Synthetic organophosphorus compounds are used as pesticides, plasticizers, air fuel ingredients and chemical warfare agents. Organophosphorus compounds are the most widely used insecticides, accounting for an estimated 34% of world-wide insecticide sales. Contamination of soil from pesticides as a result of their bulk handling at the farmyard or following application in the field or accidental release may lead occasionally to contamination of surface and ground water. Several reports suggest that a wide range of water and terrestrial ecosystems may be contaminated with organophosphorus compounds. These compounds possess high mammalian toxicity and it is therefore essential to remove them from the environments. In addition, about 200 000 metric tons of nerve (chemical warfare) agents have to be destroyed world-wide under Chemical Weapons Convention (1993). Bioremediation can offer an efficient and cheap option for decontamination of polluted ecosystems and destruction of nerve agents. The first micro-organism that could degrade organophosphorus compounds was isolated in 1973 and identified as Flavobacterium sp. Since then several bacterial and a few fungal species have been isolated which can degrade a wide range of organophosphorus compounds in liquid cultures and soil systems. The biochemistry of organophosphorus compound degradation by most of the bacteria seems to be identical, in which a structurally similar enzyme called organophosphate hydrolase or phosphotriesterase catalyzes the first step of the degradation. organophosphate hydrolase encoding gene opd (organophosphate degrading) gene has been isolated from geographically different regions and taxonomically different species. This gene has been sequenced, cloned in different organisms, and altered for better activity and stability. Recently, genes with similar function but different sequences have also been isolated and characterized. Engineered microorganisms have been tested for their ability to degrade different organophosphorus pollutants, including nerve agents. In this article, we review and propose pathways for degradation of some organophosphorus compounds by microorganisms. Isolation, characterization, utilization and manipulation of the major detoxifying enzymes and the molecular basis of degradation are discussed. The major achievements and technological advancements towards bioremediation of organophosphorus compounds, limitations of available technologies and future challenge are also discussed.
The excessive use of natural resources and large scale synthesis of xenobiotic compounds have generated a number of environmental problems such as contamination of air, water and terrestrial ecosystems, harmful effects on different biota, and disruption of biogeochemical cycling. At the present time, the most widely used pesticides belong to the organophosphorus group. The first organophosphorus insecticide, tetraethyl pyrophosphate, was developed and used in 1937 (Dragun et al., 1984). At the same time, two chemical warfare agents (also called nerve agents), Tabun and Sarin, were developed and produced. Later, several other organophosphorus pesticides were developed and commercialized. These pesticides are widely used world-wide to control agricultural and household pests. Overall, organophosphorus compounds account for ∼38% of total pesticides used globally (Post, 1998). In the USA alone over 40 million kilos of organophosphorus are applied annually (Mulchandani et al., 1999a; EPA, 2004). Glyphosate and chlorpyrifos are the most widely used in the US and account for 20% and 11% of total pesticide use, respectively (EPA, 2004). Organophosphorus compound poisoning is a world-wide health problem with around 3 million poisonings and 200 000 deaths annually (Karalliedde & Senanayake, 1999; Sogorb et al., 2004). The compounds have been implicated in several nerve and muscular diseases in human beings. Their acute adverse effects have been discussed by Colborn et al. (1996) and Ragnarsdottir (2000). Immunotoxicity of organophosphorus compounds towards human beings and wild-life has been reviewed by Galloway & Handy (2003).
Continuous and excessive use of organophosphorus compounds has led to the contamination of several ecosystems in different parts of the world (EPA, 1995; McConnell et al., 1999; Cisar & Snyder, 2000; Tse et al., 2004). For an example, surveys revealed that 100% of sampled catchments in Scotland and 75% of sampled aquatic sites in Wales were contaminated with organophosphorus compounds used in sheep dips (Boucard et al., 2004). Several organophosphorus compounds are used on animals for the control of body pests as several of them are fat soluble and can thus enter the body readily through the skin and potentially find their way into meat and milk (MAFF/HSE, 1995). Contamination of grains, vegetables and fruits with organophosphorus compounds is also well documented (Pesticide Trust 1996; National Consumer Council 1998). Another potential and more dangerous source of organophosphorus contamination comes from chemical warfare agents. About 200 000 tons of extremely toxic organophosphorus chemical warfare agents such as Sarin, Soman, and VX were manufactured and are stored. As required by the Chemical Weapon Convention (CWC) 1993, these stocks must be destroyed within 10 years of ratification by the member states. Use of micro-organisms in detoxification decontamination of organophosphorus compounds is considered a viable and environment friendly approach.
The available literature on the microbial degradation of xenobiotics indicates that most studies have considered three aspects:
1The fundamental basis of biodegradation.
2Evolution and transfer of such activities among micro-organisms.
However, the use of micro-organisms for bioremediation requires an understanding of all physiological, microbiological, ecological, biochemical and molecular aspects involved in pollutant transformation (Iranzo et al., 2001).
There are two types of xenobiotics that cause environmental concerns: (1) compounds that are persistent and therefore provide long exposure to non-target organisms such as lindane and DDT, and (2) compounds that are biodegradable but mobile in soil and are toxic and therefore have the potential to pollute ground water, such as carbofuran. Extensive and repeated use of the same pesticide without any crop or pesticide rotation for a number of years has occasionally resulted in unexpected failures to control the target organisms. It has been demonstrated that a fraction of the soil biota can develop the ability rapidly to degrade certain soil-applied pesticides. This phenomenon has been described as enhanced or accelerated biodegradation (Walker & Suett, 1986). The first evidence of biodegradation of pesticide affecting its efficacy was reported in 1971 (Sethunathan, 1971). However, it was not until the early to mid 1980s that the wider implication of enhanced bio-degradation became observable in the field (Walker & Suett, 1986) and since then this phenomenon has been reported for several other pesticides such as isofenphos (Chapman et al., 1986), fenamiphos (Stiriling et al., 1992) and ethoprophos (Karpouzas et al., 1999).
The practical significance of enhanced bio-degradation depends on a number of interactive factors like the use of the pesticides (soil or foliage applied), the frequency of use, the interval between successive applications and the stability of the active microflora without the presence of pesticides (Kaufman et al., 1985). Recently, soil pH has been implicated as a factor in enhanced degradation of atrazine in different soils (Houot et al., 2000). This hypothesis has been supported by recent reports of high enzymatic activity (Acosta-Martinez & Tabatabai, 2000) and higher bacterial activity at higher soil pH (Vidali, 2001). Sims et al. (2002) suggested that soil pH may influence the rate of degradation by affecting the uptake of the herbicide by soil micro-organisms. The problem of enhanced bio-degradation became more acute, following the observation that a pesticide can be degraded rapidly in soil from a field to which it had never been applied before but which had been exposed to a pesticide from the same chemical group (Prakash et al., 1996). This phenomenon is known as cross-adaptation. Cross-adaptation of enhanced biodegradation has been reported within many groups of pesticide, such as the carbamates (Morel-Chevillet et al., 1996), dicarboximides (Mitchell & Cain, 1996) and isothiocyanates (Warton et al., 2002). On the other hand, only limited cross-adaptation for enhanced biodegradation within the organophosphorus class has been reported (Racke & Coats, 1988; Singh et al., 2005). Cross-adaptation within groups is unpredictable and may occur only in one direction. The positive side of this problem is that micro-organisms isolated for degradation of one compound can be used for bioremediation of other compounds for which no known degrading microbial system is known. This aspect is well established for organophosphorus compounds where a parathion-degrading bacterium was able to degrade a wide range of other structurally similar compounds including chemical warfare agents. Isolation of pesticide degrading microorganisms is important for three main reasons:
1To determine the mechanism of the intrinsic process of microbial metabolism.
2To understand the mechanisms of gene/enzyme evolution.
3To use these microbes for the detoxification and decontamination of polluted aquatic and terrestrial environments (bioremediation).
Several microorganisms have been isolated which are able to utilize pesticides as a source of energy. There are some examples of fungi including Trametes hirsutus, Phanerochaete chrysosporium, Phanerochaete sordia and Cyathus bulleri that are able to degrade lindane and other pesticides (Singh & Kuhad, 1999, 2000; Singh et al., 1999). However, most evidence suggests that soil bacteria are the principal components responsible for enhanced bio-degradation (Walker & Roberts, 1993). Several pure bacterial isolates with the ability to use specific pesticides as a sole source of carbon, nitrogen or phosphorus have been isolated (Singh et al., 1999, 2000).
On numerous occasions, mixed bacterial cultures with pesticide degradation ability are isolated but their individual components are unable to utilize the chemical as an energy source when purified (Shelton & Somich, 1988; Mandelbaum et al., 1993; De Souza et al., 1993; Roberts et al., 1993); an example is the organophosphorus nematicide fenamiphos (Ou & Thomas, 1994; Singh et al., 2003b). Several other studies failed to obtain micro-organisms capable of growing on specific chemicals. However, this failure does not exclude biological involvement in degradation and could be attributed to the selection and composition of the liquid media under artificial environments, strains requiring special growth factors, or a major role of non-culturable microorganisms (Walker & Roberts, 1993). A recent report of growing previously non-culturable bacteria in the laboratory with a simulated natural environment (Kaeberlein et al., 2002) may lead to isolation and characterization of several new chemical-degrading bacteria.
The main aim of this article is to review the metabolic pathways involved in organophosphorus compound degradation. Our understanding of the molecular basis of organophosphorus degradation has progressed dramatically in recent years. Additional information has become available by genome sequencing of several microorganisms and advancement in molecular techniques. There is growing interest in developing biotechnological methods for clean up of contaminated water and soil with organophosphorus compounds and to aid in the destruction of large amounts of nerve agents. In this article we also critically review recent biotechnological advancements in the development of bio-catalysts and bio-sensors for organophosphorus compounds and their possible application in bioremediation of contaminated ecosystems.
Chemistry and toxicology of organophosphorus compounds
Most organophosphorus compounds are ester or thiol derivatives of phosphoric, phosphonic or phosphoramidic acid. Their general formula is presented in Fig. 1. R1 and R2 are mainly the aryl or alkyl group, which can be directly attached to a phosphorus atom (phosphinates) or via oxygen (phosphates) or a sulphur atom (phosphothioates). In some cases, R1 is directly bonded with phosphorus and R2 with an oxygen or sulfur atom (phosphonates or thion phosphonates, respectively). At least one of these two groups is attached with un-, mono- or di-substituted amino groups in phosphoramidates. The X group can be diverse and may belong to a wide range of aliphatic, aromatic or heterocyclic groups. The X group is also known as a leaving group because on hydrolysis of the ester bond it is released from phosphorus (Fig. 1) (Sogorb & Vilanova, 2002).
The mode of action of organophosphorus compounds includes inhibition of neurotransmitter acetylcholine breakdown. Acetylcholine is required for the transmission of nerve impulses in the brain, skeletal muscles and other areas (Toole & Toole, 1995). However, after the transmission of the impulse, the acetylcholine must be hydrolyzed to avoid overstimulating or overwhelming the nervous system. This breakdown of the acetylcholine is catalyzed by an enzyme called acetylcholine esterase. Acetylcholine esterase converts acetylcholine into choline and acetyl CoA by binding the substrate at its active site at serine 203 to form an enzyme substrate complex. Further reactions involve release of choline from the complex and then rapid reaction of acylated enzymes with water to produce acetic acid and the regenerated acetylcholine esterase. It has been estimated that one enzyme can hydrolyze 300 000 molecules of acetylcholine every minute (Ragnarsdottir, 2000).
Organophosphorus compounds inhibit the normal activity of the acetylcholine esterase by covalent bonding to the enzyme, thereby changing its structure and function. They bind to the serine 203 amino acid active site of acetylcholine esterase. The leaving group binds to the positive hydrogen of His 447 and breaks off the phosphate, leaving the enzyme phosphorylated. The regeneration of phosphorylated acetylcholine esterase is very slow and may take hours or days, resulting in accumulation of acetylcholine at the synapses. Nerves are then overstimulated and jammed (Manahan, 1992). This inhibition causes convulsion, paralysis and finally death for insects and mammals (Ragnarsdottir, 2000).
Microbial degradation of organophosphorus compounds
Use of organochlorine pesticides such as dichloro-diphenyl-trichloroethane (DDT), lindane, etc., has been reduced drastically in developed countries due to their long persistence, tendency towards bioaccumulation and potential toxicity towards non-target organisms. This group of compounds has been replaced by the less persistent and more effective organophosphorus compounds. However, most of the organophosphorus compounds possess high mammalian toxicity. Among the organophosphorus compounds, glyphosate, chlorpyrifos, parathion, methyl parathion, diazinon, coumaphos, monocrotophos, fenamiphos and phorate have been used extensively and their efficacy and environmental fate have been studied in detail. The chemical and physical properties of some of these compounds are listed in Table 1. The phosphorus is usually present either as a phosphate ester or as a phosphonate. Being esters they have many sites which are vulnerable to hydrolysis. The principal reactions involved are hydrolysis, oxidation, alkylation and dealkylation (Singh et al., 1999). Microbial degradation through hydrolysis of P-O-alkyl and P-O-aryl bonds is considered the most significant step in detoxification (Fig. 1). Both co-metabolic and bio-mineralization of organophosphorus compounds by isolated bacteria have been reported. A list of micro-organisms capable of degrading these compounds is presented in Table 2.
Table 1. History, toxicity and half-life of some organophosphorus pesticides
Year of introduction
Mammalian LD50 (mg kg−1)
Half-life soil (days)
Table 2. Microorganisms isolated for the degradation of organophosphorus compounds
Mode of degradation
Symbol in brackets after mode of degradation represents the type of nutrient that the pesticide provides to degrading microorganisms. C, carbon; N, nitrogen; P, phosphorus.
Hydrolysis of organophosphorus compounds leads to a reduction in their mammalian toxicity by several orders of magnitude. Since most of the research has been directed towards detoxification, studies on the further metabolism of the phosphorus containing products have not been extensive. Hypothetical phospho-ester hydrolysis steps can be postulated, yielding mono-ester and finally inorganic phosphate, but this pathway has not been specifically studied. Analogous phospho-monoesterase and diesterase, which degrade methyl and dimethyl phosphate, respectively, have been reported in Klebsiella aerogenes (Wolfenden & Spence, 1967) and are produced only in the absence of inorganic phosphate from the growth medium. The final enzyme in the postulated degradative pathway is bacterial alkaline phosphatase, which can hydrolyze simple monoalkyl phosphates and is also regulated by the level of phosphate available to the cell (Wolfenden & Spence, 1967). A similar mechanism of metabolism has been reported for phosphonates (Kertesz et al., 1994a). The way in which metabolism is regulated depends very strongly on what role the organophosphorus compound plays for the particular organisms studied. Most often these compounds are used to supply only a single element (carbon, phosphorus or sulfur) and the relevant gene cannot be expressed as a response to starvation for another of these elements (Kertesz et al., 1994a). For example, a strain of Pseudomonas stutzeri isolated to utilize parathion as a carbon source released the diethylphosphorothioanate products quantitatively and could not metabolize them further, even when alternative source of phosphorus or sulfur were removed (Daughton & Hsieh, 1977). Similarly, a variety of isolates that could use phosphorothionate and phosphorodithionate pesticides as a sole source of phosphorus were unable to utilize these compounds as a source of carbon (Rosenberg & Alexander, 1979). Shelton (1988) isolated a consortium that could use diethylthiophosphoric acid as a carbon source but was unable to utilize it as a source of phosphorus or sulfur. Kertesz et al. (1994a) explained possible underlying reasons for this phenomenon. They suggested that the conditions under which environmental isolates enriched were crucial in selecting for strains not only with the desired degradative enzyme systems but also with specific regulation mechanisms for the degradation pathways.
Chlorpyrifos (O,O-diethyl O-(3,5,6-trichloro-2-pyridyl) phosphorothioate) is one of the most widely used insecticides effective against a broad spectrum of insect pests of economically important crops. It is effective by contact, ingestion and vapour action but is not systemically active. It is used for the control of mosquitoes (larvae and adults), flies, various soil and many foliar crop pests and household pests. It is also used for ectoparasite control on cattle and sheep. It has low solubility in water (2 mg L−1) but is readily soluble in most organic solvents. It has a high soil sorption co-efficient (Racke, 1993) and is stable under normal storage conditions. Chlorpyrifos is defined as a moderately toxic compound having acute oral LD50; 135–163 mg kg−1 for rat and 500 mg kg−1 for guinea pig.
The environmental fate of chlorpyrifos has been studied extensively. Degradation in soil involves both chemical hydrolysis and microbial activity. The half-life of chlorpyrifos in soil varies from 10 to 120 days (Getzin, 1981; Racke et al., 1988) with 3,5,6-trichloro-2-pyridinol (TCP) as the major degradation product. This large variation in half-life has been attributed to different environmental factors, the most important of which are soil pH, temperature, moisture content, organic carbon content and pesticide formulation (Getzin, 1981a, b; Chapman & Chapman, 1986). Initially, the high rate of chlorpyrifos degradation in soils with alkaline pH was attributed to chemical hydrolysis. Later, Racke et al. (1996) concluded that the relationship between high soil pH and chemical hydrolysis was weak and that other factors like soil silt content might be important in determining environmental fate.
Unlike other organophosphorus compounds, chlorpyrifos has been reported to be resistant to the phenomenon of enhanced degradation (Racke et al., 1990). There have been no reports of enhanced degradation of chlorpyrifos since its first use in 1965 until recently. It was suggested that the accumulation of TCP, which has anti-microbial properties, acts as a buffer in the soil and prevents the proliferation of chlorpyrifos degrading microorganisms (Racke et al., 1990). However, Robertson et al. (1998) suggested that chemical hydrolysis of chlorpyrifos and enhanced degradation of TCP can result in loss of efficacy of the insecticide against termites in sugar cane fields in Australia. Attempts to introduce enhanced degradation in the laboratory or in the field by repeated application have failed (Racke et al., 1990; Mallick et al., 1999).
In recent experiments, we found that the degradation of chlorpyrifos was very slow in acidic soils but that the rate of degradation increased considerably with an increase in soil pH. However, in 90 days of incubation, there was no difference between soils in release of 14CO2 from the pyridine ring despite the large differences in degradation rate. Repeated applications of chlorpyrifos did not affect either the degradation rate or degradation kinetics, suggesting that repeated treatment did not result in enhanced degradation. Fumigation of soil samples completely inhibited hydrolysis of chlorpyrifos, suggesting an involvement of soil micro-organisms (Singh et al., 2003c). Chlorpyrifos has been reported previously to be resistant to enhanced degradation. Given the tremendous adaptability of the soil microbial community for degradation of a wide variety of synthetic compounds, Racke et al. (1990) cited three possible reasons why a specific pesticide might not be susceptible to enhanced degradation. One possibility is an inability of the microflora to initiate degradation of the parent pesticide easily. This may be due to factors such as steric hindrance of enzymes by functional groups, electronic stability against hydrolysis or lack of weak links in the molecule (Alexander, 1965; Niemi et al., 1987). The pesticide may also be unavailable for uptake and degradation by soil microorganisms due to strong sorption to organic surfaces in the soil (Orgam et al., 1985). However, these reasons cannot explain the present results because chlorpyrifos is rapidly hydrolyzed by the soil bacterial community in alkaline soils. The second possibility is that the soil environmental conditions may in some way inhibit the development or expression of enhanced degradation. This also cannot explain the present results because repeated treatment of the same soil samples resulted in enhanced degradation of fenamiphos (Singh et al., 2003b). A third possibility is that the soil micro-organisms cannot beneficially catabolize pesticide metabolites. In these circumstances co-metabolism may occur (e.g. hydrolysis of parent pesticides), but the microbial metabolism of the degradation products is not possible. This is the case with such relatively recalcitrant pesticides as DDT and alachlor, which are converted to products that are themselves quite resistant to further metabolism (Tiedje & Hagedorn, 1975). From our experiments we concluded that in high pH soils, the microbial community transforms chlorpyrifos co-metabolically into TCP. However, TCP contains three chlorine atoms on the pyridinol ring. To break this ring, chlorine atoms have to be removed (Feng et al., 1997), and free chlorine has toxic effects on the micro-organisms. Thus TCP metabolism may be toxic to micro-organisms. Similar results were obtained by Price et al. (2001) in a field where degradation of chlorpyrifos was strongly related with soil pH but degradation was mediated by soil micro-organisms. Later, Singh et al. (2003c) suggested that chlorpyrifos is degraded by non-specific and non-inducible enzyme systems produced in high pH soils. This suggests that chlorpyrifos is co-metabolically hydrolyzed to TCP and that because the TCP has toxic effects, normally enhanced degradation does not occur. Although Shelton & Doherty (1997) in their model proposed a significant role of bioavailability in degradation of xenobiotics, the toxic effect of TCP seems to be a realistic explanation of its resistance to enhanced degradation because TCP has high water solubility and therefore is bioavailable for the degradation. However, repeated treatment with chlorpyrifos over many years in an Australian soil resulted in development of some opportunist microorganisms with the capability to use the toxic compound as has been reported with organochlorine compounds (Robertson et al., 1998; Singh et al., 2000). This adaptation can provide them with a competitive advantage over other microbes in terms of sources of energy. Further studies found higher copy numbers of opd (organophosphate degrading) gene in higher pH soils (Singh et al., 2003a, c).
In most cases described to date, the aerobic bacteria tend to transform chlorpyrifos by hydrolysis to produce diethylthiophosphoric acid (DETP) and TCP, which in turn accumulate in the culture medium without further metabolism. This transformation reaction removes chlorpyrifos and its mammalian toxicity but yields compounds that are not metabolized by the microorganisms that produce them (Richins et al., 1997; Mallick et al., 1999; Horne et al., 2002b; Wang et al., 2002b).
Chlorpyrifos has been reported to be degraded co-metabolically in liquid media by Flavobacterium sp. and Pseudomonas diminuta, which were initially isolated from a diazinon treated field and by parathion enrichment, respectively (Sethunathan & Yoshida, 1973; Serdar et al., 1982). However, these microbes do not utilize chlorpyrifos as a source of carbon. A Micrococcus sp. was isolated from a malathion enriched soil which was later reported to degrade chlorpyrifos in liquid media (Guha et al., 1997). We have isolated an Enterobacter sp. from a soil from Australia showing enhanced degradation of chlorpyrifos. This bacterium degrades chlorpyrifos to DETP and TCP and utilizes DETP as a source of carbon and phosphorus (Singh et al., 2003c, 2004). Cook et al. (1978a) isolated several bacteria from sewage sludge that were able to use dialkylthiophosphonic acid as a sole source of phosphorus. One of these organisms, Pseudomonas acidovorans, was able to use DETP as a sole source of sulfur (Cook et al., 1980). Another significant observation was the utilization of organophosphorus insecticides as a source of phosphorus by Enterobacter sp. (Singh et al., 2003c, 2004). Sethunathan & Yoshida (1973) isolated a Flavobacterium sp. that could use diazinon as a source of carbon. However, Flavobacterium was not able to use other organophosphorus pesticides as a source of either phosphorus or carbon. Similarly, a variety of isolates that could use phosphorothionate or phosphorodithionate compounds as a sole source of phosphorus were unable to degrade these compounds as a source of carbon (Rosenberg & Alexander, 1979). Shelton (1988) isolated a consortium that could use DETP as a carbon source but was unable to degrade it when presented as source of phosphorus or sulfur. It is believed that the conditions under which environmental isolates are enriched are crucial in selecting for strains not only with the desired degradative enzymes systems, but also with the specific regulation mechanisms for the degradation pathways (Kertesz et al., 1994a).
Studies on further metabolism and identification of intermediate products of the phosphorus containing products have not been extensive. The postulated pathway steps include hydrolysis, yielding monoester and finally inorganic phosphate (Fig. 2). Bacterial phosphodiesterase has been purified from a wide range of organisms including Escherichia coli (Imamura et al., 1996), Haemophilus influenzae (Macfadyen et al., 1988), and Burkholderia caryophylli PG2982 (Dotson et al., 1996). The phosphodiesterase from the first two bacteria are similar in sequence and both moderate intracellular cyclic AMP levels. However, the phosphodiesterase from B. caryophilli has a different sequence from that in the first two bacteria (Dotson et al., 1996). This enzyme could not be assigned a clear function but was thought to play a role in xenobiotic degradation pathways because it degraded glycerol glyphosate. However, until recently no phosphodiesterase had been isolated or characterized which could utilize xenobiotic degradation products such as diethyl phosphate and diethyl phosphonate. A novel phosphodiesterase was isolated and cloned from Delftia acidovorans which has both mono- and di-esterase activity (Tehara & Keasling, 2003). This enzyme allows D. acidovorans to use diethyl phosphonate as a sole source of phosphorus under phosphorus limiting conditions. The final enzyme in the postulated degradative pathway is alkaline phosphatase, which can hydrolyze simple monoalkyl phosphates (Neidhardt et al., 1996).
Since only one bacterium has been isolated so far which can degrade TCP in liquid medium, little literature is available on microbial metabolism of TCP. Feng et al. (1997) isolated a Pseudomonas sp. which can mineralize TCP in liquid medium. Later the same group, on the basis of combined experiments with photolysis and microbial degradation, suggested that TCP was metabolized by a Pseudomonas sp. by a reductive dechlorination pathway (Feng et al., 1998). In this pathway, TCP is first reductively dechlorinated into chlorodihydro-2-pyridone, which is further dechlorinated to tetra-hydro-2-pyridone. Ring cleavage of this compound resulted in formation of maleamide semialdehyde, which is metabolized to water, carbon dioxide, and ammonium ions. Microbial degradation of analogous compounds such as pyridine and hydroxypyridine has been researched and reviewed extensively (Shukla, 1984; Sims & O'Loughlin, 1989; Kaiser et al., 1996). Several micro-organisms were reported to degrade hydroxypyridine (Kaiser et al., 1996). Cain et al. (1974) reported that 2- or 3-hydroxypyridine was oxidized to 2,5-dihydroxypyridine and production of maleamic acid occurred later through ring cleavage. Oxygen atoms used to transform 4-hydroxypyridine via 3,4-dihydroxypyridine were derived from water molecules by hydroxypyridine hydrolase (Watson et al., 1974). It is likely that TCP is metabolized in a similar manner as one of the metabolites of TCP was identified to have similar structure to 2-hydroxypyridine.
Fungal mineralization of chlorpyrifos by Phanerochaete chrysosporium was reported by Bumpus et al. (1993). Chlorpyrifos was hydrolyzed and then the pyridinyl ring underwent cleavage before being converted to carbon dioxide and water. Degradation of chlorpyrifos in ‘biobed’ composting substrate by two other white-rot fungi, Hypholoma fascicularae and Coriolus versicolor, was observed (Bending et al., 2002). Degradation of a wide range of xenobiotic compounds by white-rot fungi is well documented (Kuhad et al., 1997; Singh & Kuhad, 1999, 2000; Singh et al., 1999). These organisms have been reported to degrade several persistent aromatic compounds by ring cleavage (Armenante et al., 1994; Reddy & Gold, 2000). The multi-step pathway of pentachlorphenol degradation by the white-rot fungus Phanerochaete chrysosporium is initiated by lignin peroxidase and manganese peroxidase, producing tetrachloro-1-4-benzoquinone, which is further metabolized by two parallel but cross-linked pathways. The tetrachlorobenzoquinone is reduced to tetrachlorodihydroxybenzene, which can undergo four successive dechlorinations to produce 1,4-hydroquinone. This is then hydroxylated to produce the final aromatic metabolite, 1,2,4-trihydroxybenzene. Alternatively the tetrachlorobenzoquinone converts to 2,3,5-trichlorotrihydroxybenzene, which undergoes successive reductive dechlorination to produce 1,2,4-trihydroxybenzene. At several points, hydroxylation reaction converts chlorinated dihydroxybenzene to chlorinated trihydroxybenzene, linking two pathways. The 1,2,4-trihydroxybenzene is ring cleaved to produce CO2 and water (Reddy & Gold, 2000). Mineralization of TCP by white-rot fungi is possible via reductive de-chlorination. White-rot fungi have been reported previously to use this transformation step to degrade other chlorinated compounds such as pentachlorophenol (Aiken & Logan, 1996) and hexachlorocyclohexane (Mougin et al., 1996; Singh & Kuhad, 1999, 2000). Degradation of several polychlorinated compounds by white-rot fungi suggests that they produce a range of isoenzymes with a wide range of substrate specificity. Several species of Aspergillus, Trichoderma harzianum and Penicillium brevicompactum were reported to utilize chlorpyrifos as sources of phosphorus and sulfur (Omar, 1998) (Table 1). On the basis of the above discussion, the authors propose possible pathways for microbial degradation of chlorpyrifos (Fig. 2).
Parathion (O,O-diethyl-O-p-nitrophenyl phosphorothioate) is one of the most toxic insecticides registered with the US Environmental Protection Agency (EPA). Extreme toxicity with ease of exposure has resulted in numerous human and non-target species deaths in several developing countries (McConnell et al., 1999). The microbial degradation of parathion has received extensive attention among the organophosphorus compounds because of its widespread use and the ready detection of its hydrolytic product (p-nitrophenol). Parathion is rapidly degraded in biologically active soil. A proportional increase in the bacterial population in soils was observed with an increase in the concentration of parathion added (Nelson, 1982). Flooded soil conditions favoured hydrolysis of parathion and release of 14CO2 from ring labelled parathion in the rhizosphere of rice seedlings (Reddy & Sethunathan, 1983).
Several species of Bacillus and Arthrobacter have been isolated that were capable of hydrolyzing parathion; one of the Arthrobacter strains was also able to utilize p-nitrophenol as a sole source of carbon (Nelson, 1982). A Pseudomonas sp. and a Xanthomonas sp. were isolated which can hydrolyze parathion and can further metabolize p-nitrophenol (Tchelet et al., 1993). A Moraxella sp. can use p-nitrophenol as the sole source of carbon and nitrogen (Spain & Gibson, 1991). This bacterium degrades p-nitrophenol to p-benzoquinone using the enzyme p-nitrophenol monooxygenase. p-Benzoquinone is transformed to hydroquinone by a reductase (Spain & Gibson, 1991). Candida parapsilosis has been reported to produce hydroquinine 1,2-dioxygenase, which converts hydroquinone to cis,trans-4-hydroxymuconic semialdehyde. This is then metabolized to maleylacetate by semialdehyde dehydrogenase. Maleylacetate is converted to 3-oxoadipate by a reductase, which is finally metabolized to intermediary metabolites of the tricarboxylic acid (TCA) cycle (Carnett, 2002). A Pseudomonas putida strain was found to metabolize p-nitrophenol to hydroquinone and 1,2,4-benzenetriol, which was further cleaved by benzenetriol oxygenase to maleylacetate (Rani & Lalitha-kumari, 1994). A similar pathway of p-nitrophenol degradation was reported in Pseudomonas cepacia that can utilize p-nitrophenol as a source of carbon and nitrogen (Prakash et al., 1996).
A different pathway of degradation was reported in Arthrobacter sp. strain JS443 and Arthrobacter protophormiae RHJ100 where p-nitrophenol was mineralized via p-nitrocatechol. Nitrocatechol is converted to 1,2,4-benzenetriol by benzotriol dehydrogenase, which in turn is directly converted to maleylacetate by benzotriol dioxygenase (Jain et al., 1994; Bhushan et al., 2000a; Chauhan et al., 2000). Recently, a consortium of two Pseudomonas ssp. (strains S1 and S2) was isolated which can also metabolize p-nitrophenol via p-nitrocatechol (Qureshi & Purohit, 2002). The analogous compound 3-methyl-4-nitrophenol has also been reported to be metabolized by Ralstonia sp. via catechol formation (Bhushan et al., 2000b). A Nocardia sp. was reported to produce p-nitrophenol-2-hydroxylase, which catalyzes transformation of p-nitrophenol to p-nitrocatechol (Mitra & Vaidyanathan, 1984). A mono-oxygenase from a Moraxella sp. that releases nitrite from p-nitrophenol has been partially purified (Spain & Gibson, 1991). A soluble nitrophenol oxygenase was purified from P. putida B2 that converts ortho-nitrophenol to catechol and nitrite (Zeyer & Kocher, 1988). A novel monooxygenase was characterized from Bacillus sphaericus that catalyzes the first two steps of the degradation of p-nitrophenol via p-nitrocatechol and benzotriol. This enzyme consists of two components, a reductase and oxygenase, and catalyzes two sequential mono-oxygenation reactions that convert p-nitrophenol to benzotriol. The first reaction converts p-nitrophenol to p-nitrocatechol and the second removes the nitro group (Kadiyala & Spain, 1998). A pentachlorophenol degrading Sphingomonas sp. UG30 was found to degrade p-nitrophenol. A pentachlorophenol-monooxygenase was purified from this bacterium that can catalyze the hydroxylation of p-nitrocatechol to benzotriol (Leung et al., 1999). A hydroxyquinol (benzotriol) ring cleavage dioxygenase was isolated and characterized from p-nitrophenol degrading Arthrobacter sp. strain JS443. The gene encoding this dioxygenase (npdB) was found to be in the same gene cluster as reductase (npdA1) and oxygenase (npdA2) components of the p-nitrophenol mono-oxygenase, maleylacetate reductase (npdC), and a regulatory protein (npdR) (Zylstra et al., 2000; Parales et al., 2002). Rhodococcus strain PN1 and Rhodococcus erythropolis HL PM-1, which degrade 2,4-dinitrophenol and p-nitrophenol, were reported to contain an npd gene cluster including npdC (encoding hydride transferase I), npdG (encoding the NADPH-dependent F420 reductase) and npdI (encoding hydride transferase II). It was observed that npdG and npdI genes have the same function as the homologous genes (Heiss et al., 2003). Recently, a novel gene called orf243 was reported from Flavobacterium sp. orf243 which is transposon based and is linked with the opd gene (Siddavattam et al., 2003). This gene encodes a protein with homology to a family of aromatic compound hydrolases and is able to degrade p-nitrophenol.
Although in most of the studies on microbial degradation of parathion, the first reaction was hydrolysis of the phosphotriester bond, there have been reports of different degradation pathways. In one study, degradation of parathion by a mixed culture and a Bacillus sp. (Sharmila et al., 1989) was shown to occur by reduction of the nitrogroup that was later hydrolyzed to p-aminophenol. Another report of conversion of parathion to paraoxon before hydrolysis of phosphotriester bond was reported in a mixed bacterial culture (Tomlin, 2000).
Studies on the degradation of methyl parathion (O,O-dimethyl-O-p-nitrophenyl phosphorothioate) have also been reported. Methyl and ethyl parathion have identical chemical structures except for the ethyl groups of the P chain of parathion, which are replaced by methyl groups as evident by the name of the compound. A Pseudomonas sp. was isolated that can co-metabolically degrade methyl parathion (Chaudry et al., 1988). Rani & Lalitha-kumari (1994) isolated P. putida that could hydrolyze methyl parathion and utilize p-nitrophenol as a source of energy. A Bacillus sp. was reported to degrade methyl parathion by both hydrolysis and nitro group reduction (Sharmila et al., 1989). Utilization of methyl parathion by Flavobacterium balustinum as the sole source of carbon was observed earlier (Somara & Siddavattam, 1995). In this bacterium the opd gene was found to be linked with a novel gene involved in degradation of p-nitrophenol (Siddavattam et al., 2003). Degradation of methyl parathion by a Pseudomonas sp. in soil and on sodium alginate beads was reported (Ramanathan & Lalithakumari, 1996). Co-metabolic degradation of methyl parathion by Plesimonas sp. strain M6 was observed (Zhongli et al., 2001) which was mediated by a novel degrading gene. They also isolated Pseudomonas sp. A3 which can utilize p-nitrophenol as sole source of carbon and nitrogen. This isolate can also utilize a series of aromatic compounds as a sole source of carbon (Zhongli et al., 2002). Another strain of Pseudomonas sp. WBC was isolated from polluted soils around a Chinese pesticide factory. The isolate was capable of complete degradation of methyl parathion and could utilize it as sole source of carbon and nitrogen (Yali et al., 2002). The hydrolysis product of methyl parathion is also p-nitrophenol, for which the degradation pathways have already been described. The different proposed pathways of parathion and methyl degradation are presented in Fig. 3.
Glyphosate (N-(phosphonomethyl) glycine) is a globally used broad-spectrum herbicide. It is a representative of the phosphonic acid group of compounds, which is characterized by a direct carbon to phosphorus (C–P) bond. The C–P linkage is chemically and thermally very stable and renders the molecule much more resistant to non-biological degradation in the environment than its analogues with O-P linkage (Hayes et al., 2000). Mode of action of glyphosate includes inhibition of the plant enzyme 5-enol-pyruvyl-shikimate-3-phosphate synthase, which catalyzes synthesis of aromatic amino acids (Fisher et al., 1984; Cole, 1985). Glyphosate is moderately persistent with a half-life of 30–170 days (Tomlin, 2000). Microbial degradation is considered to be the most important of the transformation processes controlling its persistence in soil (Araujo et al., 2003). It was observed that mineralization of glyphosate is related to both the activity and biomass of soil micro-organisms (Wiren-Lehr et al., 1997). Microbial degradation of glyphosate produces the major metabolite aminomethyl phosphonic acid and ultimately leads to the production of CO2, phosphate and water (Forlani et al., 1999; Araujo et al., 2003). Several species of bacteria have been isolated from previously treated and untreated environments, which can degrade glyphosate either co-metabolically or as a source of phosphorus. There has been no report of the utilization of glyphosate as a source of carbon or nitrogen (Dick & Quinn, 1995). Several species of Pseudomonas have been isolated which can degrade glyphosate (Moore et al., 1983; Tolbot et al., 1984; Jacob et al., 1988; Quinn et al., 1989). Similarly, a Flavobacterium sp. (Balthazor & Hallas, 1986), an Alcaligenes sp. (Tolbot et al., 1984), Bacillus megaterium strain 2BLW (Quinn et al., 1989), several species of Rhizobium (Liu et al., 1991), three species of Agrobacterium (Wacket et al., 1987; Liu et al., 1991) and an Arthrobacter sp. (Pipke et al., 1987) have also been reported to degrade this herbicide (Table 2).
Three different pathways for C–P bond cleavage have been reported for the use of phosphonate as a source of phosphorus for growth.
The phosphonatase pathway is involved in degradation of alpha carbon substituted phosphonates, which are primarily naturally occurring phosphonates such as 2-aminoethylphosphonates that have been reported in Bacillus cereus (Lee et al., 1992b), and Pseudomonas aeruginosa (Lacoste et al., 1993), Salmonella typhimurium and several other organisms (Jiang et al., 1995). In a two-step process, this pathway leads to the cleavage of the C–P bond by a hydrolysis reaction requiring an adjacent carbonyl group. 2-Aminoethylphosphonate is converted to phosphonoacetaldehyde by a specific transaminase, which is further degraded to acetaldehyde by phosphonatase. The C–P lyase pathway is involved in the cleavage of both substituted and unsubstituted phosphonates such as methylphosphonates (Lee et al., 1992b).The phosphonoacetate hydrolase pathway specifically degrades phosphonoacetate and appears to have evolved for phosphonate use as a carbon source. This enzyme catalyzes the hydrolysis of phosphonoacetate; to acetate and inorganic phosphonates via metal cation-assisted P–C bond cleavage (McMullan & Quinn, 1994; McGrath et al., 1995). Glyphosate has been found to be degraded by the second of these pathways.
Two different pathways of glyphosate degradation are presented in Fig. 4. Arthrobacter sp. GLP-1 and Pseudomonas sp. PG2982 degrade glyphosate by initial cleavage of the C–P bond, resulting in the production of sarcosine (N-methylglycine) by C–P lyase activity (Moore et al., 1983; Shinabarger & Braymer, 1984; Pipke et al., 1987; Liu et al., 1991; Dick & Quinn, 1995). Rhizobium meliloti has also been reported to degrade glyphosate by this pathway but, unlike other bacteria, it has only one C–P lyase, which is able to degrade a wide range of phosphonates (Park & Hausinger, 1995). The sarcosine formed is further degraded to the amino acid glycine and a C1-unit, which is incorporated into purines, and the amino acids serine, cysteine, methionine and histidine (Pipke et al., 1987). The second pathway involves the conversion of glyphosate to aminomethylphosphonic acid (AMPA) by the loss of a C2 unit. This compound is then dephosphorylated by C–P lyase and further broken down by subsequent steps to methylamine and formaldehyde (Pike & Amrhein, 1988; Lerbs et al., 1990). An identical pathway has been observed in Arthrobacter atrocyaneus (Pike & Amrhein, 1988) and Flavobacterium sp. (Balthazor & Hallas, 1986; Pipke et al., 1987). Recently, a thermophile, Geobacillus caldoxylosilyticus T20 was isolated from a central heating system which also degrades glyphosate by this pathway, utilizing the compound as a sole source of phosphorus (Obojska et al., 2002). A halophilic bacterium, Chromohalobacter marismortui, isolated from soil beneath a road gritting salt pile was capable of utilizing several organophosphonates including aminomethyl phosphonic acid as a source of phosphorus (Hayes et al., 2000). Utilization of aminoalkylphosphonates as a source of nitrogen by different bacterial isolates has been reported (McMullan & Quinn, 1994; Ternana & McMullan, 2000). Pseudomonas fluorescens was reported to utilize a diverse range of organophosphonates as sources of carbon, nitrogen and phosphorus (Zboinska et al., 1992a). A strain of Kluyveromyces fragilis has been shown to utilize AMPA as a source of nitrogen (Ternana & McMullan, 2000). Strains of Streptomyces were also reported to degrade and utilize several organophosphonate compounds as sources of carbon and nitrogen. These strains were capable of degrading glyphosate in phosphate-free media via C–P bond cleavage accompanied by sarcosine formation (Obojska et al., 1999). Streptomyces morookaensis DSM 40565 could degrade aminoalkylphosphonate as a sole source of nitrogen and phosphorus (Obojska & Lejczak, 2003). Alkyl amines are intermediate degradation products for several xenobiotics such as carbofuran, atrazine, and monocrotophos and have been reported to serve as a source of energy for different micro-organisms (Strong et al., 2002). Use of methylamine as a source of carbon is widespread in nature (Hanson & Hanson, 1996; Trabue et al., 2001).
Fungi play an important role in degradation of xenobiotics and biospheres (Pothuluri et al., 1998, 1992) including glyphosate. Probably the first fungal degradation of glyphosate by Penicillium citrinum was reported by Zboinska et al. (1992b). Penicillium notatum can utilize the herbicide as a source of phosphorus and can degrade it by the aminomethyl phosphonic acid pathway (Bujacz et al., 1995). Strains of Trichoderma harzianum, Scopulariopsis spand and Aspergillus niger were able to degrade glyphosate and aminomethyl phosphonic acid in the laboratory (Krzysko-Lupicka et al., 1997). The first report of utilization of glyphosate as a source of nitrogen by a microorganism was reported for Penicillium chrysogenum (Klimek et al., 2001). The fungal cells were found to lack detectable nitrogen reductase activity and therefore this isolate seemed to lack the ability to convert nitrate to ammonium. Recently, Alternaria alternata, a plant pathogen, was found to utilize glyphosate as a source of nitrogen (Lipok et al., 2003).
The above observations suggest that glyphosate is degraded by several soil microorganisms, and different steps of the degradation involve different microorganisms which utilize different degradation products as different sources of energy. The possible pathways of glyphosate degradation are presented in Fig. 4.
Coumaphos (O,O-diethyl-O-(3-chloro-4-methyl-2-oxo-2H-1-benzo-pyran-7-yl) phosphorothioate) is used as an acaricide for the control of cattle ticks. It is widely used by different government agencies for tick eradication and quarantine purposes. The primary tool used in the eradication programme is a series of dipping vats placed at border crossing points. The cattle are induced to jump into the deep end of the vat, resulting in their complete immersion in coumaphos. They then swim the length of the vat and climb out to other end. There are around 42 vats in the USA alone and each vat contains about 15 000 L of coumaphos suspension at the rate of 1600 mg L−1 (42% active ingredient, a.i.) (Shelton & Somich, 1988; Mulbry et al., 1998). The vats are cleaned and recharged every 2 years to keep the concentration of acaricide at a desirable level. These operations generate approximately 460 000 L of concentrated insecticide waste yearly in USA alone (Mulbry et al., 1996). A similar programme within Mexico is thought to produce a much larger volume. Coumaphos is comparatively persistent in soil, with a half-life of about 300 days (Kearney et al., 1986) and it possesses a very high mammalian toxicity. Because of these characteristics, it requires a safe and effective method for disposal. Rapid degradation of coumaphos was observed in several cattle-dipping vats, resulting in loss of efficacy against cattle ticks (Shelton & Karns, 1988). Under aerobic conditions, experiments with radiolabelled coumaphos demonstrated that the aromatic portion of the molecule is susceptible to mineralization by bacteria in problematic vat dips (loss of efficacy). Three morphologically distinct bacteria (designated B-1, B-2 and B-3) that could metabolize coumaphos were isolated from a problem vat dip (Shelton & Somich, 1988). All these bacteria hydrolyzed coumaphos to DETP and chlorferon. Chlorferon was further metabolized by B-1 and B-2 to α-chloro-β-methyl-2,3,4-trihydroxy-trans-cinnamic acid (CMTC). Further experiments demonstrated that B-1 was capable of mineralizing and incorporating the aromatic portion of the coumaphos molecule into biomass, but this was inhibited by the accumulation of metabolites that was due apparently to the inefficient metabolism of a chlorinated intermediate. Combination of B-1 with another organism from the vat, designated strain B-4, which metabolized these inhibitory products, yielded a stable two-member consortium able to grow at the expense of coumaphos (Shelton & Haperman-Somich, 1991). No further study on the degradation pathway or metabolite identification has been carried out.
Ralstonia sp. LD35 has been reported to degrade an analogous compound, 3,4-dihydroxycinnamic acid via benzoic acid (Gioia et al., 2001). A similar breakdown pathway for the propenoic side chain of substituted cinnamic acid molecule, p-coumaric acid, has been observed in Pseudomonas sp. (Tse et al., 2004) and Acinetobacter strains (Delneri et al., 1995). These bacteria use p-coumaric acid as the source of carbon. In the first step, they convert p-coumaric acid into p-hydroxybenzoic acid which is then transformed to protocatechuic acid and integrated to the TCA cycle via the β-ketodipate pathway. Many bacteria degrade substituted cinnamic acid by decarboxylation of side chains. Enzymes and genes responsible for such degradation have been purified and characterized (Degrassi et al., 1995; Barthelmebs et al., 2000). Streptomyces setonii (Sutherland et al., 1983) and Rhodopseudomonas palustris (Harwood & Gibson, 1988) have been shown to degrade cinnamic and 4-coumaric acids to their corresponding benzoic acid derivatives. Several other bacteria follow the same pathway for degradation of substituted cinnamic acids. Monooxygenase and dioxygenase catalyze the formation of the 2-, 3-, and 4-hydroxy derivatives as substituted acid and/or substituted catechol (Peng et al., 2003).
The β-oxidation pathway has been proposed for the degradation of substituted cinnamic acids by Pseudomonas putida (Zenk et al., 1980). This pathway, which is analogous to the β-oxidation of fatty acids, is thought to include thiolytic cleavage of 4-hydroxy-3-methoxy-β-ketopropinyl-CoA to yield acetyl CoA and vanillyl CoA, which is catalyzed by β-ketoacyl CoA thiolase. The pathway subsequently leads to ring fission and requires several co-factors including ATP, CoA and NAD+ (Zenk et al., 1980). Under anaerobic conditions, coumaphos undergoes reductive dechlorination to form potasan (Mulbry et al., 1998).
Nocardia sp. strain B-1 was reported to degrade coumaphos by a different gene enzyme system to the known opd gene (Mulbry, 1992). Another microorganism, Nocardiodes simplex NRRL B-24074, was found to have a distinct enzymes system for coumaphos degradation (Mulbry, 2000). Horne et al. (2002b) isolated an Agrobacterium radiobacter P230 capable of hydrolyzing coumaphos from an enrichment culture containing organophosphorus as the sole source of phosphorus. This bacterium degrades coumaphos by hydrolysis of the phosphotriester bond. Pseudomonas monteilli was isolated which can hydrolyze coumaphos as well as its oxo analogue coroxon but it can utilize only coroxon as a sole source of phosphorus, not coumaphos or its hydrolysis product DETP. This bacterium degrades coumaphos and diazinon but not parathion (Horne et al., 2002a). Coumaphos is degraded by the other microorganisms like Flavobacterium sp. (Sethunathan & Yoshida, 1973), P. diminuta (Serdar et al., 1982), and Enterobacter sp. B-14 (Singh et al., 2004), which were isolated for their ability to degrade other organophosphorus compounds. This observation suggests that these microorganisms produce several isoenzymes or broad-specificity enzymes that can degrade a range of organophosphorus compounds. The proposed pathway of microbial degradation of coumaphos is shown in Fig. 5.
Fenamiphos (ethyl 4-methylthio-m-tolyl isopropylphosphoramidate) is an organophosphorate used extensively for the control of soil nematodes. It is systemic, active against ecto- and endo-parasitic, cyst forming and root-knot nematodes, and is recommended for application at 5–20 kg a.i.ha-1. Its solubility at room temperature is 700 mg L−1 water. The acute oral LD50 is 15.3–19.4mg kg−1 for rats, 10 mg kg−1 for dogs and 75–100 mg kg−1 for guinea pigs (Tomlin, 2000).
Although, there have been reports of enhanced degradation of fenamiphos, the mechanism of degradation has received little attention. Fenamiphos is oxidized rapidly to fenamiphos sulfoxide (FSO) which in turn is oxidized to fenamiphos sulfone (FSO2). As FSO and FSO2, have nematicidal activity and toxicity similar to fenamiphos (Waggoner & Khasawinah, 1974), degradation and persistence studies usually include estimation of total toxic residue, which is the combination of the two oxidation products along with parent compound. The half-life in soil for fenamiphos and its metabolites (total toxic residues) varies from 30 days to 90 days (Johnson, 1998). More rapid rates of degradation in soil repeatedly treated with the fenamiphos in the laboratory have been reported (Chung & Ou, 1996) and enhanced degradation of fenamiphos in the field has been observed in many countries (Stiriling et al., 1992; Smelt et al., 1996; Meghraj et al., 1999). It was suggested that 3–4 years were necessary before the accelerated degradation of fenamiphos declined in a sandy soil in a temperate region (Ou, 1991).
Fenamiphos rapidly disappears from both enhanced and non-enhanced soils but FSO2 is rarely formed in enhanced soils (Ou, 1991). This suggests that enhanced bio-degradation of fenamiphos total toxic residue was due to an increase in the disappearance rate of FSO in soil samples collected from field sites treated one or two consecutive times with fenamiphos (Davis et al., 1993). In a recent study of soil samples from a field in the UK, which had similar physical characteristics except for soil pH, the degradation rate of fenamiphos increased with the increase in pH. Repeated application of fenamiphos slowed down the rate of degradation in acidic soils, and in the neutral pH soil, three consecutive treatments did not result in the development of enhanced degradation of fenamiphos. However, in the two alkaline soils, a second treatment with fenamiphos led to enhanced degradation (Singh et al., 2003b). Chung & Ou (1996) have tried to shed light on the mechanism of fenamiphos degradation in soils that showed enhanced degradation. They reported that fenamiphos is degraded into FSO which in turn is rapidly degraded into FSO-phenol, which is subsequently mineralized into CO2. Therefore in enhanced soil, degradation of fenamiphos (total toxic residue) is rapid because it misses one step, FSO to FSO2. In enhanced UK soils, fenamiphos was rapidly oxidized to FSO, which in turn, was quickly degraded. The major fenamiphos metabolites identified were FSO and FSO-phenol. No FSO2 was detected in the enhanced soil samples (Singh et al., 2003b). However, in two Australian soils, a different mechanism of fenamiphos degradation was observed where the nematicide was directly converted to fenamiphos phenol, suggesting that the first oxidation step was replaced by hydrolysis (Singh et al., 2003b).
Ou & Thomas (1994) isolated the first microbial consortium with six different bacterial species that degraded fenamiphos in liquid culture. A pure culture of Brevibacterium sp. MM1 was isolated which hydrolyzed fenamiphos and its hydrolysis products but did not utilize these chemicals as energy sources (Megharaj et al., 2003). Two different consortia from Australian soils, made up of five and four different bacterial strains, were isolated [B. K. Singh, unpublished]. Both consortia could utilize fenamiphos as sole sources of carbon and nitrogen. In contrast to the consortium isolated by Ou & Thomas (1994), the two Australian consortia (CRF and BEP) did not require any supplementary nutrient source for fenamiphos degradation and were active in liquid media in the absence of mineral surfaces (Singh et al., 2003b). These microbial systems were found to mineralize fenamiphos or its oxidative metabolites by hydrolysis as a first step. The hydrolytic product fenamiphos phenol, FSO-phenol or fenamiphos sulfone phenol (FSO2-OH) can be further degraded by desulfonation. Three modes of desulfonation are reported for aromatic sulfonates: desulfonation (a) before, (b) during or (c) after ring cleavage (Kertesz et al., 1994a). Mode (a) is considered to be most common pathway of desulfonation in the environment. In this pathway, the target compound is oxygenated by a multi-component oxygenase, yielding an unstable sulfono cis-diol, which then spontaneously re-aromatizes to the corresponding catechol with the loss of sulfite. An enzyme which catalyzes this reaction in toluene sulfonate and benzene sulfonate has been isolated from an Alcaligenes sp. (Thurnheer et al., 1986, 1990). In Pseudomonas putida S-313, a broad-spectrum monooxygenolytic sulfonatase catalyzes the conversion of sulfonate to a phenol with incorporation of one oxygen atom from molecular oxygen (Kertesz et al., 1994b). Alcaligenes sp. strain O-1 is reported to contain two different desulfonative pathways where the initial desulfonation is catalyzed by different dioxygenase enzyme systems. One enzyme system can degrade 2-aminobenzenesulfonate, benzene sulfonate and 4-toluene sulfonate but the other one can degrade only the last two compounds (Junker et al., 1994). Hydrogenophaga palleronii S1 has been reported to degrade 4-carbo-4-sulfoazobenzene by the 4-sulfocatechol pathway via the formation of 4-aminobenzenesulfate (Vickers, 2002). Another proposed pathway is transformation of toluene sulfonate to hydroxy toluene by toluenesulfonate monooxygenase. Pseudomonas putida strain S-313 catalyzes toluene sulfonate desulfonation, which can serve as its sole source of sulfur and leaves 4-hydroxytoluene unmetabolized. However 4-hydroxy toluene is a metabolite that is readily catabolized by other bacteria via the toluene pathway (Eisenmaan & McLeish, 2002).
Another toluene sulfonate degrading bacterium, Comamonas testosteroni T-2, was found to contain a degrading gene on a plasmid (Hooper et al., 1990). Simple alkane sulfonates are utilized by Pseudomonas sp. as a carbon source where crude cell extract catalyzes the oxidation of the α-carbon atom of alkanesulfonate to an aldehyde bisulfite adduct. This adduct then degrades to produce the corresponding aldehyde and sulfite. The substrate range for this reaction has been reported to be relatively broad where hydroxy-, methyl-, and alkenyl-substituted compounds are all transformed (Thysse & Wanders, 1974). Degradation of alkylsulfate proceeds via initial hydrolysis of the sulfate ester linkage and subsequent oxidation of the released alkanol (Kertesz et al., 1994a). Pseudomonas sp. C12B and a strain of Comamonas terrigena were reported to utilize a range of alkylsulfates as a source of carbon (Payne & Faisal, 1963; Fitzgerald et al., 1977). Five different alkylsulfatases were characterized from Pseudomonas sp. C12B and two from C. terrigena (Dodgson et al., 1982). On the basis of the above studies, we propose the microbial degradation pathways for fenamiphos as presented in Fig. 6.
Other organophosphorus pesticides
Several other organophosphorus compounds have been used extensively for pest control. Diazinon, monocrotophos, malathion, dimethoate, etc., are being used world-wide. Several species of bacteria have been isolated and characterized that can degrade these compounds in liquid medium and soils (Table 2).
Monocrotophos ((3-hydroxy-N-methyl-cis-crotonamide) dimethyl phosphate) is widely used to control aphids, leaf hoppers, mites and other foliage pests. It has been classified as extremely hazardous, with an LD50 value of 20 mg kg−1 for mammals. The half-life of monocrotophos in soil was reported to be 40–60 days (Tomlin, 2000). Monocrotophos is easily soluble in water and therefore has potential to contaminate ground water. Together with its high mammalian toxicity, these characteristics make monocrotophos an ideal compound for decontamination and detoxification. Rangaswamy & Venkateswaralu (1992) isolated a monocrotophos degrading Bacillus sp. from previously treated soil. Megharaj et al. (1987) isolated monocrotophos degrading algae from soil. Two different algae, Aulosira fertilissima ARM 68 and Nostoc muscorum ARM 221, were found to utilize monocrotophos as a sole source of phosphorus (Subramanian et al., 1994). Pseudomonas aeruginosa F10B and Clavibacter michiganense ssp. insidiosum SBL 11 were isolated from soil. These bacteria can utilize monocrotophos as a phosphorus source but not as a carbon source (Singh & Singh, 2003). Two species of Pseudomonas, three species of Bacillus and three species of Arthrobacter were isolated from soils, which can utilize monocrotophos as a sole source of carbon (Table 2). Further studies demonstrated that Pseudomonas mendocina is the most efficient monocrotophos degrader among the isolated bacteria and its degrading capability is plasmid based (Bhadbhade et al., 2002a). The same group isolated another 17 bacterial isolates from previously exposed soils which can mineralize monocrotophos in liquid culture (Bhadbhade et al., 2002b). The two most versatile degraders, Bacillus megaterium and A. atrocyaneus, were chosen for further studies on the biochemical mechanisms and pathways of monocrotophos degradation. Phosphatase activities were observed in both cultures, and it was suggested that the phosphates identified may be mono- and dimethyl phosphates (Bhadbhade et al., 2002b). Dimethyl- and monomethyl phosphates were involved as intermediates in monocrotophos degradation in plants and animals (Menzer & Cassida, 1965; Muck, 1994). Another intermediate identified during monocrotophos degradation was methylamine, produced by an esterase enzyme. This esterase could be an amidase capable of selecting amides as substrates since esterases sometimes attack the amide bond (Hassal, 1990). Similar pathways of degradation were reported for dicrotophos, which is first demethylated to monocrotophos and then further degraded to methyl amine (Eto, 1974). As with most of the other organophosphorus compounds, the first degradation step of monocrotophos should involve hydrolysis, which could produce N-methyl acetoacetamide and dimethyl phosphate (Beynon et al., 1973). Further degradation of N-methyl acetoacetamide produced valeric acid in A. atrocyaneus and acetic acid in B. megaterium (Bhadbhade et al., 2002b). Acetic acid is the key intermediate of the glycolytic pathway in microorganisms. The pathway of dicrotophos- and monocrotophos degradation is shown in Fig. 7.
Degradation of fenitrothion (O,O-dimethyl O-4-nitro-m-tolyl phosphorothioate), a widely used insecticide, by Burkholderia sp. strain NF100 was reported (Hayatsu et al., 2000). This strain utilized fenitrothion as a source of carbon with the help of two plasmids. The first plasmid (pNF2) was found to catalyze the hydrolysis of fenitrothion to 3-methyl-4-nitrophenol. The nitro group from this compound was oxidatively removed to form methylhydroquinone, which was further metabolized by the second plasmid (pNF2) (Hayatsu et al., 2000). This bacterium was also found to degrade p-nitrophenol as a source of energy. Methylhydroquinone may be degraded by ring fission as one of the two methods described for p-nitrophenol degradation in the section dealings with parathion. p-Nitrophenol degrading Ralstonia sp. SJ98 was reported to have chemotaxis towards 3-methyl-4-nitrophenol and to utilize it as a source of carbon. This strain degrades 3-methyl-4-nitrophenol by the formation of catechol (Bhushan et al., 2000b).
Microbial degradation of various other organophosphorus compounds has been documented. Diazinon degradation by a Flavobacterium sp. was reported in 1973 (Sethunathan & Yoshida, 1973). Two Pseudomonas spp. isolated from sewage sludge were found to degrade diazinon in a culture medium (Rosenberg & Alexander, 1979). Two strains of Arthrobacter sp. were reported to hydrolyze diazinon (Barik et al., 1979). Dimethoate degradation was reported to be carried out by a plasmid based gene of P. aeruginosa MCMB-427 (Deshpande et al., 2001). A novel dimethoate degrading enzyme was purified and characterized from a strain of the fungus Aspergillus niger. This enzyme was found to degrade all compounds containing P–S linkage like malathion and fermothion but not compounds with the P–O linkage (Liu et al., 2001).
Utilization of ethoprophos as a sole source of carbon by P. putida has been observed (Karpouzas et al., 2000). Isolation and metabolism of cadusafos by Sphingomonas paucimobilis and Flavobacterium sp. have been reported recently (Karpouzas et al., 2005). Similarly, several species of bacteria were isolated from different environments which degrade organophosphorus compounds in laboratory cultures and in soils (Singh et al., 1999). Microorganisms isolated from enrichment of one organophosphorus compound can degrade other structurally similar compounds. For example, Flavobacterium sp. and P. diminuta were isolated by diazinon and parathion enrichment but they can degrade a wide range of other organophosphorus compounds such coumaphos, methyl parathion, chlorpyrifos and nerve agents (Adhya et al., 1981; Singh et al., 1999).
Chemical warfare agents
Among lethal chemical warfare agents, the nerve agents have played a dominant role since the Second World War. Nerve agents acquired their name because they affect the functioning of nerve impulses like other organophosphorus compounds. The nerve agents are a group of particularly toxic warfare agents. There are five major substances that are classified as nerve agents and they can be divided into two main groups:
1G agents, including Tabun (GA), Sarin (GB), Soman (GD) and cyclohexyl methylphosphonofluoridate (commonly referred to as cyclosarin or GF).
2V agent, represented by VX.
G agents are usually non-persistent volatile liquids whereas VX is highly persistent, non-volatile and much more active than any of the G agents. Physical and chemical properties of these nerve agents are listed in Table 3. Munro et al. (1994) described the acute and chronic toxicity of nerve agents. In another review by this group, they listed the different sources, fate and toxicity of degradation products of chemical warfare agents (Munro et al., 1999).
Table 3. Chemical, physical and biological properties of some organophosphorus chemical warfare agents
First made (Year)
Vapour pressure (mmHg)
Volatility (mg m−3)
Solubility in water (g L−1)
Breathing (mg min−1 m−3)
Skin (mg individual-1)
It is estimated that about 30 000 tons in USA and about 200 000 tons nerve agents globally have to be destroyed under the Chemical Weapons Convention (CWC), 1993. As of 30 January 2002, 175 states have made CWC commitments. CWC bans the use of chemical weapons but more significantly also bans their development, production, stockpiling and transfer and requires that all existing stocks be destroyed by the member states within 10 years of ratification. Other known chemical warfare agents concentrations include Japanese chemical weapons munitions abandoned in China in 1945, and an estimated 100 000 tons of German chemical weapons munitions that were dumped into the Baltic Sea at the end of World War II. Prior to 1969, the US army disposed of chemical weapons by open pit burning, evaporative atmospheric dilution, burial and placement of munitions in concrete coffins for ocean dumping. In the 1970s, alkaline hydrolysis replaced the above methods for destroying nerve agents. Later, due to the resistance of GB to alkaline hydrolysis, the incineration method was adopted for destroying all groups of chemical weapons. However, due to strong opposition to incineration by environmentalist and local populations, this method of chemical warfare agents destruction was stalled in the USA and republics of the former Soviet Union. Consequently, there is a need to find alternative remediation methods that can provide an environmentally safe and economically viable solution. In this section, we review the environmental fate of important nerve agents and the pathways of microbial degradation.
It is believed that Tabun (name given by its inventor), or GA (ethyl N,N-dimethylphosphoroamidocyanidate), was the first nerve agent ever discovered. It was manufactured in 1937 in Germany, although large scale production and stockpiling started in 1942 (Robinson, 1967). GA is the American denomination of Tabun. It enters the body mainly through the respiratory tract and the primary action is on the respiratory system. It can also cause vision impairment through its anti-acetylcholine esterase activity. GA has a high water solubility but is also readily soluble in organic solvents and can therefore easily penetrate skin (Munro et al., 1999).
GA contains a cyanide group and is subject to hydrolysis. Under neutral and acidic conditions, the first step, which is very rapid, includes formation of O-ethyl N,N-dimethyl amidophosphoric acid and hydrogen cyanide. The subsequent hydrolytic step, which is comparatively slow, is hydrolysis of O-ethyl N,N-dimethyl amidophosphoric acid to dimethylphosphoramidate and then finally to phosphoric acid. Under acidic conditions, hydrolysis to ethylphosphorylcyanide and dimethylamine occurs. The final product of all pathways is phosphoric acid. Several different metabolites were identified in soil exposed to GA. D'Agostino & Provost (1992) identified 16 different compounds from a soil contaminated with GA. However, several of them were impurities and some were degradation products.
The chemical structure of GA suggests that it contains several possible microbial degradation sites. The initial steps are potentially O-dealkylation and C-dealkylation, nitrile hydrolysis and N-dealkylation (Morrill et al., 1985). No specific microorganism has been isolated for exclusive GA degradation from natural environments but P. diminuta isolated for degradation of other organophosphorus compounds can degrade several chemical warfare agents including GA (Mulbry & Rainina, 1998). DeFrank et al. (1993) isolated several strains of Alteromonas that can effectively degrade all G nerve agents. As with other organophosphorus compounds, the complete degradation of GA is likely to produce phosphoric acid. Several bacterial species have been reported to cleave C–P bonds. The mechanism and associated micro-organisms have been described in detail under the section dealing with glyphosate.
Several intermediate metabolites of GA have been identified from soils that include dimethylamine and triethyl phosphate (Sanches et al., 1993; Verschueren, 1996), and diethyl dimethylphosphoramidate (Munro et al., 1999). These compounds are readily biodegradable (Verschueren, 1996). Degradation and utilization of alkylamine as a source of energy is widespread in natural environments as discussed for glyphosate degradation. On the basis of the above details, a proposed pathway for GA degradation is presented in Fig. 8.
GB (isopropyl methylphosphonofluoridate) is a highly toxic nerve agent first produced in Germany in 1937 (Bakshi et al., 2000). The term Sarin is an acronym of its discoverers (Gerhard Schrader, Ambros Rudriger and Van der Linde). Immediate death from exposure occurs because of respiratory tract failure (Rickett et al., 1986). Other routes of exposure include the gastro-intestinal tract and skin absorption (Spruit et al., 2000). GB was implicated in terrorist attacks in 1994 and 1995 in Japan, which caused death and injured many people (Abu-Qare & Abou-donia, 2002). People exposed to GB during the incident in Japan reported symptoms such as darkness of vision, ocular pain, dyspnoea and headache. A review on GB effects on health is available elsewhere (Abu-Qare & Abou-donia, 2002). GB is non-persistent, volatile and completely soluble in water and subject to acid and alkaline hydrolysis (Munro et al., 1999).
Like other xenobiotics, the fate of GB in soil includes biotic and abiotic degradation, evaporation and leaching. More than 90% of GB added to soil was reported to be degraded within 5 days (Small, 1984); however, degradation is comparatively slow at low temperature (Morrill et al., 1985; Sanches et al., 1993). As discussed for GA, several bacteria have been reported to degrade G agents including GB. The major metabolites identified for GB degradation are isopropylmethylphosphonic acid (IMPA) and methyl phosphonic acid (MPA) (Mulbry & Rainina, 1998; Munro et al., 1999). Chemically, IMPA is extremely stable and is predicted to have a half-life of over 1900 years (Rosenblatt et al., 1975). IMPA is relatively resistant to bacterial degradation. However, two bacterial species, P. testosteroni and Pseudomonas melophthora have been reported to degrade IMPA to MPA (Daughton et al., 1979). These bacteria metabolize IMPA via cleavage of the C–P bonds to methane and inorganic phosphorus compounds. Zhang et al. (1999) reported biotransformation of IMPA by a microbial consortium. Four mixed cultures were acclimated to IMPA. Two of these cultures, namely APG and SX microorganisms, used IMPA as the sole source of phosphorus. The intermediate metabolites were identified as MPA and inorganic phosphates. Although attempts to use IMPA as a source of carbon to support microbial growth were not successful, in a bioreactor 85 mg L−1 IMPA was decreased to non-detectable level within 60–75 h. MPA is also susceptible to C–P lyase producing bacteria (Zhang et al., 1999). Use of MPA as a source of phosphorus by a P. putida has been observed (Cook et al., 1978b). Several other bacteria were reported to possess C–P lyases and they have been described for glyphosate degradation. The microbial degradation pathway for GB is presented in Fig. 9.
Soman, or GD (pinacolyl methylphosphonofluoridate), is structurally similar to GB. GD was the given identifier of Soman post war (American denomination, GC was already in medical use) when the information relating to Soman was recovered by old Soviet Union in 1949. Its volatility is intermediate between GA and GB. It is less water soluble and more lipid soluble than the other two G agents, which results in more rapid skin penetration and greater toxicity (Munro et al., 1999). Like other G agents, GD is subject to hydrolysis but the rate of hydrolysis is five times slower than GA (Hambrook et al., 1971). The first step in hydrolysis is fluoride removal to form pinacolyl methylphosphonic acid (PMPA), which is then slowly degraded to MPA and pinacolyl alcohol (Kingery & Allen, 1995). No data were found on the fate of PMPA or pincolyl alcohol in the environment. It is assumed that, like alkyl methylphosphonic acid, PMPA is probably resistant to degradation (Munro et al., 1999). However, like other G agents, GD is hydrolyzed by P. diminuta and several strains of Alteromonas (DeFrank et al., 1993). Similarly, IMPA degrading consortia were able to degrade PMPA as a sole source of phosphorus. In successive batch experiments using immobilized cells, PMPA level decreased from 164 mg L−1 to below the detection limit within 60 h (Zhang et al., 1999). The proposed microbial degradation pathway is presented in Fig. 9.
O-ethyl-S[2-(di-isopropylamino) ethyl] methylphosphonothioate was first discovered by British scientists. Later, the US produced it in large quantities under code name VX. It is a moderately persistent nerve agent characterized by a P–S bond and, therefore, it belongs to the phosphorothiolates group. It is less volatile than G agents and does not evaporate easily (Munro et al., 1999). VX is soluble in water (30 g L−1 at 25°C) and is relatively resistant to hydrolysis. However, at acidic and extreme alkaline pH, cleavage of the P–S bond predominates, resulting in formation of ethyl methylphosphonic acid (EMPA) and diisopropylethyl mercaptoamine (DIEM). The latter can be oxidized to bis (2-diisopropylaminoethyl) disulfide (BIAEDS) (Yang et al., 1993). At neutral and alkaline pH, the common pathway of hydrolysis includes cleavage of C–O bonds to ethanol and S-(2-diisopropyl aminoethyl) methyl phosphonothioate (DIAEMP). The half-life of VX in water at pH 7 and 25°C is 17–42 days (Clark, 1989). Laboratory and field studies on the fate of nerve agents demonstrated that loss is due to a combination of abiotic and biotic processes such as evaporation, hydrolysis and microbial degradation. According to one study, 90% of added VX was lost from soil in 15 days (Small, 1984). Diethyl methyl phosphonate and BIAEDS were extracted when VX was added to soil (Sanches et al., 1993). In other studies, EMPA and DIEM were found to be major metabolites in soils (Kaaijk & Frijlink, 1977; Omar, 1998). Further degradation of EMPA resulted in formation of MPA (Omar, 1998). EMPA can be used as a phosphorus source for natural microbial systems (Cook et al., 1978b; Mulbry & Rainina, 1998). Diethyl dimethylpyrophosphonate, diisopropylaminoethanol (DIPAE), diisopropylamine (DIPA) ethylmethylphosphonothioic acid (EMPTA) were reported as other possible intermediate metabolites (Munro et al., 1999).
VX is resistant to microbial hydrolysis. It cannot be hydrolyzed by any of the strains of Alteromonas isolated for the degradation of G agents. OPH of P. diminuta is found to be active against VX and Russian VX but its activity is less than 0.1% toward VX as compared with parathion. However, site-specific mutagenesis in OPH resulted to a 33% increased activity against VX (Gopal et al., 2000). It was noted that VX could be rapidly degraded by chemical oxidation of the P–S bond using various peroxides (Yang et al., 1993) and monomagnesium perphthalate (Amitai et al., 1998). Oxidative hydrolysis of VX produces EMPA and dialkylaminoethanesulfonate as compared with the corresponding alkylthion hydrolytic product formed via the hydrolysis pathway. EMPA has been reported to be degraded as a source of phosphorus by two glyphosate-degrading bacteria, Burkholderia caryophilli and Pseudomonas testosteronis (Elashvili & DeFrank, 2001). A partially purified enzyme from B. caryophilli bacterium has shown a broad specificity towards neutralized nerve agents, including GF, GB, GD, VX and Russian VX (Elashvili & DeFrank, 2001).
Oxidative hydrolysis of VX by the enzyme laccase from a white-rot fungus, Pleurotus ostreatus was observed. The mechanism of such degradation is not fully understood. It was suggested that the sulfur atom is oxidized followed by cleavage of the P–S bond (Yang et al., 1990). The nitrogen atom at the β position to the carbon bound to the sulfur atom was assumed to play an important role in the enzymatic reaction. One suggested pathway is the formation of an N-oxide intermediate in the N,N dialkyl aminoethyl moiety at alkaline pH that may affect the cleavage of the P–S bond. Cleavage of the S–C bond may also occur forming O-ethyl methyl phosphorothoic acid and 2-diisopropylaminoethanol (Yang et al., 1990). The proposed pathways of VX degradation are shown in Fig. 10.
Microbial enzymes that can hydrolyze organophosphorus compounds have been identified and characterized from different microbial species. The hydrolysis of organophosphorus compounds leads to a decrease in mammalian toxicity by several order of magnitudes and therefore this step is also called detoxification. An excellent review on the role of bacterial enzymes in detoxification of organophosphorus nerve agents has been published recently (Raushel, 2002). Consequently, in the present article, we only review the characteristics, improvement and utility of a few of the most extensively studied organophosphorus hydrolyzing enzymes. Several bacterial and fungal isolates with novel enzyme/gene systems are reported (Table 4). However, despite the apparent diversity of the enzyme systems, most studies of organophosphorus degrading enzymes have focused on organophosphorus hydrolase (OPH) and organophosphorus acid anhydrolase (OPAA), which are among the most extensively studied enzymes in the biological sciences.
+, positive activity; −, no activity; ND, not determined; MW, molecular weight.
Organophosphorus hydrolase has been isolated from several bacteria (Serdar et al., 1982; Mulbry & Karns, 1989a; Singh et al., 1999). Among bacterial enzymes, OPH from P. diminuta has the widest range of substrate specificity and, therefore, has received most attention. OPH is a dimer of two identical subunits containing 336 amino acid residues (Dumas et al., 1989) that folds into a (αβ)8-barrel motif (Gerlt & Raushel, 2003). Each subunit contains a binuclear zinc situated at the C-terminal portion. The two zinc atoms are separated by about 3.4 Å and linked to the protein through the side chain of His 55, His 57, His 201, His 230, Asp 301 and a carboxylated Lys 169. Both the Lys 169 and the water molecule (or hydroxide ion) act to bridge the two zinc ions together (Benning et al., 2001). A schematic diagram of the structure of the binuclear metal centre within the active site of OPH is presented in Fig. 11. It has a molecular weight of 72 kDa.
Organophosphorus compounds bind to the binuclear metal centre within the active site via co-ordination of the phosphoryl oxygen to the β-metal ion. This interaction weakens the binding of the linking hydroxide to the β-metal. The metal-oxygen interaction polarizes the phosphoryl oxygen bond and creates a more electrophilic phosphorus centre. Subsequent nucleophilic attack by the bound hydroxide is assisted by proton abstraction from Asp 301. The hydroxide attacks the phosphorus centre, resulting in weakening of the bond to the leaving group (Raushel, 2002). A working model for the OPH reaction mechanism is shown in Fig. 12. In summary, the role of one metal ion in the active site of OPH is to increase the electrophilicity of the phosphorus centre through co-ordination with the non-ester oxygen atom of the substrate, whereas the second metal ion acts as a promoter of the attacking nucleophile (Efrmenko & Sergeeva, 2001). However, questions regarding the mechanisms of catalytic activity remained unanswered. The pKa value of the bridging solvent is not known; it is believed to be determined by variation of the kinetic parameters with pH. In addition, it has still to be resolved whether the two Zn ions within the active site have distinct functions or whether they act in tandem (Raushel, 2002).
Organophosphorus hydrolase has a wide range of substrate specificity. It hydrolyzes P–O, P–F, and P–S bonds to different extents (Table 4). The lowest specificity is for the P–S bond. However, the enzyme does not catalyze the cleavage of carbonyl groups such as those found in p-nitrophenyl acetate. Similarly, organophosphate diesters are very poor substrates. The Kcat and VKm values for the hydrolysis of paraxon by the enzyme have been determined to be nearly 104 s−1 and 108 M−1 s−1, respectively (Omburo et al., 1992).
The effects of metal substitution on the catalytic activity of OPH were studied by removing the native metal (Zn) from purified OPH and reconstitution with a series of divalent cations which include Co, Cd, Cu, Fe, Mn and Ni (Omburo et al., 1992; Di Sioudi et al., 1999; Benning et al., 2001). Further enzymatic assays showed that Co2+ had the greatest activity against paraoxon (Omburo et al., 1992). It was suggested that divalent cations increased the activity of enzyme by assisting folding of expressed enzyme in the medium (Manavathi et al., 2005). Site specific mutagenesis was used to substitute the original histidinyl residues at positions 254 and 257. Of these mutant enzymes, H254R (histidine at 254 was replaced with arginine) and H257L (histidine at 257 was substituted with leucine) demonstrated a four- to five-fold higher catalytic activity against the P–S bond (VX and demeton-S). Other mutants also showed higher activity against Soman and other nerve agents (Di Sioudi et al., 1999).
The hydrolysis of asymmetric organophosphorus compounds catalyzed by OPH is stereoselective (Lewis et al., 1988; Chae et al., 1994). For example, OPH degrades the Sp isomer of EPN 21 times faster than the Rp isomer. Similarly Sp isomers of acephate and methamidophos are catalyzed preferentially by OPH (Hong & Raushel, 1999). To achieve a practical solution to nerve agent contamination, the enzyme should be able to degrade the racemic mixture, as both isomers are usually present in compounds such as Soman, GB and VX. This aim was achieved by rational modification to the substrate binding activities. The size and shape of these binding subsites were remolded through a rational restructuring via site-directed modification (Wu et al., 2001; Raushel, 2002). Preferential degradation of the Sp isomer of the EPN by OPH presumably arises because the bulkier phenyl substituent is better accommodated in the large subsite and the ethyl group within the small subsite of OPH for the Sp–enantiomer. To increase the degradation of the Rp-enantiomer, the small subsite was expanded. This was achieved by the substitution of Phe132, Ser308 and Ile106 to glycine and/or alanine. With these mutants, the stereoselectivity for the Sp and Rp enantiomers for EPN decreased from 21 : 1 to 1 : 1.3 (Chen-Goodspeed et al., 2001; Wu et al., 2001).
The course of evolution of OPH in bacteria remains an unresolved question. Organophosphorus compounds were released as pesticides after World War II and it is difficult to understand how these OPH catalytic activities appeared so quickly. One suggestion is that this enzyme was already present in the environment before the application of organophosphorus pesticides. This hypothesis has gained importance following isolation of similar OPH from phylogenetically and geographically different micro-organisms (Mulbry et al., 1986; Somara et al., 2002). The recent finding of OPH encoding genes in a field soil which has never been exposed to this group of pesticides supports this hypothesis (Singh et al., 2003c). Another possibility is that this enzyme has evolved new substrate specificity from pre-existing enzymes as it has been shown that OPH (phosphotriesterase) could acquire phosphodiesterase activity by alteration of only one amino acid (Shim et al., 1998). Urease has been found to have carbamylated lysine as a bridging ligand with binuclear Ni at the active site (Park & Hausinger, 1995). The binuclear centre of urease and OPH was found to be remarkably similar. However, the chemical nature of the active sites of these enzymes is quite different (Raushel, 2002). A larger group of enzymes with similar active site architecture has been identified (Holm & Sander, 1997). Interestingly this superfamily also includes atrazine chlorohydrolase.
A similar enzyme, OPDA, has been isolated from A. radiobacter and was found to have 90% homology to OPH at the amino acid level and a very similar overall secondary structure (Horne et al., 2002b; Yang et al., 2003). Despite these similarities, the two enzymes have different substrate specificities. There is about a 30-sequence difference between OPH and OPDA. The largest group consists of 19 residues at the C-terminus. In addition, two regions with significant difference in OPDA from OPH are a large pocket at the active site and another in the region of the protein that is responsible for binding phenyl ethanol (an inhibitor) in OPH. Apart from the sequence difference, the water structure in this region differs in the two enzymes. There are two water molecules that form a network of hydrogen bonds in OPDA. The equivalent residues in OPH could not form the same hydrogen bonds but were found to be stabilized by the presence of phenyl ethanol. These differences at the active site of OPDA likely to give it a preference for substrates that have shorter alkyl substituents. Further studies using site-specific mutagenesis in OPH gave a series of mutants that had activities similar to those of OPDA. Yang et al. (2003) argued that this alteration in the active site gave substrate specificity and represented the progressive natural evolution of the enzyme from OPH to OPDA. However, the observation that alterations of amino acids in the active site increases the activity of enzymes emphasizes that enzyme efficiency depends on several factors and that evolution can take place in many ways.
Another enzyme that has received considerable attention recently, for detoxification of organophosphorus nerve agents is OPAA. A highly active OPAA from Alteromonas undina was isolated and purified and is composed of a single polypeptide with molecular weight 53 kDa (Cheng et al., 1993). However, another OPAA isolated from Alteromonas sp. JD6.5 is composed of 517 amino acids with molecular weight of 60 kDa. However, one from Alteromonas haloplanktis contains 440 amino acids (Cheng et al., 1996, 1997) with molecular weight 50 kDa. The 10 kDa difference between OPAAs of these two Alteromonas spp was found to be due to the presence of an extended C-terminal region in the JD 6.5 enzyme (DeFrank & White, 2002). The three-dimensional structure of this enzyme is not yet known. It possesses low catalytic activity against P–O but high activity against P–F bonds. OPAA also displays stereoselectivity towards the chiral phosphorus centre by displaying preference for the Rp-enantiomers. It also exhibits an additional preference for the stereochemical configuration at the chiral carbon centre of the Soman analogue (Hill et al., 2001). This observation was confirmed by a recent study, which found that OPAAs along with wild-type phosphotriesterase catalyze preferentially the hydrolysis of (+) GF isomer in a racemic mixture (Harvey et al., 2005). However, OPAAs from different species of Alteromonas have demonstrated wide variation in catalytic activity, with the highest activity observed with the enzyme obtained from Alteromonas sp. J.D.6.5 (DeFrank & White, 2002). OPAA from Alteromonas sp. JD6.5 has high degree of homology at amino acid level with E. coli X-Pro dipeptidase (48%) and E. coli aminopeptidase P (31%). Further molecular and biochemical analyses of OPAA have established that this enzyme is a prolidase; a type of dipeptidase cleaving dipeptide bond with a prolyl residue at the carboxyl terminus (Cheng et al., 1999). Although the native function of OPAA is not yet known, it has been suggested to play an important role in cellular dipeptide metabolism because all OPAAs were found to have activity against several dipeptides (DeFrank & White, 2002). Molecular modelling studies with Soman and Leu-Pro revealed that the three-dimensional structure and electrostatic density maps of the two are nearly identical (DeFrank & White, 2002). This explains why several dipeptidase enzymes have catalytic activity against organophosphorus compounds. Determination of the crystal structure of OPAA may confirm this hypothesis. Both prokaryotes and several eukaryotes have been found to possess this enzyme (Mazur, 1946; Hoskin et al., 1999), which suggests that OPAA is not a newly evolved enzyme (Cheng et al., 1999). In a recent study, aminopeptidase P (AMPP) was found to catalyze the hydrolysis of a wide range of organophosphate triesters. AMPP belongs to the family of protein-specific peptidases and catalyzes the cleavage of amino-terminal X-Pro peptide bonds (Jao et al., 2004). This enzyme possesses many similarities with OPAA, such as two divalent (Mn2+) ions, which are critical for maximal activity of AMPP. Replacement of the Mn2+ ion with other divalent ions except Co2+ resulted in the loss of catalytic activity (Jao et al., 2004). AMPP is a tetramer with each sub-unit composed of a ‘pita bread’ fold of the C-terminus domain (Wilce et al., 1998). The active site of AMPP is located at the C-terminal portion of the β-sheet with the two manganese ions separated by 33 Å. The two ions are co-ordinated by Asp 260, Asp 271, His 354, Glu 383, Glu 406 and two water molecules. A water molecule or hydroxide ion bridging between the two metal ions is believed to be strongly activated, and acts as the nucleophile in the attack on the scissile peptide bond Xaa–Pro. A single amino acid mutation with hydrophobic side chains such as R153W, R153L, R370L increased the hydrolysis rates towards most of the organophosphorus substrates compared to the wild-type enzyme (Jao et al., 2004). This result suggests that the further protein engineering of AMPP may significantly enhance the cleavage of P–O bond in a variety of organophosphorus compounds.
Other structurally and functionally different organophosphorus degrading enzymes have been reported. Three unique parathion hydrolases were isolated, purified and characterized from gram-negative bacterial isolates. One cytosolic hydrolase described as an ADPase (aryldialkylphosphatase) from Nocardia sp. strain B-1 was composed of a single sub-unit of approximately 43 kDa (Mulbry, 1992). Another hydrolase from strain SC was membrane bound and is composed of four identical sub-units of 67 kDa. While having some common features such as constitutive production and similar temperature optima around 40°C, the substrate specificity and structure of these enzymes differ one from another, and also from the other known OPHs (Mulbry & Karns, 1989a). A unique phosphotriesterase has been characterized from Nocardioides simplex NRRL B-24074. The purified enzyme is monomeric, has a native molecular weight of 45 kDa, is constitutively expressed and located in the cytoplasm. This enzyme is quite distinct with respect to its activity towards different substrates and also in its stimulation or inhibition by divalent cations and dithiothreitol (Mulbry, 2000). Another novel phosphotriesterase HocA (hydrolysis of caroxon) was isolated from P. monteilli (Horne et al., 2002c). This enzyme was required by the host for phosphate metabolism and was suggested to be evolved from phosphodi- or mono-esterase. HocA (19 kDa) does not require a metal ion for its catalytic activity but was reported to be less efficient at hydrolyzing organophosphorus compounds than other reported microbial phosphotriesterases (Horne et al., 2002c). HocA is not a metalloenzyme and its activity is controlled by the presence of phosphate in the medium.
There have been a number of reports on isolation and purification of organophosphorus hydrolyzing enzymes from pure isolates or from mixed cultures of bacteria. Only a few organophosphorus hydrolyzing enzymes have been reported from fungi. Degradation of phosphorothiolates by a broad-spectrum fungal enzyme, laccase (phenol oxidase) from a white-rot fungus P. ostreatus was reported (Amitai et al., 1998). This is a significant observation as this enzyme attacks P–S bond, which is comparatively resistant to OPH and OPAA cleavage. Laccase was observed to be capable of complete and rapid degradation of VX and Russian VX (Amitai et al., 1998). Several white-rot fungi are capable of organophosphorus degradation (Table 2), and it will be interesting to know if the degradation capability of all white-rot fungi towards organophosphorus compounds is mediated by the presence of laccase, or whether different fungi possess different enzyme systems. A novel organophosphorus degrading fungal enzyme (A-OPH) was isolated from A. niger ZHY256 that could hydrolyze a range of P–S bonds containing organophosphorus compounds (Liu et al., 2001). This 67-kDa enzyme has found to have optimal pH at 7 with thiol and sulfhydryl groups in the active catalytic site. A-OPH does not require divalent cations for activation; however, Cu2+ was found to activate its activity. Another novel organophosphorus-hydrolyzing enzyme was purified from Penicillium lilacinum BP303. Interestingly, this penicillium OPH (P-OPH) was found to degrade various organophosphorus compounds by cleaving both P–O and P–S linkages (Liu et al., 2004). The molecular mass of P-OPH is 60 kDa with optimal activity at pH 7.5. The purified enzyme was reported to be a member of a cysteine hydrolase group and similar to A-OPH and OPAA. Despite having several similar structural components, P-OPH is different from OPH, OPAA and A-OPH in its catalytic activity. P-OPH degraded all organophosphorus compounds with P–O and P–S linkage, whereas OPH in Flavobacterium sp. only attacks P–O bond and A-OPH splits only P–S linkage.
The first reported enzyme able to degrade the phosphonates, 2-phosphonoacetaldehyde hydrolase (phosphonatase), was isolated from B. cereus (La Nauze et al., 1970). The isolated and purified phosphonatase showed optimal activity at pH 8, required Mg2+ for its activity, and was inhibited by sulfhydryl reagents (La Nauze et al., 1970). Phosphonatase resembles alkaline phosphatase in many properties but has narrow substrate specificity. Phosphonatase does not degrade phosphomonoesters and is not a metalloenzyme (Kononova & Nesmeyanova, 2002). This enzyme has been reported from several bacterial species and can degrade a range of phosphonates including glyphosate (Baker et al., 1998). Further analysis suggests that phosphonatases belong to a new family of hydrolase having a high conservative aspartate residue in their active site to which the phosphoryl group from a lysine residue of the enzyme is transferred (Baker et al., 1998). Phosphonatase is a homodimer of 33–37 kDa subunits, and its active site is mainly comprised of polar amino acid residues, which suggests that phosphonatase may have a common origin to the NAD dependent superfamily of dehalogenase, phosphotase and phosphomutase (Kononova & Nesmeyanova, 2002). Another interesting enzyme that can degrade phosphonates is C–P lyase. There is one report suggesting partial purification of this enzyme from Pseudomonas sp. GLC11 (Selvapandiyan & Bhatnagar, 1994). The molecular mass of this enzyme was reported to be approximately 200 kDa and it was found to be localized in the periplasmic space of bacteria. However, subsequent reports suggest that C–P lyase manifests its activity only in cells and has never been reliably found in cell-free extracts (Kononova & Nesmeyanova, 2002). This obstacle considerably limits the possibility of understanding the mode of catalytic action of C–P lyase. A good review on the proposed mechanism of phosphonates degradation by C–P lyase on the basis of computer modelling and gene structures is available (Kononova & Nesmeyanova, 2002). In brief, the action is initiated by the generation of a phosphonyl radical. Subsequent cleavage of this reactive intermediate would lead to metaphosphate and alkyl moieties as the corresponding alkenes. Abstraction of hydrogen by an alkyl radical would yield the corresponding alkanes as products, which is a specific feature of phosphonate degradation.
Several microbial isolates have been reported to have further novel enzyme/gene systems but most of these were not isolated or purified such as C–P lyase (Kertesz et al., 1994a), methyl parathion hydrolase (Zhongli et al., 2001) and chlorpyrifos degrading enzyme (Singh et al., 2004). Most of the enzymatic studies were carried out to improve the catalytic activity of OPH and OPAA by protein or genetic engineering. Nonetheless, a few novel enzymes have been purified recently and these enzymes differed in molecular mass, substrate specificity and, sensitivity to chemicals (Table 4). Improvements of known enzymes should continue but this trend of characterizing new and diverse enzymes from prokaryotes and eukaryotes needs to be sustained to find the best bioremedial enzymes with optimal activity in detergents and in the presence of metal ions, and which have a broad pH and temperature optima. Discovery of diverse microbial enzymes will also facilitate understanding of the evolutionary structure–function relationship of organophosphorus-degrading enzymes.
Genetic basis of organophosphorus degradation
The first described organophosphorus degrading (opd) gene was found in P. diminuta, and was shown to be present on a plasmid (Serdar et al., 1982). The plasmid size was 66 kb and was termed pCMS1. By cloning into different plasmids and into the broad range cloning vector, it was shown that a 1.5 kb BamHI fragment with single restriction sites for SalI, PstI and XhoI encoded this enzyme (Serder & Gibson, 1985). Mulbry et al. (1987) found that the opd gene from Flavobacterium sp. strain ATCC 27551 was encoded on a 43-kb plasmid (pPDL2) and had a similar restriction map to the opd gene from P. diminuta. Southern hybridization experiments demonstrated that the opd gene from the two bacteria possessed significant homology. This finding of homologous genes on two non-homologous plasmids from two phylogenetically and temporally different bacteria isolated from different geographical regions suggests that the gene may be a mobile genetic element or transposon (Mulbry et al., 1987). Sequencing of the opd gene proved that the gene from both bacteria had identical sequences (Harper et al., 1988). Later, the nucleotide sequence of the opd gene from P. diminuta was determined and a single open read frame located (Serder et al., 1989). The opd gene from Flavobacterium sp. consists of 1693 base pairs with one open reading frame (Mulbry & Karns, 1989b). The opd gene has been cloned into various bacterial strains (Serder & Gibson, 1985), actinomycetes (Steiert et al., 1989), fungi (Xu et al., 1996) and insect cells (Dumas et al., 1989). Several other bacteria have opd genes with almost identical nucleotide sequences (Chaudry et al., 1988; Somara et al., 2002). Horne et al. (2002b) reported a similar opd gene from A. radiobacter P230 isolated in Australia for coumaphos degradation, which was chromosome based. This gene, called opdA, was approximately 88% identical at the nucleotide level to opd (Horne et al., 2002b). Sequencing of the whole genome revealed the presence of opd-like genes in Mycobacterium tuberculosis (Philipp et al., 1996) and E. coli (Blattner et al., 1997), which supports the hypothesis that the opd gene may be transposon based. However, this evidence has been provided only recently (Siddavattam et al., 2003) where a complete sequence of a region of plasmid pPDL2 from Flavobacterium sp. is reported, which has identical restriction patterns to opd containing plasmid pCMS1 of P. diminuta (Fig. 13). The opd gene was found to be flanked by an insertion sequence, ISF1sp1 (encoding a complete istAB operon), which is a member of the IS21 family, and downstream by a Tn3-like element (tnpA and tnpR) encoding a transposase and a resolvase. It was also observed that adjacent to opd, but transcribed in the opposite direction, is an open reading frame (orf243) which encodes a polypeptide of 27 kDa that plays a role in the degradation of p-nitrophenol (the major degradation product of parathion and methylparathion). A 2.5-kb region upstream of the opd gene contains two ORFs transcribed in the same direction as opd, which have significant homology to the IstA and IstB genes. Similarly, Horne et al. (2003) reported that opdA in A. radiobacter P230 is transposable. A tnpA gene was found upstream of the opdA. The two genes are flanked by insertion sequences which resembles Tn 610 transposon from Mycobacterium fortuitum. Two additional putative ORFs separate opdA and tnpA, and the deduced translation products show similarity to two proteins encoded on the Geobacillus stearothermophilus IS5376 (Horne et al., 2003). These observations of linkage of the opd/opdA genes to IS elements and transposase genes supports the idea that the widespread distribution of the opd gene could be due to its lateral transfer by a combination of transposition and plasmid transfer (Siddavattam et al., 2003).
The evolutionary origin of the opd gene is presently not known. However, it has been argued that the closest homologues of the IstA, IstB, and tnpR genes are all found in strains of Agrobacterium tumefaciens and an opd gene has recently been reported from a strain of A. radiobacter (Horne et al., 2002b). These similarities could indicate an evolutionary origin for these genes in Agrobacterium (Siddavattam et al., 2003). Another hypothesis is that the gene was present in the environment long before organophosphorus compounds were commercialized. The presence of genes similar to opd in several bacteria that have never been exposed to this group of compounds also supports this argument (Philipp et al., 1996; Blattner et al., 1997; Richins et al., 1997). Recently, higher copy numbers of the opd gene were observed in higher pH soils (Singh et al., 2003a, c). This observation has considerable significance because this field had not been exposed to organophosphorus compounds. It is worth mentioning that OPH has optimal activity at higher pH. It is possible that opd or its ancestor has some important function to play at high pH. It is unlikely that such gene/enzyme systems evolved to protect against anticholinesterase compounds since bacteria do not contain acetylcholinesterases but their widespread distribution suggests that the gene/enzyme system does serve an important function. One suggested function is the role of an opd-like gene in phosphate metabolism (Horne et al., 2002b). These genes may have evolved from pre-existing mono-phosphatase or phosphodiesterase as it has been shown that OPH (phosphotriesterase) could acquire phosphodiesterase activity by the change of only one amino acid (Shim et al., 1998). Isolation, characterization and cloning of a novel phosphodiesterase (which degrades organophosphate xenobiotics) gene from Delftia acidovorans is strong evidence to support this hypothesis (Tehara & Keasling, 2003). This gene shows sequence similarity to cyclic AMP (cAMP) phosphodiesterase and cyclic nucleotide phosphodiesterases and exhibits activity on cAMP in vivo when the gene is expressed in E. coli, suggesting that it may have evolved from a common ancestor of the cAMP gene and may regulate cAMP levels in bacterial cells (Tehara & Keasling, 2003). Further evidence in the form of the gene structure and function relationships is required to reach to a definitive conclusion.
Another organophosphorus degrading gene which has received considerable attention is opaA, first isolated and cloned from Alteromonas sp. JD6.5 (Cheng et al., 1996, 1997). In spite of functional similarity with the opd gene, no sequence homology was found between them. One ORF of 1552 nucleotides was identified that codes for OPAA. The OPAA enzyme has amino acid sequence similarity with that of E. coli AMPP and human prolidase. It is believed that opaA and the prolidase gene may have evolved from the same ancestral gene and may play a role in bacterial peptide metabolism (Cheng et al., 1996). However, the role of pepP gene (encodes for AMPP) in organophosphorus degradation has also been reported (Jao et al., 2004).
Genes for phosphonate degradation have received considerable attention due to their commercial exploitation in genetically modified crops. Seventeen open reading frames (phnA to phnQ) were reported to be involved in phosphonate uptake and degradation by E. coli (Chen et al., 1990). Further study suggested that the genes phn ABQ were not involved in degradation and that the phn operon therefore consisted of 14 genes which are transcribed from a single promoter preceding the phnC gene. On the basis of sequence analysis and comparison with known motifs, reading frames phnCDE have been proposed to form a phosphonate transport complex, whereas phnF and phnO may be involved in regulation. Two genes (phnNP) are not required for phosphonate use and may encode accessory proteins for the C–P lyase. The remaining seven genes (phnG-M) were suggested to be involved in the C–P lyase complex itself (Metcalf & Wanner, 1991). Parker et al. (1999) reported the presence of a homologous part of the E. coli phn gene cluster in Sinorhizobium meliloti. By cloning, phnGHIJK genes were identified in S. meliloti. However, several genes from phn cluster of E. coli were not detected in S. meliloti, despite the fact that S. meliloti appeared to have a broader substrate specificity. This observation suggests that not all genes in phn cluster may be required for phosphonate metabolism or that these genes are functionally redundant in S. meliloti. The molecular-genetic analyses suggest that the process of phosphonate degradation involves a multi-component system with constituents localized in the membrane and periplasm. This may explain the failure by several research groups to isolate and purify cell free C–P lyase despite considerable effort. Expression of the genes for phosphonate degradation is controlled by phosphorus supply to the cell and they have been suggested to be the part of the pho regulon on the basis of their similarity to promoter sequences (Mulbry & Karns, 1989b) and the requirement of regulatory genes (Wacket et al., 1987). Two other novel genes involved in the degradation and utilization of glyphosate, glpA and glpB, were isolated and sequenced from Pseudomonas pseudomallei. The gene glpA (1260 bp long) encodes an enzyme (phosphotransferase) of 420 amino acids, which confers increased tolerance to glyphosate. The gene glpB encodes a protein (309 amino acids long) with the ability to break the N–C bond of glyphosate to yield aminomethyl phosphonic acid (Penaloza-Vazquez et al., 1995). Another gene involved in glyphosate metabolism, pehA (encodes PEH), was cloned and sequenced from B. caryophilli PG2982 (Dotson et al., 1996).
Several other genes with similar or identical function but totally different nucleotide sequences have been reported (Table 5). A methyl parathion degrading (mpd) gene was isolated from Plesiomonas sp. strain M6 (Zhongli et al., 2001). Sequencing and cloning of mpd revealed that this gene is 1061 bp long and encodes a 35-kDa product. When the mpd nucleotide sequence and predicted protein sequence were compared with those in the Genbank database, no region of extensive DNA homology was observed. The highest similarity with predicted protein sequence was found to be 31% with beta-lactamase, suggesting significant novelty of the gene-enzyme system. Another novel gene, adpB (which encodes ADPase), for organophosphorus degradation was obtained from Nocardia strain (Mulbry, 1992). This 1600 bp long gene does not share homology with any of the other known genes involved in organophosphorus degradation. Horne et al. (2002c) isolated and cloned a gene called hocA (hydrolysis of caroxon) gene from P. monteilli, consisting of 501 bp; it has a different sequence from all other known organophosphate degrading genes. The hocA gene encodes a 19-kDa protein that can degrade a range of oxon and thion organophosphorus compounds. Increased expression of hocA was observed from an integrative hocA–lacZ fusion when the culture was grown in the absence of phosphate, suggesting that it might be part of the pho regulon. This structural diversity and functional similarity suggests that these genes have evolved from different ancestors, but that the structural similarity of xenobiotics to natural compounds, and the constant mutation in the bacterial genome and rapid doubling time play an important role in the evolution of the gene/enzyme systems for xenobiotic degradation.
Table 5. List of genes, their origin, vector, and gene products involved in degradation of organophosphorus compounds
Biotechnological aspects of the degradation of organophosphorus compounds have received considerable attention recently due partly to their high mammalian toxicity and partly to the requirements of the CWC. Several compounds need sensible detoxification and disposal techniques because of their bulk usage, storage and widespread use. Current methods for detoxifying these compounds mainly rely on incineration and landfills. Incineration of chemical warfare agents has received strong and sustained opposition from the public and environmental groups because of potentially toxic emissions. This process is also very costly, as it requires considerable amounts of energy to reach the high temperatures needed to destroy the pollutants. Landfills provide an adequate short term solution but leaching of pollutants to ground water is a major source of concern. Bioremediation with micro-organisms is therefore an attractive alternative to these conventional techniques for pollutant disposal.
Munneck (1976) first reported the potential use of parathion hydrolase producing bacteria for the detoxification and disposal of organophosphorus compounds. Later, successful use of OPH producing bacteria for complete destruction of coumaphos in cattle-dip waste was reported (Kearney et al., 1986; Karns et al., 1987). The use of a consortium of microbes in a filter bioreactor for destruction of coumaphos has been very successful. Two units, each capable of treating 15 000 litres of waste cattle-dip at a time, have been operational since 1996. The US Department of Agriculture has been using these units for treatment of coumaphos waste generated under its cattle fever tick eradication programme (Mulbry et al., 1998). The use of whole living cells for bioremediation presents some difficulties such as delivery of fresh inocula and nutrient composition. To avoid these difficulties, the use of cell free OPH was carried out successfully (Karns et al., 1998). It was observed that addition of non-ionic detergents and cobalt salts increased the efficiency of OPH in waste cattle-dips. Both native and recombinant OPHs, immobilized on a nylon membrane, powder and tubing (Caldwell & Raushel, 1991a), silica beads and glass (Caldwell & Raushel, 1991b) have been used for the detoxification of organophosphorus compounds. OPH from P. diminuta was immobilized based on the formation of non-composite protein-silicone polymers and was a highly active, stable and versatile biocatalyst for the liquid and gas phase detoxification of organophosphorus compounds. It was fabricated as monoliths, sheets, thick films, granulates or monoporous foams (Gill & Ballesteros, 2000). OPH and OPAA were also incorporated into an aqueous fire-fighting foam which was active and stable (Cheng et al., 1999; Raushel, 2002).
Recently, considerable efforts have gone into the development of OPH biosensors to detect contamination of organophosphorus compounds. This bioanalytical technique provides rapid, cost effective and in-field monitoring of contaminants. The use of OPH in this technique has advantages over acetylcholine esterase and enzyme-linked immunoassays (ELISA). ELISA can be sensitive but like most immunoassays, targets a single compound and requires multisteps. Acetylcholine esterase based enzyme inhibitions are well suited for screening applications; however, due to their irreversible inhibition by organophosphorus compounds, they are not particularly well suited for process control monitoring applications that require rapid and repeated measurements (Chough et al., 2002). Because OPH based biosensors respond to organophosphorus compounds as substrates rather than as inhibitors, they show considerable potential for applications that require repetitive analysis (White & Harmon, 2005). Two assay formats were used with OPH for biosensors: potentiometric measurement of local pH change (Mulchandani et al., 1998, 1999b) and amperometric measurement of electroactive enzyme products (Wang et al., 1999; Chough et al., 2002). A dual amperometric and potentiometric flow-injection biosensor detection system was developed recently which used different physical transducers simultaneously in connection with OPH. This enhanced the output and allowed discrimination between various organophosphorus compounds (Wang et al., 2002a). This biosensor, which combines the advantages of both the amperometric device and potentiometric detection, displays well-defined signals from the oxidized leaving group and has been accomplished with silicon-based pH sensitive electrolyte-insulation-semiconductor transducers (Wang et al., 2002a, 2003). While offering a fast response, such enzyme biosensors have limitations in terms of the number of samples that can be handled and discriminated among organophosphorus compounds. To overcome this problem, an on-chip enzymatic assay for screening organophosphorus nerve agents, based on pre-column reaction of OPH, electrophoretic separation of the phosphonic acid products, and their contactless conductivity detection has been developed (Wang et al., 2004). On-chip enzymatic assays combine the selectivity and amplification features of biocatalytic reactions with the analytic features and versatility of microchip devices. This new microsystem holds promise for field screening of organophosphorus compounds with the advantages of speed/warning, efficiency, portability, sample size and cost.
Information regarding the structure and function of enzymes and pathways involved in biodegradation will provide opportunities for improving enzyme activities. Catalytic mechanism and enzyme properties can be manipulated by site-directed mutagenesis guided by computer-modelled three-dimensional structure of enzymes (Cheng et al., 1999). Site-directed mutagenesis was successfully used to enhance the activity of OPH against racemic mixtures of organophosphorus enantiomers. The size and shape of the substrate binding subsites were remoulded through rational restructuring via site-directed mutagenesis (Wu et al., 2001; Raushel, 2002). However, rational design can fail sometimes due to unexpected influences exerted by substituted amino acids. Another limitation imposed by this rational approach is that only a limited sequence space can be explored at one time. Irrational approaches such as DNA shuffling, random priming and staggered extension processes have been suggested as preferable alternatives to direct the evolution of enzymes (Cheng et al., 1999). DNA shuffling was successfully used to isolate an improved variant of opd cloned E. coli, which can degrade methyl parathion 25 times faster than the wild type (Cho et al., 2002). However use of enzymes for detoxification of pesticides is not a cost-effective process. It was expected that genetic engineering would provide a means for cheaper production of microbial enzymes. The OPH encoding gene opd has been cloned under different promoters to increase the amount of OPH produced by cloned bacteria (Cheng et al., 1999).
It was suggested that the use of growing or non-growing whole cells immobilized onto supports could offer cheaper and more effective options. The major problem associated with whole cell bioreactors is mass transport limitation of substrate across the cell membrane where OPH resides (Mulchandani et al., 1999a). Uptake of organophosphorus compounds as a rate limiting factor has been reported by several groups (Hung & Liao, 1996; Elashvili et al., 1998). This barrier of substrate transport can be overcome by treating cells with permeabilizing agents such as EDTA and DMSO. However, several enzymes are sensitive to such treatment and immobilized viable cells cannot be subjected to permeabilization (Mulchandani et al., 1999a). To overcome this difficulty, OPH was successfully anchored and displayed onto the surface of E. coli using the same Lpp–OmpA fusion system used for beta-lactamase (Richins et al., 1997). Whole cells with surface expressed OPH had seven times higher activity than whole cells expressing similar amounts of OPH intracellularly. A genetically engineered E. coli expressing both OPH and cellulose-binding domain on the cell surface was constructed, enabling the simultaneous hydrolysis of organophosphorus nerve agents and immobilization via specific adsorption to cellulose (Wang et al., 2002b). OPH was expressed on the surface by the use of a truncated ice-nucleation protein-fusion system while the cellulose-binding domain was surface anchored by the Lpp–OmpA fusion system.
Microorganisms that can completely degrade and mineralize whole molecules of organophosphorus compounds have not yet been reported, but a diverse set of organisms has been isolated that are capable of collectively mineralizing these compounds. For example, some microbes can rapidly hydrolyze parathion and methyl parathion and utilize DETP as a source of carbon or phosphorus but quantitatively produce p-nitrophenol as a by-product. However, other bacteria can utilize p-nitrophenol as a source of energy. A micro-organism engineered to complete mineralization of organophosphorus compounds would avoid the generation of toxic hydrolytic products. This goal was achieved by the genetic engineering of p-nitrophenol mineralizing Moraxella sp., which was transformed with the opd gene. The truncated ice nucleation protein anchor was used to express the OPH onto the surface of the bacterium, overcoming the potential substrate uptake limitation (Shimazu et al., 2001). In another study, Walker & Keasling (2002) engineered P. putida KT 2442 to use parathion as a source of carbon and energy. Two separate plasmids, one harbouring a native opd gene (pAWW04) and another harbouring an operon encoding enzymes for p-nitrophenol transformation to β-ketoadipate (pSB337), were introduced into P. putida; the plasmids enabled the bacterium to utilize 0.8 mM parathion as a source of carbon. Recently, an E. coli expressing phosphotriesterase from A. radiobacter (opdA) and glycerolphosphodiesterase from Enterobacter aerogenes (GpdQ), which can use methyl parathion as a source of phosphorus, was used to screen for mutants with enhanced activity. This process of directed evolution produced a variant with increased protein expression and increased activity against organophosphorus (McLoughlin et al., 2005). The introduction of all degradative genes into a single organism allows for future optimization of gene expression and the potentials to utilize further directed evolution to optimize degradation rates and minimize the metabolic burden placed on the cell.
Degradation of organophosphorus compounds has attracted considerable attention because of their widespread use as pesticides, their high mammalian toxicity, and the CWC (1993). A large number of microorganisms have been isolated and characterized that can degrade organophosphorus compounds by mineralization or co-metabolism. Some microorganisms can degrade several compounds and some can degrade only one or few structurally similar organophosphorus compounds. Because hydrolysis of organophosphorus compounds reduces mammalian toxicity by several orders of magnitude, the environmental fate of degradation products has not received much attention from the scientific community. Complete pathways of parathion and glyphosate degradation are known but the pathways for several other organophosphorus compounds are not yet fully understood. This area of research needs concerted efforts as degradation products of several compounds are pollutants and may have deleterious effects on the environment and non-target organisms.
Organophosphorus degrading enzyme OPH has been characterized, its three-dimensional structure determined and its catalytic activity elucidated. Site-specific mutagenesis has been carried out successfully to increase the catalytic activity against poor substrates, and to decrease the stereoselectivity of the enzyme. In addition to the potential bioremedial use of microbes and enzymes for dealing with organophosphorus contamination in the environment, there has been considerable interest in the use of organophosphorus degrading enzymes prophylactically and therapeutically for organophosphorus poisonings (Sogorb et al., 2004; Petrikovics et al., 1999, 2000a, b). Future areas of research include increasing enzyme activity against poor substrates and improving enzyme catalytic activities in mixtures of chemicals.
Sequencing and structure determination of new proteins will provide missing links to relate and elucidate evolution mechanisms. Determining the three-dimensional structure of OPAA would be a major boost for furthering studies on the manipulation of enzymatic activity, which in turn may help in developing efficient enzymatic destruction methods for chemical warfare agents, as this enzyme has higher catalytic activity towards G-agents.
The biodegradation of chemical warfare agents has recently been a major area of research because of the urgency to destroy all stocks by 2007. Recent efforts have provided successful laboratory results. However, only a few isolated microorganisms have the capacity to degrade chemical warfare agents. A comprehensive screening of microbial dipeptidase activity from different sources may provide new gene/enzyme systems with higher activities against G-agents or V-agents. For example, screening of anaerobic microorganisms and extremophiles may be useful but this so far has received little attention for organophosphorus compound degradation. Another challenge for the scientific community is scaling-up of laboratory success to the field because of differential behaviour of isolated micro-organisms in the environment and also because stockpiles of chemical warfare agents contain a mixture of different chemical contaminants and degradation products. The application of genetic engineering and biochemical techniques to improve and evolve natural biodegradative capabilities will ultimately create strains capable of degrading complex mixtures of compounds. For example, micro-organisms were isolated which can utilise several neutralized chemical warfare agents as a phosphorus source but require addition of excess nitrogen and carbon which are rate limiting. Introducing cells containing C–P lyase activity in consortia or C–P lyase gene in degrading microorganisms might accelerate the overall degradation process.
Work in BKS laboratory is supported by a grant from the Scottish Executive Environment and Rural Affairs Department (SEERAD). We are grateful to Ms Pat Carnegie for drawing chemical structures and pathways. We also thank Drs Pete Millard, Charlie Shand, Colin Campbell (MI), Alun Morgan, Gary Bending (HRI), and Michael Kertesz (University of Manchester) for helpful discussions.