Heterologous expression of the gene for chlorite dismutase from Ideonella dechloratans is induced by an FNR‐type transcription factor

Abstract Regulation of the expression of the gene for chlorite dismutase (cld), located on the chlorate reduction composite transposon of the chlorate reducer Ideonella dechloratans, was studied. A 200 bp upstream sequence of the cld gene, and mutated and truncated versions thereof, was used in a reporter system in Escherichia coli. It was found that a sequence within this upstream region, which is nearly identical to the canonical FNR‐binding sequence of E. coli, is necessary for anaerobic induction of the reporter gene. Anaerobic induction was regained in an FNR‐deficient strain of E. coli when supplemented either with the fnr gene from E. coli or with a candidate fnr gene cloned from I. dechloratans. In vivo transcription of the suggested fnr gene of I. dechloratans was demonstrated by qRT‐PCR. Based on these results, the cld promoter of I. dechloratans is suggested to be a class II‐activated promoter regulated by an FNR‐type protein of I. dechloratans. No fnr‐type genes have been found on the chlorate reduction composite transposon of I. dechloratans, making anaerobic upregulation of the cld gene after a gene transfer event dependent on the presence of an fnr‐type gene in the recipient.

They are highly efficient in decomposing chlorite to chloride and molecular oxygen (ClO 2 − →Cl − + O 2 ) and therefore essential in PCRB for detoxification. Also, the oxygen produced serves as an additional respiratory electron acceptor, increasing the utility of (per)chlorate as respiratory substrate.
(Per)chlorate-reducing bacteria are thought to play important roles in the biogeochemical cycle of chlorine on Earth (Atashgahi et al., 2018). Most of the perchlorate and chlorate found in the environment today have an anthropogenic origin. Chlorate appears in wastewaters from the pulp and paper industry, and perchlorate is used in several applications, for example, in the manufacture of munitions. Contamination of soils, food, and freshwater reservoirs has become a threat to public health in, for example, the USA, India, and China (Kumarathilaka, Oze, Indraratne, & Vithanage, 2016), and bioremediation by the use of PCRB seems to be one of the best ways to decrease (per)chlorate load in the environment (Hatzinger, 2005;Ma, Bonnie, Yu, Che, & Wang, 2016). Other interesting biotechnological applications have also been suggested for PCRB based on their ability to produce molecular oxygen in anaerobic environments (Wang & Coates, 2017).
(Per)chlorate-reducing bacteria are phylogenetically diverse, found mainly in Proteobacteria  but also in Firmicutes (Balk, Gelder, Weelink, & Stams, 2008;Balk et al., 2010) and Archaea (Liebensteiner et al., 2013). Several lines of evidence suggest that dissimilatory (per)chlorate reduction has been spread by horizontal gene transfer. PCRB are distributed over different classes, phyla, and even domains whereas their closest relatives typically are non-PCRB. The phylogenetic tree of Cld does not overlap with that of 16S rDNA of PCRB (Bender, Rice, Fugate, Coates, & Achenbach, 2004;Maixner et al., 2008). In 13 perchlorate reducers examined, the Pcr operon (pcrABCD) and the gene for Cld (cld) were found on perchlorate reduction genomic islands (PRIs). Localization of PRIs in tRNA genes, presence of mobility genes close to the PRI core and inverted repeats at possible integration sites indicate integration of the PRI into the host genome (Melnyk & Coates, 2015;Melnyk et al., 2011). In a study of six chlorate reducers, the Clr operon (clrABDC) and the cld gene were found to be flanked by insertion sequences, which in five of the six strains were identified as sequences known to form composite transposons in other systems (Clark, Melnyk, Engelbrektson, & Coates, 2013).
Respiration of (per)chlorate is dependent on several proteins besides (per)chlorate reductase and chlorite dismutase. The integration of this metabolism into a new host requires that the necessary proteins either are carried on the transposable element or preexist in the recipient. Interestingly, the presence of accessory genes seems to differ between the PRIs of the perchlorate reducers and the chlorate reduction composite transposons of the chlorate reducers. While the PRIs studied by Melnyk (Melnyk & Coates, 2015) contained several accessory genes, some of which have been proven necessary for perchlorate reduction (Melnyk, Clark, Liao, & Coates, 2014), the chlorate reduction composite transposons examined by Clark (Clark et al., 2013) contained just a few genes in addition to chlorate reductase and chlorite dismutase. Most of the accessory genes identified on PRI cores belong to one of the four functional groups: transcriptional regulation, electron transport, oxidative stress resistance, or molybdenum cofactor biogenesis (Melnyk & Coates, 2015). The nature of the accessory genes reflects the functions needed in the host to be able to exhibit perchlorate reduction capacity, and the same functions should be required for chlorate reduction. Expression of the key enzymes is expected to be regulated by, for example, the availability of different electron acceptors. Suitable redox components must be present in the cell to deliver electrons from the membrane to the periplasmic reductase. Hypochlorite is produced as a byproduct of Cld activity (Hofbauer et al., 2014), and it is likely that the need for protection against oxidative stress increases during (per)chlorate reduction. A system for biogenesis and integration of the molybdenum cofactors of Pcr or Clr has to be present in the cell.
Examination of required accessory genes and their genomic localization will give insights into the evolution of the transposable elements of these complex metabolic traits and facilitate the understanding of the requirements of a non-PCRB recipient.
In this study, we have addressed the regulation of the expres-  (Clark et al., 2013). However, most members of this family are involved in metal sensing making this regulator a less likely candidate for oxygen-dependent regulation of cld. The results of the present study suggest a role for an FNR-type regulator, not included in the chlorate reduction composite transposon, in activating the cld gene of I. dechloratans under anaerobic growth conditions. This is, to our knowledge, the first report of how a gene on a chlorate reduction composite transposon is regulated and also the first report of a functional FNR transcription factor in I. dechloratans.

| Strains, plasmids, and growth conditions
Bacterial strains listed in Table A1 were used as follows: Escherichia coli XL-1 Blue and JM109 for cloning; E. coli RM101 (Sawers & Suppmann, 1992) as an fnr-negative background for expression studies; Ideonella dechloratans (culture collection of Göteborg University, Göteborg, Sweden, CCUG 30977; Malmqvist et al., 1994) as a source of the cld promoter region (AJ296077.1) and an fnr-type gene and its mRNA (img: 2510552075) and E. coli MG1655 (DSM 18039) as a source of an fnr gene (GeneID: 945908). The broad-host-range promoterless reporter vector pBBR1MCS-2-lacZ (Kan R ; Table A2) was fused with different parts of the upstream region of the cld gene of I. dechloratans and used in RM101. pBR322 (Table A2) was used for cloning and expression of the fnr genes in RM101.
All liquid cultures were grown in shake incubator at 37°C and 200 rpm. Antibiotics were used when appropriate to a final concentration of 100 µg/ml ampicillin and/or 50 µg/ml kanamycin. For β-galactosidase assay, E. coli RM101 was grown in a medium described in Constantinidou et al. (2006) containing minimal salts (Pope & Cool, 1982)  dechloratans was grown aerobically and anaerobically as described in Lindqvist, Nilsson, Sundin, and Rova (2015). E. coli MG1655, XL-1 Blue, and JM109 were grown in Luria-Bertani medium.

| Promoter constructs
A series of plasmids, p2cld-I-IV (Table A2; Figure 1), was created by insertion of different parts of the upstream region of the cld gene (AJ296077.1) from I. dechloratans into the reporter vector pBBR1MCS-2-lacZ (Fried, Lassak, & Jung, 2012). Genomic DNA from I. dechloratans was amplified by PCR primers listed in

| Cloning of fnr genes from E. coli and I. dechloratans
The fnr gene from E. coli (GeneID: 945908) was cloned from strain K-12, substrain MG1655. A sequence starting 226 bp upstream of transcription start and ending 251 bp downstream of stop codon was amplified with PCR primers shown in Table A3,

| Quantitative real-time PCR
The relative amount of mRNA from the fnr-type gene of I. dechloratans (img: 2510552075) was estimated by qRT-PCR. I. dechloratans was grown under aerobic and anaerobic conditions as in Lindqvist et al. (2015). Isolation of RNA and performance of qRT-PCR was as described in Hellberg Lindqvist et al. (2012) using the gene-specific primers listed in Table A3 and with each sample analyzed in duplicate. The specificity of primers could be confirmed by the observation of single bands after agarose gel electrophoresis of PCR products. The amount of mRNA from the target gene fnr was normalized to the reference gene 16S rRNA and presented as ΔC T = (C T target − C T reference ). In addition to nontemplate controls, samples without reverse transcriptase were used as negative controls to verify successful genomic DNA removal.

| β-Galactosidase assays
RM101 cells were grown and harvested as described in Section 2.1.
β-galactosidase assays were performed according to Miller (1972) with centrifugation of the samples at 10,000 g for 3 min before measuring OD 420 instead of recording OD 500 . For the assay, 25-500 µl of each cell culture was used. The β-galactosidase activity in Miller units (MU) was calculated with the following formula: The OD 600 denotes the cell density before the assay, OD 420 the absorbance of the o-nitrophenol, t is the reaction time in minutes, and v is the culture volume in milliliters.

| FNR-dependent expression from the cld promoter
We have previously shown that the expression of the cld gene of I.
dechloratans increases 5-10 times in a chlorate independent manner when cultures are transferred from aerobic to anaerobic conditions (Hellberg Lindqvist et al., 2012). This suggests regulation by an oxygen-or redox-sensing regulator. We have identified a 14 bp sequence centered 105.5 bp upstream of the start codon of cld that is identical in 9 out of 10 nucleotides with the canonical FNR box of E. coli (TTGACTTAAATCAA vs. TTGATNNNNATCAA), and which may serve as a regulatory sequence for an FNR-type transcriptional regulator.
To explore a possible role for this sequence and FNR as a regu-

| A putative binding site of RNA polymerase in the cld promoter
The most common position of a single activating FNR-binding sequence in E. coli appears to be at class II sites, which is centered around −41.5 (Myers et al., 2013). A previous attempt to identify to the E. coli consensus sequence TATAAT (Figure 1). Interestingly, two of these, −7 T and −11 A (bold), are identical to the two positions found to be of greatest relevance for σ 70 RNAP binding in E.
coli (Heyduk & Heyduk, 2014). Spaced by the optimal distance of 17 bp from the −10 hexamer and overlapping by 1 nt with the FNR site is the hexamer AACACA with 3 positions (underlined) corresponding to the consensus sequence TTGACA of E. coli. Promoters with weak −35 regions will often have TG in position −15 to −14, so-called extended −10 promoters (Mitchell, 2003). TG is found at −15 to −14 in the analyzed sequence. Thus, the described se- p2cld-IV and pBR322 or pBR322(fnr Ec ), and transcription of each construct was measured as β-galactosidase activity. It was found that p2cld-III could not support transcription in any of the tested conditions, that is, all combinations of aerobic or anaerobic growth with or without FNR ( Figure 2). This shows that the functional promoter was lost in this construction. On the contrary, p2cld-IV followed the same pattern as p2cld-I although the absolute Miller values were lower (Figure 2). It can be concluded that the region hypothesized to contain a −10 sequence and a TSS is necessary for transcription. Based on homology with FNR-and RNAP-binding sites of known class II promoters, it seems likely that this sequence binds RNAP also in I. dechloratans. The sequence downstream of the suggested TSS is not a requirement for transcription but seems to increase transcription since β-galactosidase activity was lower for cells containing p2cld-IV compared to p2cld-I.

| Cloning and characterization of an fnr-type gene of I. dechloratans
The capability of FNR Ec to recognize binding sequences from I. dechloratans and to regulate the expression in the reporter constructs raises the question of whether a corresponding protein is present in I. dechloratans.  of FNR protein is known to be present independent of oxygen level (Sutton, Mettert, Beinert, & Kiley, 2004;Unden & Duchene, 1987).
fnr Ec is negatively autoregulated due to an FNR-binding site spanning the TSS (Spiro & Guest, 1990). It could be hypothesized that also fnr Id is autoregulated since an analysis with Virtual Footprint (Munch et al., 2005)  The level of expression of the reporter gene resulting from the action of FNR Id was only about 25% of that seen with FNR Ec but the difference between aerobic and anaerobic expression was more pronounced since the addition of pBR322(fnr Id ) did not affect aerobic expression (Figure 2

| Implications for horizontal gene transfer of chlorate respiration
The capacity for chlorate respiration is widely distributed in the  (Clark et al., 2013). However, the complex process of chlorate respiration also requires specific biogenesis and electron delivery pathways as well as protection and regulatory systems, depending on several genes not included in the proposed transposable element. The chlorate composite transposon of I. dechloratans contains only three genes in addition to the clr operon and the cld gene that may be of relevance for chlorate respiration. Those are cyc, a c-type cytochrome (Bohlin, Bäcklund, Gustavsson, Wahlberg, & Nilsson, 2010;Lindqvist et al., 2015), mobB, that may have a role in molybdopterin cofactor synthesis (Bohlin et al., 2010), and arsR, a putative transcriptional regulator (Clark et al., 2013). The cyc gene has been cloned and its gene product characterized and tested as electron donor to Clr in vitro (Bohlin et al., 2010). However, a function for it in chlorate respiration could not be established. Instead, cyt c-Id1, a c-type cytochrome not included in the chlorate composite transposon, was shown to be able to donate electrons to Clr in vitro (Bäcklund & Nilsson, 2011).
The present study suggests that the cld gene of I. dechloratans It can be noticed that another chlorate-reducing β-Proteobacterium, Alicycliphilus denitrificans, contains a chlorate composite transposon nearly identical to that in I. dechloratans (Clark et al., 2013). This shows that at least one relatively recent and successful transfer event of chlorate-reducing capability has been enabled by a chlorate composite transposon similar to that in I. dechloratans.
Further studies to reveal the complete set of genes needed for chlorate respiration will give a deeper understanding of the physiological prerequisites for and evolution of chlorate and perchlorate respiration.

CO N FLI C T O F I NTE R E S T
None declared. writing -review and editing (equal).

E TH I C S S TATEM ENT
None required.

DATA AVA I L A B I L I T Y S TAT E M E N T
All data generated or analyzed during this study are included in this published article.