Adaptation of Cupriavidus metallidurans CH34 to Toxic Zinc Concentrations Involves an Uncharacterized ABC-Type Transporter

Cupriavidus metallidurans CH34 is a well-studied metal-resistant β-proteobacterium and contains a battery of genes participating in metal metabolism and resistance. Here, we generated a mutant (CH34ZnR) adapted to high zinc concentrations in order to study how CH34 could adaptively further increase its resistance against this metal. Characterization of CH34ZnR revealed that it was also more resistant to cadmium, and that it incurred seven insertion sequence-mediated mutations. Among these, an IS1088 disruption of the glpR gene (encoding a DeoR-type transcriptional repressor) resulted in the constitutive expression of the neighboring ATP-binding cassette (ABC)-type transporter. GlpR and the adjacent ABC transporter are highly similar to the glycerol operon regulator and ATP-driven glycerol importer of Rhizobium leguminosarum bv. viciae VF39, respectively. Deletion of glpR or the ABC transporter and complementation of CH34ZnR with the parental glpR gene further demonstrated that loss of GlpR function and concomitant derepression of the adjacent ABC transporter is pivotal for the observed resistance phenotype. Importantly, addition of glycerol, presumably by glycerol-mediated attenuation of GlpR activity, also promoted increased zinc and cadmium resistance in the parental CH34 strain. Upregulation of this ABC-type transporter is therefore proposed as a new adaptation route towards metal resistance.


Introduction
Metal homeostasis is important for all bacteria since they have to react swiftly to both the scarcity and excess of either essential or toxic metals [1][2][3]. To defend themselves against high metal toxicity, bacteria depend on multiple resistance mechanisms, including efflux pumps, proteins changing the oxidation state of metals, and intra-or extracellular sequestration of metals [4][5][6][7]. In addition, extracytoplasmic function (ECF) sigma factors play an important role in the response to environmental stressors as well as in metal homeostasis [8].
One of the essential metals is zinc, which occurs naturally in air, water, rocks, and soil. The average natural zinc level in the Earth's crust is 70 mg/kg (dry weight), generally ranging between 10 and 300 mg/kg [9]. However, zinc has been concentrated to much higher levels at some locations, either by natural geological and chemical processes or through anthropogenic interventions because of the wide use of zinc compounds in industry, agriculture, and medicine [10][11][12][13][14][15][16]. Despite its essential role as a trace element in various biological processes, including proper functioning of specific enzymes, stabilization

Isolation of Zinc-Resistant Mutants
C. metallidurans CH34 was first cultivated in MM284 at 30 • C until stationary phase and subsequently diluted 1:100 in MM284. Then Zn 2+ was added to a final concentration of 12 mM. After four days of growth at 30 • C, 10 9 cells were pelleted and cell suspensions (100 µL) of a serial tenfold dilution in saline (0.85% NaCl) were spread on MM284 agar plates containing a final concentration of 24 mM Zn 2+ and incubated at 30 • C. Colonyforming units (CFU) were counted after day two and survival frequency was calculated as viable cell count on MM284 24 mM Zn 2+ agar plates divided by viable cell count on MM284 agar plates.

Assessment of Zinc-Resistant Phenotype
The susceptibility of CH34 and CH34 ZnR to a metal ion were determined by the minimal inhibitory concentration (MIC). The strains were cultivated in biological triplicates by inoculating 2 mL MM284 or MM284 supplemented with different metals with 20 µL of a stationary phase C. metallidurans CH34 or CH34 ZnR culture. The MIC is defined as the lowest metal ion concentration that inhibits visible growth of the culture after two days of incubation at 30 • C.
In addition, the resistance of C. metallidurans strains was further assessed with doseresponse growth curves, which were conducted in MM284 supplemented with different zinc or cadmium concentrations. Precultures were incubated at 30 • C until the stationary phase, diluted 1:100 in fresh medium with increasing zinc/cadmium concentrations, and incubated at 30 • C for 72 h. Next, the optical density at 600 nm (OD 600 ) was determined in a 24-well cell culture plate (Flat-bottom, Greiner Bio-One, Vilvoorde, Belgium) which was placed into a CLARIOstar ® (BMG LABTECH, De Meern, The Netherlands). It is noteworthy that high Zn 2+ concentrations resulted in precipitation (of zinc hydroxide) in an abiotic non-inoculated control, but did not impact optical density measurements ( Figure S1).
To further analyze the phenotype of CH34 ZnR , the cell survival of both CH34 and CH34 ZnR at a high zinc concentration (25 mM Zn 2+ ) were determined. Stationary phase cultures were 1:100 diluted in fresh MM284 or MM284 supplemented with 0.3 mM Zn 2+ (pre-induction assay) and these subcultures were allowed to grow for two days. Afterwards, 2-mL cell suspensions (biological triplicate) were transferred to a 24-well cell culture plate and Zn 2+ was added to a final concentration of 25 mM. Next, 20 µL aliquots were withdrawn at different time points and cell suspensions (100 µL) of a serial dilution were spread on LB agar. The CFU/mL was determined after two days of incubation at 30 • C.

Construction of Plasmids
The glpR gene (Rmet_2235) from C. metallidurans CH34 and its zinc-resistant derivative CH34 ZnR were amplified by PCR (Phusion High-Fidelity DNA polymerase, Thermo Scientific, Aalst, Belgium) with primer pair Rmet_2235_Fw-Rv (Table S1), providing HindIII/EcoRI recognition sites. Afterwards, these PCR products were cloned as a HindIII/EcoRI fragment into pBBR1MCS2. The resulting pBBR-glpR and pBBR-glpR R plasmids from E. coli DG1 transformants selected on LB Km 50 were further confirmed by sequencing prior to conjugation (triparental with E. coli HB101 pRK600 as helper) to C. metallidurans CH34 and CH34 ZnR .

Whole-Genome Gene Expression Analysis and qRT-PCR
Whole-genome gene expression analysis of CH34 and CH34 ZnR was performed to examine which genes played a role in the increased zinc resistance. The strains were cultivated by inoculating 30 mL MM284 in biological triplicates with 300 µL of an exponentially growing C. metallidurans CH34 or CH34 ZnR culture at 30 • C. These subcultures were allowed to grow until an OD 600 value of around 0.6 was reached. Next, each subculture was immediately subdivided in 15 microcentrifuge tubes of 2 mL and cells were harvested by centrifugation for 2 min at 10,000 rpm. Supernatant was removed and the bacterial pellets were flash frozen by immersion into liquid nitrogen and kept frozen at −80 • C at all times. RNA extraction, labeling and hybridization, microarray spotting, scanning, and data analysis were performed according to the work of [48]. The full description of the microarray data has been deposited at the Gene Expression Omnibus website (http://www.ncbi.nlm.nih.gov/geo/) under accession number GSE156826.
In addition, total RNA was extracted from CH34, CH34 ∆glpR, and CH34 ZnR (similar conditions as above) and single-stranded complementary DNA (cDNA) was synthesized from 1 µg total RNA using random hexamers as primers and the TaqMan Reverse Transcription Reagents (Thermo Scientific). The quantity of synthesized cDNA was measured with a NanoDrop ® 2000 spectrophotometer (Thermo Scientific). The expression of Rmet_2229 was analyzed by qRT-PCR using the QuantiNova SYBR Green RT-PCR kit (Qiagen, Venlo, The Netherlands) according to the manufacturer's protocol with primers qF_2229 and qR_2229. Expression was compared with that of the 16S rRNA gene (Table S1). RT-qPCRs were performed with a Rotor-Gene Q (Qiagen).

Genome Sequencing
Whole-genome sequencing was performed to identify mutations responsible for the increased zinc-resistance of CH34 ZnR . The strain was cultivated by inoculating 4 mL LB at 30 • C, and total DNA was extracted using the QIAamp DNA Mini Kit (Qiagen). The quantity and quality of extracted DNA was measured using a NanoDrop TM 1000 spectrophotometer (Thermo Scientific). Ten micrograms of DNA were sent for Illumina sequencing (Baseclear, Leiden, The Netherlands). Large insertions and deletions (>200 bp) were identified using an in-house developed software able to exploit the paired-end characteristics of the sequencing data, where either an unexpected increase of unpaired sequencing reads was used as an indicator for insertions, and an increase in the distribution of insert sizes interpreted as deletion. The Genome Analysis Toolikt (GATK) was used to identify point mutations, i.e., single-nucleotide polymorphisms (SNPs) and small indels [49,50]. Sequencing data are available within the Sequencing Read Archive (SRA) of NCBI using the accession number PRJNA658861.

Isolation and Characterization of Zinc-Resistant CH34 Derivatives
Direct exposure of C. metallidurans CH34 to a high Zn 2+ concentration (24 mM Zn 2+ ) is bactericidal, as no survivors were observed even after prolonged incubation (over two weeks). However, CH34 derivatives could be isolated on MM284 agar plates with 24 mM Zn 2+ when it was first cultivated in liquid MM284 medium with 12 mM Zn 2+ (i.e., parental MIC). Growth in liquid MM284 medium with 12 mM Zn 2+ was observed after four days of incubation (no visible growth after two days of incubation) and subsequent plating of this culture on MM284 agar plates with 24 mM Zn 2+ yielded zinc-resistant CH34 mutants at a frequency of 2.37 ± 0.30 × 10 −7 (calculated after two days of incubation at 30 • C). One mutant was further purified on MM284 and retested for its resistance. The latter, designated as CH34 ZnR , exhibited an inheritable resistance phenotype and was further characterized phenotypically. Compared to its parental strain, CH34 ZnR exhibited a two-fold increased resistance to both Zn 2+ and Cd 2+ . No increased resistance to Ni 2+ and Co 2+ was observed ( Table 3), indicating that the involved resistance mechanism(s) can detoxify both Zn 2+ and Cd 2+ . In addition to MIC determination, inactivation dynamics in the presence of a high zinc concentration (25 mM) were examined for the parent and its CH34 ZnR derivative ( Figure 1). The fraction of surviving CH34 cells drastically decreased in the presence of 25 mM Zn 2+ and no survivors could be detected after 24 h of incubation. Although a drastically decreased cell survival in the first hours of incubation was also shown for CH34 ZnR , a fraction of the cells did survive (2.85 ± 2.07 × 10 −4 CFU/mL after 24 h). Pre-inducing the czc operon with 0.3 mM Zn 2+ for 48 h [30,51,52] resulted in an increased cell survival of both the parental and zinc-resistant derivative ( Figure 1). This derepression enables a fraction of the wild type cells to survive a very high concentration of Zn 2+ because the Czc efflux pumps are already active before the cell encounters the toxic Zn 2+ concentration [30,51,52]. However, cell survival for pre-induced CH34 ZnR was still higher than for pre-induced CH34. These observations suggest that the mutation(s) in CH34 ZnR do not target the czc operon and that its genetic background works synergistically with a zinc-induced czc operon, leading to more cells that survive. As the increased resistance stemmed from natural selection of resistance-conferring 262 mutations [53], the global shift in transcriptome resulting from the altered genotype of the 263

Whole-Genome Expression Profile and Genome Sequence Analysis of CH34 ZnR
As the increased resistance stemmed from natural selection of resistance-conferring mutations [53], the global shift in transcriptome resulting from the altered genotype of the evolved strain (CH34 ZnR ) as compared to the parental strain was examined in non-selective conditions [54][55][56]. This revealed 61 coding sequences (CDSs) that were significantly differentially expressed in comparison with the parental CH34 (>1 log 2 fold with an adjusted p-value < 0.05), of which 21 were upregulated and 40 downregulated (Table S2). The ABCtype transporter (Rmet_2229_2234) (Figure 2) and the flagellar filament structural protein (fliC2) were the most upregulated genes. The cupRAC operon was the most downregulated, with copK and copM also being downregulated, but no alteration in copper resistance was observed (data not shown). Interestingly, no overlap was observed between the transcriptional profile of CH34 ZnR in non-selective conditions and the previously established transcriptional response of the parental CH34 strain to zinc stress [48].

261
As the increased resistance stemmed from natural selection of resistance-conferring 262 mutations [53], the global shift in transcriptome resulting from the altered genotype of the 263 evolved strain (CH34 ZnR ) as compared to the parental strain was examined in non-selec-264 tive conditions [54][55][56]. This revealed 61 coding sequences (CDSs) that were significantly 265 differentially expressed in comparison with the parental CH34 (>1 log2 fold with an ad-266 justed p-value < 0.05), of which 21 were upregulated and 40 downregulated (Table S2). 267 The ABC-type transporter (Rmet_2229_2234) (Figure 2) and the flagellar filament struc-268 tural protein (fliC2) were the most upregulated genes. The cupRAC operon was the most 269 downregulated, with copK and copM also being downregulated, but no alteration in cop-270 per resistance was observed (data not shown). Interestingly, no overlap was observed be-271 tween the transcriptional profile of CH34 ZnR in non-selective conditions and the previ-272 ously established transcriptional response of the parental CH34 strain to zinc stress [48]. Subsequent whole genome sequencing of strain CH34 ZnR revealed seven mutations, all of which were caused by transposition of insertion sequences (IS), more specifically ISRme5, ISRme15 and IS1088 (Table 4). Taking into account the expression profile, the inactivation of Rmet_2235 (glpR) by insertion of IS1088 stands out. Therefore, we hypothesized that inactivation of Rmet_2235 resulted in derepressed transcription of the adjacently encoded ABC-type transporter (Rmet_2229-2234) and subsequent increased zinc and cadmium tolerance (Figure 2). This derepression was confirmed by qRT-PCR as transcription of Rmet_2229 was increased 3-fold in CH34 ∆glpR compared to the parental CH34 strain ( Figure S2). Existing tagRNA-seq data used to identify the 5 ends of RNAs in C. metallidurans CH34 corroborated that Rmet_2229-2234 was an operon transcribed from the same promoter into one polycistronic mRNA [57]. Furthermore, this dataset indicated that the main transcription start site of the glpR gene (for strain CH34 in non-selective growth conditions) was found 9 bp downstream of the currently annotated start codon, indicating translation of a 3-aa shorter GlpR from a leaderless transcript ( Figure S3).
A similar gene cluster, also containing a glycerol kinase (glpK) and glycerol 3-phosphate dehydrogenase (glpD), was detected in various α-proteobacteria enabling utilization of glycerol as a carbon source [58]. In fact, the cluster is similar to the plasmid-borne locus responsible for glycerol utilization from plasmid pRleVF39c in Rhizobium leguminosarum bv. viciae VF39 [58]. In this cluster, the glpR gene codes for the repressor of the glycerol-3-phosphate regulon and negatively regulates the adjacent genes involved in glycerol metabolism. Response regulator receiver domain protein (CheY-like) IS1088 Rmet_4521 Transcriptional regulator ISRme5 Rmet_4574 Acyl-CoA dehydrogenase protein IS1088

ABC-Type Transporter (Rmet_2229-2234) Is Responsible for Increased Zinc Resistance
The Rmet_2229_2234 gene cluster was subsequently deleted by replacing it with a tetracycline resistance cassette (referred to as ∆29-34) in both CH34 and CH34 ZnR to test the hypothesis that constitutive derepression of the ABC-type transporter is responsible for CH34 ZnR 's increased zinc resistance. First of all, no differences in zinc and cadmium resistance could be observed between CH34 ∆29-34 and its parental strain, indicating that the native ABC-type transporter exerts no role in the detoxification of zinc or cadmium in the parental CH34 strain (Figure 3). However, when the Rmet_2229-2234 gene cluster was deleted in CH34 ZnR , the zinc and cadmium resistance of the resulting CH34 ZnR ∆29-34 mutant dropped to levels similar to those of the parental CH34, indicating that the ABC-type transporter is required for its increased zinc and cadmium resistance (Figure 3). In turn, deletion of the glpR gene in CH34 resulted in increased zinc resistance, while plasmid-based complementation of glpR reduced zinc resistance back to the parental level in both CH34 ∆glpR and CH34 ZnR (Figure 4). The latter further confirms that loss of GlpR function results in increased zinc resistance.
Finally, CH34 ∆29-34 was still able to utilize glycerol as a sole carbon source (data not shown), demonstrating that the ABC-type transporter is not essential for glycerol uptake in C. metallidurans (contrary to R. leguminosarum bv. viciae VF39).

334
In E. coli, direct interaction with glycerol or glycerol-3-phosphate impedes the action 335 of the GlpR repressor, and thus relieves the repression of the glycerol metabolism [59]. 336 Therefore, C. metallidurans CH34 and CH34 ∆29-34 were grown in MM284 containing 1% 337 (w/v) glycerol to examine if pre-induction with glycerol could promote derepression of 338 the ABC-type transporter and consequently result in increased zinc resistance. Indeed, 339 CH34 displayed a higher resistance to zinc in the continued presence of glycerol (Figure 340 5), putatively indicating that binding of GlpR to the operator of the ABC-type transporter 341 is diminished in the presence of glycerol. Furthermore, glycerol had no direct effect on 342 zinc resistance as no increase in zinc resistance was observed for CH34 ∆29-34 ( Figure 5). 343 However, a lower level of zinc resistance was reached for glycerol-induced CH34 com-344 pared to CH34 ZnR or CH34 ∆glpR, respectively, suggesting that the presence of glycerol 345

Glycerol Induces Increased Zinc Resistance
In E. coli, direct interaction with glycerol or glycerol-3-phosphate impedes the action of the GlpR repressor, and thus relieves the repression of the glycerol metabolism [59]. Therefore, C. metallidurans CH34 and CH34 ∆29-34 were grown in MM284 containing 1% (w/v) glycerol to examine if pre-induction with glycerol could promote derepression of the ABC-type transporter and consequently result in increased zinc resistance. Indeed, CH34 displayed a higher resistance to zinc in the continued presence of glycerol ( Figure 5), putatively indicating that binding of GlpR to the operator of the ABC-type transporter is diminished in the presence of glycerol. Furthermore, glycerol had no direct effect on zinc resistance as no increase in zinc resistance was observed for CH34 ∆29-34 ( Figure 5). However, a lower level of zinc resistance was reached for glycerol-induced CH34 compared to CH34 ZnR or CH34 ∆glpR, respectively, suggesting that the presence of glycerol was not sufficient to bind with all available GlpR repressors or that glycerol-GlpR interaction allowed a reduced repression of the ABC-type transporter. was not sufficient to bind with all available GlpR repressors or that glycerol-GlpR inter-346 action allowed a reduced repression of the ABC-type transporter.

353
This study presents evidence that C. metallidurans CH34 can readily improve its ex-354 treme zinc and cadmium resistance through adaptive evolution. It is noteworthy that the 355

Discussion
This study presents evidence that C. metallidurans CH34 can readily improve its extreme zinc and cadmium resistance through adaptive evolution. It is noteworthy that the high Zn 2+ concentrations used resulted in precipitation (of zinc hydroxide) in an abiotic non-inoculated control. Therefore, in the presence of cells there will be a dynamic equilibrium between free Zn 2+ ions, protein-bound Zn 2+ , and precipitation. As correct measurement of free Zn 2+ is not straightforward [60], such analyses were out of the scope of this study. However, even without these analyses, the central observation of this study, which is the significant and consequent growth difference between the parental and adapted strain in the presence of high zinc concentrations, remains valid. Moreover, whole genome sequencing, expression profiling, and genetic analysis unexpectedly revealed that insertion sequence-mediated inactivation of the GlpR transcriptional repressor and subsequent upregulation of a neighboring ABC transporter (not previously involved in metal resistance) were causally responsible for the resistance phenotype.
GlpR belongs to the DeoR-family of transcriptional regulators that are widespread in bacteria and commonly function as specific regulators of carbon source uptake and catabolism, often playing a role in catabolite repression [59,61]. It shares 46% and 47% amino acid identity with the GlpR repressor of the glycerol 3-phosphate regulon in E. coli and the glycerol regulon in R. leguminosarum bv. viciae VF39, respectively. Glycerol uptake in R. leguminosarum bv. viciae VF39 is mediated by ABC-type active transport, in contrast to the GlpF-mediated facilitated diffusion in E. coli [62]. Uptake is followed by the enzymatic activity of glycerol kinase (glpK) and glycerol-3-phosphate dehydrogenase (glpD), for which expression is negatively regulated by GlpR binding to their operator sequences, enabling the use of glycerol as sole carbon source. In E. coli, glpD, but not glpK, is located near glpR. In contrast, in R. leguminosarum bv. viciae VF39, glpD and glpK flank the glpSTPQUV genes coding for an ABC transporter and form (together with the directly upstream located glpR) its plasmid-borne glp operon [58]. In both E. coli and R. leguminosarum, the glycerol utilization regulon is inducible by glycerol and glycerol-3-phosphate via the specific and direct interaction of these metabolites with the GlpR repressor, thereby causing derepression of the glp genes [58,62]. In C. metallidurans CH34, glpK (Rmet_2238) and glpD (Rmet_2239) are found in close proximity to glpR and to the genes coding for the ABC-type transporter (Rmet_2229-2234), although they do not form an operon with the transporter, reflecting a different genomic arrangement than in R. leguminosarum bv. viciae VF39 (Figure 2). Only a slight differential expression of glpD (1.3-fold) was detected in CH34 ZnR (expression of glpK could not be determined). As glpK (Rmet_2238) and glpD (Rmet_2239) likely form an operon (transcription start sites not detected by the existing tagRNA-seq data [57]), it is conceivable that expression of glpK is also only slightly affected in CH34 ZnR . In contrast to the derepression of the ABC-type transporter (Rmet_2229-2234), these data do not univocally show GlpR-mediated regulation of glpKD (Rmet_2238-39) in C. metallidurans. In addition, CH34 carries a second locus containing a glpK (Rmet_5445) and glpD (Rmet_5444) gene, as well as genes coding for a putative glycerol uptake operon antiterminator regulatory protein (Rmet_5442), a putative akylglycerone-phosphate synthase (Rmet_5443), and a Major Facilitator Superfamily (MFS) transporter (Rmet_5446). However, none of the genes in this second glp locus are differentially expressed in CH34 ZnR . Furthermore, as glycerol can still be used as sole carbon source in the absence of the ABC-type transporter (CH34 ∆29-34), other glycerol transporters exist, e.g., the MFS transporter in the second glp locus (Rmet_5446).
The ABC-type transporter responsible for increased zinc and cadmium resistance in CH34 ZnR comprises six proteins that are encoded by Rmet_2229 up to Rmet_2234. More specifically, one periplasmic protein, one small integral membrane protein, two permeases, and two ATPases form the ABC transporter. The six genes encoding the ABC transporter resemble the glycerol ABC transporter operon of R. leguminosarum bv. viciae VF39, with for instance 63% amino acid identity between the periplasmic protein encoded by Rmet_2229 of CH34 and glpV of VF39 [58]. These similarities clearly indicate a role in glycerol uptake for this uncharacterized ABC-type transporter. In fact, addition of glycerol to the growth medium improved zinc resistance in a Rmet_2229-2234-dependent fashion, indicating a role for glycerol-mediated relief of repression of the ABC-type transporter. This observation therefore reveals that bacterial metal resistance can also depend on the exact carbon sources in the environment.
ABC transporters are found in all taxa and form one of the largest transporter superfamilies, containing both uptake and efflux transport systems. These transporters couple hydrolysis of ATP to the translocation of various substrates, ranging from single ions to entire protein toxins, across cell membranes [63]. The best-studied metal ABC transporter is the high-affinity Zn 2+ uptake system encoded by the znuABC genes and regulated by Zur, which was initially reported in E. coli [20]. ZnuB is the membrane permease and ZnuC is the ATPase component of the transporter, whereas ZnuA is a soluble periplasmic metallochaperone that efficiently captures zinc in this cellular compartment and then delivers the metal to the transmembrane component of the transporter [64]. In contrast to eukaryotes, no examples of ABC transporters involved in metal export are known in bacteria. Nevertheless, ABC transporters that export substrates (which are still to be elucidated) to the periplasm could mediate the repair of metal-induced protein damage [65].
Based on its similarity to the glycerol utilization operon of R. leguminosarum bv. viciae VF39, the Rmet_2229-2234 transporter is likely an uptake system, and the exact mechanism underlying the observed zinc and cadmium resistance requires further study. Nevertheless, a few hypotheses could be formulated that might explain preventing excess zinc and cadmium in the cytoplasm. The imported substrates or the metabolism of these substrates might yield binding sites for metals [66,67]. However, no glycerol or sugars were added to the selective medium. Alternatively, the increased pool of periplasmic binding proteins could bind Zn 2+ (Cd 2+ ) and prevent entry into the cytoplasm. Lastly, released inorganic phosphate (P i ) by ATP hydrolysis might bind excess Zn 2+ (Cd 2+ ), resulting in phosphatezinc conjugates [68].
Finally, it is noteworthy that all of the mutations incurred by CH34 ZnR were caused by the transposition of different insertion sequences (more specifically ISRme5, ISRme15, and IS1088), underscoring the dynamic nature of these elements. Moreover, we have recently shown zinc-induced promoter activity for the transposases of ISRme5 and IS1088 [69], indicating that IS dynamics can be boosted in times of stress to provide evolutionary escape routes for the host cell [69,70].

Conclusions
To conclude, we have demonstrated that derepression of the inconspicuous Rmet_2229-2234 ABC-type transporter (which is likely involved in glycerol uptake) can serve as an evolutionary road towards extreme zinc and cadmium resistance in C. metallidurans CH34. Moreover, Rmet_2229-2234-mediated protection against zinc can also become environmentally triggered in the presence of glycerol. Finally, our observations also underscore the importance of IS elements in the adaptive potential of C. metallidurans CH34.
Supplementary Materials: The following are available online at https://www.mdpi.com/2076-2 607/9/2/309/s1, Figure S1: Zinc dose-response experiments with CH34 (blue) and CH34 ZnR (red). The abiotic (non-inoculated) control is shown in green. The average values of three independent experiments with standard deviations are shown. Figure S2: Relative fold change determined by quantitative real-time PCR (qRT-PCR) analysis of Rmet_2229 in CH34 DglpR and CH34 ZnR . All data were normalized with 16S rRNA gene expression and given as relative to the parental CH34 strain. Figure S3: Transcription profile analysis of the Rmet_2234-glpR intergenic region from C. metallidurans CH34 (coordinates for the chromosomal region are shown at the bottom). Combined transcription start site (TSS) read counts of the three biological replicates are shown for positive (green) and negative (red) strands. The small black arrows indicate clearly identified primary and internal TSSs. The green arrow represents a re-annotated CDS. Data from [57], Table S1: Primers used in this study, Table S2: Differential gene expression analysis results.