ICEAfe1, an Actively Excising Genetic Element from the Biomining Bacterium Acidithiobacillus ferrooxidans

Acidithiobacillus ferrooxidans is an environmental acidophilic, chemolithoautotrophic Gram-negative γ-pro-teobacterium (although some discrepancies have been reported [Williams et al., 2010]) that obtains its energy from the oxidation of ferrous ions or reduced sulfur compounds [Rawlings, 2001]. It belongs to the consortium of microorganisms that participate in the bioleaching of minerals with economical interest, such as copper. Although a couple of gene-specific knock-outs have been reported in the literature [Liu et al., 2000; Wang et al., 2012], many more unsuccessful attempts to genetically manipulate A. ferrooxidans have hampered the development of a genetic system for this microorganism and delayed our understanding of its physiology and genetics. Progress in this respect has resulted mainly from bio-informatic, transcriptomic and proteomic approaches [Quatrini et al., 2009; Valdés et al., 2008; Valenzuela et al., 2006] and from gene analysis in heterologous systems [Quatrini et al., 2005; Salazar et al., 1994, 2001]. Recent comparative genomic analyses between sequenced strains of A. ferrooxidans have shown that these genomes have acquired large genomic islands and/or integrative conjugative elements (ICEs) through horizontal gene transfer (HGT) [Holmes et al., 2009; Levicán et al., 2009; Orellana and Jerez, 2011].

HGT plays an essential role in bacterial evolution [O'Brien and Fraser, 2005]. As a consequence, many bacterial genomes contain, in addition to core genes, accessory genes acquired by HGT. ICEs are mobile genetic elements that, like temperate bacteriophages, integrate into and replicate with the host chromosome. Also, like conjugative plasmids, ICEs are transferred by conjugation to the new host contributing to HGT [Wozniak and Waldor, 2010]. Integration of ICEs into the host chromosome is promoted by an integrase encoded in the element that catalyzes a site-specific recombination through single-crossover between a circular form of the element and chromosomal target sequences (attP and attB sites, respectively) [Esposito and Scocca, 1997]. As a consequence, two ICE-chromosome junctions are formed one on each side of the integrated element (attL and attR sites). A tRNA gene is often the integration target site [Hou, 1999]. The integrase also catalyzes the excision of the ICE from its chromosomal localization in a reaction often facilitated by an excisionase encoded in the element [Lewis and Hatfull, 2001]. In this reaction, site-specific recombination occurs between attL and attR of the integrated element, reconstituting attB and attP, and generating the circular form of the ICE. This circular element is an intermediate in the conjugal transfer between bacteria.

Knowledge of the impact of ICEs on their host biology derives mainly from the analysis of pathogenicity islands [Schmidt and Hensel, 2004]. Less is known about the organization of ICEs from environmental microorganisms and their role in the adaptation to variable environmental conditions or colonization of new ecological niches [Bordeleau et al., 2010; Dobrindt et al., 2004].

In this report, we analyze the genetic organization and excision capabilities of ICEAfe1, a 291-kbp genetic element present in the A. ferrooxidans type strain ATCC 23270. Excision levels and potential conjugal transfer of ICEAfe1 are consistent with the measured levels of mRNAs from genes involved in these processes. Identification of mobile genetic elements in industrially relevant acidophilic microorganisms, and characterization of the genes that they carry, will contribute to our current understanding of their genetics and ecophysiology in the short run.

A putative ICE is inserted in the chromosome of A. ferrooxidans type strain ATCC 23270 in a tRNA^Ala gene with recognizable flanking 48-bp direct repeats, 291 kbp apart from each other (from bp 909,225 to bp 1,200,495; fig. 1a). Bioinformatic whole genome comparison of the two available sequences (ATCC 23270: GenBank CP001219 and ATCC 53993: GenBank CP001132) revealed that this element is not present in the ATCC 53993 strain [Holmes et al., 2009; Levicán et al., 2009]. A total of 352 open reading frames (ORFs) were originally annotated in this genomic segment [Valdés et al., 2008]. As in most mobile genetic elements [Hsiao et al., 2005], approximately 64% of the ORFs encode proteins with unknown or hypothetical functions and several others encode transposases (Tn5648 [Peters and Craig, 2001], two ISAfe3 and two ISAfe8 [Valdés et al., 2008]). ORFs with assigned functions include integration/excision, conjugative DNA processing, transfer, maintenance and regulation proteins (table 1), indicating that this element is an ICE. Being the first ICE identified in A. ferrooxidans we have named it ICEAfe1. To our knowledge, this element together with the mycobacteriophage Ms6 of Mycobacterium spp. [Freitas-Vieira et al., 1998] and a putative genomic island in the bioleaching acidophile Leptospirillum sp. group II [Boyd et al., 2009] are the only mobile genetic elements that use a tRNA^Ala gene as integration target site.

Integration-Excision Functions

The predicted protein products of the first two genes adjacent to the tRNA^Ala gene are an integrase Int (COG4974; AFE_0995) that belongs to the tyrosine recombinase family [Esposito and Scocca, 1997] and an excisionase Xis (COG1257; AFE_0996). Next to int and xis, there are genes encoding for orthologs of the cI (AFE_0998) and Cro (AFE_0999) phage regulators (COG2932 and COG1609, respectively), the relative levels of which are known to control the lysogenic/lytic switch of phages such as lambda [Court et al., 2007] and to modulate the excision and transfer of some ICEs [Beaber et al., 2004; Bellanger et al., 2007; McGrath et al., 2005]. All these proteins are likely candidates to control ICEAfe1 integration into, and SOS-regulated excision out of, the A. ferrooxidans chromosome.

Conjugative Transfer Functions

Twenty-one ORFs present in ICEAfe1 encode bacterial type IV secretion system functions possibly involved in the conjugative transfer of this genetic element. Predicted ORFs are organized as four isolated genes and four gene clusters (TRAI-TRAIV; online suppl. table S1; see www.karger.com?doi=10.1159/000346669 for all online suppl. material) whose general architecture closely resembles those of mobile genetic elements such as R391, SXT, pMERPH and R997 [Böltner and Osborn, 2004]. Cluster I encodes transfer genes and clusters II-IV encode transferosome components for pilus assembly. Isolated genes include the relaxase, the coupling protein, an additional VirB4-like ATPase and a product of unknown function similar to the central domain of TraN (online suppl. table S1). Most of the conjugation proteins encoded by ICEAfe1 resemble the Tra-type proteins of the IncF plasmids [de la Cruz et al., 2010; Frost et al., 1994; Lawley et al., 2003]. Occurrence and predicted gene clustering and motif conservation of all these gene products suggest that ICEAfe1 encodes functions for the required channel across the bacterial envelope and the extended F-type pilus for secretion of the ICEAfe1 nucleoprotein complex or relaxosome.

Stabilization Functions

At least three modules consisting of pairs of genes -encoding a stable toxin (T) and a labile antitoxin (A) that belong to the type II systems [Gerdes et al., 2005] are predicted in ICEAfe1 according to TADB data- base (AFE_1098/AFE_1099, AFE_1361/AFE_1362 and AFE_1367/AFE_1368) [Shao et al., 2011]. All these putative TA systems are transcribed in A. ferrooxidans ATCC 23270 under basal growth conditions [Quatrini et al., 2009; P. Bustamante (data not shown)]. TA modules are likely candidates for stable maintenance of ICEAfe1 as it is the case with the mosTA gene pair in the SXT ICE [Wozniak and Waldor, 2009] and other TA systems in previously characterized genomic islands [Pandey and Gerdes, 2005].

Accessory or Adaptation Functions

An entire set of 37 putative tRNA genes clustered in a 10-kbp segment was found within ICEAfe1 [Levicán et al., 2009]. All twenty tRNA species are represented in this cluster. Some of the encoded tRNAs have been shown to be functional in aminoacylation [Levicán et al., 2005, and data not shown]. Existence of tRNA genes has been suggested to contribute to better adapt the translation machinery of host bacteria to the expression of genes encoded in integrated mobile genetic elements, particularly in cyanophages [e.g. Bailly-Bechet et al., 2007; Enav et al., 2012]. Whether the presence of tRNA genes in ICEAfe1 alters the proteome of A. ferrooxidans is still to be explored. Also a Clustered Regularly Interspaced Short Palindromic Repeat locus (CRISPR) is encoded in ICEAfe1 [Holmes et al., 2009]. This locus is likely to have a role in bacterial immunity against bacteriophages and/or other types of mobile genetic elements [Bhaya et al., 2011]. The genes present in the distal region of ICEAfe1 encode functions that are predicted to be involved in the biosynthesis and export of exopolysaccharides rich in glucose, galactose, mannose and glucuronic acid (EPS/CPS) of the Wzy-dependent class [Cuthbertson et al., 2009]. It corresponds to a gene locus encoding for the EPS/CPS export apparatus, several glycosyltransferases and sugar interconversion functions. These genes represent a biosynthesis pathway described previously in A. ferrooxidans that could be linked to biofilm formation in this model bacterium [Barreto et al., 2005]. The presence of these genes in ICEAfe1 might contribute to the colonization of new ecological niches upon transfer of this mobile genetic element to other microorganisms [Rohwerder et al., 2003]. How all these genes might contribute to the fitness of A. ferrooxidans and other possible recipient microorganisms in both natural and experimental settings are interesting questions requiring further detailed studies.

The capacity of ICEAfe1 to excise from its chromosomal location was investigated. Polymerase chain reaction (PCR) primers pairs, A1A2 and A3A4 (fig. 1a; table 2), were designed to amplify attL and attR sites, respectively, when ICEAfe1 is inserted in the chromosome at the tRNA^Ala gene. Primer pair A2 and A3 should render a PCR product only if ICEAfe1 is excised from the chromosome as a circular intermediate. All three PCR products were obtained when the DNA from A. ferrooxidans ATCC 23270 was used as template (fig. 1b, top and middle), confirming the existence of the integrated and excised form of the element. The existence of an empty attB chromosomal site was evidenced using primers A1 and A4 (fig. 1b, bottom). As expected, in the case of A. ferrooxidans ATCC 53993 only the integration target site attB was detected using this approach. Sequencing of the obtained PCR products confirmed the identity of all amplicons and evidenced that site-specific recombination, occurring within the 48-bp direct repeats flanking ICEAfe1, gave rise to precisely excised products without indels (data not shown). Primers A1-A2 and A3-A4, as well as 10 other pairs designed to amplify internal sequences were used to detect the presence of ICEAfe1 in 12 other strains from A. ferrooxidans. None of them gave a positive signal for the entire ICEAfe1.

Determination of ICEAfe1 excision levels out of the chromosome was carried out by real-time PCR (fig. 2), quantifying the empty attB chromosomal site that resulted after the ICEAfe1 excision. The results revealed that when A. ferrooxidans was grown to logarithmic phase in medium supplemented with ferrous ion as energy source, there were approximately 0.034 ± 0.001 attB sites per 100 chromosomes (fig. 2a). Excised ICEAfe1 was higher (0.12 ± 0.01 attB sites per 100 chromosomes) when the bacterium was grown in medium supplemented with sulfur. There were no major differences between phases of growth in both culture media [compare logarithmic (L) vs. stationary (S) in fig. 2a]. However, >300-fold increase in attB sites was observed when these cells were treated with mitomycin C (MMC, a DNA damage inducer; 19.34 ± 0.11 vs. 0.06 ± 0.03 attB sites in MMC and control cultures, respectively; fig. 2b). These results further suggest that a phage-type mechanism underlies ICEAfe1 site-specific excision and this process might be regulated by the SOS response. Since we detected a higher copy number of attP than attB sites in all measured conditions (data not shown), we cannot rule out the eventual replication of excised ICEAfe1, as in other ICEs [Lee et al., 2010; Ramsay et al., 2006].

Transcriptional profiling of A. ferrooxidans ATCC 23270 using microarray analysis (online suppl. fig. S1) [Quatrini et al., 2009] and qRT-PCR (data not shown) revealed that the genes encoding the integrase, excisionase, cI, Cro as well as some of the conjugal transfer proteins, are transcribed in both logarithmic and stationary cultures grown in the presence of ferrous ion or sulfur as energy source. A marked increase in the level of the mRNA encoding the excisionase upon treatment of the cells with MMC (fig. 3a) is consistent with the elevated excision of ICEAfe1 in this culture condition (fig. 2b). Also, the repression of cI and induction of cro orthologs in MMC cultures (fig. 3a) are in agreement with a phage-type mechanism for integration/excision of ICEAfe1 [Court et al., 2007]. Similar changes in the expression profiles were obtained for the conjugal transfer genes. Upon MMC treatment of the cell culture, the mRNA level of these genes was enhanced (fig. 3b). These results show that conditions that enhanced the excision of ICEAfe1 also promote the expression of the type IV secretion system encoded within the ICEAfe1 and further suggest that ICEAfe1 may be capable of self-transfer to adequate recipient cells. However, these experiments remain a challenge until adequate selectable markers and conjugation protocols are developed for this microorganism.

The data presented in this report evidenced the existence of a remarkably large genetic element in the acidophilic bacterium A. ferrooxidans whose genetic organization resembles that of other well-characterized ICEs. This element is active in excision out from the host chromosome both constitutively and upon induction by DNA-damaging agents. The resulting extrachromosomal excision intermediate is most likely relevant for its own dissemination and for the genetic transfer of the functions encoded in it to other potential recipient microorganisms through conjugative transfer. The presence in this element of several toxin-antitoxin modules is predicted to contribute to its maintenance in the cell in spite of active excision and dispersal.

Our findings, and those of others [Andersson and Banfield, 2008; Orellana and Jerez, 2011; Rawlings, 2005], strongly suggest that in the acidic environments where these organisms exist, there is an active exchange of genetic material between members of the microbial community. The presence in ICEAfe1 of a large number of genes with as yet unassigned function represents a source of novel activities that, once elucidated, might contribute to a better understanding of the requirements of these organisms to adapt to their natural extreme environment. For instance, existence of a cluster of tRNA genes might contribute to the efficient translation of the genes en-coded in the ICEAfe1, as has been described in bacteriophages [Bailly-Bechet et al., 2007], and the existence of a CRISPR system might be advantageous to defend the organism against viral infections [Holmes et al., 2009]. The presence in the ICEAfe1 of a number of genes predicted to encode proteins involved in the synthesis of exopolysaccharides that contribute to the biofilm formation, could enhance the persistence of these microorganisms in the environment and/or help to improve mineral dissolution [Rohwerder et al., 2003].

These and other potential advantages for cells carrying ICEAfe1 and the elucidation of environmental conditions that enhance the excision and transfer of this element are relevant topics to be addressed to improve our understanding of the physiology and ecology of these industrially and environmentally relevant microorganisms.

Completely sequenced genomes and predicted proteomes of A. ferrooxidans ATCC 23270 (GenBank CP001219) and ATCC 53993 (GenBank CP001132) were obtained from the NCBI website (http://www.ncbi.nlm.nih.gov). To identify exclusive regions in the genome of either strain, nucleotide sequences were cross-compared using ACT [Carver et al., 2005]. The following bioinformatic resources were used to further characterize candidate genes (and the predicted protein products) found within ICEAfe1: BLASTp and PSI-BLAST [Altschul et al., 1997], the suite of protein comparison and classification programs available in InterproScan [Mulder and Apweiler, 2007], the COG database [Tatusov et al., 2000] and the POG database [Kristensen et al., 2011]. Categorizations were made using Perl scripts and manual revision. Putative ICE sequence was searched for tRNAs using tRNAscan-SE [Schattner et al., 2005] and ARAGORN [Laslett and Canback, 2004] and CRISPRs using CRISPI [Rousseau et al., 2009], CRISPR Finder [Grissa et al., 2007] and CRT1 CRISPR Recognition Tool version 1.1 [Bland et al., 2007] which are available online. Transposases were predicted using TnpPred (http://www.mobilomics.cl).

Potential transmembrane regions and putative sub-cellular localization were predicted using PSORTB (http://www.psort.org/psortb/ [Yu et al., 2010]) and TMHMM (http://www.cbs.dtu.dk/services/TMHMM/ [Krogh et al., 2001]).

Type strain A. ferrooxidans ATCC 23270 and A. ferrooxidans ATCC 53993 were grown at 30°C in modified 9K medium [Silverman and Lundgren, 1959], pH 1.6, supplemented with 120 mM FeSO₄ × 7H₂O or in modified 9K medium, pH 3.5, supplemented with 1% S⁰.

Late exponential-phase cells grown in modified 9K medium supplemented with sulfur were collected by centrifugation, the pellet was divided in two parts and each one resuspended in a half of the original volume with modified 9K medium, pH 3.5, containing 1% S⁰. One culture served as the control and the other was treated with 2 µg/ml mitomycin C (Sigma). Both cultures were incubated for 27 h at 30°C with agitation. After that time, genomic DNA and RNA were extracted as describe below.

DNA from A. ferrooxidans was isolated using Wizard Genomic DNA Purification Kit (Promega) with the following modification for cell lysis: bacterial pellet was resuspended in Nucleic Lysis Solution, frozen at -80°C for 10 min and immediately heated at 80°C for 10 min. This procedure was repeated three times and then allowed to cool down to room temperature. The rest of the lysis protocol was performed as recommended by the manufacturer. RNA was isolated using RNeasy Mini Kit (Qiagen) from bacterial pellet previously stored at 4°C in RNAlater (Qiagen). The removal of genomic DNA from RNA preparations was carried out by digestion with DNase I (Fermentas). DNA and RNA quality was evaluated by 1.0% agarose gel electrophoresis and its concentration was measured by the absorbance at 260 nm in an Epoch Microplate Spectrophotometer (BioTek).

Amplification of sequences to detect excision of ICEAfe1 from the chromosome was performed using PfuUltra II Fusion HS DNA Polymerase (Stratagene) according to the protocol provided by the manufacturer. The cycling conditions were as follows: initial denaturation (95°C, 2 min); 30 cycles consisting of denaturation (95°C, 20 s), primer annealing [TA (at the estimated primer annealing temperature), 20 s], and extension (72°C, 15 s/kbp); followed by a final extension step (72°C, 5 min). PCR products were visualized on 2% agarose gels stained with ethidium bromide. PCR products containing attL, attR, attP and attB sites were purified with SpinPrep™ Gel DNA Kit (Novagen) and sequenced by Macrogen, USA.

Amplification of DNA to generate the standard curve for qPCR was performed using Paq5000 DNA Polymerase (Stratagene) according to the protocol provided by the manufacturer. The cycling conditions were the same as before, except that the extension step in each cycle was 30 s/kbp. PCR products were purified using QIAquick PCR Purification Kit (Qiagen) and quantified by the absorbance at 260 nm in an Epoch Microplate Spectrophotometer (BioTek).

Oligonucleotides used in this study for PCR and real-time PCR are listed in table 2.

The first strand of cDNA to be used in the two-step qRT-PCR was prepared from 1 µg of total DNA-free RNA using random hexamer primers and RevertAid™ Reverse Transcriptase (Fermentas) according to instructions provided by the manufacturer.

The reactions were performed in the Rotor-Gene Q PCR System (Qiagen) using the Rotor-Gene SYBR Green PCR Kit (Qiagen) and the Rotor-Gene 6000 series Software 1.7 for data analysis. Each reaction was carried out in 12.5-µl volumes containing 1× Rotor-Gene SYBR Green Master Mix, 500 nM of each oligonucleotide and 10 ng of DNA or 1 µl of 1:10 diluted cDNA sample as templates. The cycling protocol was as follows: initial activation step (95°C, 5 min) and 35 cycles of denaturation (95°C, 5 s) and combined annealing/extension (60°C, 10 s). The fluorescence data collection was performed in this last step. The PCR products were subjected to a melting curve analysis between 50 and 99°C; specific amplification was confirmed by a single peak in the melting curve. Reaction efficiency (E = (10^-1/slope) - 1) for every gene was derived from the slope of the corresponding 6-point standard curve with 10-fold dilutions of corresponding PCR amplicons for qPCR or total DNA from A. ferrooxidans ATCC 23270 for qRT-PCR. The amounts of attB were normalized to the amount of chromosomal DNA in each sample by amplifying the fur gene, a chromosomal gene unaffected by ICEAfe1 excision. The amount of int, xis, cI, cro and tra transcripts were normalized to the amount of rpoC as reference genes [Nieto et al., 2009] in each sample. In each experimental condition, the reactions were carried out in duplicate in the same run from samples obtained from three independent cultures.

We thank Dr. A. Borole (Oak Ridge National Laboratory, USA) who kindly provided A. ferrooxidans strain ATCC 53993. This work was supported by grants from Fondecyt, Chile, 1070437 and 1110203 (O.O.), 1100887 (R.Q.), 11085045 (G.L.) and 1090451 (D.H.). P.B. and P.C.C. are recipients of graduate studies fellowships from Conicyt and P.B. of a Beca de Apoyo AT-24100112 from Conicyt, Chile.

P.B. and P.C.C. contributed equally to this work.