Regulation of the phototrophic iron oxidation (pio) genes in Rhodopseudomonas palustris TIE-1 is mediated by the global regulator, FixK

Summary The pioABC operon is required for phototrophic iron oxidative (photoferrotrophic) growth by the αproteobacterium Rhodopseudomonas palustris TIE-1. Expression analysis of this operon showed that it was transcribed and translated during anaerobic growth, upregulation being observed only under photoferrotrophic conditions. Very low levels of transcription were observed during aerobic growth, suggesting expression was induced by anoxia. The presence of two canonical FixK boxes upstream of the identified pioABC transcription start site implicated FixK as a likely regulator. To test this possibility, a δfixK mutant of R. palustris TIE-1 was assessed for pioABC expression. pioABC expression decreased dramatically in δfixK versus WT during photoferrotrophic growth, implying that FixK positively regulates its expression; coincidently, the onset of iron oxidation was prolonged in this mutant. In contrast, pioABC expression increased in δfixK under all non-photoferrotrophic conditions tested, suggesting the presence of additional levels of regulation. Purified FixK directly bound only the proximal FixK box in gel mobility-shift assays. Mutant expression analysis revealed that FixK regulates anaerobic phototrophic expression of other target genes with FixK binding sites in their promoters. This study shows that FixK regulates key iron metabolism genes in an αproteobacterium, pointing to a departure from the canonical Fur/Irr mode of regulation.

. Strains and plasmids used in the study.
This study pAB301 Fusion PCR was used to remove the revert the Phe to Leu mutation at position 1008 of the lacZ gene in pAB300 using primers lacZmp(F)for and lacZrev for PCR1; lacZmp(F)rev and lacZfor were used for PCR2. pAB300 was used as template. PCR3 was performed using lacZfor and lacZrev with PCR1 and PCR2 as templates. This PCR was then cloned into the NcoI-EcoRI digested pAB300.
This study pAB307 The promoter region of pioABC operon of Rhodopseudomonas palustris TIE-1 was amplified using primers PpioTIE1for and PpioTIE1rev and cloned into AscI-NdeI digested pAB301.
This study pAB314 Fusion PCR was used to amplify the 2 kb region surrounding the glmUSX-recG locus of R. palustris TIE-1 cloned into pJQ200KS. PCR1 was performed using TIE-1upfor and TIE-1fusionrev(new); PCR2 was performed using TIE-1dnrev and TIE-1fusionfor(new). R. palustris TIE-1 genomic DNA was used as template. PCR3 was performed using TIE-1upfor and TIE-1dnrev with PCR1 and PCR2 as templates. This PCR product has an internal NcoI site was then cloned into SphI-SmaI digested pJQ200KS.
This study pAB343 The putative first FixK binding site (BS1) was deleted using fusion PCR. PCR1 was performed using PpioTIE1for and PpioTIE1-FixK2BSrev. PCR2 was performed using PpioTIErev and PpioTIE1-FixK2BSfor. R. palustris TIE-1 genomic DNA was used as template. PCR3 was performed using PpioTIE1for and PpioTIE1rev with PCR1 and PCR2 as templates. This product was cloned into AscI-NdeI digested pAB322. For affinity purification using Ni-NTA affinity chromatography an N-terminal 6X-Histidine tagged version of fixK from R. palustris TIE-1 was amplified using primers and FixK2TIE1NtermHisfor(NdeI)and FixK2rev(BamHI) with R. palustris TIE-1 genomic DNA as template. This PCR product was then cloned into NdeI-BamHI digested pET-11a.
This study Table S3. Primers used in the study for plasmid construction.  Table S5. Primers used in the study for qRT-PCR.

Detailed description of putative FixK upregulated during anaerobic growth. 1
The putative FixK targets that were upregulated during anaerobic growth 2 included genes involved in gene regulation, photosynthesis, respiration and 3 transport-related functions. The regulatory proteins are AadR (Rpal_4713), a 4 CRP/FNR family regulator that has been shown to be important for anaerobic 5 degradation of benzoate and 4-hydroxybenzoate in R. palustris CGA009 6 (Dispensa et al., 1992); predicted CRP/FNR family gene (Rpal_1280), an Irr 7 homolog (Rpal_2583) and a PadR family protein (Rpal_1207). Proteins 8 belonging to the PadR family are repressors of the padA gene that encodes a 9 detoxifying decarboxylase of phenolic acids such as p-coumarate (Barthelmebs 10 et al., 2000, Gury et al., 2004. Interestingly, such compounds can be utilized by 11 R. palustris CGA009 and likely by R. palustris TIE-1 (Pan et al., 2008). The 12 following genes important for photosynthesis were upregulated during 13 photosynthetic growth: hemO (Rpal_0922), bchD (Rpal_1692) and cycH. HemO 14 is predicted to be 5-aminolevulinate synthase, an important early enzyme in the 15 porphyrin and bacteriochlorophyll biosynthetic pathway (Evans et al., 2008). 16 BchD is a protein important in porphyrin biosynthesis that precedes 17 bacteriochlorophyll biosynthesis (Willows & Kriegel, 2008). CycH is important in 18 cytochrome C biogenesis, thus playing a key role in both photosynthesis and 19 respiration (Lang et al., 1996). Genes involved in respiration that were induced 20 include: ccoN(OQP), encoding the cbb3 cytochrome C terminal oxidase with high 21 affinity for O 2 (Pitcher et al., 2002), a small hypothetical gene near ccoG, and a 22 cytochrome C accessory protein (Preisig et al., 1996). The upregulated genes encoding transport related proteins were: ompW, an outer membrane porin 1 homolog (Rpal_4994) (Lou et al., 2009); osmY, a putative periplasmic osmotic 2 shock gene (Rpal_1868) (Yim & Villarejo, 1992); and a predicted ORF in an 3 operon with a heavy metal transporter (Rpal_2582) (Sitthisak et al., 2007). In 4 addition, a small predicted ORF encoding a protein likely involved in siderophore 5 biosynthesis (Rpal_4015) and Rpal_1412, a signal peptide containing ORF with 6 transmembrane domains were also highly upregulated under anoxic 7 photosynthetic conditions. Overall these data show that R. palustris TIE-1 8 responds dramatically to a shift from aerobic chemoheterotrophic to anoxic 9 phototrophic growth changing expression of a number of potential FixK target 10 genes.

Detailed experimental procedures. 1
Construction of a single integration system for R. palustris TIE-1. The 2 intergenic region of the operon Rpal_2933-2935 (glmUS homolog followed by a 3 ORF of unknown function called gene glmX here) and Rpal_2936 (recG 4 homolog) was chosen as a permissive locus based on the usage of this locus as 5 the insertion site for the Mini-Tn7 transposon in diverse bacteria (Koch et al., 6 2001). As R. palustris TIE-1 does not harbor this commonly found Tn7 insertion 7 site, we devised a homologous recombination based single integration system 8 that was based on the pJQ200KS counter-selection plasmid used for 9 construction of mutants in R. palustris TIE-1 (Jiao & Newman, 2007). A 2 kb 10 region surrounding the glmUSX-recG intergenic region was cloned into 11 pJQ200KS giving rise to pAB314. A unique NcoI site was incorporated in the 12 middle of this 2kb region such that the DNA to be incorporated could be cloned 13 into pAB314. This system was later adapted to integrate lacZ translational 14 fusions into the chromosome of R. palustris TIE-1 as described below. The 15 pAB314-derived plasmids can be integrated on the R. palustris TIE-1 16 chromosome using selection on gentamicin to form a merodiploid. 17 Counterselection using sucrose results in segregation of the merodiploid to give 18 rise to either strains that have the lacZ fusion on the chromosome or WT R. 19 palustris TIE-1. PCR screening was performed to confirm the correct strain was 20 constructed. 21

22
Construction of a lacZ reporter system for R. palustris TIE-1. The lacZ gene 1 that encodes β-galactosidase of E. coli was amplified from pUC18-mini-Tn7-Gm-2 lacZ, engineering the fd terminator in front and the T7 terminator at the end of the 3 gene. This PCR product was cloned into the NcoI and EcoRI sites of pBBR1-4 MCS5 such that an NdeI site was engineered to overlap with the ATG of lacZ. In 5 front of this NdeI site, the restriction site for a rare 8 bp cutter AscI was 6 engineered resulting in pAB301 ( Figure S6). The inherent Plac promoter in 7 pBBR1-MCS5 was deleted during this construction resulting in a promoter-less 8 lacZ plasmid. AscI-NdeI sites can be used for cloning a desired promoter region 9 and will result in the formation of a translational fusion. As the pBBR1 plasmid 10 and its derivatives are broad host range plasmids, pAB301 can be used as a 11 plasmid of general use. The lacZ system was designed as a cassette system so 12 that it can be easily sub-cloned into other pBBR1 derivatives in case other 13 selectable markers are desired. 14 Determination of transcription start sites. 5'RLM-RACE. Total RNA was 15 isolated as described in the Experimental procedures. Reverse transcription (RT) 16 was performed using SuperScript III RNase H-reverse transcriptase (Invitrogen, 17 Carlsbad, CA), according to the manufacturer's protocol using gene specific 18 primers (Table S4). PCR amplification of the cDNA from the RT reaction was 19 performed with Taq DNA polymerase (New England Biolabs, Ipswich, MA) as 20 recommended. The PCR reactions were run on a 2% agarose gel stained with 5 21 µg/ml of ethidium bromide. Relevant bands were cut out and DNA from the 22 bands was purified using the Wizard ® SV Gel and PCR Clean-Up System (Promega, Madison, WI). The bands that were specific to the TAP plus reaction 1 represented the transcription start site and the bands common to the TAP plus 2 and minus reactions represented processed sites. The PCR products were 3 reamplified using a nested primer using Taq DNA polymerase as recommended 4 (New England Biolabs, Ipswich, MA). The primers used for amplification are 5 listed in Table S4. The products were then cloned into pCR ® 2.1-TOPO ® using 6 the TOPO-TA cloning kit as per the manufacturer's guidelines (Invitrogen, 7 Carlsbad, CA). The PCR products were sequenced at the Biopolymers 8

Laboratory in the Massachusetts Institute of Technology Center for Cancer 9
Research. 10 cRACE. Total RNA was isolated as described in the Experimental procedures 11 and contaminating chromosomal DNA was digested with TURBO DNA-free 12 (Ambion, Austin, TX). Reverse transcription was performed using the Superscript 13 II RNase H-reverse transcriptase (Invitrogen, Carlsbad, CA) as follows: 5 µg of 14 the resulting RNA was combined with 20 pmols of the gene specific primer and 15 0.5 mM dNTPs. The resulting mixture was heated to 70°C for 10 min and 16 immediately frozen on ice for at least 1 min. The first strand synthesis buffer, 10 17 mM DTT, 4 U/µl Rnasin (Promega, Madison, WI) were added to the above 18 mixture as specified and heated to 42°C for 2 min. To this mixture Superscript II 19 RNase H-reverse transcriptase was added and the reverse transcription was 20 performed as specified. Following reverse transcription 0.1 U/µl Rnase H 21 (Invitrogen, Carlsbad, CA) was added and the reaction mixture was incubated at 22 37°C for 20 min. The reaction mixture was then purified using the Wizard ® SV Gel and PCR Clean-Up System (Promega, Madison, WI) and eluted in 32 µl of 1 nuclease free water. This cDNA was ligated to 100 pmols of WNp213 (a primer 2 that has an inverted T on its 3ʹ′ end to prevent self-ligation) using T4 RNA ligase 3 (Ambion, Austin, TX) in a 100 µl reaction. Ligations were performed in 25% 4 polyethylene glycol 8000, 1 mM hexaammine cobalt (III) chloride, 0.1 mg/ml BSA 5 and T4 RNA ligase buffer at 22°C for 16 hrs (a brown precipitate forms). This 6 reaction was purified using the Wizard ® SV Gel and PCR Clean-Up System 7 (Promega, Madison, WI) and eluted in 20 µl nuclease free water. This eluate was 8 used for PCR amplification using a nested gene specific primer, primer WNp210 9 and Taq DNA polymerase (New England Biolabs, Ipswich, MA). The resulting 10 bands were separated on a 2% agarose gel and the desired bands were gel-11 eluted using the Wizard ® SV Gel and PCR Clean-Up System (Promega, 12 Madison, WI). The products were then cloned into pCR2.1-TOPO using the 13 TOPO-TA cloning kit as per the manufacturer's guidelines (Invitrogen, Carlsbad, 14 CA). The PCR products were sequenced at the Biopolymers Laboratory in the 15

Massachusetts Institute of Technology Center for Cancer Research. 16
Measurement of β-galactosidase activity. For photoheterotrophic growth 17 cultures were grown on YP medium and inoculated into FW medium at a 10 -2 18 dilution supplemented with appropriate electron donors. For photoautotrophic 19 growth on H 2 , cultures were grown on YP medium and inoculated into FW 20 medium at a 10 -2 dilution with H 2 /CO 2 . For photoferrotrophic growth cultures were 21 grown on FW medium at 10 -2 dilution with H 2 /CO 2 and inoculated into FW Model 550 Sonic Dismembrator for a total of 2 min at the amplitude of 3 (3 s 14 pulses followed by 6 s rest). The extracts were then spun at 14,000 rpm in a 15 microcentrifuge at 4°C for 10 min. 20 -100 µl of cell extract was mixed with the Z-16 buffer to achieve 900 µl volume. The reaction was started by adding 100 µl of a 4 17 mg/ml solution of o-nitrophenylgalactopyranoside dissolved in phosphate buffer 18 (0.06 M Na 2 HPO 4 .2H 2 O and 0.04 M NaH 2 PO 4 ). The rate of increase in the A420 19 due to o-nitrophenol formation was measured spectrophotometrically using a 20 Beckman Coulter DU 800 Spectrophotometer. Activity was calculated with a 21 molar extinction coefficient (ε) of 4500 liters.mol −1 .cm −1 for o-nitrophenol at 420 22 nm. Protein estimation was routinely performed using the microtiter plate method for the Biorad Protein Assay reagent (Biorad, Hercules, CA) using BSA as 1 standard (Thermo Scientific, Waltham, MA). Absorbance at 595 nm was 2 measured using the Biotek Synergy 4 microtiter plate reader. Reported values for 3 β-galactosidase activity represent mean and standard error of 9 independent 4 measurements. 5 Quantitative reverse-transcription PCR. 5-10 ml of appropriately grown 6 cultures were added to 2 volumes of RNAprotect Bacteria Reagent (Qiagen, 7 Valencia, CA), incubated for 5 min at room temperature, and centrifuged for 20 8 min at 5000 X g. RNA was isolated from the cell pellet using the RNeasy Mini Kit 9 (Qiagen, Valencia, CA) with proteinase K and lysozyme treatment as specified by 10 the manufacturer. RNA samples were treated with TURBO DNA-free DNase 11 (Ambion, Austin, TX) to remove genomic DNA contamination. All RNA samples 12 were stored at -80 o C till further use. 400 ng of RNA was used in a 100 µl cDNA 13 synthesis reaction using the iScript cDNA Synthesis Kit (Biorad, Hercules, CA). 14 The cDNA (1 µl) was used as template for quantitative reverse-transcription-PCR 15 (qRT-PCR) using the iTaq SYBR Green Supermix with Rox (Biorad, Hercules, 16 CA) on the 7500 Fast Real-Time PCR System (Applied Biosystems, Foster City, 17 CA). Primer Express v2.0 (Applied Biosystems, Foster City, CA) was used to 18 design primers targeting the transcripts to be tested (Table S5) (20, 50, and 100 mM; 10 bed volumes of each of the buffers was used for washes). His 6 -FixK protein was collected from the column after washing with 2.5 1 ml of buffer containing 500 mM imidazole. His 6 -FixK after affinity chromatography 2 was purified to apparent homogeneity, as ascertained by SDS-PAGE followed by 3 Coomassie Brilliant Blue staining ( Figure S5). High-purity His 6 -FixK was desalted 4 using disposable PD-10 Desalting columns with a 10 kDa molecular weight cut-5 off as per the manufacturer's guidelines (GE Healthcare, Piscataway, NJ) and detected by immunoassay as specified by the manufacturer followed by 22 chemiluminescence using Amersham Hyperfilm ECL films (GE Healthcare, Piscataway, NJ). The exposed films were developed and fixed using a Kodak 1 (CareStream) M35A X-Omat film processor. shown. These experiments were performed using RNA isolated from 9 photoferrotrophically grown cultures. Larger reactions of these PCRs were 10 separated, gel eluted and cloned into pCR ® 2.1-TOPO ® using the TOPO-TA 11 cloning kit as per the manufacturer's guidelines (Invitrogen, Carlsbad, CA). The 12 PCR products were sequenced at the Biopolymers Laboratory in the 13 1 Figure S3. PCR to confirm the ΔfixK mutant of R. palustris TIE-1. Five 2 independent PCR reactions were performed. The first PCR was used to confirm 3 that the upstream and downstream regions of fixK cloned in to pJQ200KS was 4 on the chromosome of the deletion strain. In the ΔfixK strain the product is be 5 1963 bp; the WT band is 2644 bp. The second PCR was to confirm the 6 upstream locus of ΔfixK; mutant product is 1581 bp; WT product is 2271 bp. The 7 third PCR was to confirm the downstream locus of ΔfixK: mutant product 1664 is 8 bp; WT product is 2354 bp. The fourth PCR was to confirm that the strain had a 9 lacZ gene; the mutant gives product 1102 bp long while WT does not yield any 10 product. The fifth PCR was to confirm that ΔfixK did not yield a product for fixK; 11 mutant does not give a product while WT gives a 702 bp product. Relevant 12 bands in the DNA markers are indicated.