Joined: May 2002
More fun with atrazine and related compounds:
Biochemistry, 40 (43), 12747 -12753, 2001. 10.1021/bi011293r S0006-2960(01)01293-4
Web Release Date: October 3, 2001
Copyright © 2001 American Chemical Society
Rapid Evolution of Bacterial Catabolic Enzymes: A Case Study with Atrazine Chlorohydrolase
Jennifer L. Seffernick and Lawrence P. Wackett*
This review discusses examples in which it is possible to sift through the complexity of the biosphere to find related enzymes which display distinct functions. The clearest example to date is atrazine chlorohydrolase, an enzyme which is shown to have evolved for the function of catabolizing atrazine. More than 2 billion pounds of the herbicide atrazine have been applied to soils globally, and this has provided selective pressure for the evolution of new metabolism. The amino acid sequence of atrazine chlorohydrolase is shown to be 98% identical with that of melamine deaminase, an enzyme that catalyzes deamination reactions. The chlorohydrolase is shown to be firmly linked with a major amidohydrolase protein superfamily.
Atrazine Catabolism and the Amidohydrolase Superfamily
The chlorinated herbicide atrazine was once considered to be poorly biodegraded in soils. The major metabolites detected in soils and groundwaters suggested that the herbicide underwent nonspecific oxidative dealkylation reactions (Figure 1). A cytochrome P450 monooxygenase from Rhodococcus strains TE1, N186/221, and B30 was subsequently discovered to catalyze this reaction (20-24). The bacterial cytochrome P450 was shown to degrade other herbicides structurally unrelated to atrazine and is likely functioning as a nonspecific oxygenative catalyst rather than an enzyme that has evolved specifically to catabolize atrazine. Starting in 1993, however, numerous bacteria were ascertained to initiate atrazine metabolism via a hydrolytic dechlorination reaction (Figure 2). More recently, the genes encoding the chlorohydrolase have been shown to be essentially identical in different genera of bacteria independently isolated from four continents by different researchers (25). This suggests that the ability to dechlorinate atrazine arose since the introduction of atrazine and that this phenotype spread quickly around the globe.
The enzymes responsible for the first three steps of the atrazine dehalogenation pathway were initially identified in Pseudomonas sp. strain ADP (Figure 2). The enzymes that catalyze these steps are atrazine chlorohydrolase (AtzA, EC 22.214.171.124), hydroxyatrazine ethylaminohydrolase (AtzB, EC 126.96.36.199), and N-isopropylammelide N-isopropylaminohydrolase (AtzC, EC 188.8.131.52), respectively. Sequence comparisons revealed that all three enzymes belong to the amidohydrolase superfamily (26). Amidohydrolase superfamily members for which structures are defined have an ()8 barrel structure (27, 28). Moreover, they share conserved features of the reaction mechanism in which one or two divalent metals are coordinated by the enzyme and serve to activate water for nucleophilic attack on the respective substrate. The amino acids serving as metal ligands are maintained across the superfamily. The majority of reactions catalyzed by the superfamily involve the hydrolytic removal of amino groups from purine and pyrimidine rings, or amide bond hydrolysis reactions (Figure 3). The former reactions are represented by enzymes such as adenosine deaminase. The latter are illustrated by urease and cyclic amidases such as hydantoinase.
Recent studies have expanded the range of reactions that are known to be catalyzed by amidohydrolase superfamily members (Figure 3). Some of the existing enzymes catabolize synthetic organic compounds (Table 1). Phosphotriesterase, for instance, catalyzes the cleavage of a phosphorus-oxygen bond of the pesticide parathion (29). It has been speculated that the true substrate for phosphotriesterase from Pseudomonas dismuta is yet to be discovered. But it is also plausible that the enzyme has evolved under selective pressure to hydrolyze phosphotriester insecticides since their introduction some decades ago.
Other data support the view that the Pseudomonas AtzA evolved under selective pressure and was maintained in soil microbial populations to metabolize s-triazine herbicides. The atzA gene was not found in randomly chosen laboratory strains but was detected in most bacteria recently isolated for their ability to metabolize atrazine (25). It is present with other genes, atzB and atzC, which encode enzymes that metabolize the AtzA reaction product in Pseudomonas sp. ADP Ralstonia, Alcaligenes, and Agrobacterium strains (Figure 2) (25). The atrazine catabolism genes are found on large catabolic plasmids in those same strains (42).
Melamine Deaminase and s-Triazine Hydrolase
Perhaps the best evidence that atzA is a recently evolved gene derives from its relationship with genes identified for the catabolism of melamine, or 2,4,6-triamino-1,3,5-triaizine. Melamine is an industrial product used since the early 1900s. Melamine was considered nonbiodegradable in the 1930s but was then reclassified as slightly biodegradable in the 1960s when atrazine was first introduced (43). Today, it is considered to be readily biodegradable in soil. Among the bacteria that metabolize melamine is Acidovorax avenae citrulli 12227 (formerly Pseudomonas sp. strain NRRL B-12227) (44). The first two metabolic reactions are sequential hydrolytic deamination reactions catalyzed by the same enzyme, melamine deaminase (TriA). The triA gene has recently been cloned and sequenced. The protein shows a remarkable identity to atrazine chlorohydrolase from Pseudomonas sp. ADP; it is the same in 466 of 475 amino acids (Figure 4) (45). It is also unusual that the nine nucleotide differences between triA and atzA give rise to these nine amino acid changes. The small number of changes and the absence of silent mutations are consistent with an intense selective pressure operating over a short evolutionary time period (46, 47). The kcat/Km of atrazine chlorohydrolase with atrazine is 1.5 × 104 s-1 M-1 per subunit. In our most recent study, the deamination activity of this enzyme was found to be undetectable (48). Melamine deaminase, however, exhibits the opposite specificity. It catalyzes deamination reactions at rates comparable to dechlorination rates of atrazine chlorohydrolase. Moreover, it shows dechorination activity 2 orders of magnitude lower than the deamination activity with comparable triazine substrates. In total, these data suggest that the nine amino acid changes represent a short evolutionary trajectory between the two activities.
The sequence of a related amidohydrolase superfamily member, s-triazine hydrolase or TrzA (49), is 41 and 42% identical with the sequences of atrazine chlorohydrolase and melamine deaminase, respectively. It catalyzes both deamination and dechlorination reactions. TrzA catalyzes the deamination of nonalkylated triazines such as melamine and the dechlorination of mono-N-alkylated triazines. The kcat for deamination of melamine is 243 s-1, while that for the dechlorination of desisopropylatrazine is 2.2 s-1. This is an approximately 100 times greater preference for demination over dechlorination and is consistent with the enzyme acting physiologically as a deaminase with a fortuitous dechlorination activity. This is not surprising given that chloride displacement is more facile, and adenosine deaminase is known to catalyze fortuitous halopurine dehalogenation. That TriA and AtzA discriminate between chloro and amino substrates so well despite their sequences being 98% identical is remarkable.
It is possible that fewer than nine amino acid changes are required to interconvert melamine deaminase and atrazine chlorohydrolase activities. There are 510 possible site-directed mutants bridging the two, a large set to generate, sequence, purify, and assay. In this context, DNA shuffling was conducted and the clonal variants were screened against a chemical library of substrates using high-throughput mass spectrometry (48). The chemical library of 15 substrates varied the leaving group and the side chains (Figure 5). Mutant enzymes were obtained that varied with respect to their activities against the different substrates. The sequences of daughter enzymes exhibiting the greatest activity for hydrolysis of atrazine analogues are displayed in Table 2. The activities of the shuffled clones were normalized to the activity of each parental enzyme. The clone with the best dechlorination activity was 1.4 times as fast as atrazine chlorohydrolase, and the clone with the best deamination activity was 3.6 times better than melamine deaminase. The small increases observed in activity upon shuffling suggest that atrazine chlorohydrolase and melamine deaminase have among the most optimal sequences for dechlorination and deamination activities, respectively.
It is also of potential evolutionary significance that shuffled mutants were obtained with 80-fold enhanced activities with substrates containing methyl thioether and methoxy substituents. These represent the commercially relevant herbicides ametryn and atraton, respectively. An enzyme purified from a Nocardioides sp. was shown to hydrolyze ametryn, but it was not tested with atraton or other methoxy-functionalized herbicides (50). DNA from the Nocardioides sp. did not hybridize to an atzA probe, suggesting that the enzyme does not closely resemble atrazine chlorohydrolase from Pseudomonas sp. ADP. However, the data in Table 2 suggest that enzymes capable of metabolizing ametryn, atraton, and related triazine herbicides could be derived from triA or closely homologous genes in nature.
With respect to the sequences that favor dechlorination versus deamination, the data show a trend in that residue 328 appears to largely control leaving group specificity. Asn328 tracks with narrow specificity enzymes that largely catalyze dechlorination. Asp328 tracks with broader specificity enzymes which catalyze deamination and the displacement of -NCH3, -OCH3, and -SCH3 groups. The hypothesis that this residue is crucial to the observed specificity difference between melamine deaminase and atrazine chlorohydrolase is currently being addressed with site-directed mutagenesis studies.
Nature must continually fine-tune enzyme substrate specificities and reaction rates over time under the aegis of biological need, usually called selective evolutionary pressure. This enzyme variability is particularly marked with soil bacteria due to their enormous numbers, large evolutionary span of 3.6 billion years, rapid reproductive rates, and great competition for scarce nutrient resources. Enzyme plasticity is important in this context, but this confounds genome annotation efforts where gene function is assigned on the basis of finding the homologue with the most identical sequence. As discussed here, enzymes with sequences that are 98% identical can catalyze different reactions. It will be imperative to flesh out a broader range of microbial enzymatic reactions, particularly for microbial catabolic enzymes where the diversity of enzymes will likely be great.