Facilitation of Bacterial Adaptation to Chlorothalonil-Contaminated Sites by Horizontal Transfer of the Chlorothalonil Hydrolytic Dehalogenase Gene

Abstract

Horizontal transfer of the chlorothalonil hydrolytic dehalogenase gene (chd) is proposed based on the high conservation of the chd gene and its close association with a novel insertion sequence, ISOcsp1, in 16 isolated chlorothalonil-dechlorinating strains belonging to eight different genera. The ecological role of horizontal gene transfer is assumed to facilitate bacterial adaptation to chlorothalonil-contaminated sites, through detoxification of chlorothalonil to less toxic 2,4,5-trichloro-6-hydroxybenzene-1,3-dicarbonitrile.

Microorganisms have evolved multiple mechanisms to adapt to environmental stresses, such as mutation, DNA rearrangement, and horizontal gene transfer (28). The horizontal transfer of genes plays a key role in the evolution of catabolic genes, thereby facilitating bacterial adaptation to pollutant-contaminated sites (8, 12, 22, 27). Bacterial dehalogenases catalyze the cleavage of carbon-halogen bonds of many man-made chlorinated compounds, which is a key step in the detoxification of these priority organic pollutants (20). Notably, many dehalogenase genes are associated with transmissible elements (3, 10, 11, 14–18, 21, 23, 25, 26, 29). However, the horizontal transfer of hydrolytic dehalogenase genes for chlorinated aromatics has not been studied extensively.

Chlorothalonil (2,4,5,6-tetrachlorobenzene-1,3-dicarbonitrile [TPN]), a tetrachlorinated benzonitrile fungicide, is commonly detected in ecosystems due to its wide use (5–7). TPN is toxic to fish, birds, aquatic invertebrates (4), and soil microbes (19, 24). In our laboratory, diverse TPN-dechlorinating strains have been isolated (9, 30, 31), and the novel chlorothalonil hydrolytic dehalogenase (Chd), which catalyzes a hydroxyl substitution at the 4-chlorine atom of TPN to form 2,4,5-trichloro-6-hydroxybenzene-1,3-dicarbonitrile (4-TPN-OH), was also identified (31). In this study, we discovered a close association between the highly conserved chd gene and a novel insertion sequence (IS), ISOcsp1, in diverse TPN-dechlorinating strains. We suggest that the ecological role of horizontal gene transfer is to facilitate bacterial adaptation to TPN-contaminated sites by allowing these bacteria to rapidly transform toxic TPN to less toxic 4-TPN-OH.

Sixteen TPN-dechlorinating bacterial strains (designated CTN-1 to CTN-16) were isolated from three geographically distinct TPN-contaminated sites. Strains CTN-1 to CTN-7 and strain CTN-11 were isolated as described previously (9) from the surface soil (0 to 10 cm) of Jiangyin Suli Fine Chemical Co. Ltd. (Jiangyin, China), which was directly exposed to TPN product (70 to 80% purity) for more than five years. Strains CTN-8 to CTN-10 were isolated from the company trench water by directly plating the water on Luria-Bertani (LB) plates containing 0.38 mM TPN (99.3% purity; Sigma-Aldrich) without any enrichment. Finally, strains CTN-12 to CTN-16 were isolated from the activated sludge of Xinyi Feihuang Chemical Co., Ltd. (Xinyi, China). We found that all of these isolates exhibit the same metabolic pathway and transform TPN to 4-TPN-OH in the absence of a carbon source, although they cannot use TPN for growth (9, 31). However, 16S rRNA gene analysis of these 16 isolates, in combination with their morphological, physiological, and biochemical properties, shows that they are highly diverse. These isolates belong to eight different genera (Ochrobactrum, Shinella, Caulobacter, Rhizobium, Bordetella, Pseudoxanthomonas, Pseudomonas, and Lysobacter) in the α, β, and γ branches of the Proteobacteria (Fig. 1). The different isolates in the same genus were distinguished by enterobacterial repetitive intergenic consensus sequences based on PCR.

Fig. 1.

Phylogenetic tree constructed by the neighbor-joining method based on 16S rRNA gene sequences of strains CTN-1 to CTN-16 and related species. Pseudomonas species are not clustered in a single group because the genus Pseudomonas has several phylogenetic groups. Bootstrap values (%) are indicated at the nodes. The scale bar represents 0.02 substitutions per site, and GenBank accession numbers are in parentheses.

The chd gene was amplified from all of the isolates with the primer pair ChdF (5′-GACATATGCCACTCAAGTTTTTGGG-3′)/ChdR (5′-ATCTCGAGAGGCCTGGCTGCGAGATCCTTGTAA-3′) (31), and these chd genes were found to be highly similar (99.4% to 100%). A 3,991-bp fragment containing the chd gene was cloned from Ochrobactrum sp. CTN-11 by the shotgun method as described previously (31). A 4,044-bp fragment was obtained by aligning the two sequences (GQ292539 and GQ485642) previously cloned in Pseudomonas sp. CTN-3 (31). These two fragments in strains CTN-11 and CTN-3 are nearly identical (99% similarity), and both contain three complete open reading frames (ORFs) and one truncated ORF (terminal part of orf167) (Fig. 2). Further sequence analysis revealed that a novel IS, ISOcsp1, is closely associated with the conserved chd gene. A BLAST search in the IS database (http://www-is.biotoul.fr) showed that ISOcsp1 belongs to the IS21 family. ISOcsp1 is flanked by 35-bp near-perfect inverted repeats (IRs) and contains two ORFs (Fig. 2 and Fig. 3), one of which is istA (1,023 bp), which encodes a transposase with high homology to ISBcen13 (84%). The other is istB (771 bp), which encodes an ATP-binding protein with high homology to ISBcen13 (94%). The left inverted repeat (IRL; TGTTGCGCGCAGAGTAAAACTGGGCCACTTCGCGC) is 57 bp upstream of istA, and the right inverted repeat (IRR; GCGCGAAGTGGCTCAGATTTGGGCTGCGCTTGGCA) is 32 bp downstream of istB. A direct repeat (DR; TGGG) was found outside the IRs. An outwardly directed −35 promoter hexamer for the chd gene is located just inside the IRR (Fig. 3). The presence of a −35 promoter box in the IRR of IS21-like elements seems to be a common feature of this family (2). We hypothesize that the fabricated fusion promoter might be much stronger than the original one, resulting in the chd gene being expressed at a relatively high level.

Fig. 2.

Genetic organization of the ISOcsp1 and chd gene in representative isolates from different genera. The truncated transposon containing the mercury resistance genes in strain CTN-11 is also shown. Arrows indicate the ORF and the direction of transcription. Truncated genes are indicated by arrows surrounded by a dashed line. The sequences lacking between the ISOcsp1 and the chd gene are indicated by broken lines. DR, direct repeat; IRL, left inverted repeat; istA, transposase gene; istB, ATP-binding protein gene; IRR, right inverted repeat; chd, chlorothalonil hydrolytic dehalogenase gene; Pchd, promoter of the chd gene; orf167, hypothetical protein gene; tniQ, transposition-related protein gene; tniR, resolvase gene; merE, mercury resistance protein gene; merD, mercury resistance transcriptional repressor protein gene; merB, alkyl mercury lyase gene; merA, mercury reductase gene.

Fig. 3.

Alignment of the sequences of the ISOcsp1 and the chd gene in the eight representative strains. The sequences of the istA, istB, and chd genes are omitted and indicated by the gene labels above them. The transcription start site of the chd gene is shaded. The lost base pairs are indicated by dashes. The −35 promoter motif in strain CTN-3 was TTGGCA instead of TTGACA. RBS, ribosome binding site.

Using self-formed adaptor PCR (32) with three specific primers (SP1, 5′-TGCCGAACTTGATGCCGTTGGAG-3′; SP2, 5′-TATGCTATGCAAGCCGCGCGTAA-3′; and SP3, 5′-GTGGACTGAAGAGGANNNNNNNNNGGCTGT-3′), a 3,757-bp fragment (containing a truncated mercury resistance transposon instead of the anticipated additional copy of ISO csp1) was amplified downstream of the chd gene in strain CTN-11. This noncomposite transposon shares high similarity with Tn5090/Tn402 and mercury resistance plasmids (>90% similarity), and it contains the tniQ (terminal region) and tniR resolvase genes and part of the mercury resistance operon (Fig. 2).

The entire fragment containing the conserved ISOcsp1 and the chd gene was successfully amplified with the primer pair ICF (5′-AAAACTGGGCCACTTCGCGC-3′)/ICR (5′-TCAAGGCCTGGCTGCGAGAT-3′) from six other representative TPN-dechlorinating strains in different genera (Fig. 2). These fragments show high conservation (99% similarity), and there are only slight variations in the sequences between the ISOcsp1 and the chd gene (Fig. 3).

Taken together, the short history of TPN application (less than 50 years), the taxonomic diversity of TPN-dechlorinating strains, the conservation of the chd gene, and the close association of the chd gene with ISOcsp1 all clearly suggest that the widespread distribution of the chd gene is due to horizontal transfer. We deduced that the chd gene might be transferred by the single copy of ISOcsp1. Although a single copy of IS1247 or ISPme1 can mobilize neighboring genomic DNA (1, 28), there have been no previous reports of a single copy of an IS21 family transposon transferring adjacent DNA. IS21 elements often contain long imperfect internal IRs, which are not terminal IRs (2). It is possible that the ISOcsp1-encoded transposase might mistakenly recognize genomic sequences that are functional analogs of the IRR of ISOcsp1. However, this hypothesis needs to be experimentally confirmed by further studies.

Because TPN can be used by none of these isolates as a carbon source for growth, another possible advantage obtained from horizontal gene transfer might be the detoxification of TPN, helping bacteria to improve their chance of survival in TPN-contaminated niches. The toxicity of TPN and 4-TPN-OH to bacterial strains was assessed by analyzing their toxic effects on the growth of Pseudomonas putida KT2440 (5/16 isolates were members of the genus Pseudomonas) and Escherichia coli DH5α (representative of Gram-negative bacteria). P. putida KT2440 and E. coli DH5α were cultured to the exponential phase, washed three times, and inoculated into 50 ml mineral salts medium (supplemented with 5.56 mM glucose) and 3-fold-diluted LB medium, respectively. Due to its low aqueous solubility (0.002 mM, 25°C), TPN was dissolved in acetone (50 mM) to enhance its stability and homogeneity. TPN was added to the cultures at concentrations of 0.2 mM, 0.5 mM, and 1 mM, respectively. The metabolite 4-TPN-OH (>99% purity, obtained from the transformation of TPN by Chd) was recovered using methanol (the final concentration is 50 mM) and was added to the cultures with the same concentration of TPN. To avoid any discrepancies caused by the solvents, the same volume of acetone (0.2, 0.5, and 1 ml, respectively) was added to the 4-TPN-OH treatment, and methanol was added to the TPN treatment. A blank control (only with inoculation) and a solvent treatment (with acetone and methanol) were also employed. At different growth phases, samples were collected, and cell numbers were counted by the viable plate counting method. We found that 1 mM TPN had a more serious toxic effect on the growth of both strains than the same concentration of 4-TPN-OH, although no significant differences were observed at lower concentrations (below 0.5 mM) (Fig. 4). From these results, we can see that TPN is more toxic to microorganisms than 4-TPN-OH at high concentrations and that the dechlorination of TPN to 4-TPN-OH is obviously beneficial to bacteria. Therefore, we suggest that the horizontal transfer of the chd gene could allow bacteria to rapidly transform toxic TPN to less toxic 4-TPN-OH, thereby facilitating bacterial adaptation to TPN-contaminated sites.

Fig. 4.

Effect of 1 mM TPN and 4-TPN-OH on the growth of E. coli DH5α (left) and P. putida KT2440 (right). The number of E. coli cells was counted at 2 h, 5 h, and 10 h, while the number of P. putida cells was counted at 5 h, 15 h, and 30 h. Blank control, inoculation alone; solvent treatment, acetone and methanol; TPN with solvents, TPN, acetone, and methanol; 4-TPN-OH with solvents, 4-TPN-OH, acetone, and methanol. The data are represented as the means ± standard deviations from triplicate incubations.

Nucleotide sequence accession numbers.

The sequences containing the ISOcsp1 and the chd gene were deposited in GenBank under accession numbers HQ144191 to HQ144197.