaminated sites under study. The specific objectives of the project were:
- Establishment of the correlation between soil/groundwater contamination and plant contamination.
- The design and utilization of DNA and specifically DNA-array technology for examining the catabolic potential of any given particulate sample.
- The access and analysis of the soil/groundwater meta-proteome as biomarker.
- The use of lipid biomarkers as general prediction instruments of stress/toxicity on soil and groundwater microorganisms.
- Elucidation of the roles of natural and chemical stresses and plant/microbe interactions on the metabolic activities of soil and groundwater microbial catalysts.
- The robustness of above diagnostic instruments will be validated in microcosms and used for assessment of contaminated sites under study.
For the project, experimental sites located in the Czech Republic and in Denmark that are polluted with chlorinated hydrocarbons or petroleum hydrocarbons have been selected. In the Czech Republic, the Hradcany site was used as a military airport from the Second Word War, founded by the German Army and, after an intermediate use by the Czechoslovak Army, the Soviet Army operated it from 1968 until 1990, when it was closed. First investigative and remedial works were commenced in 1986 and full-scale clean-up activities started in 1997. The site is a part of the Bohemian Cretaceous Basin, the most important source of high quality groundwater in the Czech Republic. Contamination limits future use and revitalization of the site. The North Bohemia Carcass Disposal Plant (SAP) Mimoň is one of the largest and most intensive chloroethenes contaminations of soil and groundwater in the Czech Republic. Perchloroethene was used in huge amounts in operation of the SAP Plant since 1963 to 1988. Frequent operational leakages caused a large plum of contamination. The sites have been characterized in detail regarding their geological and hydrogeological characteristics, soil and/or rock composition and global chemical analysis of unsaturated and saturated zone of the rock environment and thus supports background data. During the project, evolution of the experimental sites (soil and groundwater contamination changes, biodegradation activity characteristics) was followed. In addition cone penetrometry and membrane interface probing was used in order to obtain a precise geological cross section description and to evaluate vertical and horizontal distribution of contamination.
In order to exploit the translocation of indicator chemicals from below ground into above-ground vegetation as a cheap and rapid monitoring tool for subsurface contamination, a mathematical model describing uptake, translocation and volatilization of chemicals in / from trees was developed, parameterized and tested with lab and field data. The model was also adapted to simulate herbs and grasses, in order to select the best indicator species.
Generally, more polar, but persistent chemicals are translocated to the upper stem of trees growing on the polluted sites. Possible indicator chemicals would be chlorinated solvents perchloroethene, trichloroethene, dichloroethene, but not vinylchloride) and (chlorinated) phenols. Less applicable are BTEX, due to their low persistence in the root zone, and their high volatility. Not applicable are alkanes, due to their high vapour pressure. Out of the PAH, naphthalene is the only promising indicator chemical.
Laboratory studies on perchloroethene, trichloroethene, cis-dichloroethene, heavy metals, monoand dichlororphenol were undertaken, in order to determine toxicity, calibrate the model, and find the relation between subsurface- and vegetation contamination. Simultaneously, an automated method for measurement of chemical activity of indicator chemicals in tree samples has been developed. Process studies indicated that neither herbs nor grasses are well suited for monitoring, and that needle trees are better than deciduous trees for several reasons (rooting depth, wood structure, seasonal effects). Herbs and grasses cannot be used as indicator species for volatile compounds (chlorinated ethenes, BTEX), because the chemicals volatilize too fast from stem (and probably roots). Furthermore, these species do not root as deep as trees (in particular coniferes), and grow very fast, which leads to dilution of the chemicals. Heavy metals, such as copper and cadmium, can easily be found in leaves and wood. Thus, vegetation sampling can also be used to find subsurface contamination with several heavy metals.
Data from field campaigns showed that trees can clearly be used as indicators of the extent of subsurface contamination. Even though a correlation between trees and subsurface pollution could be established in some cases, the correlations are not generally valid, it will not be possible to conclude from tree core measurements alone to the concentration and depth of subsurface pollution. In conclusion, tree core sampling must rather be seen as a way to assess the presence of pollutants than as a technique that allows determining the subsurface concentrations.
The method was applied at the SAP site on a larger scale. A before undetected plume of PCE was found employing the new method, which was confirmed by MIP, an independent method of physical soil exploration. It was proven that monitoring of the tree core chloroethene contamination is in very good relation with direct groundwater sampling in the area of interest. Based upon this result, this method is recommended to all environmental firms as effective, fast and cheep method for monitoring of chloroethene in shallow groundwater aquifers. A practical guide (“A Guide to Vegetation Sampling for Screening of Subsurface Pollution“) gives guidance for the application of the method and summarizes the existing knowledge about features and limitations of the method. The guideline is distributed primarily via internet.
Initially, all necessary parameters were set for a systematic investigation of the effect of different chemical and physical stresses on growth and fatty acid composition of typical soil bacteria, with the goal to use lipid biomarkers as general prediction instruments of stress/toxicity on soil and groundwater microorganisms and to elucidate the roles of natural and chemical stresses on the metabolic activities of soil and groundwater microbial catalysts. From the data obtained from the PLFA analysis of different experimental systems investigated, it becomes evident that the trans/cis ratio of unsaturated fatty acids can be used as an excellent toxicity tool in laboratory experiments. However, this tool cannot be transferred to real bioremediation plants as it is only a real urgent stress response system and other adaptive mechanisms are in use in the case of long term adaptation. In contrast, a direct relation could be observed between the TPH (total petroleum hydrocarbon) concentration present in soils and the degree of saturation of the PLFA, describing the membrane rigidity of the microbial community as a useful stress parameter. Moreover, the analysis of the phospholipid fatty acids (PLFA) profile in contaminated soils is an elegant technique for a qualitative and quantitative assessment of the microbiota present. In the project we developed a method for rapid extraction of lipids from soil samples by the so-called Accelerated Solvent Extraction that allowed the higher throughput of samples. Using this method we were able analyze the status and follow the evolution of the experimental Hradcany site during different lifetime of the air sparging treatment. In detail, the monitoring showed that this method gives valuable information on: (i) the quantity of microorganisms present in the investigated soil (ii) the composition of the microbiota regarding the presence of eukaryotic (fungi, protozoa) and prokaryotic (bacteria) organisms (iii) the composition of the microbiota regarding the type of bacteria (Gram-negative or Gram-positive) that are predominating and (iv) the status of these bacteria with respect to the stress, usually caused by the presence of toxic pollutants. Analyses of respective taxonomical compositions showed a clear increase of bacterial diversity indicating that taxonomical diversity estimations may help to detect effective aerobic bioremediation, as the stimulation of certain members able to survive and degrade the compounds help to induce a shift towards an increasing complexity in the ecosystem under treatment.
To further evaluate pollutant stress effects, DNA array technology was used to decipher the interplay between expression of catabolic genes and stress caused in a model soil bacterium. The pollutant toluene can be regarded as a potential nutrient, as a membrane-damaging toxic drug, and as a macromolecule-disrupting agent. Expression profiles suggest that the bulk of the available transcriptional machinery is reassigned to endure general stress, whereas only a small share of the available machinery is redirected to the degradation of the aromatic compounds.
Gene arrays dedicated to achieve a fast monitoring of catabolic gene diversity and abundance and thus the complete catabolic landscape and catabolic potential of microbial communities have successfully been developed and also been validated. By using the current knowledge on biodegradation pathways of organic compounds in cultured bacteria, catabolic gene families coding for these activities and the molecular targets traced by culture-independent means in soil samples were defined. After an extensive literature mining on biodegradation pathways, an overall of eleven catabolic gene families catalyzing key steps of aerobic and anaerobic catabolic pathways have been selected. The databases to cover the microbial catabolome targeted for the Biotool microarray and other genetic fingerprinting applications comprises 1820 sequences. Validation included analysis of correctness of the array design using sequenced strains and mixtures, as well as verification and correlation with data obtained by catabolic gene fingerprinting (HZI) and metagenome surveys. Also a novel type of array capable to probe not only the catabolic potential for reductive dehalogenation, but also to identify organisms capable of reductive dehalogenation and the bacterial guilds which are potentially involved in the delivery of electron donors in such environments has been developed.
To further fine-tune catabolic gene analysis, various PCR primers for catabolic genes have been optimized and upgraded which allows targeted and detailed analysis of subgroups of important genes in study sites. The purpose of the development of molecular fingerprinting for a full array of catabolic gene families is the capability to rapidly determine the intrafamily diversity and predominance of gene variants in a extensive number of samples. Thereby obtained detailed pictures of the catabolic gene structure and sequence diversity in environmental samples are significantly increasing our knowledge of the functional potential of microbial communities and identification of shifts in catabolic gene structure allow the deduction of the evolutionary fitness of catabolic genes, operons and their respective hosts under changing environmental conditions. The methods have not only been adapted for the use with a capillary electrophoresis gene sequence analyzer, allowing very reproducible and semi-quantitative analyses but also by combination with a reverse transcriptase step to expression and thus catabolic activity studies.
However, PCR as well as hybridization based approaches to survey microbial communities involved in bioremediation and natural attenuation are limited by the number of sequences of genes available. One approach that does not rely on conserved nucleotide sequences is to use genomic libraries to retrieve genes from natural bacterial communities without cultivation. This functional environmental approach was successfully applied for screening of oxygenase enzymatic activities as key activities of aerobic BTEX degradation potential. Analysis of these libraries, which represented a wide fraction of the whole metagenome from two representative sites showed that abundance of key catechol 2,3-dioxygenase activity correlated with pollution level, indicating its usefulness as marker activity. Primer-based screening in coordination with other methods allowed identification of predominant catabolic gene families and indicated key subgroups. Activity-based meta-genome screening was thus validated as a suitable technology to analyze catabolic diversity and to characterize gene landscapes and gave indications on spatial and catabolic profile differences. This activity-based meta-genome screening was also suitable to recover biocatalysts with novel properties that have been selected by environmental pressure. Genetic analysis revealed a highly upgraded picture of catabolic gene diversity in contaminated environments, as subfamilies assumed to be important from culture-based studies and often used as marker for degradative potential were not abundant, but subfamilies expected from fingerprinting and array analysis were observed to be abundant also by the metagenome based approach.
However, one critical factor that limits the exploration of desired catabolic reactions is the recurrent difficulty of efficiently screening or selecting for genes of interest in large libraries of cloned DNA, in case the reactions are phenotypically silent. In this context, we have demonstrated a general approach to translate biotransformations lacking easily observable phenotypes into traits that can be selected for through the combination of evolved transcriptional regulator variants responsive to the product of a desired reaction with a suitable transcriptional response. By combining one Pseudomonas derived evolved transcriptional regulator variant responsive to 1,2,4 trichlorobenzene with a suitable transcriptional response, we were able to generate a genetic trap in which the action of the gene for dehydrochlorination of hexachlorocyclohexane was eventually converted into detectable phenotypes. To allow an even wider applicability of metagenome based screenings, genetic tools to implement genetic traps for surveying metagenomic libraries in a variety of Gramnegative hosts were constructed and a broad host range genetic platform to format orthogonal sensor circuits in the chromosome of Gram-negative engineered as bioindicators for the production of given chemical compounds has been developed. Microbial communities in bioremediation and natural attenuation scenarios however, do not only rely on appropriate biocatalytic activities but on a complex interaction between members of such communities. Aquifers, for example, are dynamic ecosystems showing complex interactions between physical, chemical and biotic components and can be considered as heterogeneous assemblages of discrete macro- and micro-scale habitats, providing a variety of living conditions, which influence the heterogeneous distribution of the microbial community structures and their inherent activities. Using optimized protocols for high-throughput community analyses and numerical ecology tools on various aquifers contaminated with chlorinated solvents we could shown that population diversity, and more specifically the diversity of the dehalorespiration guild alone, does not drive ecosystem stability and natural attenuation processes. The positive relationship between the presence of multiple pathways towards a product and functional stability parallels theoretical concepts in higher ecological organization. Ecosystem stability is the outcome not of population diversity per se, but of functional redundancy, which is ensured by the presence of a reservoir of species able to perform the same ecological function, even in very diverse and fluctuating environmental conditions. Recognizing the diversity and the links within each key functional group of a system can lead to better ways to model diversity and function, as well as helping to improve process stability.
In an approach complementary to molecular diagnostics we were attempting to include the application of state-of-the-art proteomic technologies for the identification of protein biomarkers, which are descriptive of soil status and predictive of its evolution. To explore and categorize proteins present in polluted sites in a fashion independent of the specific host that bears them, we set up a new technology for generation of natural single-chain antibody libraries against predetermined proteins by inoculation of camels (Camelus dromedarius). Camelidae produce immunoglobulins that have only one variable domain that could recognize, by itself, the antigen. The amplified variable domain of a heavy-chain antibody of Camelidae is the simplest affibody that could recognize an antigen. As it is a smaller molecule the advantages are not only the specialized recognition of antigen but also better stability, easier fusion to other molecules and better recognition of crevices in the three-dimensional structure of the antigen. We have shown the power of camel antibody technology for production of large numbers of antibodies that recognize specifically given catabolic enzymes. The new angle of such a technology is the ability to innoculate the animal with a mixture of proteins and the recovery of active clones against each of them. While these antibodies offer very good opportunities for identification of distinct enzymes in in vitro assays, a suitable protein extraction procedure from soil that can be employed for revealing the biodegradative landscape remains to be established.
The study of the global properties of complex metabolic networks in organisms has prompted us a way to tackle the first models about the organization and evolution of environmental catabolic networks. To organize all the available information in a coherent database, with substantial capacity for the interaction with the experimental biologist working in this domain, MetaRouter, a Bioinformatics system for maintaining heterogeneous information related with Biodegradation had been developed (http://pdg.cnb.uam.es/MetaRouter/). This knowledge on biological systems was now used to develop a machine learning approach for the prediction of the “biodegradability” of new compounds, based on their chemical descriptors. The system allows the interpretation of the chemical characteristics of the compounds that are better related with the capacity of the biological systems to metabolize them, opening the possibilities for their study in the laboratory, and for predicting the biodegradative fate of a new chemical compound before releasing it in the environment. As the extremely rapid pace of production of novel compounds by the chemical and pharmaceutical industry makes the detailed experimental assessment of their biodegradability virtually impossible, automatic predictive methods, such as the one of the Biodegradation Prediction Server (http://www.pdg.cnb.uam.es/BDPSERVER), are essential to evaluate the potential of new compounds to pollute the environment. The new predictive system may help to implement recent international regulations on the use of new chemicals.
In conclusion, BIOTOOL has established tree core analyses as innovative method, taking use of the qualitative relationship between soil/groundwater contamination and plant contamination, for rapid monitoring of subsurface contamination and subsurface biodegradation processes. The optimization of DNA as well as lipid extraction procedures in order to study the catabolic potential of contaminated environments has been successfully accomplished and a new kit for the efficient DNA extraction from polluted soil was developed. Novel or optimized tools (catabolome arrays, molecular fingerprinting, metagenomics, lipid profiling community diversity measures) were applied to the study sites and the combination of techniques derived from diverse scientific fields opens the way to a consistent strategy for predicting the actual behavior and future evolution of contaminated aquifers.