Emerging Microbes & Infections 2020, VOL. 9 https://doi.org/10.1080/22221751.2019.1700762 Transcriptional response to metal starvation in the emerging pathogen Mycoplasma genitalium is mediated by Fur-dependent and –independent regulatory pathways Carlos Martínez-Torróa*, Sergi Torres-Puig a*, Marta Mongeb, Lucía Sánchez-Albaa, Miguel González- Martína, Marina Marcos-Silvaa, Alex Perálvarez-Marín c, Francesc Canalsb, Enrique Querola, Jaume Piñola and Oscar Q. Pich a a Institut de Biotecnologia i Biomedicina and Departament de Bioquímica i Biologia Molecular, Universitat Autònoma de Barcelona, Barcelona, Spain; bVall d’Hebron Institute of Oncology (VHIO), Barcelona, Spain; cBiophysics Unit, Department of Biochemistry and Molecular Biology, School of Medicine, Universitat Autònoma de Barcelona, Barcelona, Spain ABSTRACT Transition metals participate in numerous enzymatic reactions and they are essential for survival in all living organisms. For this reason, bacterial pathogens have evolved dedicated machineries to effectively compete with their hosts and scavenge metals at the site of infection. In this study, we investigated the mechanisms controlling metal acquisition in the emerging human pathogen Mycoplasma genitalium. We observed a robust transcriptional response to metal starvation, and many genes coding for predicted lipoproteins and ABC-transporters were significantly up-regulated. Transcriptional analysis of a mutant strain lacking a metalloregulator of the Fur family revealed the activation of a full operon encoding a putative metal transporter system and a gene coding for a Histidine-rich lipoprotein (Hrl). We recognized a conserved sequence with dyad symmetry within the promoter region of the Fur-regulated genes. Mutagenesis of the predicted Fur operator within the hrl promoter abrogated Fur- and metal-dependent expression of a reporter gene. Metal starvation still impelled a strong transcriptional response in the fur mutant, demonstrating the existence of Fur-independent regulatory pathways controlling metal homeostasis. Finally, analysis of metal accumulation in the wild-type strain and the fur mutant by ICP-MS revealed an important role of Fur in nickel acquisition. ARTICLE HISTORY Received 6 July 2019; Revised 7 November 2019; Accepted 8 November 2019 KEYWORDS Mycoplasma genitalium; emerging STI pathogen; novel therapeutic targets; metal acquisition systems; Ferric uptake regulator; Histidine-rich proteins; metallome Introduction other bacterial STI in HIV-positive men who have Urethritis or inflammation of the urethra is a medical sex with men [7]. condition intimately related to sexually transmitted Remarkably, numerous studies indicate that infections (STI). Non-gonococcal urethritis is the M. genitalium is rapidly developing resistance to all most common treatable sexually transmitted syndrome current standard antibiotic treatments, and is bound in men [1,2] and it is frequently associated with infec- to become a major untreatable STI [5,8]. Supporting tions by Chlamydia trachomatis and Mycoplasma gen- this idea, prevalence of mutations associated with azi- italium. M. genitalium is an emergent human pathogen thromycin resistance, the first-line therapy to treat that has been also associated with several syndromes in M. genitalium infections, rises up to 50% in particular women including cervicitis, endometritis and pelvic contexts. Moreover, recent studies document a rapid inflammatory disease [3,4]. Appreciation of the signifi- emergence of mutations associated with resistance to cance of M. genitalium in human disease has been the fourth-generation fluoroquinolone moxifloxacin, hampered by in vitro culture limitations and lack of which is used as the second-line therapy to treat commercial molecular-based tests for rapid detection M. genitalium infections in Europe. Compounding of this bacterium [5]. Prevalence of M. genitalium this problem is the observation made years ago that infections range from 0.7% to 3.3% in the general 60% to 80% of M. genitalium infections fail to respond population, but this percentage increases dramatically to tetracyclines, which are still used in many parts of in specific populations at high risk of STI [2,6]. In the world as primary treatment for non-gonococcal this sense, M. genitalium is more prevalent than any urethritis. Overall, treatment of M. genitalium is CONTACT Oscar Q. Pich oscar.quijada@uab.cat Institut de Biotecnologia i Biomedicina and Departament de Bioquímica i Biologia Molecular, Universitat Autònoma de Barcelona, 08193 Cerdanyola del Vallès, Barcelona, Spain *Equal contribution. Supplemental data for this article can be accessed at https://doi.org/10.1080/22221751.2019.1700762 © 2019 The Author(s). Published by Informa UK Limited, trading as Taylor & Francis Group, on behalf of Shanghai Shangyixun Cultural Communication Co., Ltd This is an Open Access article distributed under the terms of the Creative Commons Attribution License (http://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited. 6 C. Martínez-Torró et al. increasingly challenging with resistant cases requiring and the mechanisms controlling metal homeostasis in costly drugs, which often have limited availability these bacteria remain essentially unknown. [9,10]. In consequence, there is an urgent need to In this study, we demonstrate that a metalloregula- identify novel therapeutic targets and to develop tor of the Fur family is involved in the regulation of alternative antimicrobial strategies to combat nickel acquisition in M. genitalium. A Histidine-rich M. genitalium infections. lipoprotein and an energy coupling factor-type ABC Metal acquisition systems are fundamental com- transporter system arise as the main nickel acquisition ponents of bacterial pathogens as they are essential to systems of this bacterium. In addition, we reveal the compete with the host for available limited resources. existence of Fur-independent regulatory pathways con- Accordingly, bacteria have developed diverse strategies trolling metal homeostasis in this pathogen. The results to acquire metals within the host. For example, many of this study may facilitate the development of novel bacteria have the ability to scavenge iron from storage therapeutic strategies to control M. genitalium infec- proteins found in the circulatory system, such as hae- tions based on molecules targeting the metal acqui- moglobin and transferrin, or on mucosal surfaces sition systems identified. such as lactoferrin. To this end, they synthesize and secrete molecules with high affinity for metals, known Materials and methods as metallophores, to confiscate these otherwise inac- Strains and culture conditions cessible essential elements [11]. On the other hand, some bacteria produce potent cytotoxins meant to Mycoplasma genitalium strains were grown in SP-4 med- destroy target host cells and release the intracellular ium in tissue culture flasks at 37°C in a 5% CO2 atmos- content to increase nutrient availability. Interestingly, phere. SP-4 plates were prepared supplementing the the expression of many of these cytolysins, including medium with 0.8% agar (BD). Puromycin (3 µg ml−1), the shiga toxin from Shigella dysenteriae, the diphtheria tetracycline (3 µg ml−1) or chloramphenicol −1 toxin from Corynebacterium diphtheriae, the exotoxin (17 µg ml ) were used for mutant selection. Escherichia A from Pseudomonas aeruginosa or the cagA cytotoxin coli strain XL1-Blue was used for cloning and plasmid of Helicobacter pylori, responds to changes in iron propagation. The strain was grown in LB broth or LB availability [12,13]. Additionally, it has been shown supplemented with 1.5% agar, 100 µg ml−1 ampicillin, that H. pylori exploits the cagA cytotoxin to perturb 40 µg ml−1 X-Gal and 24 µg ml−1 Isopropyl β-D-1-thio- host trafficking systems and increase metal supply at galactopyranoside (IPTG), when necessary. All the sites of infection [14,15]. Therefore, there is a M. genitalium strains used in this study are listed in Sup- tight connection between metal acquisition systems plemental Table S1. Additionally, a full description of the and bacterial virulence. procedure used to construct the different mutants is pro- While essential for survival, freely available metals vided as Supplemental Information. are toxic both for the host and the pathogen. Therefore, expression of metal acquisition systems is usually con- Plasmids and primers trolled by regulatory proteins that sense and coordinate All plasmids and primers used in this work are listed in the response to changes in metal availability. Tran- Supplemental Tables S2 and S3. In addition, a complete scriptional regulators of the Fur (Ferric Uptake Regula- description of all plasmids used in this study is pro- tor) family are widespread in bacteria and they control vided as Supplemental Information. the acquisition and storage of transition metals [16]. So far, Fur regulators have been shown to coordinate gene DNA manipulation expression in response to changes in iron (Fur), zinc (Zur), nickel (Nur) and manganese (Mur) availability Plasmid DNA was obtained using Fast Plasmid Mini [17]. In the presence of metal, Fur proteins bind to kit (5Prime) or GeneJET Plasmid Miniprep Kit specific DNA sequences, known as Fur boxes, located (Thermo Fisher Scientific). PCR products were purified within the promoter region of target genes and block from agarose gels using Nucleospin Gel and PCR RNA polymerase binding. Conversely, when metals Clean-up kit (Macherey-Nagel) and digested with the are scarce, Fur proteins are unable to bind their DNA corresponding restriction enzymes (Thermo Fisher targets and transcription of the regulated genes is de- Scientific) when necessary. Plasmids for electropora- repressed. Remarkably, inactivation of the fur locus tion of M. genitalium were obtained using the GenElute often leads to important colonization defects and atte- HP Midiprep Kit (Sigma). nuated virulence in many pathogenic bacteria, includ- ing H. pylori, Staphylococcus aureus, Campylobacter Transformation and screening for mutants jejuni, Listeria monocytogenes, Actinobacillus pleurop- neumoniae, Bacillus cereus and Vibrio cholerae [18]. Transformation of M. genitalium was achieved by elec- However, so far, only few studies have addressed the troporation as previously described [21] using either an response to iron starvation in mycoplasmas [19,20] ECM 630 (BTX) or a Gene Pulser Xcell (Bio-Rad) Emerging Microbes & Infections 7 electroporator. We used 30 µg of plasmid DNA to gen- manufacturer’s instructions. When necessary, cultures erate mutants by allelic exchange and 10 µg for trans- were treated with the iron chelator 2,2-bypiridyl poson mutants. Screening for mutants was performed (1 mg ml−1) (Sigma-Aldrich) for 1 h before the cell using cell lysates as template for PCR and sequencing lysis to create a transition metal limited environment. reactions, as previously described [22]. Reverse transcription was performed with iScript Reverse Transcriptase (Bio-Rad) and random primers as previously described [21]. Primers used for qPCR Sequencing reactions are listed in Supplemental Table S3 and they were DNA sequencing reactions were performed with the designed using Primer3 software. qPCR was performed BigDye® v3.1 Cycle Sequencing kit using 2.5 µl of with iTaq polymerase (Bio-Rad) and SYBR green in M. genitalium lysate as a template, following manufac- CFX96 or CFX384 PCR instruments (Bio-Rad). Rela- turer’s instructions. All reactions were analysed in an tive gene expression was calculated using the Pfaffl ABI PRISM 3130xl Genetic Analyzer at the Servei de method [27]. Differential gene expression was judged Genòmica i Bioinformàtica (UAB). based on the common arbitrary 2-fold cutoff. Data pre- sented in the manuscript correspond to the analysis of RNAs isolated from three independent biological RNA-Seq analysis repeats. Mid-log phase cultures of M. genitalium were scraped off in 1 mL of fresh SP-4 and reinoculated in two new 25 cm2 tissue culture flasks with fresh SP-4 med- Primer extension ium for 6 h. Then, cells were lysed, and total RNA Primer extension analyses were performed with 20 µg was extracted using the miRNeasy Mini Kit (Qiagen). of total RNA as previously described [21]. RNA was When necessary, cultures were treated with the iron extracted using the RNAqueous kit (Thermo Fisher chelator 2,2-bypiridyl (1 mg ml−1) (Sigma-Aldrich) Scientific) and treated with Turbo DNase (Thermo for 1 h prior to cell lysis to generate an iron limited Fisher Scientific). Primers used for this technique (Sup- environment. plemental Table S3) were labelled with the 6-FAM RNA libraries were prepared with TruSeq Stranded fluorescent dye at the 5’ end. Fragments were separated Total RNA Library Prep Kit (Illumina) and analysed using an ABI PRISM 3130xl Genetic Analyzer (Servei using a HiSeq 3000 System (Illumina) at the Genomics de Genòmica i Bioinformàtica, UAB) and then ana- Unit from the Centre for Genomic Regulation (CRG), lysed using PeakScanner v1.0 software (Thermo Fisher Barcelona. cDNA clusters were immobilized in sequen- Scientific). cing lanes of 2 × 50 reads. Reverse and complementary was computed for sequences coming from Read1 pri- mer. Data analysis and sequence alignment was per- Differential proteomic analysis by 2D gel formed using Bowtie2 [23] in the End-to-End mode electrophoresis DIGE and Forward-Forward paired-ends. Sequences were piled up using SAMtools [24] with no limited set to Total protein for 2D-DIGE analysis was extracted as the number of sequences in the alignment. Counts in previously described [28]. Briefly, cells were washed the different ORFs were performed with a standalone three times in PBS and lysed in lysis solution (8 M version of featureCounts programme [25] without urea, 2 M thiourea, 2.5% CHAPS, 2% ASB-14, counting the multi-mapping reads and disabling 60 mM DTT, 40 mM Tris-HCl pH 8.8 and protease multi-overlapping reads. inhibitor cocktail) prior to sonication. Protein sol- Counted features were then submitted to the R/Bio- utions were further purified by a modified TCA- conductor package DESeq2 [26] for statistical analysis. acetone precipitation (2D-CleanUp Kit, GE Health- DESeq2 analysis used a parametric fitType and a zero- care) and, finally, dissolved in the DIGE labelling mean normal prior on the non-intercept coefficients. buffer (8 M urea, 4% w/v CHAPS, 30 mM Tris pH Data were sorted by log2 fold change and statistical sig- 8.0). Protein concentration was determined using the nificance was set at the common threshold of P-value < Bio-Rad RCDC Protein Assay (Bio-Rad, UK), follow- 0.05. Three independent biological repeats of each ing manufacturer’s instructions. Triplicate protein strain or condition were analysed. samples of each of the strains were labelled using Cy3 or Cy5 cyanine dyes and separated in three two-dimen- sional electrophoresis gels as previously described [28]. qRT-PCR analysis Fluorescence images of the gels were obtained on a RNA was extracted from mid-log phase cultures of Typhoon 9400 scanner (GE Healthcare). Cy3 and M. genitalium using the RNAqueous Kit (Thermo Cy5 images were scanned at excitation/emission wave- Fisher Scientific) and then treated with Turbo DNase lengths of 532/580 nm and 633/670 nm, respectively. (Thermo Fisher Scientific) following the Image analysis and quantification of relative protein 8 C. Martínez-Torró et al. abundances were performed using SameSpots (Non- suspension culture. After 5 days of incubation at 37° linear Dynamics, UK). C at 100 rpm, cells were harvested by centrifugation and washed with PBS. In some experiments, PBS washes were performed in the presence of 0.5 mM Protein identification by mass spectrometric EDTA to remove extracellular metals as suggested by analysis Williams and coworkers [29]. Next, pellets were dried Protein spots were excised from the gel using an auto- overnight at 70°C on a Ultrawave microwave (Mile- mated Spot Picker (GE Healthcare). In-gel trypsin stone) and digested with HNO3 (Merck). Finally, the digestion was performed using autolysis stabilized sample amount was determined using a XPE205DR trypsin (Promega, Madison, USA). Protein digests analytic balance (Metler Toledo) and analysed in a were analysed by liquid chromatography-mass spec- 7500ce ICP-MS instrument (Agilent). trometry analysis on a Maxis Impact Q-TOF spec- trometer (Bruker, Bremen), coupled to a nano-HPLC Results system (Proxeon, Denmark). Samples were evaporated Global response of M. genitalium to iron and dissolved in 5% acetonitrile, 0.1% formic acid in starvation water, and concentrated on a 100 μm ID, 2 cm Proxeon nanotrapping column and then loaded onto a 75 μm We conducted a genome-wide RNA-Seq analysis to ID, 25 cm Acclaim PepMap nanoseparation column investigate transcriptional changes following metal (Dionex). Chromatography was run using a 0.1% for- depletion in M. genitalium. To this end, we shocked mic acid - acetonitrile gradient (5–35% in 30 min; the cells with the metal chelator 2,2’-bipyridyl (DPP) flow rate 300 nL/min). The column was coupled to and compared transcription to that of bacteria cells the mass spectrometer inlet through a Captive Spray grown under routine culture conditions. DPP has (Bruker) ionization source. MS acquisition was set to been shown to bind iron with high affinity, although cycles of MS (2 Hz), followed by Intensity Dependent it can also coordinate other transition metals, especially MS/MS (2–20 Hz) of the 20 most intense precursor when used at high concentrations. Our analysis ions with an intensity threshold for fragmentation of revealed a robust response to metal starvation, with 2500 counts, and using a dynamic exclusion time of more than eighty differentially expressed genes belong- 0.3 min. LC-MSMS data was analysed using the Data ing to diverse functional categories (Figure 1 and Sup- Analysis 4.0 software (Bruker). Proteins were identified plemental Table S4). Remarkably, we observed a using Mascot (Matrix Science, London UK) to search pronounced up-regulation of the molecular chaperone against the mycoplasma genitalium proteins in the dnaK (4-fold) and the protease genes clpB (14-fold) UniProt/SwissProt database. MS/MS spectra were and lon (7-fold), which are important to preserve searched with a precursor mass tolerance of 10 ppm, protein integrity in bacteria. Likewise, transcription fragment tolerance of 0.05 Da, trypsin specificity with of several genes involved in DNA repair including a maximum of 2 missed cleavages, cysteine carbamido- the excinuclease uvrC (MG_206), the ATP-dependent methylation set as fixed modification and methionine helicase uvrD (MG_244) or the recombinase recA oxidation as variable modification. Significance (MG_339), was induced under metal-depleted con- threshold for the identifications was set to p < 0.05, ditions. Transcript levels of the HPr kinase gene hprK minimum ions score of 20. (MG_085) and the glycerol uptake facilitator glpF (MG_033), involved in the regulation of carbon metab- olism and sugar transport, also increased upon metal Phase contrast and fluorescence microscopy depletion. In addition, we observed a marked activation M. genitalium cells were grown on 8-well IBIDI of three genes, MG_149, MG_321 and MG_439, cod- chamber slides for 16 h, washed three times with PBS ing for predicted lipoproteins. and visualized on a Nikon Eclipse TE 2000-E micro- Of note, we found that clusters of genes with related scope. Phase contrast and TRITC epifluorescence functions were consistently co-regulated. For instance, images were captured with a Digital Sight DS-SMC transcription of an operon coding for the different sub- Nikon camera using the NIS-Elements BR Software. units of a putative ABC-type metal transporter system ImageJ was used to analyse fluorescence images. (MG_304, MG_303 and MG_302) increased by 5-fold upon iron starvation. Similarly, genes coding for a putative oligopeptide ABC-type transporter system Inductively coupled plasma mass spectrometry (oppBCDF) were up-regulated by 4-fold. Furthermore, (ICP-MS) transcription of the MG_241 and MG_242 genes, M. genitalium strains were grown in 175 cm2 tissue which are co-transcribed with the uvrD locus, was culture flasks with SP-4 medium until mid-log phase. also activated 9- and 6-fold, respectively. These genes Then, attached mycoplasmas were scraped off in code for two proteins with two transmembrane seg- 1 mL of fresh medium and used to start a 1L ments, indicating a membrane-associated function. Emerging Microbes & Infections 9 Figure 1. Volcano plot of transcriptional differences in M. genitalium upon shock with the iron chelator 2,2’-bipyridyl (DPP) ident- ified by RNA-Seq analysis. Each spot represents a single gene. Red and green colours indicate activation and repression, respect- ively. Cut off for biological significance was set to log2 = 1. On the other hand, transcription of an operon cod- M. genitalium fur homologue in the regulation of ing for an ABC transporter system involved in glycerol- metal acquisition, we obtained a null mutant by allelic 3-phosphate uptake (ugpCAE genes) was down-regu- exchange. In the resulting fur mutant strain, the lated by 3-fold upon the shock with 2,2’-bipyridyl. In MG_236 gene is replaced by the tetracycline resistance addition, metal starvation inhibited transcription of marker (tetM). The terminator sequence of the dnaK several genes related to carbohydrate metabolism, gene, identified by the TranstermHP software [31], including the glycolytic enzymes glucose-6-phosphate was placed after the tetM marker to prevent overex- isomerase (pgi, MG_111) and fructose bisphosphate pression of the MG_237 gene. For control purposes, aldolase (fba, MG_023). Similarly, we detected we also generated a complemented strain carrying an decreased transcript levels of the ribulose-phosphate ectopic copy of the fur gene. Primer extension analyses 3-epimerase (rpe, MG_112) and the UTP-glucose-1- revealed that transcription of fur in the wild-type strain phosphate uridylyltransferase (galU, MG_453) genes. was driven by a promoter located in the upstream Therefore, metal depletion prompts an important region of the MG_232 gene (Figure 2(C)). Therefore, remodelling of the metabolic flux in M. genitalium. in the complemented strain, we imposed transcription Furthermore, we found that transcription of the strin- of the transposon-encoded fur copy by the MG_232 gent response regulator gene relA was also inhibited, promoter. However, all of the analysed clones overex- which is consistent with a coordinated response to pressed fur as compared to the wild-type strain (Figure nutrient deprivation. In summary, the response of 3). This is likely due to the presence of other promoters M. genitalium to metal starvation is both robust and within the insertion site that increase transcription dri- diverse, and it likely involves several regulatory pro- ven by the MG_232 promoter. teins and pathways. Transcriptional changes in the fur mutant under standard culture conditions were also assessed by RNA-Seq. We observed the activation (∼4-fold) of Role of Fur in the regulation of metal the histidine-rich lipoprotein gene (hrl, MG_149) and homeostasis in M. genitalium three genes (MG_302, MG_303 and MG_304) com- The MG236 protein of M. genitalium shows sequence prising a putative metal uptake system with homology homology to metalloregulators of the Fur family, to CbiMNQO transporter systems (Supplemental which control metal homeostasis in many bacteria. Table S5). As mentioned above, these genes were also Moreover, the predicted tridimensional structure of up-regulated in the wild-type strain upon metal star- the MG236 protein using the SWISS-MODEL protein vation. Reintroduction of an ectopic MG_236 copy to structure homology-modelling server [30] reveals a the fur mutant restored wild-type levels of transcrip- striking similarity to that of a Fur regulator from tion of the Fur-regulated genes (Supplemental Table C. jejuni (Figure 2(A)). Our analyses indicate that fur S5). Transcript levels of the MG_237 gene were slightly in M. genitalium is co-transcribed with several genes lower in both the fur mutant and the complemented coding for two ribosomal proteins (MG_232 and strain (∼2-fold), indicating some polar effects derived MG_233), the ribosomal-processing cysteine protease from the replacement of the fur gene by the tetM mar- Prp (MG_234), the apurinic endonuclease IV Nfo ker. qRT-PCR analysis confirmed the Fur-regulated (MG_235) and the unknown protein MG_237 (Figure expression of the hrl gene and the putative ABC-type 2(B)). To assess the possible implication of the metal transporter (Figure 3). 10 C. Martínez-Torró et al. Figure 2. Homology of the M. genitalium MG236 protein to transcriptional repressors of the Fur family and determination of the MG_236 transcriptional start site (TSS). (A) SWISS-MODEL predicted 3D dimeric structure of MG236 (left) using Campylobacter jejuni fur homolog S1 dimer as a template (right) [66]. Below, model-template alignment is shown, highlighting residues with an exact match and predicted secondary structure using DSSP. (B) Scheme of the M. genitalium operon comprising the fur gene. The asterisk indicates the experimentally identified TSS. The predicted −10 box is shown in black and underlined. The translational start codon of the MG_232 gene is shown in black. The determination of the MG_236 TSS was achieved with primer PEmg232. Previous exper- iments with other primers (marked in red) did not reveal additional TSS. (C) Primer extension analyses to identify the MG_236 TSS. Electropherogram was generated with Peak Scanner v1.0 (Applied Biosystems) analysis software. Red peaks represent ROX size standards and primer extension product is indicated in blue (233 bases). Control of metal homeostasis by Fur- of genes that respond to Fur and metal starvation, and independent pathways many others that only respond to variations in metal availability, was confirmed by qRT-PCR analyses To get further knowledge on the regulatory mechan- (Figure 3). Overall, our results demonstrate the exist- isms controlling metal homeostasis in M. genitalium, ence of a global Fur-independent response to metal we investigated the existence of transcriptional changes starvation in M. genitalium. in the fur mutant upon shock with the metal-chelator DPP. Despite the absence of the transcriptional regula- tor, a strong transcriptional response was still observed, Proteomic analysis of the fur mutant and we identified up to one hundred differentially expressed genes upon metal depletion. (Figure 1 and Then, we wondered whether the transcriptional Supplemental Table S6). We observed a good overlap changes observed in the fur mutant were conserved between the response to metal starvation in the fur at the protein level. To this end, we used 2D-DIGE mutant and that in the wild-type strain. However, tran- and LC-MS to examine the differences in protein scription of hrl and the cbio transporter operon abundance between the wild-type strain and the fur (MG_302-MG_304) remained unchanged, indicating mutant under standard culture conditions. (Figure that activation of these genes in response to metal 4). We identified sixteen protein spots with significant deprivation is entirely dependent on Fur. The existence changes in expression and above the 2-fold arbitrary Emerging Microbes & Infections 11 fluorescent marker. The resulting Hrlwt:CatCh cassette was introduced to M. genitalium by transposon deliv- ery. All transformants analysed, designated G37-Hrlwt- CatCh, exhibited marginal mCherry fluorescence (Figure 5(A)). This is consistent with low levels of expression of the mCherry marker in a wild-type back- ground driven by the hrl promoter. Then, we tested the expression of the Hrlwt:CatCh cassette in a fur mutant background. To this end, we first determined the insertion site of the minitranspo- Figure 3. qRT-PCR analysis of M. genitalium G37, a fur mutant son carrying the mcherry reporter in the different and its complemented strain derivative (fur COM) under stan- G37-HrlwtCatCh mutants described above (Sup- dard growth conditions and upon the addition of 2,2’-bipyridyl. plemental Table S8). Clone 1, with the minitransposon Bars represent the mean log2 fold-changes of three indepen- inserted in the intergenic region within the MG_339 dent biological repeats. Statistical significance of mean fold- and MG_340 genes, was selected for further analysis. changes above the arbitrary cut off >1 for biological signifi- cance was assessed using a paired t test. P-values are located In order to allow direct comparison of mCherry above each column when differences to the G37 strain are stat- expression in different strain backgrounds, the Hrlwt: istically significant (p < 0.05). The ectopic location of the fur CatCh cassette was introduced into the fur mutant at gene in the complemented strain leads to a marked fur the same chromosomal location as in clone 1 by hom- overexpression. ologous recombination. In the resulting fur-Hrlwt- CatCh transformants, we observed mCherry cutoff in the fur mutant compared to the wild type fluorescence in all the clones analysed (Figure 5(A)). strain (Supplemental Table S7). In all cases, changes Using the same procedure, an hrl promoter with a were consistent with higher expression levels in the scrambled Fur box was also introduced to the wild- fur mutant. Mass spectrometric analyses determined type strain in the intergenic region within the the presence of the Hrl (MG149) polypeptide in MG_339 and MG_340 genes by homologous recombi- two spots (log2 fold change 2.22 and 2.60, respect- nation. mCherry fluorescence was observed in both ively). Remarkably, the apparent molecular mass of mutant backgrounds (Figure 5(A)). Overall, these this lipoprotein (16 kDa or below) was lower than data demonstrate the participation of the identified the expected (32 kDa); this suggests a proteolytic clea- conserved sequence in Fur-regulation. vage at the centre of the protein. A similar obser- To obtain quantitative data, we assessed mcherry vation was previously documented by Shimizu and expression in these mutants by qRT-PCR (Figure 5 coworkers [32]. (C)). As expected, transcript levels of the mcherry repor- The different subunits of the metal transporter sys- ter were higher in the fur mutant than in the wild-type tem (MG304-MG303-MG302) were not identified in background (∼6-fold). Similarly, the presence of a any of the analysed spots. However, in other spots, we mutated hrl promoter increased transcription of the also identified the presence of the MG338 lipoprotein mcherry reporter in a wild-type background (∼4-fold). (log2 fold change 2.45 and 2.60, respectively) and the In addition, we also tested the effect of the chelator Beta subunit of the Pyruvate dehydrogenase component 2,2’-bipyridyl on mcherry expression. We found that E1 (MG273, PdhB). In addition, we identified the Meth- mcherry transcription was metal dependent in the ionine sulfoxide reductase A in three different spots wild-type strain (Figure 5(C)). Transcriptional analysis (log2 fold change from 2.23–2.49) and the 30S riboso- confirms the fluorescence data described earlier. mal proteins S7 (log2 fold change 2.54) and S9 (log2 fold change 2.19 and 2.08, respectively), in one and Determination of the metallome of M. genitalium two spots, respectively. Analysis of five protein spots did not produce any significant identification results. The study of metal acquisition systems of M. genitalium prompted us to determine the metal content (metal- lome) of this pathogen under routine culture con- Identification of a regulatory element in Fur- ditions. Culture medium of this bacterium is regulated promoters extraordinarily rich and it contains a significant M. genitalium Fur-regulated genes seem to be con- amount of transition metals as supported by ICP-MS trolled by σ70-dependent promoters. Of note, we recog- analysis (Supplemental Table S9). In pellets of the nized a conserved sequence with dyad symmetry near wild-type strain, zinc was the most abundant tran- the putative Pribnow boxes of these promoters (Figure sition metal identified (∼100 µg g−1), followed by 5(B)). To assess the possible contribution of this con- iron (∼20 µg g−1) and copper (∼3 µg g−1) (Figure 6 served sequence to Fur-regulation, we created a tran- (A)). Other metals such as nickel, manganese or cobalt scriptional fusion of the hrl promoter to the mcherry were scarce or undetectable in our analysis. Treatment 12 C. Martínez-Torró et al. Figure 4. 2D-DIGE analysis of the M. genitalium wild-type strain and a fur mutant. (A) Superimposed images in pseudocolour from Cy3 (green, WT proteins) and Cy5 (red, fur proteins) labelled samples run on a 2D-DIGE gel. The positions of some of the differential proteins are marked and labelled. (B) Fluorescence images corresponding to each of the 3 replicate samples of the WT and fur samples for some of the differential proteins identified. of the pellets with EDTA did not reveal any difference observe any transcriptional differences upon sup- in metal accumulation. In contrast, cells from the fur plementation of the SP-4 medium with 1 mM CoCl2, mutant contained about 10-fold more nickel than FeCl2 or NiCl2. However, the addition of 1 mM the wild-type strain. In contrast, the intracellular ZnCl2 triggered a significant down-regulation of hrl metal content was effectively restored in the comple- and a slight decrease on MG_304 transcription (Figure mented fur mutant. Cobalt was also detected at higher 6(B)). Remarkably, these changes were Fur-dependent levels in the fur mutant (∼5-fold), but this result var- as an excess of zinc did no modify transcription of hrl ied largely among the different repeats. Therefore, our or MG_304 in a fur background. Altogether, our data data indicate that Fur may be implicated in the regu- indicate that Fur could use zinc as a cofactor to regulate lation of nickel acquisition in M. genitalium. gene expression in M. genitalium. Effect of transition metals on Fur regulation Discussion Finally, we evaluated the impact on Fur-regulation of Sequestration of micronutrients represents one of the an excess of some transition metals. We did not first lines of defense of the host against bacterial Emerging Microbes & Infections 13 Figure 5. Expression of a gene reporter (catcherry) under the control of the hrl promoter of M. genitalium. (A) Fluorescence microscopy analysis. Each row contains a series of three images corresponding to the phase contrast, the Texas Red channel and the resulting overlay, respectively. (B) Identification of a putative Fur box sequence in M. genitalium and M. pneumoniae. MPN162, MPN433 and MPN043 are the respective homologs of MG_149 (hrl), MG_304 and MG_033 (glpF) in M. pneumoniae. Sequences in bold correspond to the conserved palindromic regions identified within the promoters of Fur-regulated genes of M. genitalium and M. pneumoniae. The graphic is a sequence logo generated with the conserved sequences. The putative −10 promoter elements are underlined. The overall height of each stack indicates the sequence conservation at that position. The scrambled promoter sequence used to generate the mutant strain characterized in panel 5A is located below the consensus Fur box sequence and the mutated residues are marked in red. (C) Transcriptional analysis by qRT-PCR. Bars represent the mean log2 fold-changes of three independent biological repeats. Statistical significance of mean log2 fold-changes above the arbitrary cut off >1 for biological significance was assessed with Student’s T test. Statistically significant values (p < 0.05) are indicated above the columns. Transcription of hrl and MG_304 was also analysed for control purposes. 14 C. Martínez-Torró et al. Figure 6. Acquisition and response to transition metals. (A) Analysis of metal accumulation in M. genitalium by inductively coupled plasma mass spectrometry (ICP-MS). Statistical significance was assessed with Student’s T test. Statistically significant values (p < 0.05) are indicated with their corresponding p-values. (B) qRT-PCR analysis of key Fur-regulated genes in the wild-type strain upon addition of 1 mM FeCl2, NiCl2, ZnCl2 or CoCl2. (C) qRT-PCR analysis of key Fur-regulated genes in the wild-type strain upon addition of 1 mM ZnCl2. Bars indicate the mean log2 fold-changes of three independent biological repeats. Statistical significance of mean log2 fold-changes above the arbitrary cut off >1 was assessed with Student’s T test. Statistically significant values (p < 0.05) are indicated above the corresponding columns. infection. To counteract this defensive strategy, termed Lactoferrin is an iron-binding protein found at high nutritional immunity [33], bacteria display a large concentrations in mucosal secretions that shows anti- repertoire of dedicated transporters to guarantee a bacterial activity both in vitro and in vivo [35]. How- sufficient supply of essential elements. Several years ever, the same study concluded that M. genitalium ago, Tryon and Baseman reported the ability of the res- did no bind either lactoferrin or transferrin. Therefore, piratory pathogen M. pneumoniae to acquire human despite the clinical relevance of this emerging mucosal- lactoferrin by a saturable and specific manner [34]. surface pathogen, studies addressed to identify the Emerging Microbes & Infections 15 metal acquisition strategies of M. genitalium are essen- conditions [47]. However, we successfully deleted the tially non-existent. MG_236 gene by allelic exchange using standard pro- Herein, we show that metal starvation elicits a broad cedures, indicating an incomplete saturation of the transcriptional response in M. genitalium, suggesting transposon library characterized to assess gene essenti- that regulation of metal homeostasis in this bacterium ality. Transcriptional analysis of the fur mutant involves different regulatory pathways. In this sense, we revealed the activation of a gene coding for a Histi- found that iron regulates transcription of several cha- dine-rich lipoprotein (MG149) and a putative metal perone and protease genes under the control of the transporter system with homology to CbiMNQO HrcA repressor, which coordinates the heat stress uptake systems (MG302-MG303-MG304), which are response in M. genitalium [36]. Similarly, transcript putatively implicated in cobalt import in association levels of relA, a major regulator of the nutrient-star- with the cobalamin (vitamin B12) biosynthetic path- vation response in bacteria [37], decrease upon metal ways. The hrl gene was previously shown to be mark- depletion. On the other hand, we found a significant edly induced upon hyperosmotic shock in activation of an operon coding for an oligopeptide per- M. genitalium [48]. The overlap between osmotic stress mease (Opp) transporter system (MG_077-MG_078- and metal starvation in bacteria has been previously MG_079 and MG_080). Oligopeptide and dipeptide documented [49,50]. In the upstream region of the permeases have been found to be significantly up-regu- Fur-regulated genes, we recognized a conserved lated during iron starvation in other pathogenic bac- sequence with dyad symmetry with the consensus teria [12,38] and they have been implicated in metal gcTTATTTtgAA-N3-TTcaAAATAAgc located near acquisition in S. aureus [39] and E. coli [40]. Moreover, the predicted −10 promoter elements (Figure 5(B)). some oligopeptide-binding proteins can accommodate Proximity of the predicted Fur binding sites to key pro- haeme in its binding pocket [41], hence providing a moter elements is in agreement with the classic repres- new strategy to acquire iron from the host. In addition, sor role of Fur regulators, which bind to the DNA and the oligopeptide transport ATP-binding protein OppD prevent RNA polymerase binding. Of note, the putative has been shown to be critical for survival of the avian Fur binding sites identified in M. genitalium differ con- pathogen Mycoplasma gallisepticum within the respir- siderably from the conventional Fur boxes originally atory tract [42], supporting the relevance of oligopep- described in E. coli [51,52]. However, divergent Fur tide permeases in mycoplasma infection. operators have been identified in mycobacteria and Additionally, our results also show that metal streptomycetes [53]. Despite the presence of a Fur depletion activates transcription of the glycerol uptake metalloregulator in M. genitalium, our analyses facilitator gene (glpF). This finding is notable because demonstrate that the transcriptional response to glycerol metabolism is involved in hydrogen peroxide metal deprivation in this emerging pathogen is mainly production in mycoplasmas and it represents a wide- Fur-independent. spread virulence factor of these unique pathogens In the respiratory pathogen M. pneumoniae, the [43–45]. Therefore, an increased glycerol uptake presence in the genome of a pseudopalindromic capacity could enhance the cytotoxic potential of sequence resembling the putative Fur boxes identified M. genitalium under metal limited conditions. On the in this work was already pinpointed in two comprehen- other hand, we found that metal starvation inhibits sive transcriptional studies [20,54]. Candidate Fur transcription of several genes involved in carbohydrate binding sites in M. pneumoniae are located in the metabolism. Moreover, we also observed decreased upstream region of the MPN162 (hrl homologue), transcript levels of the gene coding for the Ribulose- MPN433 (cbiO) and MPN043 genes (glpF) (Figure 5 phosphate 3-epimerase, an enzyme that has been (B)). Fur overexpression in M. pneumoniae induces shown to use zinc as a cofactor for catalysis along transcriptional changes in genes containing the ident- with cobalt and manganese [46]. Thus, our results indi- ified Fur boxes [54]. Unlike in M. pneumoniae, our cate that metal deprivation prompts an important results indicate that glpF expression is regulated by metabolic reprogramming in M. genitalium, with metal deprivation but not Fur in M. genitalium, altered expression of several genes encoding for puta- which likely reflects the exquisite adaptation of these tive transporters and metabolic enzymes. human pathogens to their respective infection niches. The MG_236 gene of M. genitalium codes for a Fur On the other hand, we demonstrate that homologue, which are widespread metalloregulators M. genitalium cells grown in a nutrient rich environ- controlling metal homeostasis in bacteria. Other ment preferentially accumulate zinc, iron and copper, human mycoplasmas including the respiratory patho- likely through constitutive metal transporters. This gen M. pneumoniae (MPN329) and the STI pathogen metal accumulation profile is similar to that described Mycoplasma penetrans (MYPE1200), also contain in E. coli by Zhao and co-workers [55]. According to putative Fur proteins. In a previous report, the this report, E. coli and Enterococcus faecium acquire MG_236 gene was classified as essential for preferentially iron, manganese and zinc, while nickel M. genitalium growth under laboratory culture and cobalt are predominant in S. aureus and Klebsiella 16 C. Martínez-Torró et al. pneumoniae. Remarkably, ICP-MS analysis of a of the CbiMNQO-like uptake system. Alternatively, M. genitalium fur mutant revealed increased concen- although metallophores have not been described in myco- trations of nickel compared to the wild-type strain, plasmas, Hrl may exhibit metallophore-like activity and establishing an important role for Fur in the acqui- cooperate with a yet to be identified EcfS protein to facili- sition of nickel in this urogenital pathogen. In addition, tate nickel uptake. Histidine-rich proteins are present in we found that high levels of zinc inhibit the expression other mycoplasma species, where they are usually anno- of genes under the control of Fur, suggesting that this tated as ZIP zinc transporter proteins. Experiments are transition metal might regulate Fur function in under way to establish the exact role of Hrl and the M. genitalium. Indeed, crystal structures of Fur pro- CbiMNQO-like transporter in metal uptake. teins from other bacteria have revealed coordination Given the tight link between metal acquisition sys- of zinc ions functioning as co-regulators to metal bind- tems and virulence, Fur-regulated proteins constitute ing sites [56–58]. Furthermore, zinc coordination in good therapeutic candidates for drug development. the Fur homologue of E. coli has been proposed to be Transporters from the ECF family, which are not pre- essential for dimerization, highlighting the importance sent in humans, have been proposed as potential anti- of metals for structural maintenance and activity of Fur microbial targets. Assessment of the likelihood to find a regulators [59]. This is also in agreement with a recent selective, low-molecular weight molecule that binds report, showing that thiolutin, a sulfur-containing anti- with high affinity to the target, which is usually referred biotic that has been used to chelate zinc[60], modifies to as druggability assessment, identified up to twelve transcription of Fur-regulated genes in druggable pockets within a reference ECF transporter M. pneumoniae [20]. Therefore, it is tempting to specu- system [63]. An alternative approach to prevent late that DPP treatment chelates zinc ions that are metal acquisition could be either the inhibition of bind- essential for Fur activity in M. genitalium, which ing of the S-components to the ECF module or the leads to de-repression of the Fur regulon. inhibition of their ability to bind metals. In this Nickel participates in different biological processes sense, compounds inhibiting Hrl activity may also pre- and they are considered essential micronutrients. vent M. genitalium growth in vivo. Of note, it has been Nickel is a cofactor of urease, Ni-Fe hydrogenases shown that Hrl from M. genitalium [32] and some superoxide dismutases, enzymes that do not M. pneumoniae [64] elicit a prominent pro-inflammatory seem to be present in M. genitalium. However, response and stimulate cytokine production through acti- M. genitalium is a facultative anaerobic organism, vation of NF-KB. Therefore, Hrl represents important which is relevant because many Ni2+-containing virulence protein and a key therapeutic drug target of enzymes are active under low-oxygen conditions. M. genitalium. In addition, Fur regulators are absent in Thus, we anticipate that M. genitalium encounters oxy- eukaryotes and they also constitute attractive antibacterial gen depletion during infection, which opens the possi- targets. Several attempts have been made to develop bility of the existence of unknown Ni2+-containing therapies based on inhibition of Fur activity in other enzymes in this bacterium. pathogens. Remarkably, several small peptides that inhi- A search for conserved domains within Hrl using bit Fur function have been shown to decrease pathogenic the MOTIF tool (www.genome.jp) reveals up to 13 E. coli strain virulence in a fly infection model [65]. hits, most of them belonging to known proteins In summary, in this study, we describe the Fur reg- involved in nickel, cobalt and zinc transport such as ulon of M. genitalium and identify the proteins that are RcnA, ZnuA, CbtA, ZntC, CzcD or Zip (Supplemental likely mediating nickel acquisition in this emerging Figure S1). For example, RcnA belongs to the Nickel/ human pathogen. Our results also suggest that zinc Cobalt Transporter (NicO) Family and it is believed ions might regulate Fur activity in M. genitalium, to catalyze Co2+ and Ni2+ efflux in E. coli [61]. As men- although further work is necessary to determine the tioned earlier in this text, the putative metal transporter biochemical nature of this regulator. Furthermore, regulated by Fur in M. genitalium shows homology to our results show a complex, multilayered transcrip- CbiMNQO uptake systems [62]. The CbiMNQO tional response to metal deprivation in this bacterium, protein complex belongs to the ECF (Energy Coupling suggesting a central role for metal regulatory systems in Factor) subtype of ABC transporters, which are com- the survival of this urogenital pathogen within the host. posed of two ATP-binding cassette ATPases (EcfA- MG304 and EcfA′ -MG303) and a transmembrane coupling subunit (EcfT-MG302). These three core Acknowledgements ECF proteins should interact with cognate substrate- binding subunits (EcfS), which confer ligand specifi- We are very grateful to Dr D. Scott Merrell and Caitlin Wil- liams (USUHS) for carefully reading and commenting on the city. In M. genitalium, no ecfS gene has been identified manuscript. We are also thankful to the Servei de Genòmica i in the vicinity of the metal responsive ECF-type operon Bioinformàtica (UAB) and the Genomics Unit (CRG) for identified in this study. Therefore, it is tempting to specu- performing DNA sequencing, and to the Servei de Química late that Hrl may function as a substrate-binding subunit Analítica (UAB) for performing ICP-MS analysis. STP and Emerging Microbes & Infections 17 CMT want to acknowledge a PIF fellowship from the Uni- Consultation and consensus Recommendations for versitat Autònoma de Barcelona. Future Research Priorities. J Infect Dis. 2017;216: S427–S430. doi:10.1093/infdis/jix147. [9] Deguchi T. Proposed treatment strategies for non- Author contributions gonococcal urethritis. Lancet Infect Dis. 2017;17:1121–1122. doi:10.1016/s1473-3099 CMT, STP and OQP conceived and designed the (17)30571-6. experiments; CMT, STP, MM, LSA, MGM, MMS and [10] Jensen JS, Cusini M, Gomberg M, et al. 2016 European OQP performed the experiments; CMT, STP, MM, guideline on Mycoplasma genitalium infections. J Eur APM, FC, JP and OQP analysed the data; FC, EQ Acad Dermatol Venereol. 2016;30:1650–1656. doi:10. 1111/jdv.13849. and JP contributed reagents/materials/analysis tools; [11] Neumann W, Gulati A, Nolan EM. Metal homeostasis and CMT, STP, APM, FC and OQP wrote the paper. in infectious disease: recent advances in bacterial metallophores and the human metal-withholding response. Curr Opin Chem Biol. 2017;37:10–18. Disclosure statement doi:10.1016/j.cbpa.2016.09.012. No potential conflict of interest was reported by the authors. [12] Merrell DS, Thompson LJ, Kim CC, et al. Growth phase-dependent response of Helicobacter pylori to iron starvation. Infect Immun. 2003;71:6510–6525. Funding doi:10.1128/IAI.71.11.6510-6525.2003. [13] Litwin CM, Calderwood SB. Role of iron in regulation This work was supported by the Ministerio de Ciencia, Inno- of virulence genes. Clin Microbiol Rev. 1993;6:137– vación y Universidades (Grant BIO2017-84166-R). The Pro- 149. doi:10.1128/CMR.6.2.137. teomics Unit at VHIO is a member ProteoRed, PRB3 (Grant [14] Tan S, Tompkins LS, Amieva MR. Helicobacter pylori IPT17/0019 – ISCIII-SGEFI/ERDF). usurps cell polarity to turn the cell surface into a repli- cative niche. PLoS Pathog. 2009;5:e1000407. doi:10. 1371/journal.ppat.1000407. ORCID [15] Tan S, Noto JM, Romero-Gallo J, et al. Helicobacter Sergi Torres-Puig http://orcid.org/0000-0002-8976-6488 pylori perturbs iron trafficking in the epithelium to Alex Perálvarez-Marín http://orcid.org/0000-0002-3457- grow on the cell surface. PLoS Pathog. 2011;7: 0875 e1002050. doi:10.1371/journal.ppat.1002050. Oscar Q. Pich http://orcid.org/0000-0003-3561-630X [16] Troxell B, Hassan HM. Transcriptional regulation by ferric uptake regulator (Fur) in pathogenic bacteria. Front Cell Infect Microbiol. 2013;3. doi:10.3389/ References fcimb.2013.00059. [17] Lee JW, Helmann JD. Functional specialization within [1] Moi H, Blee K, Horner PJ. Management of non-gono- the fur family of metalloregulators. BioMetals. coccal urethritis. BMC Infect Dis. 2015;15, doi:10. 2007;20:485–499. doi:10.1007/s10534-006-9070-7. 1186/s12879-015-1043-4. [18] Pich OQ, Merrell DS. The ferric uptake regulator of [2] Horner PJ, Blee K, Falk L, et al. 2016 European guide- Helicobacter pylori: a critical player in the battle for line on the management of non-gonococcal urethritis. iron and colonization of the stomach. Future Int J STD AIDS. 2016;27:928–937. doi:10.1177/ Microbiol. 2013;8:725–738. doi:10.2217/fmb.13.43. 0956462416648585. [19] Madsen ML, Nettleton D, Thacker EL, et al. [3] McGowin CL, Totten PA. The unique microbiology Transcriptional profiling of Mycoplasma hyopneumo- and molecular pathogenesis of Mycoplasma genita- niae during iron depletion using microarrays. lium. J Infect Dis. 2017;216:S382–S388. doi:10.1093/ Microbiology. 2006;152:937–944. doi:10.1099/mic.0. infdis/jix172. 28674-0. [4] Jensen JS. Mycoplasma genitalium : yet another chal- [20] Yus E, Lloréns-Rico V, Martínez S, et al. Determination lenging STI. Lancet Infect Dis. 2017;17:795–796. of the gene regulatory network of a genome-reduced doi:10.1016/S1473-3099(17)30364-X. bacterium highlights alternative regulation indepen- [5] Bradshaw CS, Horner PJ, Jensen JS, et al. Syndromic dent of transcription factors. Cell Syst. 2019;9:143– management of STIs and the threat of untreatable 158.e13. doi:10.1016/j.cels.2019.07.001. Mycoplasma genitalium. Lancet Infect Dis. [21] Torres-Puig S, Broto A, Querol E, et al. A novel sigma 2018;18:251–252. doi:10.1016/s1473-3099(18)30080-x. factor reveals a unique regulon controlling cell-specific [6] Anagrius C, Loré B, Jensen JS. Mycoplasma genitalium: recombination in Mycoplasma genitalium. Nucleic prevalence, clinical significance, and transmission. Sex Acids Res. 2015;43:4923–4936. doi:10.1093/nar/ Transm Infect. 2005;81:458–462. doi:10.1136/sti.2004. gkv422. 012062. [22] Torres-Puig S, Martínez-Torró C, Granero-Moya I, [7] Soni S, Alexander S, Verlander N, et al. The prevalence et al. Activation of σ20-dependent recombination of urethral and rectal Mycoplasma genitalium and its and horizontal gene transfer in Mycoplasma genita- associations in men who have sex with men attending lium. DNA Res. 2018;25:383–393. doi:10.1093/ a genitourinary medicine clinic. Sex Transm Infect. dnares/dsy011. 2010;86:21–24. doi:10.1136/sti.2009.038190. [23] Langmead B, Salzberg SL. Fast gapped-read alignment [8] Martin DH, Manhart LE, Workowski KA. Mycoplasma with Bowtie 2. Nat Methods. 2012;9:357–359. doi:10. genitalium from basic science to public health: sum- 1038/nmeth.1923. mary of the results from a National Institute of [24] Li H, Handsaker B, Wysoker A, et al. The sequence Allergy and Infectious Disease Technical alignment/Map format and SAMtools. 18 C. Martínez-Torró et al. Bioinformatics. 2009;25:2078–2079. doi:10.1093/ urinary tract infection. Mol Microbiol. 2010;77:1246– bioinformatics/btp352. 1260. doi:10.1111/j.1365-2958.2010.07287.x. [25] Liao Y, Smyth GK, Shi W. Featurecounts: An efficient [41] Tanaka KJ, Pinkett HW. Oligopeptide-binding protein general purpose program for assigning sequence reads from nontypeable Haemophilus influenzae has ligand- to genomic features. Bioinformatics. 2014;30:923–930. specific sites to accommodate peptides and heme in the doi:10.1093/bioinformatics/btt656. binding pocket. J Biol Chem. 2019;294:1070–1082. [26] Love MI, Huber W, Anders S. Moderated estimation of doi:10.1074/jbc.RA118.004479. fold change and dispersion for RNA-seq data with [42] Tseng C-W, Chiu C-J, Kanci A, et al. The oppD gene DESeq2. Genome Biol. 2014;15. doi:10.1186/s13059- and putative peptidase genes may be required for viru- 014-0550-8. lence in Mycoplasma gallisepticum. Infect Immun. [27] Pfaffl MW. A new mathematical model for relative 2017;85. doi:10.1128/iai.00023-17. quantification in real-time RT-PCR. Nucleic Acids [43] Pilo P, Vilei EM, Peterhans E, et al. A metabolic enzyme as Res. 2001;29:e45. a primary virulence factor of Mycoplasma mycoides subsp. [28] Párraga-Niño N, Colomé-Calls N, Canals F, et al. A mycoides small colony. J Bacteriol. 2005;187:6824–6831. comprehensive proteome of Mycoplasma genitalium. doi:10.1128/JB.187.19.6824-6831.2005. J Proteome Res. 2012;11:3305–3316. doi:10.1021/ [44] Hames C, Halbedel S, Hoppert M, et al. Glycerol pr300084c. metabolism is important for cytotoxicity of [29] Williams CL, Neu HM, Michel SLJ, et al. Measuring Mycoplasma pneumoniae. J Bacteriol. 2009;191:747– intracellular metal concentration via ICP-MS following 753. doi:10.1128/JB.01103-08. copper exposure. Methods Mol Biol. 2019;1946:195– [45] Großhennig S, Schmidl SR, Schmeisky G, et al. 205. doi:10.1007/978-1-4939-9118-1_19. Implication of glycerol and phospholipid transporters [30] Waterhouse A, Bertoni M, Bienert S, et al. SWISS- in Mycoplasma pneumoniae growth and virulence. MODEL: homology modelling of protein structures Infect Immun. 2013;81:896–904. doi:10.1128/iai. and complexes. Nucleic Acids Res. 2018;46:W296– 01212-12. W303. doi:10.1093/nar/gky427. [46] Liang W, Ouyang S, Shaw N, et al. Conversion of d [31] Kingsford CL, Ayanbule K, Salzberg SL. Rapid, accu- -ribulose 5-phosphate to d -xylulose 5-phosphate: rate, computational discovery of Rho-independent new insights from structural and biochemical studies transcription terminators illuminates their relationship on human RPE. FASEB J. 2010;25:497–504. doi:10. to DNA uptake. Genome Biol. 2007;8:R22. doi:10. 1096/fj.10-171207. 1186/gb-2007-8-2-r22. [47] Glass JI, Assad-Garcia N, Alperovich N, et al. Essential [32] Shimizu T, Kida Y, Kuwano K. A triacylated lipoprotein genes of a minimal bacterium. Proc Natl Acad Sci. from Mycoplasma genitalium activates NF-κB through 2006;103:425–430. doi:10.1073/pnas.0510013103. Toll-like receptor 1 (TLR1) and TLR2. Infect Immun. [48] Zhang W, Baseman JB. Transcriptional regulation of 2008;76:3672–3678. doi:10.1128/IAI.00257-08. MG_149, an osmoinducible lipoprotein gene from [33] Hood MI, Skaar EP. Nutritional immunity: transition Mycoplasma genitalium. Mol Microbiol. 2011;81:327– metals at the pathogen-host interface. Nat Rev 339. doi:10.1111/j.1365-2958.2011.07717.x. Microbiol. 2012;10:525–537. doi:10.1038/ [49] Gancz H, Merrell DS. The Helicobacter pylori ferric nrmicro2836. uptake regulator (Fur) is essential for growth under [34] Tryon VV, Baseman JB. The acquisition of human lac- sodium chloride stress. J Microbiol. 2011;49:294–298. toferrin by Mycoplasma pneumoniae. Microb Pathog. doi:10.1007/s12275-011-0396-7. 1987;3:437–443. doi:10.1016/0882-4010(87)90013-1. [50] Gancz H, Jones KR, Merrell DS. Sodium chloride [35] Bullen JJ, Rogers HJ, Spalding PB, et al. Iron and infec- affects Helicobacter pylori growth and gene expression. tion: the heart of the matter. FEMS Immunol Med J Bacteriol. 2008;190:4100–4105. doi:10.1128/JB. Microbiol. 2005;43:325–330. doi:10.1016/j.femsim. 01728-07. 2004.11.010. [51] De Lorenzo V, Wee S, Herrero M, et al. Operator [36] Musatovova O, Dhandayuthapani S, Baseman JB. sequences of the aerobactin operon of plasmid colV– Transcriptional heat shock response in the smallest K30 binding the ferric uptake regulation (fur) repres- known self-replicating cell, Mycoplasma genitalium. J sor. J Bacteriol. 1987;169:2624–2630. doi:10.1128/jb. Bacteriol. 2006;188:2845–2855. doi:10.1128/JB.188.8. 169.6.2624-2630.1987. 2845-2855.2006. [52] Escolar L, Pérez-Martín J, De Lorenzo V. Binding of [37] Hauryliuk V, Atkinson GC, Murakami KS, et al. Recent the Fur (ferric uptake regulator) repressor of functional insights into the role of (p)ppGpp in bac- Escherichia coli to arrays of the GATAAT sequence. J terial physiology. Nat Rev Microbiol. 2015;13:298– Mol Biol. 1998;283:537–547. doi:10.1006/jmbi.1998. 309. doi:10.1038/nrmicro3448. 2119. [38] Hempel K, Herbst FA, Moche M, et al. Quantitative [53] Milano A, Forti F, Sala C, et al. Transcriptional regu- proteomic view on secreted, cell surface-associated, lation of furA and katG upon oxidative stress in and cytoplasmic proteins of the methicillin-resistant Mycobacterium smegmatis. J Bacteriol. 2001;183:6801– human pathogen Staphylococcus aureus under iron- 6806. doi:10.1128/JB.183.23.6801-6806.2001. limited conditions. J Proteome Res. 2011;10:1657– [54] Güell M, Van Noort V, Yus E, et al. Transcriptome 1666. doi:10.1021/pr1009838. complexity in a genome-reduced bacterium. [39] Letoffe S, Delepelaire P, Wandersman C. The house- Science. 2009;326:1268–1271. doi:10.1126/science. keeping dipeptide permease is the Escherichia coli 1176951. heme transporter and functions with two optional pep- [55] Zhao S, Cao S, Luo L, et al. A preliminary investigation tide binding proteins. Proc Natl Acad Sci. of metal element profiles in the serum of patients with 2006;103:12891–12896. doi:10.1073/pnas.0605440103. bloodstream infections using inductively-coupled [40] Hiron A, Posteraro B, Carrière M, et al. A nickel ABC- plasma mass spectrometry (ICP-MS). Clin Chim transporter of Staphylococcus aureus is involved in Acta. 2018;485:323–332. doi:10.1016/j.cca.2018.07.013. Emerging Microbes & Infections 19 [56] Pohl E, Haller JC, Mijovilovich A, et al. Architecture of [62] Rodionov DA, Hebbeln P, Gelfand MS, et al. a protein central to iron homeostasis: crystal structure Comparative and functional genomic analysis of pro- and spectroscopic analysis of the ferric uptake regula- karyotic nickel and cobalt uptake transporters: evi- tor. Mol Microbiol. 2003;47:903–915. doi:10.1046/j. dence for a novel group of ATP-binding cassette 1365-2958.2003.03337.x. transporters. J Bacteriol. 2006;188:317–327. doi:10. [57] Jacquamet L, Aberdam D, Adrait A, et al. X-ray 1128/JB.188.1.317-327.2006. absorption spectroscopy of a new zinc site in the Fur [63] Bousis S, Setyawati I, Diamanti E, et al. Energy-coup- protein from Escherichia coli. Biochemistry. ling factor transporters as novel antimicrobial targets. 1998;37:2564–2571. doi:10.1021/bi9721344. Adv Ther. 2018;2:1800066. doi:10.1002/adtp. [58] Sheikh MA, Taylor GL. Crystal structure of the Vibrio 201800066. cholerae ferric uptake regulator (Fur) reveals insights [64] Shimizu T, Kida Y, Kuwano K. Triacylated lipoproteins into metal co-ordination. Mol Microbiol. 2009;72:1208– derived from Mycoplasma pneumoniae activate nuclear 1220. doi:10.1111/j.1365-2958.2009.06718.x. factor-κB through toll-like receptors 1 and 2. [59] D’Autréaux B, Pecqueur L, De Peredo AG, et al. Immunology. 2007;121:473–483. doi:10.1111/j.1365- Reversible redox- and zinc-dependent dimerization 2567.2007.02594.x. of the Escherichia coli fur protein. Biochemistry. [65] Mathieu S, Cissé C, Vitale S, et al. From peptide apta- 2007;46:1329–1342. doi:10.1021/bi061636r. mers to inhibitors of FUR, bacterial transcriptional [60] Lauinger L, Li J, Shostak A, et al. Thiolutin is a zinc regulator of iron homeostasis and virulence. ACS chelator that inhibits the Rpn11 and other JAMM Chem Biol. 2016;11:2519–2528. doi:10.1021/ metalloproteases. Nat Chem Biol. 2017;13:709–714. acschembio.6b00360. doi:10.1038/nchembio.2370. [66] Sarvan S, Charih F, Askoura M, et al. Functional [61] Rodrigue A, Effantin G, Mandrand-Berthelot M-A. insights into the interplay between DNA inter- Identification of rcnA (yohM), a nickel and cobalt resist- action and metal coordination in ferric uptake reg- ance gene in Escherichia coli. J Bacteriol. 2005;187:2912– ulators. Sci Rep. 2018;8. doi:10.1038/s41598-018- 2916. doi:10.1128/JB.187.8.2912-2916.2005. 25157-6.