Original Research ARTICLE
Pleiotropic Clostridioides difficile Cyclophilin PpiB Controls Cysteine-Tolerance, Toxin Production, the Central Metabolism and Multiple Stress Responses
- 1Institut für Mikrobiologie, Technische Universität Braunschweig, Braunschweig, Germany
- 2Moleküler Biyoteknoloji Bölümü, Türk-Alman Üniversitesi, Istanbul, Turkey
- 3Braunschweig Integrated Centre of Systems Biology, Braunschweig, Germany
- 4Cellular Proteomics Research, Helmholtz Centre for Infection Research, Braunschweig, Germany
- 5Helmholtz Centre for Infection Research, Braunschweig, Germany
The Gram-positive pathogen Clostridioides difficile is the main bacterial agent of nosocomial antibiotic associated diarrhea. Bacterial peptidyl-prolyl-cis/trans-isomerases (PPIases) are well established modulators of virulence that influence the outcome of human pathologies during infections. Here, we present the first interactomic network of the sole cyclophilin-type PPIase of C. difficile (CdPpiB) and show that it has diverse interaction partners including major enzymes of the amino acid-dependent energy (LdhA, EtfAB, Had, Acd) and the glucose-derived (Fba, GapA, Pfo, Pyk, Pyc) central metabolism. Proteins of the general (UspA), oxidative (Rbr1,2,3, Dsr), alkaline (YloU, YphY) and cold shock (CspB) response were found bound to CdPpiB. The transcriptional (Lrp), translational (InfC, RFF) and folding (GroS, DnaK) control proteins were also found attached. For a crucial enzyme of cysteine metabolism, O-acetylserine sulfhydrylase (CysK), the global transcription regulator Lrp and the flagellar subunit FliC, these interactions were independently confirmed using a bacterial two hybrid system. The active site residues F50, F109, and F110 of CdPpiB were shown to be important for the interaction with the residue P87 of Lrp. CysK activity after heat denaturation was restored by interaction with CdPpiB. In accordance, tolerance toward cell wall stress caused by the exposure to amoxicillin was reduced. In the absence of CdPpiB, C. difficile was more susceptible toward L-cysteine. At the same time, the cysteine-mediated suppression of toxin production ceased resulting in higher cytotoxicity. In summary, the cyclophilin-type PPIase of C. difficile (CdPpiB) coordinates major cellular processes via its interaction with major regulators of transcription, translation, protein folding, stress response and the central metabolism.
The multi-resistant Gram-positive obligate anaerobe Clostridioides (Clostridium) difficile is the main nosocomial bacterial agent of antibiotic treatment associated diarrhea (Hopkins and Wilson, 2018). The rise of C. difficile infection (CDI) coincides with the introduction of clindamycin, the first broad-band lincosamin against anaerobic Gram-negative pathogens (Bartlett et al., 1977; Lusk et al., 1978). Currently, first line of treatment is the use of antibiotics like vancomycin, metronidazole and fidaxomicin. However, high recurrence rates, especially after treatment with vancomycin or metronidazole, make this disease difficult to manage (Mullane, 2014; Hopkins and Wilson, 2018).
A major factor influencing the course and outcome of CDI is the composition of the host intestinal microbiome and, connected to it, the bile acid status of the infected individual. Patients undergoing a CDI typically have an antibiotic dependent change of their microbiome followed by a transition from antibacterial to less antibacterial bile acids in their gastrointestinal tract (Theriot et al., 2016). Most commonly, CDI manifests in form of pseudomembranous colitis, which is a strong inflammation of the large intestine. This is caused by massive tissue destruction and leukocytosis due to the production of two large glycosylating toxins, TcdA and TcdB. These exert their cytotoxic activity by inactivating small GTPases and disturbing the actin cytoskeleton dynamics leading to disruption of the gut epithelial barrier (Barbut et al., 2007; Rupnik et al., 2009; Chandrasekaran and Lacy, 2017).
The production of TcdA and TcdB typically coincides with the entry of the bacteria into the stationary phase. Nevertheless, it is controlled by complex regulatory processes on multiple levels and by diverse environmental cues, including subinhibitory concentrations of antibiotics, temperature or the availability of carbon sources and amino acids (Dupuy and Sonenshein, 1998; Bouillaut et al., 2015). One central metabolite in this respect is L-cysteine that represses toxin production (Karlsson et al., 2000). Apart from that, L-cysteine also affects the expression of genes of amino acid biosynthesis, fermentation, energy metabolism, iron acquisition and stress response (Dubois et al., 2016; Gu et al., 2018). Furthermore, there is a strong interconnectedness between iron and L-cysteine dependent gene regulation circuits. Several of the genes of the latter contain Fur boxes in their promoters, which are binding sites for the global iron-responsive transcriptional ferric uptake regulator (Fur) (Dubois et al., 2016). If not metabolized effectively L-cysteine also has an inhibitory effect on growth on C. difficile as it was nicely shown in case of a mutant lacking the cysteine desulfidase CdsB (Gu et al., 2017).
Although there is strong correlation between CDI severity and the capability of the bacteria to produce TcdA and TcdB, this alone does not explain the virulence spectrum of C. difficile in its entirety. Accordingly, several other virulence factors such as extracellular proteases, surface layer proteins, a fibronectin binding protein (Fbp68), a collagen binding protein (CbpA), type IV pili and flagella have been shown or are suspected to contribute to disease severity and host colonization (Geric et al., 2004; Janoir, 2016; Ünal and Steinert, 2016; Péchiné et al., 2018). Especially, the regulation of flagellation and its main structural component flagellin (FliC) is crucial for modulating adherence to host cells, promoting colonization and eliciting immune responses via toll-like receptor 5 (Tasteyre et al., 2001; Dingle et al., 2011; Batah et al., 2016). Moreover, flagellation and toxin production are co-regulated by an intricate genetic switch adding up to the complexity of toxin-related pathogenicity of C. difficile (Aubry et al., 2012; Anjuwon-Foster and Tamayo, 2017).
Currently, many single components and mechanisms of CDI that contribute to host colonization, disease outbreak and dissemination are known. However, their concerted action is still not fully understood. Bacterial PPIases (FKBPs, parvulins and cyclophilins) have in many instances been shown to modulate infectious processes (Ünal and Steinert, 2014). Among those are prominent representatives like Mip and Mip-like FKBPs, or parvulins, like SurA in Gram-negative as well as PrsA and PrsA2 in Gram-positive bacteria (Jacobs et al., 1993; Heikkinen et al., 2009; Behrens-Kneip, 2010; Alonzo et al., 2011; Obi et al., 2011; Ikolo et al., 2015; Ünal and Steinert, 2015). Additionally, recent studies have shown the participation of cyclophilins in the maturation of a secreted nuclease in Staphylococcus aureus as well as in virulence traits, like biofilm formation in Escherichia coli or in sliding motility and amoeba infection by Legionella pneumophila (Skagia et al., 2016; Wiemels et al., 2017; Rasch et al., 2018). The aim of our study was to identify the interaction partners of the sole cyclophilin of C. difficile (CdPpiB) and to characterize its contribution to bacterial physiology as well as virulence. Our results show an unprecedented relationship between bacterial cysteine metabolism and CdPpiB with far reaching implications on virulence. Furthermore, a protein network controlling major cellular function including transcription, translation, protein folding, metabolism, and stress responses was uncovered.
Materials and Methods
Bacterial Strains and Culture
Bacterial strains and plasmids used in this study are listed in Table 1. C. difficile 630Δerm (hereafter referred to as wild type or wt) and its derivatives were cultured in BHIS medium (brain-heart infusion broth (Carl Roth GmbH, Germany) supplemented with 5 g/L yeast extract (BD BactoTM, United States) and 1 g/L L-cysteine (Sigma-Aldrich®, Germany) or TY-medium (30 g/L tryptone, 20 g/L yeast extract) under anaerobic conditions (95% N2/5% H2). If necessary, 5 μg/mL erythromycin or 15 μg/mL thiamphenicol were added. Escherichia coli and Bacillus megaterium were cultured in LB medium supplemented with 100 μg/mL ampicillin, 10 μg/mL tetracycline or 30 μg/ml chloramphenicol at 37°C and 200 rpm. All antibiotics were purchased from Sigma-Aldrich®. Media were solidified by adding 1.5% (wt/vol) agar when needed.
In vivo Cross-Linking for the Identification of PpiB Interaction Partners
Possible interaction partners of PpiB were identified by in vivo cross-linking as previously described (Borrero-de Acuña et al., 2015). Briefly, C. difficile str. 630ΔermΔppiB + PpiB-NStrep was grown for 8 h in 25 mL of BHIS and refreshed 1:1000 in 2 L BHIS. After 16 h 0.125 % (vol/vol) formaldehyde (AppliChem, Germany) was added for chemical cross-linking (30 min at 37°C). Excess formaldehyde was quenched for 5 min at 37°C by the addition of glycine at a final concentration of 130 mM. Cells were harvested and washed twice with PBS (4000 g for 20 min at 4°C). Finally, the cells were resuspended in 25 mL PBS supplemented with complete Protease Inhibitor Cocktail (Roche, Switzerland), and homogenized using FRENCH® Press (Thermo, United States) or glass beads and FastPrep® (MP Biomedicals, United States). Cell debris was removed by centrifugation (10,000 g, 10 min, 4°C), and filtering the supernatant through a 0.45 μm syringe filter. The supernatant was applied onto a 3 mL Strep-Tactin® column (iba Lifesciences, Germany) in order to purify PpiB and its cross-linked partners according to manual instructions. The elution fraction was concentrated using Amicon Ultra-0.5 mL centrifugal filters (Merck, Germany) with a molecular cut-off of 10 kDa. For analysis, the concentrated eluate was mixed with 4x SDS loading buffer and subjected to decrosslinking by heating at 95°C for 30 min. Proteins were separated on a 12% acrylamide gel, and stained using Coomassie stain [100 g/L (NH4)2SO4, 100 mL/L H3PO4, 20% vol/vol methanol]. Bands of interest were analyzed by LC-MS as previously described (Borrero-de Acuña et al., 2016). Resulting protein identifications (hits) were filtered according to their overall coverage by the LC/MS approach and the overall number of unique peptides. Hits with at least 15% coverage and two or more unique peptides were taken into account as this indicated robust enrichment. Ribosomal proteins were discarded because of their intrinsically high abundance in the cell and their known unspecific protein-protein interactions due to their overall strong basic character.
Bacterial Two Hybrid (BACTH) Screening
Protein-protein interactions were confirmed using the BACTH-system of Euromedex (France), which bases on an adenylate cyclase (CyaA) dependent reporter system. For this, ppiB, fliC (CD630_02390), lrp (CD630_35440) and cysK (CD630_15940) were cloned in frame with in two different vectors, each containing one subunit of the adenylate cyclase of Bordetella pertussis (Tables 1, 2). After confirmation of correct cloning, the CyaA-deficient strain E. coli BTH101 was transformed with combinations of plasmids carrying ppiB and the respective interaction partners, and transformants were selected on LB-agar containing 50 μg/mL kanamycin and 100 μg/mL ampicillin. Single transformants were then streaked out on MacConkey-agar (BD DifcoTM, United States) supplemented with 50 μg/mL kanamycin and 100 μg/mL ampicillin, incubated for 2 days at 30°C, and checked for pink coloring due to CyaA-activity.
Measuring the Interaction of PpiB and Lrp and Their Amino Acid Substitution Mutants by β-Galactosidase Assay
Escherichia coli BTH101 carrying plasmid combinations that had been confirmed on MacConkey-agar were used in a β-galactosidase assay for quantitative comparison of interactions between amino acid mutants of PpiB and Lrp, respectively. For this, at least six clones of each transformation were picked and grown o/n (30°C, 200 rpm) in 5 mL LB medium supplemented with 50 μg/mL kanamycin, 100 μg/mL ampicillin and 0.5 mM IPTG (GERBU, Germany). Prior to the assay, bacterial growth was stalled by chilling the cultures for at least 30 min at 4°C and the OD600 was determined. Bacteria of 200 μL culture were pelleted (3500 rpm, 5 min, 4°C), and resuspended in 900 μL of Z-Buffer (60 mM Na2HPO4, 40 mM NaH2PO4, 10 mM KCl, 1 mM MgSO4, and freshly added 50 mM β-mercaptoethanol, pH 7.0). Bacteria were lysed by the addition of 50 μL chloroform and 50 μL 0.1% SDS (wt/vol) for 5 min at 30°C. After this, 200 μl of o-nitrophenyl-β-D-galactoside (ONPG, Roth GmbH, Germany) dissolved at 4 mg/mL in phosphate buffer (60 mM Na2HPO4, 40 mM NaH2PO4, pH 7.0) was added. The reactions were kept for 12 min at 30°C and subsequently stopped by the addition of 500 μL 1 M Na2CO3. After removal of cell debris and chloroform by centrifugation (14000 rpm, 5 min, 4°C) absorption at 420 and 550 nm were recorded with a VarioSkanTM (Thermo Fisher, United States) plate reader. β-galactosidase activity, expressed in Miller Units, was calculated according to the following formula:
Construction of a ppiB Destruction Mutant (ΔppiB) in C. difficile 630Δerm Using ClosTron
A synthetic vector containing the region of ppiB (CD630_03310/CDIF630erm_00459) was designed with the help of the ClosTron website1 using the Perutka algorithm (Perutka et al., 2004). E. coli CA434 was transformed with customized vectors for mating with C. difficile 630Δerm cells as described previously (Heap et al., 2009, 2010). Mutants were selected on BHIS containing 15 μg/mL erythromycin and confirmed by PCR and sequencing using gene specific primers (Table 2).
Assessing Growth and Susceptibility Toward Amoxicillin
Bacterial growth was monitored by refreshing cultures of C. difficile grown overnight (o/n) in 25 mL medium to a starting optical density (OD600 nm) of 0.05 in 25 mL fresh medium. Increase in OD600 nm was measured starting with t0 and every 60 min until the cultures reached stationary phase. For assessing amoxicillin susceptibility, o/n cultures of wild type C. difficile and its ΔppiB mutant were prepared in 20 mL BHIS medium. The next morning the cultures were adjusted to an OD600 nm of 0.1 in BHIS medium. Amoxicillin stock solution was prepared fresh (2 mM in ddH2O) and a 1:1 serial dilution starting at 256 μM was prepared in 100 μl of BHIS distributed in the wells of a 96-well plate. Following this, 100 μL of the bacterial suspension were added into each well resulting in a final OD of 0.05. Bacteria were grown for 6 h at 37°C under anaerobic conditions, and the final OD was measured at 600 nm using a VarioSkanTM (Thermo Fisher, United States) plate reader.
Complementation of C. difficileΔppiB With StrepII-Tagged PpiB and Implementing a Tetracycline- Inducible Promoter
In order to achieve a complementation of the mutant with wild type PpiB carrying a N-terminal StrepII-tag for affinity purification, ppiB was amplified including its 250 bp up- and 100 bp-downstream regions using the primer combination PpiB_KompF1/R1. This fragment was cloned into the shuttle vector pMTL82151 with HindIII and BamHI yielding pMTL_ppiB. This plasmid was used as a template for an inverse PCR by which an N-terminal StrepII-tag was added to PpiB with the primer combination PpiB_NStrep_F/R. The resulting PCR-product was gel purified, phosphorylated with T4 polynucleotide kinase (NEB) and re-ligated with T4 ligase (NEB) according to manufacturer’s instructions. Plasmids carrying the desired mutations were selected in E. coli DH10β and verified by sequencing. For complementation of the ΔppiB mutant under the control of a tetracycline-inducible promoter (Ptet), the ORF of CdPpiB was cloned into pDSW1728 utilizing the restriction sites SacI and BamHI. The ΔppiB mutant was transformed by conjugation and selected using 15 μg/ml thiamphenicol. For complementation studies the bacteria were cultured with 10 ng/mL anhydrotetracycline (AT), and a dose-dependent complementation was achieved by performing the assays with decreasing concentrations of TA starting at 400 ng/mL.
Recombinant Protein Production
For recombinant production, ppiB or cysK were amplified with Q5-Polymerase (NEB GmbH, Germany) and cloned into the production vector pN-STREPXa1622 using the restriction sites BglII and SphI yielding N-terminally StrepII-tagged proteins (Biedendieck et al., 2007). For recombinant protein production, an o/n culture of transformed B. megaterium MS941 was prepared, and the next morning refreshed 1:100 in 300 ml LB medium containing 10 μg/mL tetracycline (Biedendieck et al., 2011). The cells were induced at an OD600 nm of ∼0.4 by the addition of D-xylose (Carl Roth GmbH, Germany) at a final concentration of 0.5 % (w/v). After 6 h, the cells were harvested (3000 g, 15 min, 4°C), and the pellet was once washed with 1x PBS (5000 g, 15 min, 4°C). Cells were resuspended in 25 mL ddH2O, and disrupted by FRENCH® Press (Thermo, United States). Immediately after the disruption, 10x wash buffer was added to a final concentration of 1x. Cell debris was removed by centrifugation (10000 g, 20 min, 4°C) and the supernatant was filtered through a 0.45 μm syringe filter. StrepII-tagged PpiB or CysK were purified from the supernatant using Strep-Tactin® sepharose (iba Lifesciences, Germany) according to manufacturer’s instructions. Concentration and buffer change of purified proteins were achieved by centrifugation (5000 g, RT) using Vivaspin® 20 ultrafiltration units (Sartorius, Germany) with 30 or 10 kDa molecular weight cut offs. Proteins were stored in 100 mM HEPES (pH 7.0) at −20°C. Protein yield and purity were analyzed by SDS-page, and protein concentration was determined by absorption at 280 nm.
CysK Activity Assay
Enzymatic activity of recombinantly produced CysK of C. difficile was measured photometrically as previously described with slight modifications (Tai et al., 1993; Saavedra et al., 2004). Briefly, the assays were performed in a 96-well format in 100 μL final volume. As substrates O-Acetyl-L-serine hydrochloride (OAS; Sigma-Aldrich®, A6262) and 5-Thio-2-nitrobenzoic acid (TNB) were used. 20 mM OAS stock solution was freshly prepared in 100 mM HEPES (pH 7.0). Fresh TNB-stock solution was prepared by dissolving 10 mM 5,5′-Dithiobis(2-nitrobenzoic acid) (DTNB; Sigma-Aldrich®, D8130) and 15 mM dithiothreitol in 100 mM HEPES (pH 7.0). Each reaction contained 2 mM OAS and 50 μM TNB. Following the addition of CysK at a final concentration of 7 μM, the reactions were monitored in a VarioSkanTM (Thermo Fisher, United States) plate reader at 30°C by measuring the OD412 nm every 4 min. If needed CysK was denatured by incubating aliquots for 30 min at 56°C.
Determination of Toxin Concentration in Supernatants
For determining the toxin production, exponentially growing cultures were adjusted to an OD600 nm of 0.05 in 10 mL of fresh BHIS and grown at 37°C under anaerobic conditions. After 48 h, the culture supernatants of 2 mL of each culture were harvested (5000 g, 5 min, RT) and sterile filtered using 0.2 μm syringe filters. Toxin concentrations in culture supernatants were determined using the Clostridium difficile Toxin A OR B ELISA Kit (tgc-E002-1) of tgcBIOMICS (Germany) following the manual instructions.
In order to visualize the cytotoxic activity of C. difficile culture supernatants, NIH-3T3 mouse fibroblast cells (ATCC® CRL-1658TM, United States) were cultured in DMEM (Gibco, United States) supplemented with 10% FBS at 37°C and 5% CO2. For the assay, 105 cells/mL were seeded on glass cover slips in 24-well plates and cultured for 2 days. In parallel, C. difficile were grown in anaerobic chamber for 24 h, and the cell free supernatant containing secreted toxins was harvested by centrifugation (8000 g, 5 min, RT). This supernatant was diluted to 0.1 % (vol/vol) in cell culture medium. Adherent NIH-3T3 cells were washed once with 1 mL PBS, and 1 mL medium with bacterial culture supernatant was added. After incubation for 2 h, medium was removed and the cells were fixed for 15 min at RT with 2% (wt/vol) PFA in 1x PBS, and subsequently permeabilized with 0.1 % (vol/vol) Triton X-100 in 1x PBS with 1% (wt/vol) BSA for 5 min. Cells were stained with PBS containing 1 μg/mL DAPI (Sigma-Aldrich®) and 1:1000 diluted Phalloidin-Alexa488 (Abcam, United Kingdom). The mounted samples were analyzed with a confocal microscope (Leica® SP8, Germany) using an 63x oil immersion objective.
Identification of the Protein Interaction Partners of CdPpiB
A protein blast search using the PpiB sequence of Bacillus subtilis str. 168 (Uniprot ID: P35137) against the most recently annotated C. difficile str. 630Δerm genome identified a single potential cyclophilin encoded by the gene locus CDIF630erm_00459 (CD630_00330 in C. difficile 630). This gene is flanked upstream by a gene potentially encoding for a patatin-like phospholipase, and downstream by a gene encoding the putative exosporium glycoprotein BclA1 (Dannheim et al., 2017a). An alignment of the potential C. difficile cyclophilin with bacterial cyclophilins of E. coli str. K12 (EcPpiB; P23869), B. subtilis (BsPpiB; P35137) and Staphylococcus aureus str. NCTC8325 (SaPpiB; Q2FZU9) as well as eukaryotic representative members of yeast (ScCyp1; P14832) and human (HsCypA; P62937) revealed high amino acid sequence similarities between the different proteins. An exceptional high degree of conservation was observed for amino acid residues involved in PPIase activity (Figure 1). Accordingly, this gene was selected for destruction by ClosTron, which resulted in the generation of a ΔppiB mutant in the C. difficile str. 630Δerm background.
Figure 1. CdPpiB is a classical cyclophilin-type PPIase. The alignment of amino acid sequences of cyclophilin type PPIases of the Gram-positive species S. aureus (SaPpiB; Q2FZU9), C. difficile (CdPpiB; Q18D70) and B. subtilis (BsPpiB; P35137) as well as the Gram-negative model organism E. coli (EcPpiB; P23869), and the yeast (ScCyp1; P14832) and human CypA (HsCypA; P62937) reveals highly conserved amino acids along the PPIase active surface (marked by a black line). The alignment was performed using the T-Coffee webserver and visualized by Jalview 2.10.3b1. Amino acids that are conserved to 100% are highlighted by color coding according to Taylor (Taylor, 1986; Notredame et al., 2000; Waterhouse et al., 2009). Amino acids that have been mutated in CdPpiB are marked by an asterisk.
Next, we identified in vivo protein interaction partners of CdPpiB using an interactomics approach by using N-terminally StrepII-tagged PpiB as bait and formaldehyde as a cross-linker. A preparative SDS-PAGE of the purified protein complexes revealed in total, seven distinct protein bands that were cut out and prepared for LC-MS analysis (Figure 2).
Figure 2. In vivo cross-linking reveals putative interaction partners of CdPpiB. Shown is a representative analytical SDS-PAGE with the elution fraction after the in vivo cross-linking and the subsequent Strep-tag affinity purification of PpiB-StrepII and its putative interaction partners. Bands denoted with an asterisk were cut out analyzed by mass spectrometry. The bracket indicates bands that also occurred in control experiments without cross-linking and were excluded from the analysis.
This yielded 39 unique hits, and as expected, PpiB was found to be with 83% the best covered protein among those. The majority of the remaining 38 hits that resembled proteins interacting either directly or indirectly in larger complexes with CdPpiB belonged to the functional groups metabolism (19 hits, 50%) and stress response (8 hits, 21%). The rest of the hits were distributed among transcription/translation, protein folding and transport (3 hits and 8% each) as well as motility/surface (2 hits 5%) (Table 3). Among the retrieved potential interaction partners were LdhA, AcdB, HadA, EtfA1, EtfA3, and EtfB1. These enzymes represent the complete pathway for phenylalanine and leucine fermentation (Kim et al., 2006). Similarly, the also detected pyruvate carboxylase Pyc, pyruvate-ferredoxin oxidoreductase Pfo, aldehyde-alcohol dehydrogenase AdhE2 and pyruvate kinase Pyk as well as the glyceraldehyde-3-phosphate GapA and fructose-1,6-bisphosphate aldolase Fba are central enzymes of carbon metabolism (Dannheim et al., 2017b). Finally, the O-acetylserine sulfhydrylase/cysteine synthase (CysK), was among the metabolic proteins the one with the highest sequence coverage. It represents the last enzyme of cysteine biosynthesis (Dubois et al., 2016).
A whole battery of stress response proteins was also found attached to CdPpiB, including the two rubyerythrins Rbr1 and Rbr2 and the desulfoferrodoxin Dsr, which are involved in oxidative stress response (Kawasaki et al., 2009; Riebe et al., 2009). Furthermore, the cold shock protein CspB and the two alkaline shock proteins YloU and YqhY as well as the universal stress protein UspA were among the interacting proteins (Nyström and Neidhardt, 1993; Söderholm et al., 2011; Derman et al., 2015; Tödter et al., 2017). The chaperones GroS and DnaK are part of multiple stress responses (LaRossa and Van Dyk, 1991). Key players of translational quality control, the ribosome-recycling factor RRF, translation initiation factor IF-3 InfC and most interestingly, the pleiotropic transcriptional regulator Lrp were further proteins associated with CdPpiB (Brinkman et al., 2003; Ohashi et al., 2003; Laursen et al., 2005). The screen also yielded flagellin (FliC, 54% coverage) and surface layer protein (SlpA, 19% coverage), two major virulence associated proteins (Tasteyre et al., 2001; Merrigan et al., 2013). Finally, a sugar-family solute binding protein, a putative heavy metal binding protein (UPF0145) and a predicted biotin carrier protein were identified as transporter or transporter associated proteins cross-linked to CdPpiB (Table 3).
CdPpiB Interacts With Lrp, FliC and CysK in a Bacterial Two Hybrid System
In order to verify direct interactions between candidate proteins identified by our interactomic studies, we applied bacterial two hybrid (BACTH) based protein-protein interaction tests. For this, we chose Lrp as it is a global transcriptional regulator, and as there are no reports on functional interactions between bacterial PPIases and transcriptional regulators. Another candidate was the metabolic enzyme with the highest coverage, CysK, that participates in cysteine metabolism which is central to virulence of C. difficile (Dubois et al., 2016). FliC was chosen as it is a well-established virulence factor in many pathogenic bacteria including C. difficile (Tasteyre et al., 2001). For all the tested proteins a clear-cut interaction with CdPpiB was observed, indicated by pink coloring of the resulting clones on MacConkey-agar verifying our findings from the interactomics study (Figure 3).
Figure 3. Confirmation CdPpiB interaction partners with BACTH on MycKonkey-Agar. (A) E. coli BTH101 cells transformed with wt ppiB carrying pKNTB in combination with pUT18_FliC carrying wt fliC or pUT18_Lrp carrying wt lrp were streaked out on MacKonkey-agar and grown for 2 days at 30°C. (B) E. coli BTH101 cells transformed with wt ppiB carrying pKNTB in combination with pUT18_CysK carrying wt cysK were streaked out on MacKonkey-agar and grown for 2 days at 30°C. Purple coloring was indicative of protein interaction due to adenylate cyclase activity. On each plate E. coli BTH101 carrying empty pUT18 plasmid in combination with pKNTB served as negative control, whereas colonies carrying the pUT18-zip and pKT25-zip plasmids delivered by the manufacturer served as positive control. At least three colonies of each transformation were tested. Representative plates of three separate transformations are shown.
An Intact PPIase Domain of CdPpiB Is Required for the Interaction With P87 Substrate of Lrp
Following the confirmation of the interaction via the BACTH screen, we wanted to analyze whether these interactions may be dependent on the PPIase activity of CdPpiB. For this purpose, we chose Lrp for a detailed study as it is with 15 kDa comparably small and has only five proline residues. As PPIases act on the peptidyl-prolyl bonds in proteins, we exchanged all five proline residues in Lrp by alanine generating the single amino acid substitution mutants P32A, P56A, P72A, P87A and P135A. The interactions between CdPpiB and the generated Lrp-variants were quantified in β-galactosidase assays using the outlined BACTH system. Among the five Lrp-variants only variant P87A showed a significantly reduced β-galactosidase activity. It was with 868.8 ± 43.55 MU 30% less than the interaction with wild type Lrp which had an average β-galactosidase activity of 1246 ± 59.06 MU indicating a specific interaction between the two proteins (Figure 4A). In order to prove that the PPIase domain is involved in the overall process, we went on by exchanging highly conserved single amino acid residues within the active site of CdPpiB that are known from homologous cyclophilins of Gram-positive and Gram-negative bacteria (Göthel et al., 1996; Skagia et al., 2017a; Wiemels et al., 2017). Accordingly, we generated the CdPpiB variants R50A, F109A and F110A (s.a. Figure 1). For all three mutants, significant differences were observed regarding the interaction with Lrp as assessed by β-galactosidase activity (Figure 4B).
Figure 4. CdPpiB targets P87 in CdLrp and this interaction depends on conserved amino acids. (A) β-galactosidase activities of E. coli BTH101 clones carrying combinations of wt CdPpiB and single proline exchange mutants of CdLrp. (B) β-galactosidase activities of E. coli BTH101 clones carrying combinations of wt CdLrp and single amino acid exchange mutants of CdPpiB. Clones carrying wt CdPpiB and the empty companion vector pKNT25 served as negative control in both experiments. Shown are mean and SEM of three independent experiments with at least five clones each. Statistical significance was calculated by unpaired t-Test with Welch’s correction (∗p ≤ 0.05, ∗∗p ≤ 0.01, ∗∗∗∗p ≤ 0.0001, n.s., not significant).
Destruction of ppiB Leads to Sensitivity to Envelope Stress and L-Cysteine in C. difficile
As our interactomic data revealed that CdPpiB interacts with several stress response proteins, we performed a literature survey of available transcriptomic data. By this, we found that ppiB was the only PPIase of C. difficile that was significantly upregulated under cell wall stress caused by the beta-lactam amoxicillin (Emerson et al., 2008). Accordingly, we tested the susceptibility of the ΔppiB mutant in a serial dilution assay, and registered that the mutant was significantly more susceptible to amoxicillin concentrations ≥ 32 μM supporting our findings regarding the involvement of CdPpiB in stress tolerance (Figure 5).
Figure 5. CdPpiB-deficiency increases susceptibility toward amoxicillin. Wild type and its isogenic ΔppiB-mutant were grown in the presence of decreasing concentrations of amoxicillin (Amx) and bacterial growth was measured at 600 nm. The graph depicts mean ± SEM of three separate experiments performed in duplicates. Statistical significance was calculated by unpaired t-Test (∗p ≤ 0.05, ∗∗p ≤ 0.01, n.s., not significant).
We further went on characterizing the generated ΔppiB mutant by performing growth assays using the standard rich medium BHIS, which is supplemented with 0.5% (wt/vol) yeast extract and 0.1% (wt/vol) L-cysteine. Here, the ΔppiB mutant displayed a significantly delayed growth in BHIS medium containing L-cysteine. Slower growth in the exponential phase resulting in a later entry into stationary phase (9 h in wt vs. 12 h in ΔppiB) was observed. The final OD after 22 h, were comparable for both strains (Figure 6A). When the assays were repeated with plain BHI or with BHIS lacking L-cysteine, no difference between wild type and the ΔppiB mutant were observed. These tests identified L-cysteine as the cause for the observed phenotype. In order to verify the ppiB mutation as sole cause for the observed differences, we complemented the mutant with an intact ppiB gene in trans and brought ppiB under the control of the tetracycline promoter (Ptet) allowing fine-tuned expression under the control of anhydrotetracycline (aTc). The wild type strain and its isogenic ΔppiB mutant were transformed, and grown in the presence of different concentrations of L-cysteine (2–62 mM) in combination of increasing concentrations of aTc (3,625-400 ng/mL) (Figures 6B,C). When fully induced with 400 ng/mL, the ΔppiB mutant was able to cope with 8 mM L-cysteine, which equals to the concentration in BHIS. Its growth under full induction was significantly better compared to reduced induction with 200 ng/mL or less aTc (Figure 6B). Overexpressing ppiB in the wild type resulted in an even higher tolerance of up to 32 mM L-cysteine and a significantly better growth compared to lower induction levels. In case of 16 mM L-cysteine the wild type overexpressing ppiB reached optical densities comparable to the wild type gown at 8 mM L-cysteine without ppiB overexpression (Figure 6C). These observations corroborated the interactomic data where CysK was the major metabolic protein interacting with CdPpiB.
Figure 6. CdPpiB confers cysteine tolerance to C. difficile. (A) Destruction of ppiB causes growth defect in BHIS as assessed by the change in the OD600 nm over time. (B) Complementation of the ΔppiB mutant through induction with 400 ng/mL anhydrotetracycline allows the bacteria to cope with 8 mM L-cysteine when grown in BHI supplemented with different concentrations of cysteine. (C) Stepwise over-expression of ppiB with increasing amounts of anhydrotetracycline raises the tolerance of the wild type toward 16 and 32 mM L-cysteine. Shown are means ± SEM of three independent experiments performed in duplicate. Statistical significance was calculated by unpaired t-Test (∗p ≤ 0.05, ∗∗∗∗p ≤ 0.0001).
CdPpiB Restores CysK Activity After Heat Inactivation
One important function of PPIases is the restoration of enzyme activity after protein inactivation by unfavorable chemical or physical conditions. In our interactomic study, we identified and verified CysK, the terminal enzyme in L-cysteine biosynthesis, as an interaction partner of CdPpiB. Accordingly, we tested the influence CdPpiB on heat inactivated CysK, the last enzyme of cysteine biosynthesis. Activity of recombinantly produced CysK was measured in a photometric assay at 412 nm, and was shown to be active at 7 μM concentration. Next, CysK was mildly heat denatured at 56°C for 30 min and its enzymatic activity was measured with or without five-fold excess of CdPpiB. As expected, heat denatured CysK showed no activity. But, the addition of 35 μM CdPpiB to the reaction efficiently restored its enzymatic activity (Figure 7). This indicates that CdPpiB functionally interacts with CysK and promotes its structural stability.
Figure 7. CdPpiB restores CysK-activity after heat denaturation. O-acetyl-sulfhydrylase activity of recombinant CysK was measured by the decrease of OD412 at 30°C. Activity of 7 μM recombinant CysK diminished after inactivation at 56°C for 30 min, and could be restored by the addition of recombinant PpiB in 5-times molar excess. Shown are mean and SEM of three independent measurements proteins of two different productions. Statistical significance was calculated by unpaired t-Test with Welch’s correction (∗p ≤ 0.05, n.s., not significant).
Deletion of ppiB Increases Toxin Production and Cytotoxic Activity
Considering that CysK is a central enzyme in L-cysteine metabolism and that this amino acid is also a potent suppressor of toxin production, we tested the influence of CdPpiB on the cellular amounts of TcdA and TcdB under in vitro conditions (Karlsson et al., 2000; Dubois et al., 2016). We quantified both toxins in stationary cultures of wild type and the ppiB mutant using ELISA. Both toxins were found in significantly higher concentrations in the culture supernatants of the ppiB mutant. While the wild type strain revealed in average 309 ng/mL TcdA in its culture supernatant, the ppiB mutant produced 2.8-fold more TcdA (868 ng/mL) (Figure 8A). Similarly, TcdB was 2.3-fold higher concentrated in the growth supernatant of the mutant (89 ng/mL) compared to the wild type (40 ng/mL) (Figure 8B). In accordance with the amounts of toxins found in the culture supernatant, the broth from mutant growth was considerably more active at 0.1% (vol/vol) dilution, causing stronger actin depolymerization and cell rounding of NIH-3T3 cells (Figures 8C–E). Similar results were obtained for the mutant complemented with the ppiB gene in trans, however, without gene induction by aTc (Figure 8G). Wild type levels of toxins and corresponding cytotoxic effect similar to wild type were observed when the complemented mutant was induced with 400 ng/mL aTc (Figure 8H). Cells treated with supernatants from wild type bacteria overexpressing ppiB exhibited no visual abnormalities of the cellular morphology (Figure 8F). Hence, CdPpiB not only interacts with components of the L-cysteine metabolism and supports the growth in the presence of L-cysteine, but it also affects toxin production in C. difficile.
Figure 8. Deletion of ppiB leads to higher toxin titres and cytotoxicity. (A) TcdA and (B) TcdB concentrations were measured by ELISA in 48 h old culture supernatants. Both toxins accumulated to significantly higher titres in culture supernatants in the isogenic ΔppiB mutant. Shown are means ± SD of three independent experiments performed in duplicate. Statistical significance was calculated by unpaired t-Test (∗p ≤ 0.05). (C–H) PpiB influences the cytotoxic activity in the culture supernatant. (C) In untreated control NIH-3T3 cells the actin cytoskeleton is organized in stressfibers as it is typical for epithelial cells. (D) Cell-free supernatants from 24 h old wild type cultures that were diluted to 0.1% (vol/vol) in cell culture medium have visually no detrimental effect on the actin cytoskeleton. (E) Supernatants of the ΔppiB mutant cause disintegration and rounding up of the cells at the same dilution. (F) Overexpression of ppiB by the addition of 400 ng/mL anhydrotetracycline (aTc) in a tetracycline-inducible system has no detrimental effect on the cells in the wild type background. (G) The complemented strain exerts comparable cytotoxic activity as the ΔppiB mutant in the absence of aTc. (H) Overexpression of ppiB by the addition of aTc reduces the cytotoxic effect in the ΔppiB background to wild type levels. Shown are representative views of two separate experiments performed in triplicates. Actin cytoskeleton was stained green with Alexa488-coupled phalloidin, while nuclear DNA was stained blue with DAPI. Scale bars correspond to 10 μm.
The aim of this study was to gain insights into the physiological and virulence related functions of the cyclophilin CdPpiB of the nosocomial pathogen C. difficile. Our in vivo interactomics approach in combination with bacterial two hybrid testing allowed us the identification and verification of a specific set of interaction partners that are major control proteins of cell physiology and virulence. The majority of the interaction partners belonged to the category of energy and central metabolism (Table 3). Here, several functionally related proteins were identified, including LdhA, EtfA1, AcdB, EtfA3, HadA, and EtfB1. These enzymes are part of the same Stickland reaction involving L-leucine catabolism. By this, C. difficile harnesses metabolic energy from the oxidation of one equivalent of L-leucine to isovalerate, CO2 and NH4+. In return, two equivalents of L-leucine are reduced to isocapronate and NH4+ (Stickland, 1934; Britz and Wilkinson, 1982; Kim et al., 2006). This pathway is coupled via NAD+ and ferredoxin to the RNF complex responsible for proton/sodium gradient formation required for ATP generation in C. difficile (Kim et al., 2004; Aboulnaga et al., 2013).
Another group of proteins of putative interaction partners belonged to the primary carbon metabolism. These included pyruvate carboxylase Pyc, pyruvate-ferredoxin oxidoreductase Pfo, aldehyde-alcohol dehydrogenase AdhE2 and pyruvate kinase, which control the flux around the pyruvate knot of the central metabolism. On the other hand, the glyceraldehyde-3-phosphate GapA and fructose-1,6-bisphosphate aldolase represent key control points of glycolysis (Dannheim et al., 2017b). Interestingly, gapA and adhE2 were also up-regulated under cysteine rich conditions (Dubois et al., 2016; Gu et al., 2018). Thus, CdPpiB potentially controls the major fluxes of the central metabolism of C. difficile.
The second largest group of interaction partners belonged to the group of stress response, including proteins protecting against temperature, oxidative and alkaline shock. In line with this, the ΔppiB mutant was more susceptible to amoxicillin, which causes cell wall stress. This finding was also in accordance with a previous microarray study, where ppiB was significantly upregulated in amoxicillin treated C. difficile (Emerson et al., 2008). Currently, we do not know how CdPpiB may influence this phenotype. However, the well-known chaperones DnaK and GroS as well as YloU and YqhY were also upregulated in case of amoxicillin stress (Emerson et al., 2008). In S. aureus DnaK is part of the cell wall stress stimulon, and dnaK mutants have a higher susceptibility toward the beta-lactam oxacillin (Pechous et al., 2004; Singh et al., 2007). Recently, DnaK was identified as a binding partner of EcPpiB in functional complementation and co-precipitation studies in E. coli, where it positively influenced the enzymatic activity and correct localization of DnaK (Skagia et al., 2016). Hence, it is conceivable that, also in C. difficile, CdPpiB is part of the DnaK stress response network. This feature might further be supported by the interactions with YloU and YqhY, which are close homologs of the alkaline shock protein 23 (Asp23) of S. aureus that is also associated with cell envelope stress (Müller et al., 2014; Tödter et al., 2017). EcPpiB is connected to another cell envelope related phenomenon, namely, cell division by directly interacting with FtsZ and probably further client proteins (Skagia et al., 2017b). Our interactomic study did not identify FtsZ as interaction partner in C. difficile. This might indicate that cyclophilins perform species-specific duties despite their high degree of conservation.
Several oxidative stress response proteins of the group of (Rbr1, Rbr2, and Rbr3) and a desulfoferrodoxin (Dsr) were found to bind to CdPpiB. Rubrerythrins are small proteins that have NADH-dependent peroxidase activity and act in concert with rubredoxins and desulfoferrodoxins (Mishra and Imlay, 2012). The homologs of Rbr1 and Dsr are important for conferring O2-tolerance to C. acetbutylicum under the control of the transcriptional regulator PerR, which located between these two genes in C. difficile (Hillmann et al., 2008; Kawasaki et al., 2009; Dannheim et al., 2017a). As these proteins act in close vicinity for efficient electron transfer, it is likely that CdPpiB is involved in complex formation or stability supporting their physiological function. Moreover, oxidative stress genes are upregulated in the presence of L-cysteine in C. difficile (Dubois et al., 2016; Gu et al., 2018).
Two well-established virulence partners, namely the flagellar subunit FliC and the surface layer protein SlpA, were among the CdPpiB interaction partners. In case of FliC this interaction was confirmed by BACTH indicating a functional relationship between these proteins. Two oligomeric states of FliC seem to co-exist: the monomer which has been found in diverse cellular compartments and the polymeric organelle which is employed for motility, attachment and penetration (Campodónico et al., 2010). In fact, the implications of FliC monomers in invasiveness and virulence to the host are not restrained to C. difficile but apply to many other bacteria. For instance, in Pseudomonas aeruginosa or Salmonella typhimurium FliC appears to be the most prominent virulence factor in terms of inflammasome and immune response activation (Duncan and Canna, 2018; Garcia et al., 2018). Interestingly, FliC was shown to interact with DnaK in the periplasm of P. aeruginosa, and in C. difficile a ΔdnaK mutant had four-fold diminished fliC and four-fold increased groS expression (Borrero-de Acuña et al., 2016; Jain et al., 2017). This supports our idea of CdPpiB being a crucial part of the chaperone network of C. difficile.
Next to some transport related proteins, very crucial factors of translational fidelity, IF-3 and RRF, were detected by interactomics. IF-3 is involved in the stabilization of the 30S ribosomal subunit, enabling mRNA binding to the 30S subunit and warranting the accuracy of the first aminoacyl-tRNA binding (Hua and Raleigh, 1998). The RRF, on the other hand, facilitates the dissociation of ribosomes from mRNA after termination of translation. RRF was together with elongation factor P one of the two highly upregulated translation factors in B. subtilis (Ohashi et al., 2003). As a foldase CdPpiB might, indeed, be in close vicinity of ribosomes and assist folding of newly synthesized proteins. However, there are currently no reports supporting this. Also, there is with trigger factor a known ribosome associated protein with a FKBP-domain that was shown to accompany proteins during their maturation following translation (Göthel et al., 1998; Kawagoe et al., 2018). This ubiquitous protein is also present in C. difficile (CDIF630erm_03607/CD630_33060) and is accordingly expected to be the major PPIase that facilitates protein folding. Nevertheless, it could be that CdPpiB does not associate with newly synthesized proteins but is involved in the recycling of translation factors for efficient ribosomal fidelity.
One of the most interesting binding partners was the transcriptional regulator Lrp. Hence, we analyzed the putative interaction between CdPpiB and Lrp in more detail and showed that this interaction depends on the PPIase active site residues. Mutating the highly conserved R50 and F109 weakened, while a F110A substitution significantly strengthened the association between both proteins. In return, out of the five proline residues of CdLrp only mutating P87 reduced the interaction of both proteins indicating a specific PPIase client protein relationship. Lrp-type proteins are global transcriptional regulators that consist of a classical N-terminal helix-turn-helix DNA-binding domain and a C-terminal substrate binding/ activation domain and typically act as specific regulators of amino acid metabolism-related genes (Brinkman et al., 2003). However, in B. subtilis LrpC, a close homolog of CdLrp, was also shown to be important for DNA-repair (López-Torrejón et al., 2006). More interestingly, deleting decR (STM0459), a very close homolog of CdLrp in Salmonella typhimurium, rendered the bacteria more susceptible to L-cysteine (Oguri et al., 2012). Hence, the cysteine-dependent phenotypes of the ΔppiB mutant might be due to the participation of CdPpiB on multiple levels ranging from gene regulation to enzymatic activity. The proline residue at position 87 is in the vicinity of the substrate binding/activation domain. Considering that peptidyl-prolyl-cis/trans-isomerization results in conformational changes in proteins, it is possible that CdPpiB assists CdLrp in changing its conformation upon binding its substrate or that it regulates the conformational equilibrium of CdLrp for a fine-tuned gene expression. Currently, only for eukaryotic PPIases examples of gene regulatory actions in the form of transcription complex stabilization or mRNA maturation are known (Hanes, 2015; Thapar, 2015). To the best of our knowledge, this is the first time a direct interaction between a PPIase and a transcriptional regulator with direct implications on gene regulation has been shown in bacteria.
Interestingly and also in accordance with the interactomic findings, the ΔppiB mutant showed two L-cysteine dependent phenotypes: a higher susceptibility toward L-cysteine and accumulation of more toxin in the supernatant. Typically added to the medium to promote growth, L-cysteine has also some detrimental effects because of the generated H2S during its degradation. As a result, deleting cysteine catabolizing genes results in decreased cysteine tolerance in bacteria (Oguri et al., 2012; Gu et al., 2017). Cysteine is also among several metabolic stimuli that inhibit toxin production in C. difficile on transcriptional level, and it does so via its degradation byproducts pyruvate and H2S (Karlsson et al., 2000; Dubois et al., 2016). That cysteine tolerance and regulation of toxin expression converges via the cysteine metabolism was recently shown in a mutant lacking the cysteine desulfhydrase CdsB. This mutant was impaired in its growth in the presence of 5 mM L-cysteine and its toxin production was unaffected by L-cysteine (Gu et al., 2017). These observations overlap with the ΔppiB mutant in our study, and are further supported by the identification of CysK as an interaction partner of CdPpiB. As a O-acetylserine sulfhydrylase CysK catalyzes the synthesis of L-cysteine from O-acetyl-L-serine and sulfide (Kredich et al., 1979). But it was also speculated that at higher L-cysteine concentrations CysK catalyzes the opposite reaction and contributes to L-cysteine degradation by desulfhydration (Auger et al., 2005; Awano et al., 2005; Dubois et al., 2016). This was further supported by the observation that in C. difficile CysK was together with CysE one of the two transcriptionally most up-regulated genes of the cysteine metabolism in L-cysteine treated C. difficile (Dubois et al., 2016; Gu et al., 2018). In our study, we showed that CdPpiB interacts with CysK and it restores the enzymatic activity of denatured CysK. Considering this and the similarities between the phenotypes, we propose that CdPpiB contributes to L-cysteine tolerance of C. difficile by stabilizing the metabolic network on the enzymatic level. Here again, a cooperative action with other chaperones cannot be excluded as dnaK and groS were also among the highly up-regulated genes in the presence of L-cysteine (Dubois et al., 2016). Furthermore, a direct or indirect influence on the transcription of toxin genes by CdPpiB cannot be excluded and would need further analysis.
Taken together, CdPpiB is a key regulator that acts at the protein folding and modification level and controls central cellular processes. We also suggest that targeting CdPpiB as a therapeutic strategy aiming at virulence reduction may prove intricate, since in the absence of PpiB the bacteria become more toxigenic. Our findings confirm the contribution of the PPIase domain to its regulatory activity. But recent findings also suggest that cyclophilin function in bacteria does not solely rely on PPIase activity (Skagia et al., 2017b; Keogh et al., 2018). Accordingly, whether the PPIase-activity itself or some domains other than the PPIase domain are crucial for a certain interaction or even biological function still need to be evaluated in the future.
CÜ, JBdA, LJ, DJ, and MS conceived the experiments. CÜ, MK, MB, and CP conducted the experiments. JW performed the MS/LC analysis. CÜ, MK, MB, LJ, DJ, and MS analyzed the data. CÜ, DJ, and MS drafted and finalized the manuscript. All authors reviewed and approved the final manuscript.
This work was funded by the Federal State of Lower Saxony, Niedersächsisches Vorab CDiff and CDInfect projects (VWZN2889/3215/3266) as well as by the German Research Foundation and the Open Access Publication Funds of the Technische Universität Braunschweig.
Conflict of Interest Statement
The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.
We would like to thank Drs. Rebekka Biedendieck, Sarah Wienecke, and Tobias Knuuti for pN-STREPXa1622 and their technical support regarding recombinant production in B. megaterium, and B.Sc. Lisa Maria Held for her technical support.
Aboulnaga, E.-H., Pinkenburg, O., Schiffels, J., El-Refai, A., Buckel, W., and Selmer, T. (2013). Effect of an oxygen-tolerant bifurcating butyryl coenzyme A dehydrogenase/electron-transferring flavoprotein complex from Clostridium difficile on butyrate production in Escherichia coli. J. Bacteriol. 195, 3704–3713. doi: 10.1128/JB.00321-13
Alonzo, F., Xayarath, B., Whisstock, J. C., and Freitag, N. E. (2011). Functional analysis of the Listeria monocytogenes secretion chaperone PrsA2 and its multiple contributions to bacterial virulence. Mol. Microbiol. 80, 1530–1548. doi: 10.1111/j.1365-2958.2011.07665.x
Aubry, A., Hussack, G., Chen, W., KuoLee, R., Twine, S. M., Fulton, K. M., et al. (2012). Modulation of toxin production by the flagellar regulon in Clostridium difficile. Infect. Immun. 80, 3521–3532. doi: 10.1128/IAI.00224-12
Awano, N., Wada, M., Mori, H., Nakamori, S., and Takagi, H. (2005). Identification and functional analysis of Escherichia coli cysteine desulfhydrases. Appl. Environ. Microbiol. 71, 4149–4152. doi: 10.1128/AEM.71.7.4149-4152.2005
Barbut, F., Gariazzo, B., Bonné, L., Lalande, V., Burghoffer, B., Luiuz, R., et al. (2007). Clinical features of Clostridium difficile-associated infections and molecular characterization of strains: results of a retrospective study, 2000-2004. Infect. Control Hosp. Epidemiol. 28, 131–139. doi: 10.1086/511794
Bartlett, J. G., Onderdonk, A. B., Cisneros, R. L., and Kasper, D. L. (1977). Clindamycin-associated colitis due to a toxin-producing species of Clostridium in hamsters. J. Infect. Dis. 136, 701–705. doi: 10.1093/infdis/136.5.701
Batah, J., Denève-Larrazet, C., Jolivot, P.-A., Kuehne, S., Collignon, A., Marvaud, J.-C., et al. (2016). Clostridium difficile flagella predominantly activate TLR5-linked NF-κB pathway in epithelial cells. Anaerobe 38, 116–124. doi: 10.1016/j.anaerobe.2016.01.002
Biedendieck, R., Borgmeier, C., Bunk, B., Stammen, S., Scherling, C., Meinhardt, F., et al. (2011). Systems biology of recombinant protein production using Bacillus megaterium. Methods Enzymol. 500, 165–195. doi: 10.1016/B978-0-12-385118-5.00010-4
Biedendieck, R., Yang, Y., Deckwer, W.-D., Malten, M., and Jahn, D. (2007). Plasmid system for the intracellular production and purification of affinity-tagged proteins in Bacillus megaterium. Biotechnol. Bioeng. 96, 525–537. doi: 10.1002/bit.21145
Borrero-de Acuña, J. M., Jänsch, L., Rohde, M., Timmis, K. N., Jahn, D., and Jahn, M. (2015). “Interatomic characterization of protein–protein interactions in membrane-associated mega-complexes,” in Hydrocarbon and Lipid Microbiology Protocols Springer Protocols Handbooks, eds T. McGenity, K. Timmis, and B. Nogales (Berlin: Springer).
Borrero-de Acuña, J. M., Rohde, M., Wissing, J., Jänsch, L., Schobert, M., Molinari, G., et al. (2016). Protein network of the Pseudomonas aeruginosa denitrification apparatus. J. Bacteriol. 198, 1401–1413. doi: 10.1128/JB.00055-16
Britz, M. L., and Wilkinson, R. G. (1982). Leucine dissimilation to isovaleric and isocaproic acids by cell suspensions of amino acid fermenting anaerobes: the Stickland reaction revisited. Can. J. Microbiol. 28, 291–300. doi: 10.1139/m82-043
Campodónico, V. L., Llosa, N. J., Grout, M., Döring, G., Maira-Litrán, T., and Pier, G. B. (2010). Evaluation of flagella and flagellin of Pseudomonas aeruginosa as vaccines. Infect. Immun. 78, 746–755. doi: 10.1128/IAI.00806-09
Dannheim, H., Riedel, T., Neumann-Schaal, M., Bunk, B., Schober, I., Spröer, C., et al. (2017a). Manual curation and reannotation of the genomes of Clostridium difficile 630Δerm and C. difficile 630. J. Med. Microbiol. 66, 286–293. doi: 10.1099/jmm.0.000427
Dannheim, H., Will, S. E., Schomburg, D., and Neumann-Schaal, M. (2017b). Clostridioides difficile 630Δerm in silico and in vivo - quantitative growth and extensive polysaccharide secretion. FEBS Open Bio 7, 602–615. doi: 10.1002/2211-5463.12208
Derman, Y., Söderholm, H., Lindström, M., and Korkeala, H. (2015). Role of csp genes in NaCl, pH, and ethanol stress response and motility in Clostridium botulinum ATCC 3502. Food Microbiol. 46, 463–470. doi: 10.1016/j.fm.2014.09.004
Dingle, T. C., Mulvey, G. L., and Armstrong, G. D. (2011). Mutagenic analysis of the Clostridium difficile flagellar proteins, FliC and FliD, and their contribution to virulence in hamsters. Infect. Immun. 79, 4061–4067. doi: 10.1128/IAI.05305-11
Dubois, T., Dancer-Thibonnier, M., Monot, M., Hamiot, A., Bouillaut, L., Soutourina, O., et al. (2016). Control of Clostridium difficile physiopathology in response to cysteine availability. Infect. Immun. 84, 2389–2405. doi: 10.1128/IAI.00121-16
Emerson, J. E., Stabler, R. A., Wren, B. W., and Fairweather, N. F. (2008). Microarray analysis of the transcriptional responses of Clostridium difficile to environmental and antibiotic stress. J. Med. Microbiol. 57, 757–764. doi: 10.1099/jmm.0.47657-0
Garcia, M., Morello, E., Garnier, J., Barrault, C., Garnier, M., Burucoa, C., et al. (2018). Pseudomonas aeruginosa flagellum is critical for invasion, cutaneous persistence and induction of inflammatory response of skin epidermis. Virulence 9, 1163–1175. doi: 10.1080/21505594.2018.1480830
Geric, B., Rupnik, M., Gerding, D. N., Grabnar, M., and Johnson, S. (2004). Distribution of Clostridium difficile variant toxinotypes and strains with binary toxin genes among clinical isolates in an American hospital. J. Med. Microbiol. 53, 887–894. doi: 10.1099/jmm.0.45610-0
Göthel, S. F., Herrler, M., and Marahiel, M. A. (1996). Peptidyl-prolyl cis-trans isomerase of Bacillus subtilis: identification of residues involved in cyclosporin A affinity and catalytic efficiency. Biochemistry 35, 3636–3640. doi: 10.1021/bi9520803
Göthel, S. F., Scholz, C., Schmid, F. X., and Marahiel, M. A. (1998). Cyclophilin and trigger factor from Bacillus subtilis catalyze in vitro protein folding and are necessary for viability under starvation conditions. Biochemistry 37, 13392–13399. doi: 10.1021/bi981253w
Grant, S. G., Jessee, J., Bloom, F. R., and Hanahan, D. (1990). Differential plasmid rescue from transgenic mouse DNAs into Escherichia coli methylation-restriction mutants. Proc. Natl. Acad. Sci. U.S.A. 87, 4645–4649. doi: 10.1073/pnas.87.12.4645
Gu, H., Shi, K., Liao, Z., Qi, H., Chen, S., Wang, H., et al. (2018). Time-resolved transcriptome analysis of Clostridium difficile R20291 response to cysteine. Microbiol. Res. 215, 114–125. doi: 10.1016/j.micres.2018.07.003
Gu, H., Yang, Y., Wang, M., Chen, S., Wang, H., Li, S., et al. (2017). Novel cysteine desulfidase cdsb involved in releasing cysteine repression of toxin synthesis in Clostridium difficile. Front. Cell Infect. Microbiol. 7:531. doi: 10.3389/fcimb.2017.00531
Heap, J. T., Kuehne, S. A., Ehsaan, M., Cartman, S. T., Cooksley, C. M., Scott, J. C., et al. (2010). The ClosTron: mutagenesis in Clostridium refined and streamlined. J. Microbiol. Methods 80, 49–55. doi: 10.1016/j.mimet.2009.10.018
Heap, J. T., Pennington, O. J., Cartman, S. T., Carter, G. P., and Minton, N. P. (2007). The ClosTron: a universal gene knock-out system for the genus Clostridium. J. Microbiol. Methods 70, 452–464. doi: 10.1016/j.mimet.2007.05.021
Heikkinen, O., Seppala, R., Tossavainen, H., Heikkinen, S., Koskela, H., Permi, P., et al. (2009). Solution structure of the parvulin-type PPIase domain of Staphylococcus aureus PrsA–implications for the catalytic mechanism of parvulins. BMC Struct. Biol. 9:17. doi: 10.1186/1472-6807-9-17
Hillmann, F., Fischer, R.-J., Saint-Prix, F., Girbal, L., and Bahl, H. (2008). PerR acts as a switch for oxygen tolerance in the strict anaerobe Clostridium acetobutylicum. Mol. Microbiol. 68, 848–860. doi: 10.1111/j.1365-2958.2008.06192.x
Hua, Y., and Raleigh, D. P. (1998). On the global architecture of initiation factor IF3: a comparative study of the linker regions from the Escherichia coli protein and the Bacillus stearothermophilus protein. J. Mol. Biol. 278, 871–878. doi: 10.1006/jmbi.1998.1736
Hussain, H. A., Roberts, A. P., and Mullany, P. (2005). Generation of an erythromycin-sensitive derivative of Clostridium difficile strain 630 (630Deltaerm) and demonstration that the conjugative transposon Tn916DeltaE enters the genome of this strain at multiple sites. J. Med. Microbiol. 54, 137–141. doi: 10.1099/jmm.0.45790-0
Ikolo, F., Zhang, M., Harrington, D. J., Robinson, C., Waller, A. S., Sutcliffe, I. C., et al. (2015). Characterisation of SEQ0694 (PrsA/PrtM) of Streptococcus equi as a functional peptidyl-prolyl isomerase affecting multiple secreted protein substrates. Mol. Biosyst. 11, 3279–3286. doi: 10.1039/c5mb00543d
Jacobs, M., Andersen, J. B., Kontinen, V., and Sarvas, M. (1993). Bacillus subtilis PrsA is required in vivo as an extracytoplasmic chaperone for secretion of active enzymes synthesized either with or without pro-sequences. Mol. Microbiol. 8, 957–966. doi: 10.1111/j.1365-2958.1993.tb01640.x
Jain, S., Smyth, D., O’Hagan, B. M. G., Heap, J. T., McMullan, G., Minton, N. P., et al. (2017). Inactivation of the dnaK gene in Clostridium difficile 630 Δerm yields a temperature-sensitive phenotype and increases biofilm-forming ability. Sci. Rep. 7:17522. doi: 10.1038/s41598-017-175830-9
Karlsson, S., Lindberg, A., Norin, E., Burman, L. G., and Akerlund, T. (2000). Toxins, butyric acid, and other short-chain fatty acids are coordinately expressed and down-regulated by cysteine in Clostridium difficile. Infect. Immun. 68, 5881–5888. doi: 10.1128/IAI.68.10.5881-5888.2000
Kawagoe, S., Nakagawa, H., Kumeta, H., Ishimori, K., and Saio, T. (2018). Structural insight into proline cis/trans isomerization of unfolded proteins catalyzed by the trigger factor chaperone. J. Biol. Chem. 293, 15095–15106. doi: 10.1074/jbc.RA118.003579
Kawasaki, S., Sakai, Y., Takahashi, T., Suzuki, I., and Niimura, Y. (2009). O2 and reactive oxygen species detoxification complex, composed of O2-responsive NADH:rubredoxin oxidoreductase-flavoprotein A2-desulfoferrodoxin operon enzymes, rubperoxin, and rubredoxin, in Clostridium acetobutylicum. Appl. Environ. Microbiol. 75, 1021–1029. doi: 10.1128/AEM.01425-08
Keogh, R. A., Zapf, R. L., Wiemels, R. E., Wittekind, M. A., and Carroll, R. K. (2018). The intracellular cyclophilin Ppib contributes to the virulence of Staphylococcus aureus independently of its peptidyl-prolyl cis/trans isomerase activity. Infect Immun. 86:e00379-18. doi: 10.1128/IAI.00379-18
Kim, J., Darley, D., Selmer, T., and Buckel, W. (2006). Characterization of (R)-2-hydroxyisocaproate dehydrogenase and a family III coenzyme A transferase involved in reduction of L-leucine to isocaproate by Clostridium difficile. Appl. Environ. Microbiol. 72, 6062–6069. doi: 10.1128/AEM.00772-06
Kim, J., Hetzel, M., Boiangiu, C. D., and Buckel, W. (2004). Dehydration of (R)-2-hydroxyacyl-CoA to enoyl-CoA in the fermentation of alpha-amino acids by anaerobic bacteria. FEMS Microbiol. Rev. 28, 455–468. doi: 10.1016/j.femsre.2004.03.001
LaRossa, R. A., and Van Dyk, T. K. (1991). Physiological roles of the DnaK and GroE stress proteins: catalysts of protein folding or macromolecular sponges? Mol. Microbiol. 5, 529–534. doi: 10.1111/j.1365-2958.1991.tb00724.x
Laursen, B. S., Sørensen, H. P., Mortensen, K. K., and Sperling-Petersen, H. U. (2005). Initiation of protein synthesis in bacteria. Microbiol. Mol. Biol. Rev. MMBR 69, 101–123. doi: 10.1128/MMBR.69.1.101-123.2005
López-Torrejón, G., Martínez-Jiménez, M. I., and Ayora, S. (2006). Role of LrpC from Bacillus subtilis in DNA transactions during DNA repair and recombination. Nucleic Acids Res. 34, 120–129. doi: 10.1093/nar/gkj418
Merrigan, M. M., Venugopal, A., Roxas, J. L., Anwar, F., Mallozzi, M. J., Roxas, B. A. P., et al. (2013). Surface-layer protein A (SlpA) is a major contributor to host-cell adherence of Clostridium difficile. PLoS One 8:e78404. doi: 10.1371/journal.pone.0078404
Müller, M., Reiß, S., Schlüter, R., Mäder, U., Beyer, A., Reiß, W., et al. (2014). Deletion of membrane-associated Asp23 leads to upregulation of cell wall stress genes in Staphylococcus aureus. Mol. Microbiol. 93, 1259–1268. doi: 10.1111/mmi.12733
Nyström, T., and Neidhardt, F. C. (1993). Isolation and properties of a mutant of Escherichia coli with an insertional inactivation of the uspA gene, which encodes a universal stress protein. J. Bacteriol. 175, 3949–3956. doi: 10.1128/jb.175.13.3949-3956.1993
Obi, I. R., Nordfelth, R., and Francis, M. S. (2011). Varying dependency of periplasmic peptidylprolyl cis-trans isomerases in promoting Yersinia pseudotuberculosis stress tolerance and pathogenicity. Biochem. J. 439, 321–332. doi: 10.1042/BJ20110767
Oguri, T., Schneider, B., and Reitzer, L. (2012). Cysteine catabolism and cysteine desulfhydrase (CdsH/STM0458) in Salmonella enterica serovar typhimurium. J. Bacteriol. 194, 4366–4376. doi: 10.1128/JB.00729-12
Ohashi, Y., Inaoka, T., Kasai, K., Ito, Y., Okamoto, S., Satsu, H., et al. (2003). Expression profiling of translation-associated genes in sporulating Bacillus subtilis and consequence of sporulation by gene inactivation. Biosci. Biotechnol. Biochem. 67, 2245–2253. doi: 10.1271/bbb.67.2245
Péchiné, S., Bruxelle, J. F., Janoir, C., and Collignon, A. (2018). Targeting Clostridium difficile surface components to develop immunotherapeutic strategies against Clostridium difficile infection. Front. Microbiol. 9:1009. doi: 10.3389/fmicb.2018.01009
Pechous, R., Ledala, N., Wilkinson, B. J., and Jayaswal, R. K. (2004). Regulation of the expression of cell wall stress stimulon member gene msrA1 in methicillin-susceptible or -resistant Staphylococcus aureus. Antimicrob. Agents Chemother. 48, 3057–3063. doi: 10.1128/AAC.48.8.3057-3063.2004
Perutka, J., Wang, W., Goerlitz, D., and Lambowitz, A. M. (2004). Use of computer-designed group II introns to disrupt Escherichia coli DExH/D-box protein and DNA helicase genes. J. Mol. Biol. 336, 421–439. doi: 10.1016/j.jmb.2003.12.009
Purdy, D., O’Keeffe, T. A. T., Elmore, M., Herbert, M., McLeod, A., Bokori-Brown, M., et al. (2002). Conjugative transfer of clostridial shuttle vectors from Escherichia coli to Clostridium difficile through circumvention of the restriction barrier. Mol. Microbiol. 46, 439–452. doi: 10.1046/j.1365-2958.2002.03134.x
Ransom, E. M., Ellermeier, C. D., and Weiss, D. S. (2015). Use of mCherry Red fluorescent protein for studies of protein localization and gene expression in Clostridium difficile. Appl. Environ. Microbiol. 81, 1652–1660. doi: 10.1128/AEM.03446-14
Rasch, J., Ünal, C. M., Klages, A., Karsli, Ü, Heinsohn, N., Brouwer, R. M. H. J., et al. (2018). PPIases Mip and PpiB of Legionella pneumophila contribute to surface translocation, growth at suboptimal temperature and infection. Infect. Immun 87: e00939-17. doi: 10.1128/IAI.00939-17
Riebe, O., Fischer, R.-J., Wampler, D. A., Kurtz, D. M., and Bahl, H. (2009). Pathway for H2O2 and O2 detoxification in Clostridium acetobutylicum. Microbiol. Read. Engl. 155, 16–24. doi: 10.1099/mic.0.022756-0
Saavedra, C. P., Encinas, M. V., Araya, M. A., Pérez, J. M., Tantaleán, J. C., Fuentes, D. E., et al. (2004). Biochemical characterization of a thermostable cysteine synthase from Geobacillus stearothermophilus V. Biochimie 86, 481–485. doi: 10.1016/j.biochi.2004.06.003
Singh, V. K., Utaida, S., Jackson, L. S., Jayaswal, R. K., Wilkinson, B. J., and Chamberlain, N. R. (2007). Role for dnaK locus in tolerance of multiple stresses in Staphylococcus aureus. Microbiol. Read. Engl. 153, 3162–3173. doi: 10.1099/mic.0.2007/009506-0
Skagia, A., Vezyri, E., Sigala, M., Kokkinou, A., Karpusas, M., Venieraki, A., et al. (2017a). Structural and functional analysis of cyclophilin PpiB mutants supports an in vivo function not limited to prolyl isomerization activity. Genes Cells 22, 32–44. doi: 10.1111/gtc.12452
Skagia, A., Zografou, C., Venieraki, A., Fasseas, C., Katinakis, P., and Dimou, M. (2017b). Functional analysis of the cyclophilin PpiB role in bacterial cell division. Genes Cells 22, 810–824. doi: 10.1111/gtc.12514
Skagia, A., Zografou, C., Vezyri, E., Venieraki, A., Katinakis, P., and Dimou, M. (2016). Cyclophilin PpiB is involved in motility and biofilm formation via its functional association with certain proteins. Genes Cells Devoted Mol. Cell. Mech. 21, 833–851. doi: 10.1111/gtc.12383
Söderholm, H., Lindström, M., Somervuo, P., Heap, J., Minton, N., Lindén, J., et al. (2011). cspB encodes a major cold shock protein in Clostridium botulinum ATCC 3502. Int. J. Food Microbiol. 146, 23–30. doi: 10.1016/j.ijfoodmicro.2011.01.033
Stickland, L. H. (1934). Studies in the metabolism of the strict anaerobes (genus Clostridium): the chemical reactions by which Cl. sporogenes obtains its energy. Biochem. J. 28, 1746–1759. doi: 10.1042/bj0281746
Tai, C. H., Nalabolu, S. R., Jacobson, T. M., Minter, D. E., and Cook, P. F. (1993). Kinetic mechanisms of the A and B isozymes of O-acetylserine sulfhydrylase from Salmonella typhimurium LT-2 using the natural and alternative reactants. Biochemistry 32, 6433–6442. doi: 10.1021/bi00076a017
Tasteyre, A., Barc, M. C., Collignon, A., Boureau, H., and Karjalainen, T. (2001). Role of FliC and FliD flagellar proteins of Clostridium difficile in adherence and gut colonization. Infect. Immun. 69, 7937–7940. doi: 10.1128/IAI.69.12.7937-7940.2001
Theriot, C. M., Bowman, A. A., and Young, V. B. (2016). Antibiotic-induced alterations of the gut microbiota alter secondary bile acid production and allow for Clostridium difficile spore germination and outgrowth in the large intestine. mSphere 1:e00045-15. doi: 10.1128/mSphere.00045-15
Tödter, D., Gunka, K., and Stülke, J. (2017). The highly conserved Asp23 family protein YqhY plays a role in lipid biosynthesis in Bacillus subtilis. Front. Microbiol. 8:883. doi: 10.3389/fmicb.2017.00883
Ünal, C. M., and Steinert, M. (2014). Microbial peptidyl-prolyl cis/trans isomerases (PPIases): virulence factors and potential alternative drug targets. Microbiol. Mol. Biol. Rev. MMBR 78, 544–571. doi: 10.1128/MMBR.00015-14
Waterhouse, A. M., Procter, J. B., Martin, D. M. A., Clamp, M., and Barton, G. J. (2009). Jalview Version 2–a multiple sequence alignment editor and analysis workbench. Bioinformatics 25, 1189–1191. doi: 10.1093/bioinformatics/btp033
Wiemels, R. E., Cech, S. M., Meyer, N. M., Burke, C. A., Weiss, A., Parks, A. R., et al. (2017). An intracellular peptidyl-prolyl cis/trans isomerase is required for folding and activity of the Staphylococcus aureus secreted virulence factor nuclease. J. Bacteriol. 199:e00453-16. doi: 10.1128/JB.00453-16
Keywords: Clostridium difficile, peptidyl-prolyl-cis/trans-isomerase (PPIase), cytotoxicity, interactomics, transcription
Citation: Ünal CM, Karagöz MS, Berges M, Priebe C, Borrero de Acuña JM, Wissing J, Jänsch L, Jahn D and Steinert M (2019) Pleiotropic Clostridioides difficile Cyclophilin PpiB Controls Cysteine- Tolerance, Toxin Production, the Central Metabolism and Multiple Stress Responses. Front. Pharmacol. 10:340. doi: 10.3389/fphar.2019.00340
Received: 04 December 2018; Accepted: 19 March 2019;
Published: 05 April 2019.
Edited by:Flavio Rizzolio, Università Ca’ Foscari, Italy
Reviewed by:Laura J. Blair, University of South Florida, United States
Floriana Campanile, Università degli Studi di Catania, Italy
Copyright © 2019 Ünal, Karagöz, Berges, Priebe, Borrero de Acuña, Wissing, Jänsch, Jahn and Steinert. This is an open-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms.