Identification and characterization of smallest pore-forming protein in the cell wall of pathogenic Corynebacterium urealyticum DSM 7109

Background Corynebacterium urealyticum, a pathogenic, multidrug resistant member of the mycolata, is known as causative agent of urinary tract infections although it is a bacterium of the skin flora. This pathogenic bacterium shares with the mycolata the property of having an unusual cell envelope composition and architecture, typical for the genus Corynebacterium. The cell wall of members of the mycolata contains channel-forming proteins for the uptake of solutes. Results In this study, we provide novel information on the identification and characterization of a pore-forming protein in the cell wall of C. urealyticum DSM 7109. Detergent extracts of whole C. urealyticum cultures formed in lipid bilayer membranes slightly cation-selective pores with a single-channel conductance of 1.75 nS in 1 M KCl. Experiments with different salts and non-electrolytes suggested that the cell wall pore of C. urealyticum is wide and water-filled and has a diameter of about 1.8 nm. Molecular modelling and dynamics has been performed to obtain a model of the pore. For the search of the gene coding for the cell wall pore of C. urealyticum we looked in the known genome of C. urealyticum for a similar chromosomal localization of the porin gene to known porH and porA genes of other Corynebacterium strains. Three genes are located between the genes coding for GroEL2 and polyphosphate kinase (PKK2). Two of the genes (cur_1714 and cur_1715) were expressed in different constructs in C. glutamicum ΔporAΔporH and in porin-deficient BL21 DE3 Omp8 E. coli strains. The results suggested that the gene cur_1714 codes alone for the cell wall channel. The cell wall porin of C. urealyticum termed PorACur was purified to homogeneity using different biochemical methods and had an apparent molecular mass of about 4 kDa on tricine-containing sodium dodecyl sulfate polyacrylamide gel electrophoresis (SDS-PAGE). Conclusions Biophysical characterization of the purified protein (PorACur) suggested indeed that cur_1714 is the gene coding for the pore-forming protein in C. urealyticum because the protein formed in lipid bilayer experiments the same pores as the detergent extract of whole cells. The study is the first report of a cell wall channel in the pathogenic C. urealyticum.


Background
The outer membrane of mycolata represents a permeability barrier similar to that of Gram-negative bacteria and needs specialized pore-forming proteins, called porins, to facilitate the passage of water-soluble solutes. Selective transport of molecules through the cell wall is crucial to the life-or-death of bacterial cells [1,2]. They continuously need to exchange small molecules, nutrients and proteins with the exterior environment, while they need to keep toxic substances out [3][4][5]. Currently almost 100 species of the genus Corynebacterium are known, which are divided in three different groups: human pathogens, animal pathogens, and non-pathogens [6,7]. Many species within the group of mycolic acid containing bacteria are important either because of their medical or biotechnological relevance. Non-pathogenic corynebacteria are used for the production of amino acids such as L-glutamate and L-lysine at industrial scale [8].
In general, C. urealyticum is highly resistant to βlactams and aminoglycosides, and occasionally susceptible to fluoroquinolones, macrolides, ketolides, rifampicin, and tetracyclines [16,18,20,21]. C. urealyticum has also been detected as a rare pathogen in the urinary tract of small animals, such as cats and dogs [22,23]. The bacterium possesses a strong urease activity and this activity leads to the formation of struvite (ammonium magnesium phosphate) stones by increasing the pH and ammonium precipitation [17,24]. The efficiency of its treatment is often affected by multiple resistance of C. urealyticum to a broad range of antibiotics [18,25].
As mentioned above communication between a bacterium and its environment is essential for the survival of bacterial cells [3,5,26]. Many of these processes involve channels in the cell wall. C. urealyticum was also thought to have cell wall channels for the transport of hydrophilic solutes across the cell wall. In the closely related Corynebacterium species C. glutamicum and C. efficiens, porA and porH genes encode for PorA and PorH proteins, which assemble for large, water-filled cell wall pores [27][28][29][30]. We could also previously demonstrate that a channel-forming protein, named PorACj, was identified in the known genome of C. jeikeium by the similar chromosomal localization of its gene to the known porH and porA genes of other Corynebacterium strains [31]. However, in contrast to certain Corynebacterium species, where two polypeptides PorA and PorH are needed to form a functional cell wall pore, the pore in the cell wall of C. jekeium is formed by a single polypeptide PorACj [31].
Protein homology search allows the study of the evolutionary relationship between proteins, since homologous proteins share likely the same function. Analysis of sequence similarity of related proteins can be useful for functional annotation of proteins. For the search for porins in C. urealyticum that are homologous to the known porins from Corynebacterium species we used a similar approach here by using the NCBI BLAST-translation tool search [32,33].
This alignment allowed an interesting comparison of the localization of genes coding for PorA and PorH within the genomes. Two genes coding for porins are present in the genomes of all these Corynebacterium species with the exception of C. jeikeium, where the genome contains only one gene. In all other cases, the genes are located in tandem between the genes coding for the chaperone GroEL2 and the polyphosphate kinase PKK2 [30]. So far, we demonstrated that PorACj is the smallest polypeptide forming well defined and stable channels [31]. The complete genome sequence of C. urealyticum DSM 7109 is known [13]. The search for the gene coding for the cell wall channel in case of C. urealyticum DSM 7109 provided a very interesting result because there exist three open reading frames (ORFs) located between the genes coding for GroEL2 (cur_1716) and PKK2 (cur_1712). It is not possible to clearly identify which of the interjacent genes (cur_1713, cur_1714 or cur_1715) could code for channel-forming proteins. The additional question was if one of the three ORFs was coding for a cell wall channel alone or together with one (or two) of the other two genes. These questions were also addressed in this work. The results suggested that the gene cur_1714 within the known genome sequence of C. urealyticum codes for the cell wall channel. The study is the first report of the cell wall channel in the pathogenic C. urealyticum and could help in better understanding of passage of hydrophilic solutes across the cell wall and developing new drugs against this opportunistic pathogen causing urinary tract infections.

Bacterial strains and growth conditions
The C. urealyticum DSM 7109 strain obtained from Deutsche Sammlung von Mikroorganismen und Zellkulturen was grown in 1000 ml baffled Erlenmeyer flasks containing 250 ml of brain-heart infusion (BHI) medium (Becton) and 250 ml Erlenmeyer flasks containing 25 ml BYT medium [34]. The Corynebacterium strains C. glutamicum ΔporHΔporA [30] and C. glutamicum ΔporHΔ-porA expressing cur_1714 were grown in BHI broth and were stirred on a rotary shaker at 200 rpm and 30°C. Escherichia coli NEB5α (New England Biolabs) and E. coli BL21 DH5α (ThermoFisher Scientific) were used for cloning and grown under standard conditions in Luria broth (LB) at 37°C, shaken at 250 rpm. Agar plates and liquid media were supplemented with 25 μg/ml chloramphenicol, respectively, if required. Porin deficient BL21 DE3 Omp8 E. coli strains [30] were used for GST tagging and expressing of cur_1714 and cur_1715, and were grown at 37°C under standard conditions in Luria broth (LB) while shaking at 250 rpm. Agar plates and liquid media were supplemented with kanamycin (25 μg/ ml) or ampicillin (100 μg/ml) respectively, if necessary.

Cloning of cur_1714 and cur_1715
Construction of a C-terminal His 8 -tag The gene cur_1714 and its putative ribosome binding site was amplified by PCR from genomic DNA of C. urealyticum DSM 7109. The expression vector pXHis, a derivative of pXMJ19 was used for expression [35]. The vector is a shuttle vector which can express proteins in both Grampositive and Gram-negative bacteria and confers chloramphenicol resistance to the transformed cells. The gene cur_ 1714 was amplified by PCR with 1× Fermentas Taq buffer, 0.2 mM dNTPs, 2 mM MgCl 2 , 1 U Taq polymerase (Fermentas) and 0.2 pmol primers; Fwd_1714_XbaI and RPcur 1714-KpnI pXHis ( Table 1). The PCR product was cleaned up from other components by using Qiagen PCR clean up kit, was ligated into the pGEM-T easy vector and transformed in E. coli BL21 DH5α competent cells. After confirming the plasmid construction by sequencing with primers M13 reverse and T7 Promotor (Table 1) (GATC Biotech, Cologne, Germany), the cur_1714 gene and its flanking regions were double-digested from the pGEM-T easy vector using XbaI and KpnI (Fermentas). The gene cur_1714 was cloned and ligated into the expression shuttle vector pXMJ19 containing an octahistidine-tag at the C terminal end for efficient purification of the protein. This tag is cleaved later with Factor Xa protease at its recognition site (I-E-G-R).
Fwd_1714_XbaI and RP pX cur_1714 KpnI pairs of primers were used for cloning the cur_1714 gene in pXMJ19 without His tag under the same conditions. Transformation into C. glutamicum ΔporAΔporH was performed by electroporation using a slightly modified electro-transformation method [36]. Heterologous expression of the protein was induced by addition of 1 mM IPTG (isopropyl-beta-D-thiogalactopyranoside) to a liquid culture in the mid-exponential growth phase. Besides the C-terminal octa-histidine tag we introduced also an N-terminal GST-tag on Cur_1714 and Cur_1715. For the expression of C. urealyticum porins a pGEX-2 T expression vector (Amersham Biosciences, GE Healthcare, and UK) was used. The vector contains glutathione S-transferase (GST) as a fusion tag and a specific cleavage site (L-V-P-R) for thrombin at the N-terminus for efficient cleavage of the tag from the protein. Due to the N-terminal fusion, the first methionine was removed.

Colony PCR and selection of transformed colonies
Colony PCR of C. glutamicum ΔporAΔporH with pXMJ19 cur_1714 and pXHis cur_1714 was performed using slightly modified standard conditions with the specific primers for the insert to check colonies containing the required gene before sequencing. For this issue, single cell colonies were selected and transferred into a 1. 5 ml Eppendorf tube containing 20 μl of distilled water. Then, the culture was given seven times 10 s heat shock with 5 s interval between every heat shock. 1 μl of the sample was used as a template for the colony PCR reaction and the correct transformants were checked by sequencing (GATC Biotech, Cologne, Germany).

Expression of recombinant his tag proteins
Heterologous expression of the protein was induced by addition of 1 mM IPTG (isopropyl-beta-D- Table 1 Oligonucleotides used in this study for cloning of the genes cur_1714 and cur_1715. The sequences of the primers were derived from the prospective genes cur_1714 and cur_1715 and their flanking regions taken from the genome of C. urealyticum DSM 7109 [13] thiogalactopyranoside) to a liquid culture at the midexponential growth phase.

Isolation of recombinant cell wall proteins
Cell wall-associated proteins were isolated by methods described previously in detail [37,38]. A liquid culture of grown cells was centrifuged (4000 rpm, 15 min, 4°C in an Eppendorf centrifuge 5810R rotor A-4-81). The cell pellet was washed twice with 10% culture volume (10 mM Tris-HCl, pH 8) before cell wall proteins were extracted either by shaking the cells in detergent solution or in a 1:2 (v/v) mixture of the organic solvents chloroform and methanol. For both extraction methods one part cells (0.3 g wet weight bacterial cell pellet) was resuspended in five parts detergent solution (1.5 ml 1% LDAO (lauryldimethylamine-oxide), 10 mM Tris-HCl, pH 8) or organic solvent (1.5 ml chloroform/methanol). After 3 h agitation at room temperature, cells were sedimented in a table top centrifuge (Microcentrifuge 5415R) (10 min, 4°C, 10,000 rpm) and the pellet was discarded. The detergent supernatant was immediately applied to IMAC purification. The chloroformmethanol mixture had first to be precipitated with 9 times the volume of ice-cold diethyl ether (16 h, − 20°C) before the obtained pellet was either resolved in detergent solution (1% LDAO, 10 mM Tris, pH 8) or in loading buffer for gel electrophoresis [37].

Expression and isolation of recombinant GST-tagged cur_1714 and cur_1715
The plasmids including desired genes were sequenced and transformed into the porin deficient BL21 DE3 Omp8 E. coli. Cells were grown at 37°C in LB medium and induced by 1 mM IPTG. The culture was incubated in media overnight at 28°C after induction., cells were harvested by centrifugation at 6000×g for 20 min at 4°C and were resupended in PBS (phosphate buffered saline) (pH 7.4). Subsequently the cells were lysed using a highpressure-homogenizer (2 × 1500 bar). Unbroken cells were removed by centrifugation, the supernatant was used for purification.

GST-PorACur fusion protein purification
Purification of GST-tag proteins were performed using glutathione Sepharose 4B medium, (following batch protocol GST Gene Fusion System Handbook, GE Healthcare). Glutathione is a tripeptide (Glu-Cys-Gly) and the specific substrate for glutathione S-transferase (GST). When reduced glutathione (GSH) is immobilized through its sulfhydryl group it can be used to capture GST-tagged proteins via the enzyme-substrate binding reaction. After 5 times washing with buffer A (0.5% Genapol, 100 mM NaCl, 50 mM Tris-HCl, 1 mM CaCl 2 , 2. 5 mM DTT, pH 7.4) to remove non-bound sample components, the purified GST-fusion protein was eluted by addition of buffer 1 supplemented with 10 mM glutathione, pH 8. The protein sample was concentrated using Amicon ultra 3 kDa to a concentration of GST-fusion proteins.
Protease Xa cleavage of C-terminal His 8 of C. urealyticum proteins Subsequent to IMAC purification the sample contained high imidazol concentrations which strongly inhibit protease Xa (Qiagen) activity. Removal of imidazol was performed by dialyzing the sample over night against cleavage buffer (20 mM Tris, 50 mM NaCl, 1 mM CaCl 2 , 0.4% LDAO, pH 6.5) using a cellulose membrane with a MWCO of 2 kDa (Spectra/Por 6, Carl Roth, Germany). For cleavage of the poly-histidine-tag four units protease Xa (Qiagen) were added to the sample (23°C, overnight). The enzyme was removed with the factor Xa removal Kit according Qiagen instructions while the cleaved protein was separated from uncleaved PorACur-His 8 by a second passage through a Ni-NTA filter.

Thrombin cleavage
In order to remove the GST-tag from purified protein it was incubated for 10 h at room temperature with 100 μl of thrombin agarose resin and 100 μl of 10 X cleavage buffer (500 mM Tris-HCl, pH 8.0, 100 mM CaCl 2 ). The pure recombinant protein was purified from GST fusion protein by using preparative SDS-PAGE.

Protein electrophoresis and immunoblotting of the Cterminal his-tag
Analytical and preparative SDS-PAGE was performed according to Laemmli [39] subsequent to a denaturation step (5 min, 95°C). Because of the low resolution of this gel system for the detection of low molecular mass proteins, tricine containing gels containing 15% or 16.5% polyacrylamide were used as described by Schägger and von Jagow [40]. After electrophoresis, gels were either stained with Coomassie Brilliant Blue G-250, or by silver stain [41] or electroblotted [42]. In the latter case, proteins were transferred to a 0.1 μm nitrocellulose membrane (Protran, BA79, Whatman). The blotting was performed for 4-5 min in a wet tank blot system (Biorad) with Towbin buffer (25 mM Tris, 192 mM glycine, 20% methanol) at 350 mA current. Unspecific binding sites on the membrane were blocked with 5% skimmed milk in TBS-T buffer (20 mM Tris, 0.01 M NaCl, 0.1% Tween, pH 7.5) before probing with the first 1: 5000 diluted monoclonal Anti-his antibody (Amersham Biosciences, UK). Subsequent to multiple TBS-T washing steps the second peroxidaseconjugated Anti-Mouse antibodies (DAKO, Denmark) were added at the same dilution. According to the manufacturer's instructions the use of the ECL Western blotting detection system (GE Healthcare, UK) resulted in light emission recorded on autoradiography films (Hyperfilm™MP, GE Healthcare, UK). Dot blot immune detection was carried out identically without prior SDS-PAGE. Exposure times varied between 10 s to 5 min as required by the sample.

Protein electrophoresis and immunoblotting of the Nterminal GST-tag
Proteins were transferred to a 0.45 μm membrane. Subsequently, the membrane was shaken for 1 h in a solution of anti GST antibody (GE Healthcare, UK) at a dilution of 1: 3, 000 and then washed three times with Tris-buffer saline, each time 10 min. Anti-Goat IgG bound to alkaline phosphatase from Sigma at a dilution of 1:30, 000 was used as secondary antibody to detect the bound antibodies. The membrane was washed three times with alkaline phosphatase buffer pH 9.5 (0.1 M Tris, pH 9.5 and 0.1 M NaCl) afterward. Alkaline phosphatase activity was detected by colorimetric reaction. To detect the protein bands recognized by specific antibodies, the membranes were incubated with BCIP (5-bromo-4-chloro-3-indolylphosphate)/NBT (nitro blue tetrazolium) (Sigma) as substrate until the bands appeared.

Black lipid bilayer membranes
The device used for the black lipid bilayer experiments has been described in detail [43]. The basic part is a Teflon cell with two aqueous compartments connected by a circular hole with a small area. The lipid bilayers were formed by painting onto the hole a 1% (w/v) solution of diphytanoyl phosphatidylcholine (Diph-PC)/n-decane (Avanti Polar Lipids, Alabaster, USA). The aqueous salt solutions were used unbuffered and the temperature during all experiments was kept at 20°C. Two Ag/AgCl electrodes with salt bridges were inserted in the compartments that were connected in series with a batterydriven voltage source and a highly sensitive current amplifier. Its output signal was monitored with a digital oscilloscope and recorded with a strip chart recorder.
The zero-current membrane potential measurements were performed as it has been described earlier [44] by establishing a 5-fold salt gradient across membranes containing about 100 cell wall channels. The potentials were measured using a high impedance electrometer (Keithley 617), and analyzed using the Goldman-Hodgkin-Katz equation [44].

Molecular dynamics simulations
Initial models of hexamer and octamer channels were constructed using the CCBuilder tool [45] without prior structural information. Both channels were simulated in a POPE bilayer containing a 1 M KCl solution (hexamer system about 66.000 atoms, octamer system about 90. 000 atoms) using GROMACS 5.1.2 [46] and the CHARMM36 force field [47,48]. Long-range electrostatic potentials beyond the cut-off of 12 Å were computed using the particle mesh Ewald algorithm [49]. Lennard-Jones potentials were computed using the 10-12 Å force-switch scheme. The systems were equilibrated for 16 ns following the protocol described in the CHARMM-GUI Membrane Builder [50]. Unbiased production runs of 500 ns each were performed for the two systems in the pressure ensemble. The Nosé-Hoover thermostat [51] was used to maintain a constant temperature of 303.15 K by coupling the protein, lipid and solvent separately to a temperature bath using a coupling constant of 1 ps. Moreover, the pressure was maintained constant at 1 bar by using the Parrinello-Rahman barostat [52] with semi-isotropic scaling with a coupling constant of 5 ps. Covalent bonds containing hydrogen atoms were constrained using the LINCS algorithm [53] to be able to use an integration time step of 2 fs. Furthermore, two simulations of the octamer system were performed in the presence of an applied voltage of 100 mV to determine the ion conductance at 1 M KCl. The membrane potential (V = E/L z ) was modelled by applying a constant electric field (E) perpendicular to the membrane [54,55].

Results
Cell wall proteins from C. urealyticum effect the conductance of lipid bilayer membranes Cells of a C. urealyticum culture grown overnight were extracted with 1% LDAO. A few μl of the crude cell wall extract were tested in the lipid bilayer assay for poreforming activity. Conductance steps with 1.75 nS were observed at 20 mV applied membrane potential in 1 M KCl solution. Furthermore, the conductance increase caused by the detergent extract was a function of time after the addition of the protein to membranes in the black state. Control experiments with LDAO alone at the same concentration as in the experiments with extracts demonstrated that the membrane activity was caused by the presence of the extracts and not by the detergent. This result suggested that the channel-forming activity was an intrinsic property of the detergent extracts of whole C. urealyticum cells. The channelforming proteins were further purified by excising the enriched bands from preparative Tricine-PAGE gel and extracting the gel slice with 1% Genapol X-80. Addition of the extracted 4 kDa band to black lipid bilayer membranes resulted in reconstitution of channels. When regions at different molecular masses were excised from the same Tricine-PAGE gel, the highest channel-forming activity was always combined with the low molecular mass 4 kDa band. However, we also noticed that some minor channel-forming activity was spread across the molecular mass region between about 4 and 40 kDa (Fig. 1) of the tricine-containing SDS-PAGE.
Single-channel analysis of the pores formed by the C. urealyticum cell wall extract In single channel experiments with lipid bilayers, we observed a considerable increase of the specific membrane conductance in the presence of C. urealyticum cell wall extracted with 1% LDAO. The addition of lower concentrations of the cell wall extract to lipid bilayers made of Diph-PC/n-decane allowed the resolution of stepwise conductance increases. Figure 2a shows a single channel recording in the presence of the C. urealyticum cell wall extract, 5 min after the addition of the protein to a lipid bilayer membrane. Each step indicates the incorporation of one channel forming unit into the membrane. All steps were directed upwards, which demonstrated that the channels were always in the open state and had a long lifetime. Figure 2b shows a histogram of all conductance steps observed in reconstitution experiments with the C. urealyticum cell wall extract in 1 M KCl and a voltage of 20 mV. The most frequent value for the single-channel conductance was 1.75 nS, another less frequent conductive unit had a conductance of about 4 nS.
The single-channel conductance of the pore-forming protein from C. urealyticum cell wall extract was also studied as a function of different KCl concentrations ( Table 2) to get some insight into the biophysical properties of the cell wall porin of C. urealyticum. Additional information was obtained from single-channel experiments with salts containing other ions than K + and Cl − . The results summarized in Table 2 demonstrated that the channel was only moderately selective. This is concluded from experiments in which KCl was replaced by LiCl or KCH 3 COO. The exchange of the mobile ions K + and Cl − by the less mobile ions Li + and acetate showed that cations and anions have certain permeability through the C. urealyticum channel. The permeability of the cations through the channels followed approximately their mobility sequence in the aqueous phase. This means probably that the cell wall porin of C. urealyticum forms a wide and water-filled channel that exhibited some selectivity for cations.
Selectivity measurements with pores formed by the C. urealyticum cell wall extract Further information about the channel observed in C. urealyticum cell wall extract was obtained from zerocurrent membrane potential measurements in the presence of salt gradients. Fivefold salt gradients (100 mM versus 500 mM) were established across lipid bilayer membranes in which about 100 to 1000 channels were reconstituted. The results are summarized in Table 3 Fig . 1 Extracts from whole C. urealyticum cells visualized on a 16.5% SDS-PAGE containing tricine. The SDS-PAGE was stained with silver. Lane 1: High molecular mass markers, Lane 2: whole C. urealyticum cell wall extract, heated at 95°C for 10 min Lane 3: whole C. urealyticum cell wall extract (not heated) and Lane 4: Low molecular mass protein markers (4.6, 10, 15, 25 and 40 kDa). The arrows indicate the gel location of the putative pore-forming protein and suggested indeed that the cell wall pore of C. urealyticum is slightly cation selective.
Estimation of the diameter of the cell wall porin of C. urealyticum Experiments with non-electrolytes (NEs) were performed according to an established method to estimate the effective pore diameter of the pore in the cell wall of C. urealyticum [56][57][58][59][60][61][62]. The 1 M KCl solutions were supplemented for these experiments with 20% (w/v) NEs with different molecular masses ranging from 62 Da (ethylene glycol) to 2 kDa (PEG 2000) and molecular radii between 0.26 nm and 1.22 nm ( Table 4). The results of single-channel conductance measurements with these NEs suggested that in the presence of small NEs with hydrodynamic radii up to 0.94 nm, the single-channel conductance decreased, while PEG 2000 with hydrodynamic radii of 1.22 nm did not enter the C. urealyticum channel and showed no effect on its conductance (Table 4). This means that the pore in the cell wall of C. urealyticum has an exclusion limit below 2 kDa.

The cell wall channel of C. urealyticum is not voltage dependent
To study of the voltage-dependence of the channel from C. urealyticum, 200 ng/ml of the cell wall extracts was added to 1 M KCl solution at the cis-side of a Diph-PC/ n-decane membrane and the reconstitution of the   The potential V m , measured for fivefold gradients of different salts, is defined as the difference between the potential on the dilute side (100 mM) and the potential on the concentrated side (500 mM). The aqueous salt solutions were used unbuffered with a pH of 6, if not indicated otherwise. The temperature was 20°C. The cation/anion permeability ratio was calculated as the mean of at least 3 individual experiments [44] channels was followed for about 30 min. Then increasing positive and negative voltages were applied to the cis-side of the membrane. The C. urealyticum cell wall channel did not show any voltage dependence up to ±100 mV (data not shown).
Identification of the gene coding for the cell wall pore of C. urealyticum The extracts from whole C. urealyticum cells showed many bands on 16.5% tricine containing SDS-PAGE (Fig. 1). Excision of the low molecular mass bands from preparative SDS-PAGE suggested that they contained the channel-forming protein, but it was impossible to relate one defined single band to the channel-forming activity. Therefore, we looked for an alternative method to identify the channel-forming protein. Previously we could identify PorACj of C. jeikeium based on its homology with PorA of C. glutamicum and other Corynebacterium species [31,[63][64][65]. Therefore, we performed a similar approach here. A NCBI BLAST-translation tool search [32,33] using porA of C. glutamicum in the known genome of C. urealyticum DSM 7109 [13] did not lead to a clear indication for a homologous gene. However, search within the genome suggested that it contained three open reading frames (ORFs) between the genes coding for GroEL2 (cur_1716) and PKK2 (cur_1712) that could code for cell wall proteins similar to PorA (see Fig. 3). Primers were designed to clone cur_1714 and cur_1715 separately using DNA of C. urealyticum DSM 7109 as a template ( Table 1). The PCR-product was cloned into the pGEM-T easy vector and transformed to E. coli BL21 DH5α competent cells. The plasmid containing the desired insert was confirmed by sequencing using primers M13 reverse and T7 Promoter ( Table 1). The cur_ 1714 gene and its flanking regions were double digested from pGEM-T easy vector using XbaI and KpnI for cloning the gene into the expression vectors. Proteins coded by cur_1714 and cur_1715 of C. urealyticum DSM 7109 were purified separately by immobilized metal ion affinity chromatography (IMAC) and preparative SDS-PAGE according to Laemmli [39]. Addition of purified GST and His tagged proteins to black lipid bilayer membranes made of PC/n-decane did not result in conductance increases. This could mean that the tags had to be removed for poreforming activity of the two proteins. Histidine residues attached to the C-terminal end of CUR_1714 were cleaved with Factor Xa protease and the GST attached to the N-terminal of CUR_1715 was cleaved with thrombin protease. Pure CUR_1714 and CUR_1715 proteins where Histag and GST-tag, respectively, were cleaved did not show any channel-formation in lipid bilayer membranes despite many attempts. This result clearly indicated that the two proteins in this form (i.e. where His-tag and GSTtag were cleaved) did not represent channel-forming components. We also tried to detect channels when both proteins were added together to the membranes. However, also in this case we did not observe any poreforming activity, which means that these two proteins after cleavage of the affinity tags, neither together nor alone were channel formers.

Mature CUR_1714 is the channel former in C. urealyticum DSM 7109
Another problem that had seriously be considered is the possibility that CUR_1714 contained after cleavage with FXa protease still 6 additional amino acids at the C-terminal end that could interfere with channel formation. Four amino acids (IEGR) are leftovers from FXa protease cleavage because they are the recognition site (I-E-G-R) for FXa to facilitate removal of the fusion tags after arginine (R) residue. Additional two amino acids (GT) are the leftovers from KpnI restriction enzyme which cannot be avoided due to the cloning strategy. Therefore, we cloned cur_ 1714 into the pXMJ19 using Fwd_1714_XbaI and RP pX cur_1714 KpnI primers. The vector contained a cloning site with XbaI and KpnI restriction sites. After control of the plasmid by sequencing with primers M13 reverse and T7 Promotor (Table 1), cur_1714 gene and its flanking regions were doubledigested from pGEM-T easy vector using XbaI and KpnI. The gene cur_1714 was cloned and ligated into the shuttle vector pXMJ19. Next step was the The aqueous phase contained 1 M KCl and 20% (w/v) of different nonelectrolytes. The non-electrolytes were used to determine the diameter of the cell wall pore. The molecular masses and the hydrodynamic radii (r) of the non-electrolytes are given together with the conductance of the channel (G ± SD) in the corresponding solution and the aqueous conductivity. X is the aqueous conductivity of the salt solutions with and without the NEs and was taken from ref. [62]. The membrane voltage was 20 mV; T = 20°C expression of the protein without any tag in C. glutamicum ΔporAΔporH pXmj19 lacking the major cell wall channel proteins. The transformed cells were grown at 30°C in BHI medium to an OD 600 of around 0.6-0.8. Induction of CUR_1714 expression by adding 1 mM IPTG to the growth medium and additional overnight growth. The protein was successfully expressed and purified without any other extra amino acid. For purification analytical and preparative SDS-PAGE was performed [31,66]. Because of the low resolution of this gel system for observation of low molecular mass proteins, tricine containing gels were used as described previously [40]. Protein samples were size separated with 15% or 16.5% polyacrylamide gels containing Tris-Tricine. After gel size separation, a protein band of~4 kDa was cut out and was homogenized in 1% Genapol and 10 mM Tris-HCl, pH 8 for 20 h at 12°C under agitation (Fig. 4). Poreforming activity was present in the resulting supernatant similar to that shown in Fig. 2.
In addition, control experiments with LDAO and 1% Genapol alone at the same concentration as in the experiments with extracts demonstrated that the membrane activity was caused by the presence of the 4 kDa protein extracts and not by the detergent. In reconstitution experiments, we observed a considerable increase of the specific conductance. Singlechannel analysis of the purified cell wall porin, named here PorACur (PorA of C. urealyticum) showed a conductance of 1.5 ± 0.25 and 3 nS in 1 M KCl (Fig. 5). These channels were identical to those observed after addition of the C. urealyticum crude extracts to lipid bilayer membranes. This result suggested that PorACur (CUR_1714) and not CUR_1715 or CUR_1713 represented the cell wall channel of C. urealyticum.

Single-channel analysis of PorACur of C. urealyticum
The PorACur protein was successfully expressed and further purified by excising the enriched band from preparative Tricine-PAGE gel and extracting it with 1% Genapol X-80 from the gel slice. Addition of the extracted 4 kDa band to black lipid bilayer membranes resulted in very slow reconstitution of channels. In lipid bilayer single channel experiments, we observed two frequent conductive units of the specific membrane conductance in the presence of pure PorACur of C. urealyticum (see Fig. 5). The average conductance of the steps corresponding to the left-side maximum was 1.75 ± 0.25 nS in 1 M KCl and that of the right-side maximum was 3 nS. The single-channel conductance of pure PorACur was also studied as a function of different KCl concentrations. The results are summarized in Table 5 and demonstrate that the single-channel conductance was approximately a linear function of the aqueous salt concentration. The homologous genes of the chaperonin GroEL2 and a polyphosphate kinase PPK2 enclose a presumed conserved porin domain. The operon covering the genes CgporH and CgporA whose gene products form the main cell wall channel of C. glutamicum is assumed to exist in all strains except for C. jeikeium. Search within the genome of C. urealyticum indicated that this strain does not follow this rule and suggested instead that it contained 3 open reading frames (ORFs) between the genes coding for GroEL2 (cur_1716) and PKK2 (cur_1712). They may code for cell wall proteins similar to PorA. Possible terminator sequences of mRNA transcripts were predicted with TranstermHP (indicated by hairpins [78] or were identified manually (marked by asterisk))

Selectivity measurements with PorACur
Further information on PorACur of C. urealyticum was obtained from zero-current membrane potential measurements in the presence of salt gradients. Fivefold salt gradients (100 mM versus 500 mM) were established across lipid bilayer membranes in which about 100 to 1000 channels were reconstituted. Further singlechannel analysis indicated that the cell wall channel is wide and water-filled because it is only slightly selective for cations over anions. Its conductance followed the mobility sequence of cations and anions in the aqueous phase. Information on the size and selectivity of PorA-Cur are summarized in Tables 5 and 6.

Comparison of PorACur with PorACj of C. jeikeim K411
In the following we present a comparison of the sequences of the major cell wall proteins PorACj of C. jeikeium K411 with that of PorACur of C. urealyticum DSM 7109, because both cell wall pores were obviously formed by a single polypeptide. The multialignment was performed using Pole Bioinformatique Lyonnaise Network Protein Sequence Analysis (http://npsa-pbil.ibcp.fr). Interestingly, we observed some homology between PorACur (coded by cur_1714 gene) and PorACj as shown in Fig. 6. There exist four amino acids identical (*; 10%) and thirteen amino acids strongly similar (:; 32.5%) in the primary sequence of PorACur (37 amino acids) as compared to PorACj of C. jeikeium (40 amino acids).

Putative structure of the channel formed by PorACur
Secondary structure predictions suggested that PorACur contained a heptameric repeat motive (abcdefg, Fig. 8) indicating the existence of α-helical structures with hydrophobic and hydrophilic residues localized on different sides of the helices. Figure 8 shows also the possible arrangement of the amino acids within an α-helix PorACur. This means that this protein could form amphipathic helices similar to the possible secondary structure in the monomeric PorH and PorA proteins [28,29,64] and also PorACj of C. jeikeium [31].
The PorACur monomer is orientated with the hydrophobic residues (Fig. 8 (f c g)) to the lipid phase, while the two negative residues glutamate and aspartate (E9, D19) together with the positive lysine (K23) are oriented to the channel lumen (a d). Therefore, the negatively charged residues must take a dominant position in the determined cation selectivity. The number of PorACur monomers in the homooligomeric channel is not known yet and needs further experimental and structural information on the channel. Five monomers are suggested to form the PorB channel of C. glutamicum [67]. Thus, it is very likely that the PorACur channel is formed by six to eight monomers as it is shown in Fig. 9.

MD simulation results of PorACur
Initial models of hexameric and octameric channels were constructed as shown in Fig. 9a. Subsequently, unbiased MD simulations were performed on these oligomeric initial models to check their structural and dynamic behavior. Root mean square deviation profiles (Fig. 9b)   Fig. 4 Different preparations of PorACur visualized on a 16.5% SDS-PAGE containing tricine. The SDS-PAGE was stained with silver. Lanes 1 and 4, Low molecular mass protein ladder. Lane 2, Purified PorA-Cur with C-terminal His 8 -tag expressed in C. glutamicum ATCC 13032 ΔHA, Lane 3, Purified PorACur from preparative gel without tag expressed in C. glutamicum ATCC 13032 ΔHA, Lane 5, Purified CUR_1715 without GST-tag (around 9 KDa) expressed in E. coli BL21 DE3 (Omp8). The purified proteins with and without tags were reconstituted into lipid bilayer membranes. Only protein of lane 3 shows channel-formed activity in lipid bilayers. The channel-forming protein, named PorACur (coded by cur_1714 gene), was identified in the known genome of C. urealyticum by its similar chromosomal localization to known porH and porA genes of other Corynebacterium strains (see Fig. 3). The arrows indicate the location of the corresponding proteins indicate that the both models deviate by 3-5 Å from starting structure and reach stable conformations after about 250 ns. The octameric arrangement clearly forms a wider lumen than the hexameric channel. The kinking of the α-helices during equilibration is caused by the proline residue present which is known to induce this kind of kinks in α-helices. Furthermore, to determine which of the proposed structures of the channel is more likely, applied-field simulations were performed to compare with experimental electrophysiological data. Applied simulations were performed on the octameric channel bathed in 1 M KCl solution at 100 mV. The calculated average conductance of the octamer model is 1. 04 ± 0.14 nS which is in reasonable agreement with the experimental values though lower. Due to the channel diameter, it is obvious that the hexameric model would result in even lower conductance values and thus we refrained from performing the respective simulations. Actually, at this point we cannot rule out that a decameric model would result in conductance values even closer to the experimental findings.

C. Urealyticum contains a cell wall channel
Treatment of whole cells from the CNM-group of grampositive bacteria with different detergents provides an efficient and simple way to isolate pore-forming proteins from the cell wall [37,38]. The same method was also applied to C. urealyticum cultures. Lipid bilayer experiments in the presence of the LDAO-extracts b Histogram of the probability P(G) for the occurrence of a given conductivity unit observed with membranes formed of 1% Diph-PC dissolved in n-decane. The histogram was calculated by dividing the number of fluctuations with a given conductance step by the total number of conductance fluctuations in the presence of pure PorACur. Two frequent conductive units were observed for single events taken from different individual membranes. The average conductance of the steps corresponding to the left-side maximum was 1.75 ± 0.25 nS and that of the right-side maximum was 3 nS (65 current steps in total). The aqueous phase contained 1 M KCl, and10 ng/ml protein extract, the applied membrane potential was 20 mV, T = 20°C The membranes were formed by Diph-PC/n-decane. The single-channel conductance was measured at 20 mV applied voltage and T = 20°C. The average singlechannel conductance, G (± SD), was calculated from at least 100 single events from up to five individual membranes The potential V m , measured for a fivefold gradient of different salts, is defined as the difference between the potential on the dilute side minus the potential on the concentrated side. The aqueous salt solutions were buffered with 10 mM Tris-HCl (pH 7), and the temperature was 20°C. The cation/anion permeability ratio was calculated from at least 3 individual experiments [44] demonstrated the presence of a pore-forming protein that formed pores with a single-channel conductance of 1.75 nS in a 1 M KCl solution. This result indicated that the cell wall of C. urealyticum contained a pore similar to those of the cell wall of all members of the mycolata studied to date. The pore-forming protein responsible for the formation of the pores had obviously a very small molecular mass because low molecular mass bands in the range of 3 kDa excised from preparative SDS-PAGE showed highest pore-forming activity.
The cell wall channel of C. urealyticum is wide and waterfilled We studied the properties of the cell wall pore of C. urealyticum, termed PorACur (see below) in some detail by measuring its single channel conductance and its selectivity. Single channel data and selectivity measurements suggest that the cell wall channel is slightly cationselective for KCl (with about equal aqueous mobility for K + and Cl − ) and higher cation-selective for potassium acetate. In the case of LiCl, where the cation has approximately half of the aqueous mobility of Cl − , the pore appeared to be slightly anion-selective. This result suggested that the different ions move inside the channel in a similar way as in the aqueous phase because the pore becomes anion-selective in the case of the less mobile Li + ions. This means that the pore appears to be wide and water-filled. But what is the exact diameter of the cell wall channel PorACur of C. urealyticum? This is an interesting question important also for the development of new antimicrobial compounds precisely targeting essential intracellular components. To get some insight in the effective size of the pores formed by PorACur we performed single-channel conductance experiments with non-electrolytes (NEs) according to established methods to measure the exclusion limit of the PorACur channel [56][57][58][59][60][61][62]. The results of the measurements with NEs with different molecular masses suggested (see Table 4) that not all NEs were able to enter the channel. From the NEs listed in Table 4 only ethylene glycol (62 Da) to PEG 600 were obviously able to enter the PorACur channel to reduce its conductance. This means that the pore allows the free passage of molecules with a diameter of 1.6 nm and below because the conductance in the presence of these NEs decreased approximately proportional to that of the bulk aqueous conductivity X (see Table 4). The conductance in the presence of PEG 2000 did not change in comparison with that of 1 M KCl alone suggesting that this NE with hydrodynamic radius around 1.22 nm did not enter the PorA-Cur channel. The single-channel conductance also decreased for PEG 1000 (hydrodynamic radius 0. 94 nm) somewhat, but interestingly much less than that of the specific conductivity X of the aqueous bulk solution (see Table 4). This result indicated that there is break of entry of NEs into the PorACur channel around PEG 1000 indicating that the radius of the channel is close to that of PEG 1000, which means that the diameter of the PorACur pore formed  [79]. b The secondary structure of PorACur was predicted using a consensus method at the Pole Bioinformatique Lyonnaise network (http://npsa-pbil.ibcp.fr/cgi-bin/npsa_automat.pl?page=/NPSA/npsa_seccons.html) to form α-helices. Amino acid residues were arranged on the basis of heptameric repeats (a-g, rotation of 100 degrees per residue starting from a, in the clockwise direction) showing distinct separation in a hydrophobic domain that could be surrounded by lipid molecules (black) while the hydrophilic domain (grey) is suggested to represent the α-helices orientated to the water-filled lumen of the putative oligomeric PorACur channel is approximately 1.8 nm. Similar diameters have been found for the different cell wall channels of Corynebacterium species, such as C. glutamicum, C. efficiens, C. diphtheriae and C. jeikeium [29-31, 37, 65]. However, such a diameter is by far larger than diameters of channels in nerve and muscle tissues. This has to do with the function of the different channels. Bacteria live in relatively dilute environment indicating that the channels cannot be too small for the efficient scavenging of nutrients.
The diameter of the pore formed by PorACur is definitely large enough to account for the high susceptibility of C. urealyticum for a variety of antibiotics. Fluoroquinolones, ketolides, rifampicin, and tetracyclines are presumably small enough to pass through the cell wall channel of C. urealyticum, which means that they can enter the cells [16,18]. Beta-lactam antibiotics, macrolides and aminoglycosides are more bulky. For these antibiotics the PorACur pore could offer some barrier for their diffusion into the cells causing some sort of intrinsic resistance against them [20,21,25].

The C. Urealyticum cell wall channel is formed dependent only on Cur_1714
Earlier studies into the cell wall channels of Corynebacterium species demonstrated the co-existence of two pore-forming polypeptides PorA and PorH in the cell wall of C. glutamicum, C. diphtheriae, C. callunae and C. efficiens [28][29][30]. In addition, results from RT-PCR experiments suggested co-transcription of the corresponding genes cgporA and cgporH of C. glutamicum, which constitute an operon. The important question of this study was whether the cell wall channel of C. urealyticum is also formed by two polypeptides like in the Corynebacterium species mentioned above or whether the situation is like in C. jeikeium where the cell wall channel is formed by a single polypeptide, PorACj [31]. The analysis of the corresponding gene region in the genome of C. urealyticum demonstrated that 3 open reading frames (cur_1713, cur_1714 and cur_1715) were located between the genes coding for GroEL2 (cur_1716) and PKK2 (cur_1712), where in the genomes of other Corynebacterium species only one or two open reading frames were found (see Fig. 3). One gene (cur_1713) is transcribed in the opposite direction to cur_1714 and cur_1715 and its gene product (molecular mass 25.9 kDa) does not show any homology with that of pore formers. This suggested that this gene does not code for the cell wall channel observed in whole cell extracts. The molecular mass of proteins encoded by cur_1714 and cur_1715 are 4.1 and 8.9 kDa respectively. Both genes were separately expressed in the C. glutamicum ΔporHΔporA mutant strain. The purified proteins were studied in experiments using lipid bilayer membranes. Only Cur_1714 alone, termed PorACur, was able to form ionpermeable channels in these experiments when the protein did not contain N-or C-terminal extensions such as the His 8 -tag or the GST-tag. A comparison of the lipid bilayer data obtained with whole cell extracts and purified PorACur demonstrated that PorACur is identical to the channel-forming protein observed in detergent extracts of whole cells. Cur_1715 did not show any pore-forming activity in lipid bilayer membranes despite many attempts suggesting that it is not a channel-forming component.

Putative structure of the channel formed by PorACur
Our results demonstrate that PorACur with a molecular mass of 4.1 kDa (37 amino acids) is the channel-former in the cell wall of C. urealyticum. Its length is even smaller than PorACj of C. jeikeium that has a length of 40 amino acids. PorACur is definitely the smallest polypeptide that forms stable and voltage-independent pores in membranes. All the other low molecular mass poreforming polypeptides, such as melittin, alamethicin and the defensins form pores that are either voltagedependent or need often to be initiated by higher voltages [68][69][70]. The membrane pores formed by all these molecules are oligomers. This most likely also the case for PorACur because it is definitely too small to form a pore on its own. The single PorACur molecule forms obviously an alpha-helical structure with hydrophobic and hydrophilic residues localized on different sides of the helix as suggested by secondary structural predictions (see Figs. 7 and 8). Thus, it is very likely that the membrane pore is formed by several amphipathic PorA-Cur molecules arranged a barrel stave model with an inner diameter of about 1.8 nm. The number of monomers needed for such a structure is an open question but this could influence channel conductance. We consider the possibility of an uneven number of monomers of PorACur in a channel as rather unlikely because the subunits of the PorA/H channel in the cell wall of C. glutamicum is presumably a PorA/PorH dimer [30]. In our previous study of the pore formed by PorACj of C. jeikeium we made similar considerations and suggested that the pore is very likely formed by an octamer [31].
In this study, we performed applied-field MD simulations of 250 ns length to compare the theoretical conductance of the PorACur membrane pore with the electrophysiological data from single-channel measurements. The results suggested that the octameric pore is definitely more likely because the inner diameter of the hexameric pore is too small to account for the conductance of the cell wall channel of C. urealyticum. The conductance of the octameric pore shows reasonable agreement with the experimental values. However, it should be carefully noted that the octamer model is a predicted structure without any prior structural information.
Comparison of the cell wall channel properties of C. urealyticum and other Corynebacterium species The cell envelopes of members of the genus Corynebacterium are similar in their composition and organization [71,72]. Their mycolic acids are relatively short in comparison to those of bacteria of related genera. The known cell wall pores of Corynebacterium species share common properties. In particular, they are formed by homo-or heterooligomers of small polypeptides of less than 60 amino acids in contrast to the much bigger MspA homologs of Mycobacteria, Nocardia and Tsukamurella species [73][74][75][76]. Similarly, they all have a high propensity to form alpha-helical structures to form wide and water-filled pores [27,28,30,31,37]. Common to PorA and PorH polypeptides is also that the polypeptides contain no presequences suggesting that they are translated across the cytoplasmic membrane and integrated into the cell wall by a still unknown export system. Many of the cell wall channels of Corynebacterium species are formed by PorA/PorH subunits, i.e. they are heterooligomeric. However, two examples now exist, where the cell wall pores are homooligomeric, PorACj of C. jeikeium [31] and PorACur of C. urealyticum, studied here. Analysis of the primary sequences of the different subunits of the cell wall channels provides a clear phylogenetic relationship between them. They form a superfamily of proteins analogous to the MspA cell wall channel family of Mycobacterium smegmatis and related species [73][74][75][76] or PorACoram of Corynebacterium amycolatum that belongs to the DUF 3068 (Domains of Unknown Function 3068) superfamily of proteins, mainly found in bacteria from the family Corynebacteriaceae [77].
The subunit of a monomeric cell wall channel of Corynebacterium species, such as PorACj of C. jeikeium or PorACur studied here could be the ancestor of the PorA and PorH subunits of the heterooligomeric cell wall channels. In these cases, either PorA-homologs or PorH-homologs could have evolved by gene duplication of an ancestral gene. A phylogenetic tree of the PorA/H family of subunits of cell wall channels of Corynebacterium species together with PorACur and PorACj from C. urealyticum and C. jeikeium is shown in Fig. 10. The tree demonstrates that these two subunits are somewhat distantly related to another although they are both short polypeptides with a high propensity to form alphahelical structures (see also Fig. 7). Both polypeptides are related to subunits of the corresponding cell wall channels that are obviously formed by PorA and PorH. This could mean that gene duplication occurred several times. It could also mean that gene duplication occurred after a long time when the single cell wall channel subunit existed for already a longer period of time and spread to new Corynebacterium species. The interesting point in this relationship is that the PorA family of proteins is more closely related to PorACur and PorACj than the PorH family.

Conclusions
This paper describes the identification, purification and characterization of the cell wall channel of the pathogenic actinomycete Corynebacterium urealyticum. The pore is encoded by the gene cur_1714 that codes only for the mature protein and not for a protein precursor indicating a not yet identified mechanism for the translocation of PorACur across the cytoplasmic membrane. The channel is formed by small polypeptide subunits that comprise only 37 amino acids, which means that PorACur is the smallest subunit of a cell wall channel from Corynebacterium species that forms a stable pore. Secondary structure predictions suggest that the PorACur polypeptide forms an alpha-helical structure. The pore is slightly cation selective and voltage-independent for voltages up to ±100 mV. Single-channel experiments with different salts and non-electrolytes indicate that it is wide and water-filled. Molecular dynamics simulations suggest that 8 to 10 PorACur monomers form the homooligomeric cell wall channel.  The tree was generated with the phylogeny.fr tool [80] using protein sequences downloaded from the NCBI protein database with the indicated identifiers. The multiple sequence alignment was calculated with MUSCLE using the custom mode with a maximum number of 16 iterations. The phylogenetic analysis was performed with PhyML and the approximate likelihood-ratio test for branch support. The substitution model was used in default settings. The tree was rendered with TreeDyn and default settings