The distribution and divergence of DNA sequences related to the Tn21 and Tn501 mer operons

The mercury resistance (mer) operons of the Gram-negative bacterial transposons, Tn21 and Tn501, are phenotypically indistinguishable and have extensive DNA identity. However, Tn21 mer has an additional coding region (merC) in the middle of the operon which is lacking in Tn501 and there is also a discrete region of the mercuric ion reductase gene (merA) which differs markedly between the two operons. DNA fragment probes were used to determine the distribution of specific mer coding regions in two distinct collections of mercury-resistant (Hgr) Gram-negative bacteria. Colony blot hybridization analysis showed that merC-positive operons occur almost exclusively in Escherichia, although merC-negative operons can also be found in this genus. The merC-negative operons were found in Citrobacter, Klebsiella, and Enterobacter and in some Pseudomonas. Most of the Pseudomonas did not hybridize detectably with either of the two operons studied, indicating that they harbor an unrelated or more distantly related class of mercury resistance locus. Southern hybridization patterns demonstrated that the merC-positive mer operon is well conserved at the DNA level, whereas the merC-negative operons are much less conserved. The presence of merC also correlated with conservation of a specific variant region of the merA gene and with an antibiotic resistance pattern similar to that of Tn21. Tn501 appears to be an atypical example of the merC-negative subgroup of Hgr loci.     

Translation of merD in Tn21.

All four sequenced examples of the mercury resistance (mer) operon of gram-negative bacteria have a promoter-distal reading frame, merD, whose removal has little effect on the resistance phenotype and whose translation has not previously been observed. Using merD-lacZ protein fusions, we show that merD is translated. However, Hg(II)-induced merD expression, as measured by beta-galactosidase activity and immunoblotting, is 10- to 15-fold lower than that of fusions to the gene immediately preceding it, merA.     

Restriction pattern and polypeptide homology among plasmid-borne mercury resistance determinants

The structural and functional properties of mercury resistance determinants cloned from a series of independently isolated conjugative plasmids were compared with those of the prototype HgR determinants from Tn501 and plasmid R100 (containing Tn21). Restriction endonuclease mapping classified the HgR determinants into at least three different but related structural groups which are distantly related to those from Tn501 and R100. These relationships were confirmed by the functional analysis of sub-clones and gamma delta insertion mutations and from the polypeptides specified by the cloned HgR determinants. Each mercury resistance clone synthesized polypeptides equivalent in size to the merA, merT, and merP gene products. However, those for merA and merT showed considerable size variation. No polypeptide equivalent to merD or merC of R100 was detected.     

Polypeptides specified by the mercuric resistance (mer) operon of plasmid R100

Overlapping deletion mutations were constructed in chimaeric plasmids carrying the mer operon of plasmid R100. Polypeptides specified by the mutant plasmids in Escherichia coli minicells correlated with the mer genes as follows: merT, 17- and 16-kDa polypeptides; merP, 9.8- and 9.5-kDa polypeptides; merC, a 14-kDa polypeptide; merA, 65- and 62-kDa polypeptides. The products of the merR and merD genes were not identified. The revised nomenclature of the mer genes is explained.     

Tn5037-a Tn21-like mercury transposon, detected in Thiobacillus ferrooxidans

The 6645-bp mercury resistance transposon of the chemolithotrophic bacterium Thiobacillus ferrooxidans was cloned and sequenced. This transposon, named Tn5037, belongs to the Tn21 branch of the Tn21 subgroup, many members of which have been isolated from clinical sources. Having the minimum set of the genes (merRTPA), the mercury resistance operon of Tn5037 is organized similarly to most of the Gram-negative bacteria mer operons and is closest to that of Thiobacillus 3.2. The operator-promoter region of the mer operon of Tn5037 also has the common (Tn21/Tn501-like) structure. However, its inverted, presumably MerR protein binding repeats in the operator/promoter element are two base pairs shorter than in Tn21/Tn501. In the merA region, this transposon shares 77.4, 79.1, 83.2 and 87.8% identical bases with Tn21, Tn501, T. ferrooxidance E-15, and Thiobacillus 3.2, respectively. No inducibility of the Tn5037 mer operon was detected in the in vivo experiments. The transposition system (terminal repeats plus gene tnpA) of Tn5037 was inactive in Escherichia coli K12, in contrast to its resolution system (res site plus gene tnpR). However, transposition of Tn5037 in this host was provided by the tnpA gene of Tn5036, a member of the Tn21 subgroup. Sequence analysis of the Tn5037 res site suggested its recombinant nature.     

Mercuric reductase structural genes from plasmid R100 and transposon Tn501: functional domains of the enzyme

The nucleotide sequence for the 2240 bp of plasmid R100 following the merC gene of the mercuric resistance operon has been determined and compared with the homologous sequence of transposon Tn501. The sequences following merC and preceding the next structural gene merA are unrelated between R100 and Tn501 and differ in length, with 72 bp in Tn501 and 509 bp in R100. The R100 sequence has a potential open reading frame (ORF) for a 140 amino acid polypeptide with a reasonable translational start signal preceding it. The merA genes contain 1686 (Tn501) and 1695 (R100) bp respectively. When optimally aligned, the merA sequences differ in 18% of their positions. These differences were clustered in specific regions. In addition, there was one nucleotide triplet in the Tn501 sequence which has no counterpart in the R100 sequence and one dodecyl-nucleotide sequence in the R100 sequence without counterpart in Tn501. Thus the predicted merA polypeptide of Tn501 contains 561 amino acids and the R100 counterpart contains 564 amino acids. Comparison of the R100 mercuric reductase sequences with that for human glutathione reductase [Krauth-Siegel et al.: Eur. J. Biochem. 121 (1982) 259-267], for which there is a 2 A resolution electron density map [Thieme et al.: J. Mol. Biol. 152 (1981) 763-782] shows a strong homology, with 26% identical amino acids and many conservative substitutions. This homology allows the conclusion that the active site of these enzymes and the contact positions for flavin adenine dinucleotide (FAD) and NADPH are highly conserved, while the amino- and carboxyl-terminal sequences differ.     

Deletion mutant analysis of the Staphylococcus aureus plasmid pI258 mercury-resistance determinant

Deletion mutant analysis of the mercury-resistant determinant (mer operon) from the Staphylococcus aureus plasmid pI258 was used to verify the location of the merA and merB genes and to show the existence of mercuric ion transport gene(s). ORF5 was confirmed to be a transport gene and has an amino acid product sequence homologous to the merT gene products from several gram-negative bacteria and a Bacillus species. Deletion analysis established that inactivation of merA on a broad-spectrum mer resistance determinant resulted in a mercury-hypersensitive phenotype. Gene dosage had no apparent effect on the level of resistance conferred by the intact mer operon or on the expression of an inducible phenotype, except that when the intact pI258 mer operon was on a high copy number plasmid, uninduced cells possessed a volatilization rate that was at most only 3.5-fold less than that observed for induced cells. There was no need for mercury ion transport proteins for full resistance when the mer operon was expressed in a high copy number plasmid.     

Electrotransformation of Thiobacillus ferrooxidans with plasmids containing a mer determinant.

The mer operon from a strain of Thiobacillus ferrooxidans (C. Inoue, K. Sugawara, and T. Kusano, Mol. Microbiol. 5:2707-2718, 1991) consists of the regulatory gene merR and an operator-promoter region followed by merC and merA structural genes and differs from other known gram-negative mer operons. We have constructed four potential shuttle plasmids composed of a T. ferrooxidans-borne cryptic plasmid, a pUC18 plasmid, and the above-mentioned mer determinant as a selectable marker. Mercury ion-sensitive T. ferrooxidans strains were electroporated with constructed plasmids, and one strain, Y4-3 (of 30 independent strains tested), was found to have a transformation efficiency of 120 to 200 mercury-resistant colonies per microgram of plasmid DNA. This recipient strain was confirmed to be T. ferrooxidans by physiological, morphological, and chemotaxonomical data. The transformants carried a plasmid with no physical rearrangements through 25 passages under no selective pressure. Cell extracts showed mercury ion-dependent NADPH oxidation activity.     

Analysis of mer Gene Subclasses within Bacterial Communities in Soils and Sediments Resolved by Fluorescent-PCR-Restriction Fragment Length Polymorphism Profiling

Bacterial mer (mercury resistance) gene subclasses in mercury-polluted and pristine natural environments have been profiled by Fluorescent-PCR-restriction fragment length polymorphism (FluRFLP). For FluRFLP, PCR products were amplified from individual mer operons in mercury-resistant bacteria and from DNA isolated directly from bacteria in soil and sediment samples. The primers used to amplify DNA were designed from consensus sequences of the major subclasses of archetypal gram-negative mer operons within Tn501, Tn21, pDU1358, and pKLH2. Two independent PCRs were used to amplify two regions of different lengths (merRT(Delta)P [ca. 1 kb] and merR [ca. 0.4 kb]) starting at the same position in merR. The oligonucleotide primer common to both reactions (FluRX) was labelled at the 5(prm1) end with green (TET) fluorescent dye. Analysis of the mer sequences within databases indicated that the major subclasses could be differentiated on the basis of the length from FluRX to the first FokI restriction endonuclease site. The amplified PCR products were digested with FokI restriction endonuclease, with the restriction digest fragments resolved on an automated DNA sequencing machine which detected only those bands labelled with the fluorescent dye. For each of the individual mer operon sources examined, this single peak (in bases) position was observed in separate digests of either amplified region. These peak positions were as predicted on the basis of DNA sequence. mer PCR products amplified from DNA extracted directly from soil and sediment bacteria were studied in order to determine the profiles of the major mer subclasses present in each natural environment. In addition to peaks of the expected sizes, extra peaks were observed which were not predicted on the basis of DNA sequence. Those appearing in the restriction endonuclease digests of both study regions were presumed to be novel mer types. Genetic heterogeneity within and between mercury-polluted and pristine sites has been studied by this technique. Profiles generated were highly similar for samples taken within the same soil type. The profiles, however, changed markedly on crossing from one soil type to another, with gradients of the different groupings of mer genes identified.     

Constitutive synthesis of a transport function encoded by the Thiobacillus ferrooxidans merC gene cloned in Escherichia coli.

Mercuric reductase activity determined by the Thiobacillus ferrooxidans merA gene (cloned and expressed constitutively in Escherichia coli) was measured by volatilization of 203Hg2+. (The absence of a merR regulatory gene in the cloned Thiobacillus mer determinant provides a basis for the constitutive synthesis of this system.) In the absence of the Thiobacillus merC transport gene, the mercury volatilization activity was cryptic and was not seen with whole cells but only with sonication-disrupted cells. The Thiobacillus merC transport function was compared with transport via the merT-merP system of plasmid pDU1358. Both systems, cloned and expressed in E. coli, governed enhanced uptake of 203Hg2+ in a temperature- and concentration-dependent fashion. Uptake via MerT-MerP was greater and conferred greater hypersensitivity to Hg2+ than did uptake with MerC. Mercury uptake was inhibited by N-ethylmaleimide but not by EDTA. Ag+ salts inhibited mercury uptake by the MerT-MerP system but did not inhibit uptake via MerC. Radioactive mercury accumulated by the MerT-MerP and by the MerC systems was exchangeable with nonradioactive Hg2+.     

The merR regulatory gene in Thiobacillus ferrooxidans is spaced apart from the mer structural genes

Two distinct merR genes, which regulate expression of the mercuric ion resistance gene (mer), of Thiobacillus ferrooxidans strain E-15 have been cloned, sequenced and termed merR1 and merR2. As a result of gene walking around two merR genes, it was found that these two genes were quite close in distance. The nucleotide sequence of the region (5,001 base pairs; PstI-EcoRI fragment) containing the merR genes was determined. Between the two merR genes, there were five potential open reading frames (ORFs). Two of these were identified as merC genes, and the other three as ORFs 1 to 3. ORFs 1 to 3 show significant homology to merA, tnsA from transposon Tn7, and merA, respectively. Both merR genes consist of a 408 bp ORF coding for 135 amino acids. Their gene products, MerR1 and MerR2, differed at three amino acid positions, and shared 56-57% and 32-38% identity with the MerRs from other Gram-negative and Gram-positive bacteria, respectively. Competitive primer extension analysis revealed that both regulatory genes were expressed in the host cells. These merR genes were located more than 6 kb from either end of the mer structural genes (merC-merA). This is the first example of merR being separated from the mer structural genes. The two merC genes, each of which coded for a 140-amino-acid protein, appeared to be functionally active because Escherichia coli cells carrying these merC genes on plasmid vectors showed hypersensitivity to HgCl2. However, ORFs 1 and 3, which were homologous to merA, seemed to be inactive both structurally and enzymatically. The gene arrangement in this region took on a mirror image, with the truncated tnsA as the symmetrical centre. It is suggested that the Tn7-like factor may have participated in gene duplication events of the mer region, and in its chromosomal integration.     

Simultaneous detection and removal of organomercurial compounds by using the genetic expression system of an organomercury lyase from the transposon Tn<Emphasis Type="Italic">MERI1</Emphasis>

Using a newly identified organomercury lyase gene (merB3) expression system from Tn MERI1, the mercury resistance transposon first found in Gram-positive bacteria, a dual-purpose system to detect and remove organomercurial contamination was developed. A plasmid was constructed by fusing the promoterless luxAB genes as bioluminescence reporter genes downstream of the merB3 gene and its operator/promoter region. Another plasmid, encoding mer operon genes from merR1 to merA, was also constructed to generate an expression regulatory protein, MerR1, and a mercury reductase enzyme, MerA. These two plasmids were transformed into Escherichia coli cells to produce a biological system that can detect and remove environmental organomercury contamination. Organomercurial compounds, such as neurotoxic methylmercury at nanomolar levels, were detected using the biomonitoring system within a few minutes and were removed during the next few hours.     

Nucleotide sequence of a chromosomal mercury resistance determinant from a Bacillus sp. with broad-spectrum mercury resistance.

A 13.5-kilobase HindIII fragment, bearing an intact mercury resistance (mer) operon, was isolated from chromosomal DNA of broad-spectrum mercury-resistant Bacillus sp. strain RC607 by using as a probe a clone containing the mercury reductase (merA) gene. The new clone, pYW33, expressed broad-spectrum mercury resistance both in Escherichia coli and in Bacillus subtilis, but only in B. subtilis was the mercuric reductase activity inducible. Sequencing of a 1.8-kilobase mercury hypersensitivity-producing fragment revealed four open reading frames (ORFs). ORF1 may code for a regulatory protein (MerR). ORF2 and ORF4 were associated with cellular transport function and the hypersensitivity phenotype. DNA fragments encompassing the merA and the merB genes were sequenced. The predicted Bacillus sp. strain RC607 MerA (mercuric reductase) and MerB (organomercurial lyase) were similar to those predicted from Staphylococcus aureus plasmid pI258 (67 and 73% amino acid identities, respectively); however, only 40% of the amino acid residues of RC607 MerA were identical to those of the mercuric reductase from gram-negative bacteria. A 69-kilodalton polypeptide was isolated and identified as the merA gene product by examination of its amino-terminal sequence.     

Physical and genetic map of the organomercury resistance (Omr) and inorganic mercury resistance (Hgr) loci of the IncM plasmid R831b

Tn7 insertion mutagenesis has been used to facilitate the generation of a physical (restriction endonuclease) and genetic map of the IncM plasmid, R831b. The only selectable phenotypes carried by this 90-kb conjugative plasmid are resistances to inorganic mercury [Hg(II)] and to organomercury compounds. Mutants in the Hgr locus of R831b complemented previously described mutants in the mer operon of the IncFII plasmid R100, indicating functional homology of the locus in each of these different plasmids. However, the R831b Hgr locus is not notably similar in restriction site pattern to either the mer operon of R100 or the mercury resistance transposon, Tn501. Although the enzymes they encode are co-ordinately regulated, the Omr locus of R831b maps approx. 13.5 kb away from the Hgr locus. Three insertions which affect neither phenotype lie between the Hgr and Omr loci; thus, the loci are separated both physically and genetically. One mutant was obtained which tentatively identifies the position of the Tra locus of R831b as adjacent to the Hgr locus.