Characterization of mercury-resistant clinical Aeromonas species.

Mercury-resistant Aeromonas strains isolated from diarrhea were studied. Resistance occurs via mercuric ion reduction but merA and merR genes were only detected in some strains using PCR and dot hybridization. Results indicate a high variability in mer operons in Aeromonas. To our knowledge, this is the first report of mercury-resistant clinical Aeromonas strains.

Mercury is one of the most toxic of the heavy metals. Mercury toxicity arises from the avid ligation of its compounds to the thiol groups in proteins, which results in inhibiting macromolecule synthesis and enzyme action. Bacteria have developed several mechanisms to neutralize mercury toxicity, the most common being the conversion of highly toxic mercuric ions to less toxic and volatile elemental mercury through chemical reduction. Mercuric ion resistance involves a diverse set of genes, which are widespread in both Gram-positive and Gram-negative bacteria. The bacterial mer operons encode a cluster of genes involved in the detection, mobilization and enzymatic detoxification of mercury (Barkay and Wagner-Döbler, 2005).

Bacteria belonging to the genus Aeromonas are Gram-negative metabolically versatile bacilli, able to adapt and colonize different terrestrial and aquatic ecosystems. In accordance with this ubiquity, Aeromonas species resistant to a great variety of antibiotics and heavy metals have been described. Some species of this genus are considered pathogens responsible for various infections in humans (Janda and Abbott, 2010). This work focused on the isolation and characterization of mercury resistant Aeromonas from human diarrheic feces where no other known enteropathogenic bacterium was isolated.

Forty-one clinical strains were isolated from diarrheic feces collected between July and October 2005 from patients attending the Public Health Institute at Hidalgo State, Mexico. Strains were classified as Aeromonas by standard bacteriological tests. Mercury resistant strains (n = 28) were further identified using the 16S rDNA-RFLP pattern (Figueras ). Species assignation were completed by PCR amplification and sequencing of rpoD gene. Strains not yielding an rpoD amplicon were identified by gyrA amplification and sequencing (Beaz-Hidalgo ). Using these approaches, the following species were found: A hydrophila (n = 11), A. caviae (n = 10), A. veronii (n = 3), A. aquariorum (n = 3) and A. media (n = 1).

The susceptibility profile to five different antimicrobial agents was tested by the Kirby-Bauer disk diffusion method (Table 1) (CLSI 2007). Escherichia coli ATCC 25922, Pseudomonas aeruginosa ATCC 25923, Staphylococcus aureus ATCC 27853 and A. hydrophila ATCC 7966 were used as reference strains.

Table 1

Characterization of mercury-resistant Aeromonas species.

Strain	Specie	MIC HgCl2 (μg/mL)	Volatilization assay	merA gene amplification*	merA gene hybridization	merR gene amplification	Antibiotic resistance pattern
6610	A. hydrophila	400	+	+	ND	−	TET
6632	A. hydrophila	200	+	+	+	+	S
6479	A. hydrophila	300	+	+	ND	+	TET, NAL, TMP, STR
6698	A. hydrophila	200	+	+	ND	−	S
6689	A. hydrophila	200	+	−	+	+	S
6587	A. hydrophila	200	+	+	ND	+	TET, NAL, TMP, STR
6492	A. hydrophila	200	+	+	ND	−	S
6376	A. hydrophila	25	+	−	+	−	TET
5794	A. hydrophila	200	+	−	+	+	TET, TMP, STR
3510	A. hydrophila	200	+	+	ND	+	S
3611	A. hydrophila	600	+	+	ND	+	TET
6640	A. aquariorum	300	+	−	+	+	TET, NAL, TMP
6336	A. aquariorum	200	+	−	+	−	TET, NAL, TMP
3430-1	A. aquariorum	300	+	+	ND	+	TET, TMP, STR
6597	A. caviae	25	+	+	ND	+	TET, TMP
6626	A. caviae	300	+	+	ND	−	TET, TMP
6425	A. caviae	200	+	+	ND	+	TET, TMP
6436	A. caviae	25	+	+	ND	−	TET, TMP
6661	A. caviae	25	+	−	+	+	TMP
6582	A. caviae	25	+	−	+	+	TET, TMP
6548	A. caviae	25	+	−	+	−	TMP
5881	A. caviae	200	+	−	+	+	TMP, STR
5919	A. caviae	25	+	−	+	−	TET, TMP
3817	A. caviae	300	+	−	+	+	TET, NAL, TMP, STR
6654	A. veronii	200	+	−	+	+	S
5869	A. veronii	200	+	+	ND	−	TET, NAL, TMP
5933	A. veronii	300	+	+	ND	+	TET, TM, CAM
3925	A. media	25	+	−	+	−	TMP, STR

TET, Tetracycline; TMP, Trimethoprim; STR, streptomycin; NAL, nalidixic acid; CAF, Chloramphenicol; ND, not done; S, sensitive to all antibiotics tested.

Amplified with primers MerA1F/MerA1R and MerA1/MerA2.

In order to determine the minimum inhibitory concentration of mercury (MIC), cultures were grown overnight in trypticase soy broth (TSB) and then 10 μL of each culture was spotted onto brain heart infusion (BHI) (Oxoid Hampshire, England) agar plates containing 0 to 600 μg mL−1 HgCl2. Plates were incubated at 37 °C. The lowest concentration that inhibited growth was considered the MIC. Results are shown in Table 1. Since all strains (N = 41) were able to grow in the presence of 10 μg mL−1 HgCl2, the cutoff value for mercury resistance was set at 25 μg mL−1.

All mercury-resistant Aeromonas strains were tested in a mercury volatilization assay by the X-ray film method as described (Nakamura and Nakahara, 1988). All 28 mercury-resistant strains were able to reduce mercuric ions as shown by the appearance of dark spots on the X-ray film, resulting from the formation of a silver amalgam (Table 1).

The presence of merA and merR genes was assessed by PCR amplification with primers previously described (MerA1, MerA2, A1s-n.F and A5-n.R) (McIntosh ; Vetriani ) or designed in this work (MerA.1F: CGACTGCACCGACCAGCC, MerA.1R: AAGGTCTGCGCCGCGAGCT, MerR.1F: ATCCGBTTCTATCAGCGCAAG and MerR.1R: ACGTCCTTNRGCTTGTGCTCG). Isolation of total DNA from metal-resistant strains was done according to standard protocols (Sambrook and Russell, 2001). PCR reactions were done with PCR Master Mix (2×) (Fermentas, Lithuania) and amplifications involved standard cycles of PCR.

Using MerA.1F and MerA.1R primers 12 out of 28 strains yielded the expected 461 bp amplicon. Strains that did not yield an amplification product with these primers were subjected to a new amplification reaction with primers MerA1 and MerA2. Using these primers, amplicons with the expected size were obtained for three additional strains. In an further attempt to amplify the merA gene from total DNA of thirteen negative strains, degenerate primers A1s-n.F y A5-n.R were used in new merA amplification reactions. Five of the strains tested this way rendered an amplification product of the right size. However, sequencing of these fragments followed by alignment to the Gen Bank database revealed no similarity with any gene belonging to the mer operon. The merR gene amplification assays yielded the right amplicon in 17 mercury-resistant strains. Not all the strains positive for the merA gene rendered a product for the merR gene or viceversa (Table 1).

The presence of genes coding for MerA and MerR was also explored using a dot hybridization assay. DNA probes to detect merA and merR homologues were synthesized by PCR amplification of A. salmonicida AS03 total DNA with MerA1 and MerA2 or MerR.1F and MerR.1R primers, respectively, in the presence of digoxigenin-dUTP. Dot blot hybridization was conducted at high stringency according to the instructions of the manufacturer (Digoxigenin-High Prime DNA Labeling and Detection Starter Kit I; Roche Molecular Biochemicals, Indianapolis, IN). All strains produced positive hybridization signals for merA but none for merR gene (Table 1).

In order to know whether mercury resistance was carried by an extrachromosomal element, plasmids were extracted from all the mercury-resistant strains according to Birnboim and Doly (Birnboim and Doly, 1979) and Kado and Liu procedures (Kado, and Liu, 1981). Eight plasmid-bearing strains with sizes ranging from 3 to 53 kb were found. These strains were used as donors in conjugal transfer experiments. Matings were performed on sterile nitrocellulose filters on TSA plates using E. coli J53-1 NalR as a recipient. Transconjugants were selected on TSA plates containing 30 μg mL−1 nalidixic acid and 50 μg mL−1 HgCl2. A. salmonicida AS03 bearing a conjugative plasmid conferring mercury resistance was used as a positive control. In this case matings were performed at 28 °C. No mercury-resistant transconjugants could be detected in any of the mating experiments. In view of these results we intended to introduce the plasmids by bacterial transformation. Transformation experiments were as described by Chung , using E. coli DH5α (Invitrogen CA, USA) as host for the incoming plasmids. Transformants were selected on TSA plates supplemented with 50 μg mL−1 HgCl2. Transformation using plasmid preparations did not yield any mercury-resistant E. coli colonies. In agreement with these results, attempts to PCR amplify merA or merR genes from plasmid bands excised from agarose gels were unsuccessful. These results show that plasmids are not related with metal resistance.

Multiple reports on bacterial resistance to mercury including diverse bacterial groups exist in scientific literature. However, works on mercury resistance in Aeromonas are scarce (De J Ramaiah ; Huddleston ; McIntosh ). In this work mercury-resistant strains were characterized as: A. hydrophila, A. caviae, A. veronii, A. aquariorum and A. media. The first three species have been reported as the most frequently associated to diarrhea in Mexico (Aguilera-Arreola ). Some species of Aeromonas present high intraspecific diversity and the presence of different ecotypes has been suggested depending on the strain provenance. Therefore our strains were probably able to acquire the genes for adaptation to mercury-rich environments from diverse sources (Aguilera-Arreola ).

Different works on the isolation of bacteria resistant to heavy metals as part of normal human biota or clinical specimens exist. These works were performed using a variety of culture media (Deredjian ; Pike ). Since it has been reported that results on resistance levels are highly affected by the culture medium used (Pike ); a direct comparison of mercury resistance among strains is not easy. In our experimental conditions strains resistance ranged between 25 and 600 μg mL−1. HgCl2. Bacteria growing in the presence of 25 μg mL−1 are true resistants as confirmed by the mercury volatilization assay and by the amplification of the merA or merR genes. Some of these bacteria are particularly resistant comparing to strains described recently by McIntosh . They isolated several highly resistant A. salmonicida strains able to grow in 270 μg mL−1 HgCl2. Using the same medium and one of these strains (AS03), as a positive control, we found several clinical isolates able to grow between 300 to 600 μg mL−1 HgCl2. Similar levels of resistance as observed in our aeromonad isolates have only been reported in Bacillus subtilis strains, which are able to grow in medium containing up to 550 μg mL−1 HgCl2 (Kannan and Krishnamoorthy, 2006).

Many papers evidence the frequent association between the mer operon and antibiotic resistance determinants (Baker-Austin ; Mathema ). Given that several mercury-resistant clinical strains isolated in this work showed also resistance to tetracycline, trimethoprim, nalidixic acid and streptomycin (Table 1), a search for plasmids was implemented. We found plasmids in some mercury-resistant strains. Mercury resistance could not be transferred by conjugation or transformation with plasmid DNA, which suggests that no relationship between these mobile elements and the metal-resistance phenotype exists. This was confirmed by the lack of amplification of merA or merR genes using plasmid DNA as a template. Our results contrast with reports in literature where mer operons reside frequently in plasmids (Nakahara ). Previous work in our laboratory searching for Class 1 integrons revealed also a preferential chromosomal vs. plasmid localization of these elements in Aeromonas spp. strains (Pérez-Valdespino ).

Mercury reduction is the most common mercury detoxification mechanism in bacteria. All strains were able to reduce the mercuric ion to elementary mercury. Our results show a widespread distribution of mercury reducing activity in the clinical Aeromonas spp. strains studied. Despite these results, we could not amplify the merA gene in around half of the strains using three different primer sets, including primers A1sn.F and A5n.R, which are widely used to demonstrate the presence of the merA gene in an large diversity of bacteria (Ní Chadhain ; Skurnik ; Vetriani ). Using these primers we found amplicons of the right size derived from some of our clinical Aeromonas strains. However, these amplicons showed sequences not related to merA. We suggest that amplicons generated with these primers should always be sequenced in order to avoid false positive results. This is in accordance with Rochelle et al. (Rochelle ), who recommend confirming the presence of the mer genes using several techniques or probes. Due to the failure to amplify the merA gene in several strains, we searched for the presence of the merR gene. As observed with merA, not all strains yielded the merR amplicons. Considering these results, we resorted to search for merA and merR homologues through dot hybridization assays. All strains hybridized with the merA probe but none of them hybridized with the merR probe. Taken together, these results suggest a high variability of mer operons in clinical strains. Mercury exerts a selective pressure on microorganisms, which facilitates the selection of resistant strains carrying genetic elements conferring resistant to mercury from different bacteria to Aeromonas, accounting for this variability (Baker-Austin ; Smit ). Highly resistant microorganisms play a key role in mercury detoxification and mobilization in natural environments. Isolation of highly mercury-resistant Aeromonas from diarrheic feces might result from human exposure to water heavily contaminated with the metal, which would pose a threat to human health.