Degradation of xenobiotics: soil surfaces and biofilms and their relevance to bioavailability and bioremediation
Richard Burns
ECG bulletin July 2001
ECG bulletin July 2001
Degradation of xenobiotics: soil surfaces and biofilms and their relevance to bioavailability and bioremediation
As part of the half-day symposium, which accompanied Dr Emsley’s DGL, Professor Richard Burns from the University of Kent gave a wider-ranging talk on the science of soil microorganisms. If you have ever wondered how the work of the American abstract painter Mark Rothko (1903-1970) helps us to understand the breakdown of pollutants by soil microbes, then read on!
Introduction
In soil environments, where carbon and energy substrates are usually sub-optimal, bacteria reduce in size and surface area and migrate to clay and organic matter surfaces. Once at the solid-liquid interface bacteria attach themselves using a flagella and fimbrae and generate adhesive extracellular polysaccharides. These ensure that shear forces do not dislodge them from a comparatively favourable site and gives them time to divide and, by chance or design, accumulate cells of other species within a three-dimensional matrix known as a biofilm. These sessile multispecies communities exhibit complex reciprocal relationships and, in some situations, assume the properties of a multicellular interactive structure. The evolutionary implications of surface biofilms have been much debated, but their composition and the nature of the intercellular communication have only recently become tractable. This is because of the development and application of a wide range of novel molecular, microscopical, analytical and immunological techniques. These provide us, for the first time, with the opportunity for real time, non-destructive measurements.
Soil surfaces, microbes and xenobiotics
The soil surface offers a much more attractive environment to the microbe than the aqueous phase. This is because the high surface area and ionogenic properties of clays and humates allow them to accumulate concentrations of substrates and energy sources (as well as potential toxicants) in their vicinity far in excess of those found just a few micrometres away in the soil water. Thus, amino acids, proteins, enzymes, carbohydrates, alcohols, etc. all associate to a greater or lesser extent with surfaces. In addition many inorganic cations (e.g. ammonium, hydrogen, calcium, iron) and, in the context of this review, a plethora of xenobiotics, accumulate at soil surfaces. The strength of association of these man-made organics with soil surfaces, and their subsequent impact on co-located microbes, depends on the chemical and physical properties of both the organic and the surface as well as the species of microbe.
Microbe, substrate, enzyme and product associations with surfaces and within biofilm structures present challenges to those who wish to understand the microbial ecology of the soil microenvironment and evaluate the impact of xenobiotics. These challenges begin with the relationships between, for example, pesticide sorption kinetics and microbial response and growth kinetics and proceed to involve the study of model biofilm development and activity using a wide range of molecular techniques. Of course, the presence of putative energy sources at surfaces does not necessarily equate directly to their transformation and mineralization by microorganisms. These and other processes are determined by what is called bioavailability.
Bioavailability: a dynamic state
The term bioavailability has a number of theoretical and experimentally-determined definitions. Depending on the objectives of the experiment, xenobiotics may be bioavailable:
As far as bioremediation is concerned, it is of paramount importance to understand the determinants and dynamics of all three of these types of bioavailability. It is tempting to give a general definition of bioavailability that describes it as that proportion of the organic that is in the liquid phase and therefore, by definition, capable of uptake by suitable microorganisms in the vicinity. However, it is not that simple, as some of our recent studies of a number of different herbicides have shown.
The reported degradation of the herbicide 2,4-dichlorophenoxy acetic acid (2,4-D) associated with the highly adsorptive clay mineral chlorite [1] is a good example of the dynamic and somewhat unpredictable nature of bioavailability. At the beginning of the experiment, 254 micrograms 2,4-D were strongly adsorbed to each gram of clay. In the control, non-inoculated, flask 140 micrograms/gram (55%) remained adsorbed and 114 micrograms/gram (45%) were desorbed after shaking for 28 days. This 45% is the fraction that the simple definition of bioavailabilty, given above, would identify as being available for degradation. However, in flasks inoculated at day zero with a known 2,4-D degrading bacterium, 58% (148 micrograms/gram) of the original 2,4-D was mineralised (i.e. converted to CO2, water and microbial biomass). Furthermore, at the end of the experiment, 55 micrograms/gram of 2,4-D remained in solution (and probably would have been degraded also had the experiment been prolonged) and only 51 micrograms/gram were adsorbed. One interpretation of these data is that the Pseudomonas species was capable of accessing 89micrograms/gram (140 less 51) of the adsorbed ‘bio-nonavailable’ 2,4-D.
A second example of how microbes utilise immobilised xenobiotics comes from our studies of triazine herbicides adsorbed, not to clays, but to granular activated carbon [GAC] [2,3,4]. GAC saturated with simazine and simazine in solution were metabolised as the sole carbon sources for the known triazine degrader Rhodococcus rhodochrous SL1. After 21d at 20 °C 96% of the soluble simazine had been degraded but, in addition, 28% of the adsorbed simazine had been degraded.
How do microbes access adsorbed substrates?
There are a number of obvious and less obvious possibilities and four are suggested here:
This last mentioned possibility awaits investigation, but organics can be associated with surfaces such that rings, substituents and or side chains are accessible to the active site of the appropriate enzyme but the entire structure is not. Under these circumstances it is possible that parts of a xenobiotic are cleaved and enter the aqueous phase whilst a ‘footprint’ of the original compound is left behind. Recent developments in atomic force microscopy and the use of differential radio-labelling of complex molecules may enable us to visualise and even quantify this form of bioavailability.
There are two other properties of soils which influence bioavailability and should guide decisions about bioemediation. The first relates to the many different mechanisms of adsorption that may operate in the heterogeneous organic/inorganic matrix and the biotic and abiotic transformations that take place with time. Overall these processes determine the phenomena of pollutant ‘ageing’.
One of the classic examples of ageing is a very rapid process that occurs when highly cationic compounds (such as the bipyridylium herbicide, paraquat) are added to soils. These are adsorbed on the first anionic surface (usually a clay) with which they come in contact. Weak van der Waals forces may allow limited desorption and migration within the alumino-silicate lattices where irreversible hydrogen bonding will occur rendering the compound bio-nonavailable. If the first point of contact is a humic colloid, some greater mobility may take place before irreversible association with the clay fraction. A more complex and drawn out process concerns acylanilide herbicides such as propanil. This pesticide is degraded rapidly by microbes and the parent compound can be described as non-persistent. However, one of its metabolites 3,4-dichloroaniline can condense abiotically to form tetrachloroazobenzene or can be copolymerised enzymatically to form a humic-aniline complex – a so-called bound residue. Both of these products are persistent.
The various types of ageing tell us that bioavailabilites change with time and that great caution should be used when interpreting data arising from soils spiked with parent compounds in laboratory experiments. In the field, bioremediation strategies must take account of whether the pollution is acute or chronic and must monitor metabolites as well as parent compounds.
Pesticides, ageing and soil depth
Another obvious but often ignored property of soil is that chemical, physical and biological characteristics change, not only over small distances (i.e. colloids, microaggregates, macroaggregates), but also with depth. This is illustrated by two examples. The first involves the herbicide 2,4-D (Box 1) which was shown to be degraded much more rapidly in surface soils than in subsurface soils [5, 6]. This was certainly due to the comparatively initial low numbers of 2,4-D degraders in the sub soil but changing soil physico-chemical properties may have influenced bioavailability. Interestingly, it is likely that in soils to which herbicide has been applied over a period of years, the surface layers will contain high numbers of the parent compound degraders whilst the lower horizons (which have been enriched by the breakdown products) will have higher levels of metabolite degraders. In the case of 2,4-D these sub-soil microbes could be 2,4-dichlorophenol degraders.
The second example involves dichlorprop, which is an enantiomeric herbicide. When the effect of soil depth on the degradation of this compound was studied it was revealed that the surface layers contained only (R)-enantiomer degraders (Box 1) while the sub-surface contained both (R)- and (S)-enantiomer degraders. The enantioselectivity of microbial degraders is of interest because analysis of ground water may provide ratios of one form of the herbicide to another that is indicative of the ratio in the original application. If manufacturers were required to register the ratios in their formulations and, if enough were known about the properties of the soils and the impact of rainfall ands other climatic factors, then the origins of a pollution event could be identified.
These properties of soils have implications for rational bioremediation strategies. It must be remembered that soils change with depth, substrates change with depth, and that microbes in sub-surface horizons and deep soils are abundant and active but may behave differently to those at the surface. In other words, bioremediation (whether by biostimulation or bioaugmentation) may require or target different species and require different strategies at different depths within the same soil.
[BOX 1 HERE]
Bioremediation: conventional and molecular
One of the first and most important questions to ask when considering if bioremediation is a possible solution to a contaminated site is - do the appropriate microbes already exist in the soil? If the soil has been contaminated for a period of time it is likely that selection pressures have given rise to a population of pollutant-resistant bacteria and fungi and it is probable that at least some of them have the capacity to transform at least some of the xenobiotics. However, the site is contaminated (that is why it is in need of remediation) so that even if competent microbes exist they are not solving the problem at a significant rate. Attempts to answer the question posed above have in the past involved isolating numerous bacteria on selective media and then attempting to grow them up in pure cultures with the target pollutant(s) as the sole carbon source. This will sometimes provide a partial answer and will certainly generate bacteria that will allow the biochemical study of xenobiotic metabolism. However, many bacteria are difficult if not impossible to isolate using conventional media and growth conditions and most pollutants are degraded in the environment by a community of microorganisms rather than a single species. Another approach is to assume that all the right microbes are already in place and to manipulate the soil to encourage them to express their degradative potential at a desired rapid rate. This is known as natural attenuation and may be initiated by additions of nitrogen, potassium, phosphate and organic carbon, shifting the pH with lime or sulphur, or simple aeration and irrigation [7]. Of course, even if one or a combination of these modifications give rise to accelerated degradation the fundamental microbiology behind the success is unknown and the knowledge generated may be of little value when dealing with the next contaminated site.
In the last few years this hit-or-miss approach has been questioned and a more logical analysis of a soil’s potential has been investigated. This involves what has been termed ‘molecular site assessment’ [8] and is based on our ability to extract nucleic acids from soil and to probe that material for the presence of appropriate catabolic genes. If a complete genotypic potential is demonstrated, the next questions to ask is what proportion of that potential is being transcribed and phenotypically expressed. Once this is established, experiments can be carried out to determine precisely what manipulations are required to realise the complete potential. If, on the other hand, the genotypic potential does not exist, then inoculation (bioaugmentation) must be considered. Box 2 illustrates the molecular approach to bioremediation assessment, the various decision points and a number of ways in which the accelerated degradation of target compounds may be monitored and achieved.
[BOX 2 HERE]
Inoculating contaminated soils with catabolically-competent microbial strains that have been isolated and evaluated in the laboratory is only rarely successful in remediation. There are a number of reasons for this perhaps the most important being the presence of an established and highly evolved indigenous population that is already exploiting existing resources and occupying all suitable niches. Thus the inoculant ‘foreign’ microorganism fails to survive and proliferate. Even if it does survive, it does not necessarily express the degradative capacity demonstrated in vitro. This is because the regulation and control of transcription and expression (of a sequence of catabolic enzymes) in a complex environment such as soil is poorly understood. What this knowledge teaches us (although we have been slow to learn) is that the conventional processes utilised in selecting and isolating putative inoculants are inadequate. This is because, although catabolic potential is essential, the microbial species must have a number of other characteristics in order to compete with the indigenous population. These characteristics are listed in Box 3 and must be considered when designing a rational and appropriate enrichment strategy for generating bioremediating microorganisms.
[BOX 3 HERE]
With this in mind we have investigated the bioavailability of polycyclic aromatic hydrocarbons [9, 10] and microbial consortia derived from conventional and soil-targeted types of enrichments [11]. Conventional shake flask enrichment, using naphthalene and phenanthrene as sole carbon sources and a contaminated soil as a source of inocula, selected twelve bacterial species capable of functioning as a consortium and degrading the two PAHs in liquid culture. In contrast, by using a continuous-flow cell, the contaminated soil as an upstream source of inocula, and a continuous stream of PAHs, we were able to construct a stable degradative community composed of thirty-eight bacterial species and two fungal species. This surface-competent community only contains two naphthalene degraders but degrades the PAH’s at four times the rate of the conventional enrichment culture. We are currently studying the stability, spatial integrity and degradative capacity of biofilm components using nucleic acid probes and fluorescent stains in combination with confocal microscopy. Community diversity is being investigated using such techniques as amplified ribosomal DNA restriction analysis and single strand conformation polymorphism [12].
Soil Complexity
Soil can be described as “a heterogeneous matrix in which microbial communities and processes are discontinuous, dynamic and usually chaotic in space and time”. This is not only a daunting definition but also a challenge that drives a great deal of the fundamental research in microbial ecology. One important decision for researchers who are looking for the solution to the contaminated land problem concerns the amount of detailed knowledge that is necessary to enable the reliable remediation of a high proportion of the sites [13]. Put another way - do we need a complete understanding of all components of the soil system before it can be manipulated? Of course confidence in our ability to attain a complete understanding of soil (or any complex system) assumes a Newtonian deterministic probability that is build on order and rules (i.e. basic laws). It also accepts the principal that the component parts of a complex system should be individually definable and measurable and that they give rise to some sort of stability and homeostasis. But the study of soil is not just the study physical and chemical phenomena, it isn’t just hydrology and biology and it certainly isn’t just ion exchange and enzyme kinetics. What it is microbial ecology, and ecology has an irreducible complexity that may not be rule-based or even predictable. In other words, the sum is not necessarily defined by the properties of the component parts. In ecology the relationships between laws and behaviour is not straightforward – even if it exists at all. In a complex ecological system, experimental or theoretical reductionism tells you a lot about the properties of the component parts but little about the synergistic interactions between those parts. In fact, a reductionist approach may render the whole system unworkable and may even (somewhat heretically) suggest that complex interdependent systems can arise from functionally independent (or even functionless) component parts [14].
Some have argued that the measure of a system’s complexity is algorithmic – that is its complexity is as great as the shortest possible comprehensible message describing it. As a system becomes more complex (and the message gets longer) the closer it approaches randomness. Of course, messages may be shorter and better understood if mathematical principles and its language are applied but there is a limited value to that in as much as the mathematicians ability to visualise and manipulate and extrapolate from a complex equation is an ability shared by only a small number of biologists. Inevitably at some point complexity becomes chaos – although this is used here to define irregular yet patterned behaviour governed by non-random laws.
A system, such as a soil biofilm or root surface biofilm, becomes complex by developing from simpler stages. Therefore, it seems logical that the more complex a system is the longer or more involved must have been its evolution. Thus, we could define the complexity of a soil process or a microenvironment in terms of the amount of time and energy that has gone into organising it. This period of development has been called the depth of complexity. However, there is a real danger that intense examination of a chaotic system with a great depth of complexity will result in a description no less complicated than that used to describe the original system. So there has been no real advance in understanding.
The Nobel physicist Murray Gell-Mann based at the Sante Fe Institute in New Mexico [15] has attempted to explain these different forms of complexity using pictorial art metaphors. He has described a minimalist painting (by an artist such as Mark Rothko) as having a low algorithmic complexity because it doesn’t take many words to describe its composition or very long to reproduce it (after a fashion). An ‘action painting’ by Jackson Pollock, on the other hand, has a high algorithmic complexity (it would take a long time and be very difficult to explain over the telephone!) and is in fact close to chaos. [Interestingly fractal geometry, based on the size and frequency of the blobs of paint and the size of the canvas, has been used to give these paintings a semblance of order [16]]. As with the Rothko’s, ‘Jack the Dripper’s’ paintings appear to have a low depth of complexity (although you should, maybe, consult Mrs Pollock on that one).
Gell-Mann then compares these 20th century works with the early 16th century paintings of Heironymous Bosch. These have a very high algorithmic complexity (but are definitely not random) and require a very long descriptive message. They also have a great depth of complexity because of the time, thought, effort (endless numbers of sketches, repaintings, etc.) that has gone into their production. But then you could argue that a minimalist painting has a great depth of complexity because its developments encompass the entire history of 20th century art. Or that Bosch was completely mad and the only logic and organisation in his paintings are those imposed by the viewer determined to create order (and therefore understanding) out of chaos. What I am stating here is that everything, whether it is a painting, a soil surface, a pollutant leaching to ground water, or whatever, the processes and component parts can be simple or complex, organised or disorganised, logical or illogical - depending upon your perception and needs. Most importantly, gathering more and more information won’t necessarily allow you to resolve the issues.
Soil systems appear to have a high algorithmic complexity – we use whole books and journal volumes in an attempt to describe them. Soil also has a great depth of complexity in terms of its morphogenesis and the evolution and community structure of its indigenous microflora. Even at its most fundamental level, such as a highly weathered clay particle, there is immense complexity and heterogeneity associated with surface topology, charge distribution, diffuse double layer effects, sorbed organic and inorganic ions, and so on. However, other than as a challenging intellectual exercise and a test of experimental design and expensive pieces of equipment, is an understanding of these microenvironment complexities going to help us quantify, predict and manipulate pollutant breakdown? In this context it is probable that many of the features of a complex system have a negligible impact on the overall processes and may be redundant. Modellers might define the soil system and its bioremediation capacity according to such gross properties as microbial biomass, average temperature, rainfall and pH and might suggest that these ‘big’ factors define and predict 90% of the events that take place in soil. Understanding, for example, substrate adsorption and desorption kinetics only increases the predictability to 91% - and may not be a top research priority. Relating effort and the acquisition of knowledge to a set of defined objectives is an issue for funders, programme managers and researchers alike.
Conclusion
More than 100 years of soil microbiology research has helped us to appreciate the essential contribution that bacteria and fungi make to the continuation of life on Earth. Nonetheless, we are still struggling to understand the complex functions of the millions of microbes per gram of soil and how their activities can be predicted and manipulated. In the context of this article the manipulation is targeted towards degrading various toxic organics and returning the soil to something approaching its pristine non-contaminated state. Given that there are estimated to be as much as 360,000 hectares of contaminated land in the UK, the importance of developing rational and successful remediation processes is evident. In situ bioremediation is a cost-effective alternative to chemical and physical ex situ processes, and a significant proportion of UK brown-field sites could be cleaned up to standard appropriate for use. We are on the verge of major breakthroughs in our understanding of the fundamental processes involved in soil decontamination, and the long and sometimes frustrating gestation of bioremediation science is coming to an end.
References
Canterbury, Kent CT2 7NJ, U.K.
As part of the half-day symposium, which accompanied Dr Emsley’s DGL, Professor Richard Burns from the University of Kent gave a wider-ranging talk on the science of soil microorganisms. If you have ever wondered how the work of the American abstract painter Mark Rothko (1903-1970) helps us to understand the breakdown of pollutants by soil microbes, then read on!
Introduction
In soil environments, where carbon and energy substrates are usually sub-optimal, bacteria reduce in size and surface area and migrate to clay and organic matter surfaces. Once at the solid-liquid interface bacteria attach themselves using a flagella and fimbrae and generate adhesive extracellular polysaccharides. These ensure that shear forces do not dislodge them from a comparatively favourable site and gives them time to divide and, by chance or design, accumulate cells of other species within a three-dimensional matrix known as a biofilm. These sessile multispecies communities exhibit complex reciprocal relationships and, in some situations, assume the properties of a multicellular interactive structure. The evolutionary implications of surface biofilms have been much debated, but their composition and the nature of the intercellular communication have only recently become tractable. This is because of the development and application of a wide range of novel molecular, microscopical, analytical and immunological techniques. These provide us, for the first time, with the opportunity for real time, non-destructive measurements.
Soil surfaces, microbes and xenobiotics
The soil surface offers a much more attractive environment to the microbe than the aqueous phase. This is because the high surface area and ionogenic properties of clays and humates allow them to accumulate concentrations of substrates and energy sources (as well as potential toxicants) in their vicinity far in excess of those found just a few micrometres away in the soil water. Thus, amino acids, proteins, enzymes, carbohydrates, alcohols, etc. all associate to a greater or lesser extent with surfaces. In addition many inorganic cations (e.g. ammonium, hydrogen, calcium, iron) and, in the context of this review, a plethora of xenobiotics, accumulate at soil surfaces. The strength of association of these man-made organics with soil surfaces, and their subsequent impact on co-located microbes, depends on the chemical and physical properties of both the organic and the surface as well as the species of microbe.
Microbe, substrate, enzyme and product associations with surfaces and within biofilm structures present challenges to those who wish to understand the microbial ecology of the soil microenvironment and evaluate the impact of xenobiotics. These challenges begin with the relationships between, for example, pesticide sorption kinetics and microbial response and growth kinetics and proceed to involve the study of model biofilm development and activity using a wide range of molecular techniques. Of course, the presence of putative energy sources at surfaces does not necessarily equate directly to their transformation and mineralization by microorganisms. These and other processes are determined by what is called bioavailability.
Bioavailability: a dynamic state
The term bioavailability has a number of theoretical and experimentally-determined definitions. Depending on the objectives of the experiment, xenobiotics may be bioavailable:
- As toxicants to microbes, plants and invertebrates.
- For microbial transformation involving catabolism, co-metabolism or some partial transformation or modification of the molecule; or (following leaching or run-off).
- To inhabitants and consumers of receiving waters.
As far as bioremediation is concerned, it is of paramount importance to understand the determinants and dynamics of all three of these types of bioavailability. It is tempting to give a general definition of bioavailability that describes it as that proportion of the organic that is in the liquid phase and therefore, by definition, capable of uptake by suitable microorganisms in the vicinity. However, it is not that simple, as some of our recent studies of a number of different herbicides have shown.
The reported degradation of the herbicide 2,4-dichlorophenoxy acetic acid (2,4-D) associated with the highly adsorptive clay mineral chlorite [1] is a good example of the dynamic and somewhat unpredictable nature of bioavailability. At the beginning of the experiment, 254 micrograms 2,4-D were strongly adsorbed to each gram of clay. In the control, non-inoculated, flask 140 micrograms/gram (55%) remained adsorbed and 114 micrograms/gram (45%) were desorbed after shaking for 28 days. This 45% is the fraction that the simple definition of bioavailabilty, given above, would identify as being available for degradation. However, in flasks inoculated at day zero with a known 2,4-D degrading bacterium, 58% (148 micrograms/gram) of the original 2,4-D was mineralised (i.e. converted to CO2, water and microbial biomass). Furthermore, at the end of the experiment, 55 micrograms/gram of 2,4-D remained in solution (and probably would have been degraded also had the experiment been prolonged) and only 51 micrograms/gram were adsorbed. One interpretation of these data is that the Pseudomonas species was capable of accessing 89micrograms/gram (140 less 51) of the adsorbed ‘bio-nonavailable’ 2,4-D.
A second example of how microbes utilise immobilised xenobiotics comes from our studies of triazine herbicides adsorbed, not to clays, but to granular activated carbon [GAC] [2,3,4]. GAC saturated with simazine and simazine in solution were metabolised as the sole carbon sources for the known triazine degrader Rhodococcus rhodochrous SL1. After 21d at 20 °C 96% of the soluble simazine had been degraded but, in addition, 28% of the adsorbed simazine had been degraded.
How do microbes access adsorbed substrates?
There are a number of obvious and less obvious possibilities and four are suggested here:
- As the soluble substrate is metabolised, adsorbed substrate is desorbed in an attempt to maintain equilibrium.
- During metabolism of soluble compounds microbes externalise H+ ions, which become concentrated at the anionic clay and humic surfaces; this acidifies the microenvironment pH. As a consequence adsorbants and adsorbates that depend on their ionogenic properties for mutual association may, depending on their pKa values, undergo a change in the distribution, intensity or even sign of their exposed functional groups. Thus, what was once an attraction due to oppositely charged interactants becomes one of repulsion and substrates may pass into solution.
- Microbes are known to produce surfactants that will render hydrophobic surface associated organics more available by creating emulsions that can be transported into the cell and catabolised.
- Microbes produce extracellular enzymes that will be attracted to surfaces and will form complexes with an adsorbed xenobiotic depending upon its configuration in the adsorbed state.
This last mentioned possibility awaits investigation, but organics can be associated with surfaces such that rings, substituents and or side chains are accessible to the active site of the appropriate enzyme but the entire structure is not. Under these circumstances it is possible that parts of a xenobiotic are cleaved and enter the aqueous phase whilst a ‘footprint’ of the original compound is left behind. Recent developments in atomic force microscopy and the use of differential radio-labelling of complex molecules may enable us to visualise and even quantify this form of bioavailability.
There are two other properties of soils which influence bioavailability and should guide decisions about bioemediation. The first relates to the many different mechanisms of adsorption that may operate in the heterogeneous organic/inorganic matrix and the biotic and abiotic transformations that take place with time. Overall these processes determine the phenomena of pollutant ‘ageing’.
One of the classic examples of ageing is a very rapid process that occurs when highly cationic compounds (such as the bipyridylium herbicide, paraquat) are added to soils. These are adsorbed on the first anionic surface (usually a clay) with which they come in contact. Weak van der Waals forces may allow limited desorption and migration within the alumino-silicate lattices where irreversible hydrogen bonding will occur rendering the compound bio-nonavailable. If the first point of contact is a humic colloid, some greater mobility may take place before irreversible association with the clay fraction. A more complex and drawn out process concerns acylanilide herbicides such as propanil. This pesticide is degraded rapidly by microbes and the parent compound can be described as non-persistent. However, one of its metabolites 3,4-dichloroaniline can condense abiotically to form tetrachloroazobenzene or can be copolymerised enzymatically to form a humic-aniline complex – a so-called bound residue. Both of these products are persistent.
The various types of ageing tell us that bioavailabilites change with time and that great caution should be used when interpreting data arising from soils spiked with parent compounds in laboratory experiments. In the field, bioremediation strategies must take account of whether the pollution is acute or chronic and must monitor metabolites as well as parent compounds.
Pesticides, ageing and soil depth
Another obvious but often ignored property of soil is that chemical, physical and biological characteristics change, not only over small distances (i.e. colloids, microaggregates, macroaggregates), but also with depth. This is illustrated by two examples. The first involves the herbicide 2,4-D (Box 1) which was shown to be degraded much more rapidly in surface soils than in subsurface soils [5, 6]. This was certainly due to the comparatively initial low numbers of 2,4-D degraders in the sub soil but changing soil physico-chemical properties may have influenced bioavailability. Interestingly, it is likely that in soils to which herbicide has been applied over a period of years, the surface layers will contain high numbers of the parent compound degraders whilst the lower horizons (which have been enriched by the breakdown products) will have higher levels of metabolite degraders. In the case of 2,4-D these sub-soil microbes could be 2,4-dichlorophenol degraders.
The second example involves dichlorprop, which is an enantiomeric herbicide. When the effect of soil depth on the degradation of this compound was studied it was revealed that the surface layers contained only (R)-enantiomer degraders (Box 1) while the sub-surface contained both (R)- and (S)-enantiomer degraders. The enantioselectivity of microbial degraders is of interest because analysis of ground water may provide ratios of one form of the herbicide to another that is indicative of the ratio in the original application. If manufacturers were required to register the ratios in their formulations and, if enough were known about the properties of the soils and the impact of rainfall ands other climatic factors, then the origins of a pollution event could be identified.
These properties of soils have implications for rational bioremediation strategies. It must be remembered that soils change with depth, substrates change with depth, and that microbes in sub-surface horizons and deep soils are abundant and active but may behave differently to those at the surface. In other words, bioremediation (whether by biostimulation or bioaugmentation) may require or target different species and require different strategies at different depths within the same soil.
[BOX 1 HERE]
Bioremediation: conventional and molecular
One of the first and most important questions to ask when considering if bioremediation is a possible solution to a contaminated site is - do the appropriate microbes already exist in the soil? If the soil has been contaminated for a period of time it is likely that selection pressures have given rise to a population of pollutant-resistant bacteria and fungi and it is probable that at least some of them have the capacity to transform at least some of the xenobiotics. However, the site is contaminated (that is why it is in need of remediation) so that even if competent microbes exist they are not solving the problem at a significant rate. Attempts to answer the question posed above have in the past involved isolating numerous bacteria on selective media and then attempting to grow them up in pure cultures with the target pollutant(s) as the sole carbon source. This will sometimes provide a partial answer and will certainly generate bacteria that will allow the biochemical study of xenobiotic metabolism. However, many bacteria are difficult if not impossible to isolate using conventional media and growth conditions and most pollutants are degraded in the environment by a community of microorganisms rather than a single species. Another approach is to assume that all the right microbes are already in place and to manipulate the soil to encourage them to express their degradative potential at a desired rapid rate. This is known as natural attenuation and may be initiated by additions of nitrogen, potassium, phosphate and organic carbon, shifting the pH with lime or sulphur, or simple aeration and irrigation [7]. Of course, even if one or a combination of these modifications give rise to accelerated degradation the fundamental microbiology behind the success is unknown and the knowledge generated may be of little value when dealing with the next contaminated site.
In the last few years this hit-or-miss approach has been questioned and a more logical analysis of a soil’s potential has been investigated. This involves what has been termed ‘molecular site assessment’ [8] and is based on our ability to extract nucleic acids from soil and to probe that material for the presence of appropriate catabolic genes. If a complete genotypic potential is demonstrated, the next questions to ask is what proportion of that potential is being transcribed and phenotypically expressed. Once this is established, experiments can be carried out to determine precisely what manipulations are required to realise the complete potential. If, on the other hand, the genotypic potential does not exist, then inoculation (bioaugmentation) must be considered. Box 2 illustrates the molecular approach to bioremediation assessment, the various decision points and a number of ways in which the accelerated degradation of target compounds may be monitored and achieved.
[BOX 2 HERE]
Inoculating contaminated soils with catabolically-competent microbial strains that have been isolated and evaluated in the laboratory is only rarely successful in remediation. There are a number of reasons for this perhaps the most important being the presence of an established and highly evolved indigenous population that is already exploiting existing resources and occupying all suitable niches. Thus the inoculant ‘foreign’ microorganism fails to survive and proliferate. Even if it does survive, it does not necessarily express the degradative capacity demonstrated in vitro. This is because the regulation and control of transcription and expression (of a sequence of catabolic enzymes) in a complex environment such as soil is poorly understood. What this knowledge teaches us (although we have been slow to learn) is that the conventional processes utilised in selecting and isolating putative inoculants are inadequate. This is because, although catabolic potential is essential, the microbial species must have a number of other characteristics in order to compete with the indigenous population. These characteristics are listed in Box 3 and must be considered when designing a rational and appropriate enrichment strategy for generating bioremediating microorganisms.
[BOX 3 HERE]
With this in mind we have investigated the bioavailability of polycyclic aromatic hydrocarbons [9, 10] and microbial consortia derived from conventional and soil-targeted types of enrichments [11]. Conventional shake flask enrichment, using naphthalene and phenanthrene as sole carbon sources and a contaminated soil as a source of inocula, selected twelve bacterial species capable of functioning as a consortium and degrading the two PAHs in liquid culture. In contrast, by using a continuous-flow cell, the contaminated soil as an upstream source of inocula, and a continuous stream of PAHs, we were able to construct a stable degradative community composed of thirty-eight bacterial species and two fungal species. This surface-competent community only contains two naphthalene degraders but degrades the PAH’s at four times the rate of the conventional enrichment culture. We are currently studying the stability, spatial integrity and degradative capacity of biofilm components using nucleic acid probes and fluorescent stains in combination with confocal microscopy. Community diversity is being investigated using such techniques as amplified ribosomal DNA restriction analysis and single strand conformation polymorphism [12].
Soil Complexity
Soil can be described as “a heterogeneous matrix in which microbial communities and processes are discontinuous, dynamic and usually chaotic in space and time”. This is not only a daunting definition but also a challenge that drives a great deal of the fundamental research in microbial ecology. One important decision for researchers who are looking for the solution to the contaminated land problem concerns the amount of detailed knowledge that is necessary to enable the reliable remediation of a high proportion of the sites [13]. Put another way - do we need a complete understanding of all components of the soil system before it can be manipulated? Of course confidence in our ability to attain a complete understanding of soil (or any complex system) assumes a Newtonian deterministic probability that is build on order and rules (i.e. basic laws). It also accepts the principal that the component parts of a complex system should be individually definable and measurable and that they give rise to some sort of stability and homeostasis. But the study of soil is not just the study physical and chemical phenomena, it isn’t just hydrology and biology and it certainly isn’t just ion exchange and enzyme kinetics. What it is microbial ecology, and ecology has an irreducible complexity that may not be rule-based or even predictable. In other words, the sum is not necessarily defined by the properties of the component parts. In ecology the relationships between laws and behaviour is not straightforward – even if it exists at all. In a complex ecological system, experimental or theoretical reductionism tells you a lot about the properties of the component parts but little about the synergistic interactions between those parts. In fact, a reductionist approach may render the whole system unworkable and may even (somewhat heretically) suggest that complex interdependent systems can arise from functionally independent (or even functionless) component parts [14].
Some have argued that the measure of a system’s complexity is algorithmic – that is its complexity is as great as the shortest possible comprehensible message describing it. As a system becomes more complex (and the message gets longer) the closer it approaches randomness. Of course, messages may be shorter and better understood if mathematical principles and its language are applied but there is a limited value to that in as much as the mathematicians ability to visualise and manipulate and extrapolate from a complex equation is an ability shared by only a small number of biologists. Inevitably at some point complexity becomes chaos – although this is used here to define irregular yet patterned behaviour governed by non-random laws.
A system, such as a soil biofilm or root surface biofilm, becomes complex by developing from simpler stages. Therefore, it seems logical that the more complex a system is the longer or more involved must have been its evolution. Thus, we could define the complexity of a soil process or a microenvironment in terms of the amount of time and energy that has gone into organising it. This period of development has been called the depth of complexity. However, there is a real danger that intense examination of a chaotic system with a great depth of complexity will result in a description no less complicated than that used to describe the original system. So there has been no real advance in understanding.
The Nobel physicist Murray Gell-Mann based at the Sante Fe Institute in New Mexico [15] has attempted to explain these different forms of complexity using pictorial art metaphors. He has described a minimalist painting (by an artist such as Mark Rothko) as having a low algorithmic complexity because it doesn’t take many words to describe its composition or very long to reproduce it (after a fashion). An ‘action painting’ by Jackson Pollock, on the other hand, has a high algorithmic complexity (it would take a long time and be very difficult to explain over the telephone!) and is in fact close to chaos. [Interestingly fractal geometry, based on the size and frequency of the blobs of paint and the size of the canvas, has been used to give these paintings a semblance of order [16]]. As with the Rothko’s, ‘Jack the Dripper’s’ paintings appear to have a low depth of complexity (although you should, maybe, consult Mrs Pollock on that one).
Gell-Mann then compares these 20th century works with the early 16th century paintings of Heironymous Bosch. These have a very high algorithmic complexity (but are definitely not random) and require a very long descriptive message. They also have a great depth of complexity because of the time, thought, effort (endless numbers of sketches, repaintings, etc.) that has gone into their production. But then you could argue that a minimalist painting has a great depth of complexity because its developments encompass the entire history of 20th century art. Or that Bosch was completely mad and the only logic and organisation in his paintings are those imposed by the viewer determined to create order (and therefore understanding) out of chaos. What I am stating here is that everything, whether it is a painting, a soil surface, a pollutant leaching to ground water, or whatever, the processes and component parts can be simple or complex, organised or disorganised, logical or illogical - depending upon your perception and needs. Most importantly, gathering more and more information won’t necessarily allow you to resolve the issues.
Soil systems appear to have a high algorithmic complexity – we use whole books and journal volumes in an attempt to describe them. Soil also has a great depth of complexity in terms of its morphogenesis and the evolution and community structure of its indigenous microflora. Even at its most fundamental level, such as a highly weathered clay particle, there is immense complexity and heterogeneity associated with surface topology, charge distribution, diffuse double layer effects, sorbed organic and inorganic ions, and so on. However, other than as a challenging intellectual exercise and a test of experimental design and expensive pieces of equipment, is an understanding of these microenvironment complexities going to help us quantify, predict and manipulate pollutant breakdown? In this context it is probable that many of the features of a complex system have a negligible impact on the overall processes and may be redundant. Modellers might define the soil system and its bioremediation capacity according to such gross properties as microbial biomass, average temperature, rainfall and pH and might suggest that these ‘big’ factors define and predict 90% of the events that take place in soil. Understanding, for example, substrate adsorption and desorption kinetics only increases the predictability to 91% - and may not be a top research priority. Relating effort and the acquisition of knowledge to a set of defined objectives is an issue for funders, programme managers and researchers alike.
Conclusion
More than 100 years of soil microbiology research has helped us to appreciate the essential contribution that bacteria and fungi make to the continuation of life on Earth. Nonetheless, we are still struggling to understand the complex functions of the millions of microbes per gram of soil and how their activities can be predicted and manipulated. In the context of this article the manipulation is targeted towards degrading various toxic organics and returning the soil to something approaching its pristine non-contaminated state. Given that there are estimated to be as much as 360,000 hectares of contaminated land in the UK, the importance of developing rational and successful remediation processes is evident. In situ bioremediation is a cost-effective alternative to chemical and physical ex situ processes, and a significant proportion of UK brown-field sites could be cleaned up to standard appropriate for use. We are on the verge of major breakthroughs in our understanding of the fundamental processes involved in soil decontamination, and the long and sometimes frustrating gestation of bioremediation science is coming to an end.
References
- McGhee I et al. Bioavailability of 2,4-D sorbed to a chlorite- like complex. Chemosphere, 1999, 39: 285-291.
- Feakin SJ et al. Inoculation of granular activated carbon in a fixed bed with s-triazine-degrading bacteria as a water treatment process. Water Research, 1995, 29: 819-825.
- Feakin SJ et al. Inoculation of granular activated carbon with s-triazine-degrading bacteria for water treatment at pilot scale. Water Research, 1995, 29: 1681-1688.
- Jones LR et al. Bacterial inoculation of granular activated carbon filters for the removal of atrazine from surface waters. Water Research, 1998, 32: 2542-2549.
- Shaw LJ & Burns RG. Biodegradation of 2,4-D in a noncontaminated grassland soil profile. Journal of Environmental Quality, 1998, 27, 1464-1471.
- Shaw LJ & Burns RG. Biodegradation-transport interactions of pulse applied 2,4-D in repacked soil column microcosms. Journal of Environmental Quality, 1998, 27: 1427-1478.
- McGhee I & Burns RG. Bioremediation of soil contaminated with the herbicides 2,4-dichlorophenoxyacetic acid (2,4-D) and 2-methyl-4-chlorophenoxyacetic acid (MCPA). Applied Soil Ecology, 1995, 2: 143-154.
- Sayler GS et al. Molecular Site Assessment and Process Monitoring in Bioremediation and Natural Attenuation. Applied Biochemistry and Biotechnology
- Smith MJ et al. Bioavailability and biodegradation of polycyclic aromatic hydrocarbons in soil. FEMS Microbiology Letters, 1997, 152: 141-147.
- Smith MJ et al. Fate of phenanthrene, pyrene and benzo(a)pyrene during biodegradation of crude oil added to soils. FEMS Microbiology Letters, 1999, 173:445-452.
- Stach JEM et al. PCR-SSCP comparison of 16S rDNA sequence diversity in soil DNA obtained using different isolation and purification methods. FEMS Microbiology Ecology, 2001 (in press).
- Stach JEM & Burns RG. Functional and phylogenetic comparisons of biofilm and conventional enrichment communities using RT-PCR-SSCP. 2001 (in preparation).
- Burns RG. Bioavailability, bioremediation and biofilms in soil: how complex is complex? Mitteilung der Duetschen Bodenkundlichen Gesellschaft (in press).
- Behe MJ. Darwin’s Black Box. 1997, ISBN 0684834936, also http://www.arn.org/behe/mb_ic.htm
- Taylor RP et al. Fractal analysis of Pollock’s drip paintings. Nature, 1999, 39: 422, also http://www.sciencenews.org/sn_arc99/9_18_99/mathland.htm
Canterbury, Kent CT2 7NJ, U.K.
