Molecular mechanisms of adaptation of soil bacteria to chlorinated benzenes

The pollution of our environment with a large number of different organic compounds poses a serious threat to existing life, since many of these chemicals are toxic or are released in such quantities that exceed the potential of biological detoxification and degradation systems. Bacteria and other microorganisms play an essential role in the breakdown of xenobiotic compounds. Microbes use these compounds as carbon and energy source and metabolize them to harmless endproducts. However, not all compounds are easily metabolized and some structures resist the action of existing enzyme systems in bacteria. Nevertheless, bacterial species have been isolated which have overcome these metabolic barriers and completely metabolize chemicals that were previously considered to be persistent.The project of this thesis was initiated to study the genetic mechanisms in bacteria that cause adaptation to use xenobiotic compounds as novel growth substrates (see Chapter I for a review). The work presented here mainly focused on one class of compounds, i. e. lower chlorinated benzenes such as dichlorobenzenes (DCB) and 1,2,4- trichlorobenzene (1,2,4-TCB). These compounds were known to be very resistant to biodegradation by bacteria. A number of bacterial species was isolated by enrichment techniques which were able to use DCB's and/or 1,2,4-TCB as sole carbon and energy source for growth. One of these bacteria, Pseudomonas sp. strain P51, was investigated further in this study. We have obtained strong evidence that the pathway for chlorobenzene metabolism arose as a consequence of the unique combination of two gene clusters, each specifying part of the complete pathway. These individual gene clusters are not uncommon and probably exist separately in different bacteria. Our results suggest that one of the gene clusters is contained in a novel transposable element that may have been acquired by strain P51 and integrated into a catabolic plasmid that already contained the other gene cluster. A further fine-tuning of the new pathway may have occurred through specialization of individual enzymes towards novel metabolic intermediates and by changes in the regulatory system in response to novel inducer molecules.The degradation of DCB's and 1,2,4-TCB was studied at concentrations between 10 μg/l and 1 mg/l in soil percolation columns filled with sediment of the Rhine river, which in some cases were inoculated with Pseudomonas sp. strain P51 (Chapter 2). In the inoculated columns, DCB's and 1,2,4-TCB were instantly degraded. Strain P51 remained viable in the column as long as the chlorinated benzenes were fed in the influent. Interestingly, minimal concentrations of the chlorinated benzenes were measured in the effluent of the columns, independently of the influent concentrations used (6 ± 4 μg/l for 1,2-DCB; 20 ± 5 μg/l for 1,2,4-TCB; more than 20 μg/l for 1,3-DCB and 1,4-DCB), which could not be lowered by additional inoculations with strain P51. The native microbial population in the noninoculated columns adapted to degrade 1,2-DCB after a lag phase of about 60 days, and was then able to remove a concentration of 25 μg/l of 1,2-DCB in the influent to less than 0.1 μg/l.Detailed genetic studies were carried out with Pseudomonas sp. strain P51 to characterize the genetic determinants for chlorobenzene metabolism. A large plasmid of 110 kilobase-pairs (kb) (pP51) could be isolated from cells that were cultivated on 1,2,4- TCB (Chapter 3). This plasmid could be cured from the strain by successive inoculations on non-selective media, rendering the strain incapable of metabolizing chlorinated benzenes. Subsequent cloning and deletion experiments in Escherichia coli, Pseudomonas putida, and Alcaligenes eutrophus showed that two regions on plasmid pP51 were responsible for chlorobenzene metabolism. Expression studies in E. coli revealed that a 5-kb region encoded the activity to convert 1,2,4-TCB and 1,2-DCB to 3,4,6-trichlorocatechol and 3,4-dichlorocatechol, respectively. This activity was determined using whole cell incubations, and in analogy with other described catabolic pathways it was proposed that the activity was caused by a chlorobenzene dioxygenase multienzyme complex and a dehydrogenase (encoded by tcbA and tcbB, respectively). Separated from the chlorobenzene dioxygenase gene cluster by approximately 6 kb a region was located which contained the genes for the conversion of chlorocatechols. Different DNA fragments of this region of pP51 were cloned in expression vectors in E. coli, P. putida and A. eutrophus. Both P.putida KT2442 and A. eutrophus JMP222 could be complemented for growth on 3-chlorobenzoate by a 13-kb fragment of pP51, which indicated that a functional pathway for degradation of chlorocatechols was encoded on this fragment. Enzyme activity measurements and transformation reactions with 3,4-dichlorocatechol in cell extracts of E. coli harboring cloned pP51 DNA fragments showed the activity of three enzymes, chlorocatechol 1,2-dioxygenase (catechol 1,2-dioxygenase II), chloromuconate cycloisomerase (cycloisomerase II), and dienelactone hydrolase II. The genes encoding these activities were designated tcbC, tcbD, and tcbE, respectively, and their deduced gene order was found to be tcbC-tcbD- tcbE. It was thus proposed that 3,4-dichlorocatechol was converted via a chlorocatechol oxidative pathway (or modified ortho cleavage pathway), similar to that described in Pseudomonas sp. strain B 13 and A. eutrophus JMP134 , leading finally to the formation of 5-chloromaleylacetate. The release of one chlorine atom from 3,4- dichlorocatechol was shown to take place spontaneously during lactonization in the cycloisomerization reaction.The genes of the chlorocatechol oxidative pathway of strain P51 are organized in a single operon, comprising a region of 5.5 kb, which was fully sequenced and contained five large open reading frames (Chapter 4). The gene products of the different open reading frames were analyzed by subcloning appropriate pP51 DNA fragments in E. coli expression vectors. Expression studies confirmed the previously determined gene order and could attribute three open reading frames to the gene loci tcbC, tcbD, and tcbE, respectively. In between tcbD and tcbE an 1,022 bp open reading frame was present (ORF3), but we could not detect any protein encoded by this ORF. Immediately downstream of tcbE another ORF was found, designated tcbF, which encoded a 38 kDa protein. Until now, no clear function has been attributed for the tcbF gene product. The tcbCDEF genes and their encoded gene products showed high (50.6% - 75.7%) homology to two other chlorocatechol oxidative gene clusters, clcABD of P.putida (pAC27) and tfdCDEF of A. eutrophus JMP134(pJP4). Furthermore, a homology of 22% and 43.9% was found of TcbC and TcbD to CatA and CatB, respectively, the catechol 1,2-dioxygenase and cycloisomerase of the β-ketoadipate pathway of Acinetobacter calcoaceticus. This suggests that the chlorocatechol oxidative pathway originated from other, more common, metabolic pathways. Despite the strong DNA and amino acid sequence homology of the genes and enzymes of the chlorocatechol oxidative pathways, the substrate range of the pathway enzymes from the three organisms differed subtly. This was demonstrated for the chlorocatechol 1,2- dioxygenases TcbC, ClcA, and TfdC. In contrast to ClcA and TfdC, which showed a high relative activity for 3,5-dichlorocatechol, TcbC exhibited a strong preference for 3,4- dichlorocatechol and a weak affinity for the 3,5-isomer. This suggested that the tcb -encoded pathway enzymes had become specialized for intermediates (i.e. 3,4- dichlorocatechol) which arise in the metabolism of the novel compound 1,2- dichlorobenzene. Different genetic mechanisms may cause the divergence of genes and allow a specialization of encoded proteins (see also Chapter 1). Recently it has been proposed that slippage of short sequence repetitive motifs and subsequent mismatch repair would be the major driving force for rapid evolutionary divergence, rather than single base-pair substitutions. Detailed DNA sequence comparisons between the chlorocatechol 1,2-dioxygenase genes tcbC , clcA , and tfdC gives evidence for slippage of short sequence repetitions in regions of strong divergence in amino acid sequence.The transcription of the tcbCDEF operon was found to be regulated by the gene product of tcbR, a gene located upstream of and divergently transcribed from the tcbC gene. The tcbR gene was characterized by DNA sequencing and expression studies in E. coli and appeared to encode a 32 kDa protein (Chapter 5). The activity of the tcbR gene was analyzed in P.putida KT2442 harboring the cloned tcbR and tcbCDEF genes by determining the activity of the chlorocatechol 1,2-dioxygenase TcbC during growth on 3-chlorobenzoate. Strains of P.putida KT2442, which carried a frame shift mutation in the tcbR gene, could no longer induce tcbC expression during growth on 3-chlorobenzoate, suggesting that TcbR functions as a positive regulator of tcbC expression. A region of 150-bp is separating tcbR and tcbC, the first gene of the tcbCDEF cluster, and contains the expression signals needed for the transcriptional activation of tcbCDEF by the tcbR gene product. The transcriptional start sites of tcbR and tcbC were determined by primer extension analysis and this showed that the two divergent promoter sequences of the genes overlap. Protein extracts of both E. coli overproducing TcbR and of Pseudomonas sp. strain P51 showed specific DNA binding to this 150-bp region. TcbR probably regulates tcbCDEF expression and autoregulates its own expression, by binding the DNA region containing the promoters of tcbC and tcbR. It is likely that an inducer molecule interacts with TcbR, which may cause alterations or partially unwinding of the bound region and stimulation of RNA polymerase to start transcription of the tcbCDEF operon. Amino acid sequence comparisons indicated that TcbR is a member of the LysR family of transcriptional activator proteins and shares a high degree of homology with other activator proteins involved in regulating the catabolism of aromatic compounds, such as CatM, CatR and NahR. Detailed studies have recently been carried out to determine the precise interaction of TcbR with its operator region by DNasel footprinting. It would be interesting to see if in analogy with the specialization of TcbC, TcbR has diverged from a more common regulator protein such as CatM or CatR, and became specialized in recognizing chorinated inducer molecules.DNA sequence analysis of the start of the chlorobenzene dioxygenase cluster revealed the presence of an insertion element, IS 1066 (Chapter 6). An almost exact copy of this element, IS 1067, was discovered on the other side of this gene cluster, although oriented in an inverted position. Thus, the complete genetic element formed by IS 1066, the tcbAB gene cluster, and IS 1067, resembled a composite bacterial transposon. The functionality of this transposon, which was designated Tn 5280 , was established by inserting a 12-kb Hin dIII fragment of pP51 containing Tn 5280 , marked with a kanamycin resistance gene in between the IS-elements, into the suicide donor plasmid pSUP202 followed by conjugal transfer to P.putida KT2442. Analysis by DNA hybridization of transconjugants with acquired kanamycin resistance showed that Tn 5280 had transposed into the genome of this strain at random and in single copy. The insertion elements IS 1066 and IS 1067 were found to be highly homologous to a class of repetitive elements of Bradyrhizobium japonicum and Agrobacterium rhizogenes, and were distantly related to IS 630 of Shigella sonnei. The presence of the tcbAB genes on Tri 5280 suggested a mechanism by which a chlorobenzene dioxygenase gene cluster was mobilized as a gene module by the mediation of IS-elements. This gene module was then joined with the chlorocatechol gene cluster to form the novel chlorobenzene pathway.To obtain more information on the distribution of chlorobenzene degradation genes in the environment, different methods were applied which were based on DNA- DNA hybridization with gene probes derived from chloroaromatic metabolism (Chapter 7). A number of bacterial strains which were isolated by selective enrichment from soil samples for growth on chloroaromatic compounds .was screened for the presence of catabolic plasmids. Hybridization of these plasmid-DNA's with DNA fragments of the tcbAB or tcbCDEF genes revealed a class of plasmids which were identical or homologous to plasmid pP51 of strain P51. In other experiments soil microorganisms were directly extracted from soil samples, plated on nonselective media and screened by DNA-DNA colony hybridization for the presence of catabolic genes with a set of probes for three chlorocatechol 1,2-dioxygenase genes (tcbC, clcA, and tfdC). Positively reacting colonies were obtained under selective conditions with a frequency of 1 to 5 per 2000, which indicated that in the soil samples microorganisms were present which contained DNA sequences homologous to the used probes. However, from additional screening and hybridization experiments of these positively reacting colonies it became clear that some of these were false positives. Furthermore, positive strains were lost easily during transfer from the original agar plates due to the heterogeneity in colony types of the different soil microorganisms. In a third method the variation of chlorocatechol 1,2-dioxygenase genes among soil microorganisms was analyzed by amplifying total DNA from soil samples in the polymerase chain reaction, which was primed with degenerate oligonucleotides derived for conserved regions in tcbC, clcA, and tfdC. Discrete amplified fragments were obtained in this manner, which were cloned and analyzed by hybridization and DNA sequencing. We found six different types of fragments which had the expected size, only one of which was related significantly to the chlorocatechol 1,2-dioxygenase (and in fact was identical to the tcbC- type). This indicated that it was possible to detect and isolate chlorocatechol 1,2-dioxygenase sequences from soil DNA although the selectivity of the amplification reaction was relatively low.In this study, we have entered a field of microbial research which will have continuing evolutionary and environmental interest. A detailed genetic characterization of bacteria which adapted to use xenobiotic compounds as novel growth and energy subsrates, suggested different mechanisms by which novel metabolic pathways evolve in bacteria. Our results presented evidence for i) a specialization of enzyme systems and ii) an exchange and combination of pre-existing gene modules. Still we do not know what the capacity of microorganisms present in the natural environment is to adapt rapidly to metabolize xenobiotic substrates, nor do we know how and which environmental factors influence genetic adaptation. Astonished by the diversity of genetic mechanisms displayed in bacteria which govern evolutionary change, we shouldn't be surprised to find mechanisms which direct and regulate genetic adaptation in response to changing environmental conditions.