Atrazine is an herbicide of the s-triazine family that is used primarily as a nitrogen source by degrading microorganisms. While many catabolic pathways for xenobiotics are subjected to catabolic repression by preferential carbon sources, atrazine utilization is repressed in the presence of preferential nitrogen sources. This phenomenon appears to restrict atrazine elimination in nitrogen-fertilized soils by indigenous organisms or in bioaugmentation approaches. The mechanisms of nitrogen control have been investigated in the model strain Pseudomonas sp. ADP. Expression of atzA, atzB ad atzC, involved in the conversion of atrazine in cyanuric acid, is constitutive. The atzDEF operon, encoding the enzymes responsible for cyanuric acid mineralization, is a target for general nitrogen control. Regulation of atzDEF involves a complex interplay between the global regulatory elements of general nitrogen control and the pathway-specific LysR-type regulator AtzR. In addition, indirect evidence suggests that atrazine transport may also be a target for nitrogen regulation in this strain. The knowledge about regulatory mechanisms may allow the design of rational bioremediation strategies such as biostimulation using carbon sources or the use of mutant strains impaired in the assimilation of nitrogen sources for bioaugmentation.
Atrazine (2-chloro-4-ethylamino-6-isopropylamino-1,3,5-triazine) is an effective herbicide used worldwide for broad-leaf weed control. Because of its high mobility in soil and its massive application, atrazine has often been detected in surface and ground waters at concentrations well above the legal limits (Kolpin and Kalkhoff, 1993; Richards and Baker, 1993; Biradar and Rayburn, 1995; Hayes et al., 2002; 2003; Tappe et al., 2002). The high incidence of atrazine contamination, along with an increasing concern about the toxicological properties of atrazine, has prompted researchers to seek bioremediation options for atrazine-polluted sites (Allran and Karasov, 2001). Unlike most xenobiotics, atrazine and other s-triazines are used preferentially, and often exclusively, as nitrogen sources. The enzymology of the atrazine degradative pathways is now well understood (Ellis et al., 2006). However, gene regulation has scarcely been studied and is emerging as a main factor that limits atrazine bioremediation in the field.
Atrazine mineralization is achieved in two stages. Initially, the chlorine and N-alkyl side-chains are removed to yield cyanuric acid (2,4,6-trihydroxy-1,3,5-triazine). The enzymes involved in this stage often have broad substrate specificity and may perform a variety of reactions on different substrates (Shapir et al., 2007). The s-triazine ring is then cleaved and subsequently converted to ammonium and carbon dioxide by a set of enzymes that have a much narrower substrate range. Based on the enzymatic activities compiled at the University of Minnesotta Biocatalysis and Biodegradation Database (http://umbbd.msi.umn.edu/) (Ellis et al., 2006), conversion of atrazine to cyanuric acid can be achieved via three distinct pathways, one of which is purely hydrolytic, while the other two are mixed oxidative-hydrolytic (Fig. 1). The hydrolytic pathway has been extensively characterized in Pseudomonas sp. ADP, and consists of three enzymatic steps catalysed by the gene products of atzA, atzB and atzC (Martinez et al., 2001). AtzA is responsible for dechlorination of atrazine (de Souza et al., 1996), and AtzB and AtzC catalyse the elimination of N-alkyl substituents from the s-triazine ring to yield cyanuric acid (Boundy-Mills et al., 1997; Sadowsky et al., 1998). The atzA, atzB and atzC genes have been shown to be widespread and plasmid-borne in multiple atrazine-degrading isolates from all continents (de Souza et al., 1998a,b; Topp et al., 2000a; Rousseaux et al., 2001). In a few Gram-positive strains, trzN substitutes for atzA. TrzN is a broad substrate range hydrolase that performs dechlorination of atrazine and can remove multiple functional groups from a wide variety of s-triazines (Topp et al., 2000b; Shapir et al., 2005a). Oxidative-hydrolytic pathways involve the initial oxidative N-dealkylation of atrazine to deisopropylatrazine or deethylatrazine. The products of the above reactions may be dealkylated again to deisopropyldeethylatrazine, or may be subjected to hydrolytic dechlorination, deamination and/or dealkylation to yield cyanuric acid. Enzymes catalysing the oxidative reactions include Rhodococcus sp. N186/21 cytochrome P450 (Nagy et al., 1995) and AtrA from Rhodococcus sp. TE1 (Shao et al., 1995). The enzyme TriA displays a 98% identity with AtzA, but is active in deamination of atrazine derivatives rather than dehalogenation (Seffernick et al., 2001). On the contrary, TrzA from Rhodococcus corallinus NRRLB-15444R and the broad-specificity hydrolase AtzB catalyse the dechlorination of dealkylated atrazine (Cook et al., 1984; Mulbry, 1994; Seffernick et al., 2002). Although a number of bacteria harbouring the complete set of genes encoding for the hydrolytic pathway have been identified, no single organism studied to date accomplishes all the enzymatic reactions required for atrazine conversion to cyanuric acid via an oxidative-hydrolytic pathway. Degradation may nevertheless be achieved by a consortium of organisms harbouring the appropriate combination of enzymes, as shown for deisopropylatrazine with a mixed culture of R. corallinus NRRLB-15444R and Pseudomonas huttiensis NRRLB-12228 (Cook et al., 1984).
Cyanuric acid is a common intermediate of many s-triazine biodegradative pathways (Cook et al., 1985) and its metabolization invariably occurs via hydrolytic cleavage of the s-triazine ring and consecutive hydrolysis of biuret and allophanate to produce ammonium and carbon dioxide (Cheng et al., 2005). In the case of Pseudomonas sp. ADP and several other organisms (Martinez et al., 2001; Cheng et al., 2005), the enzymes responsible for these reactions are encoded by the atzDEF operon. Homologues to AtzD (TrzD) and AtzF (TrzF) with small differences in substrate affinity and specificity have been shown to perform the equivalent reactions in other bacteria (Karns, 1999; Shapir et al., 2002; Fruchey et al., 2003; Shapir et al., 2005b). Some of the best-characterized atrazine-degrading bacterial strains, along with their enzymatic activities and genes, are listed in Table 1.
Nitrogen control of atrazine degradation has been characterized extensively in Pseudomonas sp. ADP. Bichat and colleagues (1999) showed that this bacterial strain metabolizes atrazine rapidly in the presence of glycine, but significantly slower in the presence of ammonium, nitrate or urea. A later study (García-González et al., 2003) demonstrated that atrazine degradation is inhibited in resting cell suspensions previously cultured in the presence of nitrogen sources that support fast growth of Pseudomonas sp. ADP, such as ammonium, urea, proline or nitrate. In contrast, cells acclimated in media containing a growth-limiting nitrogen source, such as serine, or a pathway metabolite, such as cyanuric acid, biuret or atrazine, efficiently eliminated atrazine in the resting cell assay. The presence of atrazine in growth media containing other nitrogen sources did not stimulate degradation, indicating that the pathway is not induced by its substrate. Nitrogen control of atrazine degradation requires assimilation of the repressing nitrogen source, since the inhibitory effect is suppressed in the presence of a glutamine synthetase inhibitor (García-González et al., 2003).
The relevance of the negative effect of preferential nitrogen sources on atrazine degradation in cultures or cell suspensions to the bioremediation of contaminated soils has been addressed in Pseudomonas sp. ADP; García-González and colleagues (2003) showed that nitrate amendment inhibited atrazine mineralization in non-sterile soil microcosms bioaugmented with Pseudomonas sp. ADP. This effect is the result of nitrogen control, as evidenced by the fact that a mutant defective in nitrate assimilation effectively mineralized atrazine in the presence of nitrate in the soil microcosm assay. Inhibition of atrazine utilization by Pseudomonas sp. ADP mimics the results obtained with indigenous populations (Entry et al., 1993; Alvey and Crowley, 1995; Abdelhafid et al., 2000a,b; Guillén Garcés etal., 2007), indicating that nitrogen control is likely a prevalent trait among atrazine-degrading strains that may limit degradation in soils with high nitrogen content. Indeed, the success of biostimulation strategies using carbon sources as amendments (Abdelhafid et al., 2000a,b; Silva et al., 2004) may reflect the shift in carbon/nitrogen balance caused by the added carbon, as this is the signal sensed by general nitrogen control (Ninfa and Atkinson, 2000; Commichau et al., 2006).
Organization and regulation of the Pseudomonas sp. ADP atrazine catabolic genes
The Pseudomonas sp. ADP atz genes are harboured on the catabolic plasmid pADP-1. Two sets of genes can be distinguished with significant differences in organization and regulation. The genes atzA, atzB and atzC, encoding broad-specificity enzymes, are dispersed in an unstable region of the catabolic plasmid and flanked by insertion sequence elements (Martinez et al., 2001). Northern blot hybridization and real-time RT-PCR experiments showed that their transcription is constitutive (Martinez et al., 2001; Devers et al., 2004), and atrazine chlorohydrolase assays in crude extracts failed to reveal any post-transcriptional regulation (V. García-González, PhD thesis). New catabolic pathways normally exhibit unregulated or suboptimally regulated expression profiles, while in further evolved pathways, catabolic genes are assembled in operons whose expression responds to both substrate-specific and general physiological signals (Cases and de Lorenzo, 2001). Since atrazine was introduced in the environment about 50 years ago, it has been proposed that acquisition and evolution of atzA, atzB and atzC have taken place in recent times (Martinez et al., 2001; Shapir et al., 2007). In contrast, the three genes encoding narrow-specificity enzymes involved in the utilization of cyanuric acid are clustered in the atzDEF operon, which is located in a stable region of pADP-1 (Martinez et al., 2001). Cyanuric acid is not only a common metabolite of multiple s-triazine degradation pathways, but also a compound that is naturally present in soils (Fruchey et al., 2003). Cyanuric acid is used as a nitrogen source by multiple bacteria and supports fast growth of Pseudomonas sp. ADP (Neumann et al., 2004). Due to the frequent occurrence of cyanuric acid in nature, the atzDEF genes may have assembled and evolved to achieve high specificity prior to the acquisition of atzA, atzB and atzC via transposition or recombination events.
Consistent with its early assembly and evolution, the atzDEF operon displays a complex and sophisticated regulation that integrates a specific response to the substrate of the pathway with a general response to the physiological status of the cell (Fig. 2). Expression of atzDEF is positively regulated in response to nitrogen limitation and cyanuric acid (García-González et al., 2005). Transcription from the atzDEF promoter is strictly dependent on the product of the divergently transcribed gene, atzR (García-González et al., 2005). AtzR is a LysR-type transcriptional regulator (LTTR) and, as most members of this family of proteins, simultaneously activates the expression of the catabolic genes and represses its own synthesis by binding to a single site with conserved structure located at the divergent promoter region (Schell, 1993; Porrúa et al., 2007). Interestingly, AtzR can activate atzDEF expression in response to either nitrogen limitation or cyanuric acid, but also in response to both signals concurrently in a synergistic manner, which is an unusual feature for this type of regulators. Similarly to other LTTRs (Wang and Winans, 1995; Van Keulen et al., 2003), the interaction of AtzR with its effector molecule, cyanuric acid, triggers the repositioning of AtzR at its binding site to achieve a conformation of the AtzR–DNA complexes that is competent for activation (Porrúa et al., 2007). The nitrogen status modulates the activity of AtzR by two different mechanisms. First, transcription of atzR is a target for nitrogen regulation. In this case, the global regulator NtrC is responsible for activation of PatzR in response to nitrogen limitation (García-González et al., 2005). NtrC is an archetypical activator of σ54-RNA polymerase (Reitzer et al., 1989; Su et al., 1990; Rippe et al., 1997), and atzR is expressed from a σ54-dependent promoter, but unlike typical NtrC-regulated promoters, transcription activation does not require specific DNA binding by NtrC (Porrúa et al., 2007; O. Porrúa, unpubl. results). Second, the ability of AtzR to activate the PatzDEF promoter is promoted in nitrogen-limiting conditions. Recent studies suggest that the signal transduction protein GlnK is the element that completes the regulatory circuit and communicates the nitrogen limitation signal to AtzR via protein–protein interaction (V. García-González, PhD thesis). Expression of GlnK is also subjected to positive control by NtrC in enterobacteria (Zimmer et al., 2000) and in Pseudomonas putida KT2440 (Hervás et al., 2008). Since atzDEF is transcribed from a σ70-dependent promoter (García-González et al., 2005), this circuit allows the general nitrogen control system to simultaneously regulate the expression of both the σ54-dependent PatzR and the σ70-dependent PatzDEF.
The shape of regulatory circuit for the atzDEF operon is a typical example of a feed-forward loop, in which a global regulatory element (NtrC) simultaneously controls the synthesis of the specific regulator (AtzR) and expression of the structural genes downstream (atzDEF). Feed-forward loop circuits are often found in nature because they permit not only the integration of the general and specific regulation, but also a response of the regulated genes that is proportional to the intensity of the stimuli (Cases and de Lorenzo, 2005). It should be mentioned that a very similar arrangement is found in the cascade regulation of the Klebsiella pneumoniae nitrogen fixation genes, where NtrC activates transcription of both nifLA and glnK, and GlnK abolishes formation of the inhibitory NifL–NifA complex, allowing nifHDKTY activation (Stips et al., 2004).
Is atrazine transport a target for nitrogen control?
Nitrogen control of the cyanuric acid utilization operon atzDEF does not suffice to explain the inhibition of atrazine degradation by the presence of preferential nitrogen sources. Since expression of atzA, atzB and atzC is constitutive (Martinez et al., 2001; Devers et al., 2004), conversion of atrazine to cyanuric acid is expected to occur in any condition. Accordingly, growth in medium containing ammonium did not affect atrazine elimination in cell-free extracts or toluene-permeabilized Pseudomonas sp. ADP cells (V. García-González, unpubl. results). This is in sharp contrast with the inhibitory effect of ammonium on atrazine utilization by whole cells (García-González et al., 2003). Since disruption of the cell membrane eliminates nitrogen regulation, we believe that atrazine transport may also be a target for the nitrogen control system (Fig. 2). The existence of a specific transport system for atrazine was proposed first by Gebendinger and Radosevich (1999)Ralstonia brasilensis M91-3, based on a major discrepancy between the atrazine degradation rates in cell extracts and whole cells. Transcriptomic studies have shown that transport systems for nitrogen sources are highly represented among the functions of the genes subjected to general nitrogen control in Escherichia coli and P. putida (Zimmer et al., 2000; Hervás et al., 2008). An atrazine transporter may have evolved from a pyrimidine transport system, due to the structural similarity between the s-triazine ring and the diazine ring in pyrimidines (Wackett et al., 2002). Interestingly, synthesis of the cytosine permease CodB is subjected to general nitrogen control in both P. putida and E. coli (Zimmer et al., 2000; Hervás et al., 2008). Finally, the pADP-1 plasmid harbours genes encoding transport systems that may be suitable candidates for an atrazine transporter, such as orf46, annotated as a putative uracil/xanthine permease, or orf94-95-96-97-98, encoding a putative ABC-type transporter, which are located immediately downstream from and may be co-transcribed with atzR (Martinez et al., 2001). A similar arrangement of genes encoding a transport system, clustered with s-triazine utilization genes, is observed at the melamine degradative plasmid pPDL12 of Enterobacter chloacae 99 (GenBank sequence Accession No. AF342826), and the atrazine utilization genes of Arthrobacter aurescens TC1 (Sajjaphan et al., 2004).
Final remarks and future perspectives
While the catabolic potential of any microbial strain is dictated by its enzymatic complement, it is well known that most strains that perform well in a lab setting fail to behave similarly in natural conditions (Young and Burns, 1993). The reasons often mentioned are the suboptimal growth environmental conditions and competition with well-adapted indigenous organisms. A third factor may be the presence of environmental traits that repress the genes responsible for the catabolic activities. Recent findings point out that atrazine degradation in agricultural soils may be limited by the effect of nitrogen added as fertilizer on catabolic or transport gene expression (García-González et al., 2003), and a complex regulatory circuit that may involve nitrogen control of atrazine utilization at three different levels has been described (Fig. 2). Based on the new knowledge on nitrogen control of atrazine biodegradation, rational strategies to overcome this limitation may be designed. Some of these strategies may include biostimulation using carbon sources in order to increase the carbon/nitrogen balance (Abdelhafid et al., 2000a,b; Silva et al., 2004), the application of inhibitors of nitrogen assimilation (García-González et al., 2003), or the use of mutant strains impaired in the assimilation of nitrogen sources for bioaugmentation (García-González et al., 2003).
The work on atrazine degradation in our laboratory has been supported by Grants QLK3-CT-1999-00041 (European Union), BIO2004-01354 and BIO2007-63754 (Ministerio de Educación y Ciencia, Spain), and fellowships from the I3P (CSIC/Ministerio de Educación y Ciencia, Spain) and FPU (Ministerio de Educación y Cultura, Spain) programmes, awarded to O.P. and V.G.-G. respectively.