Reconstitution experiments and gene deletions reveal the existence of two-component major cell wall channels in the genus Corynebacterium

Abstract

Two small polypeptides, PorA and PorH, are known to form cell wall channels in Corynebacterium glutamicum and in Corynebacterium efficiens. The genes coding for both polypeptides are localized in close proximity to one another between the genes coding for GroEl2 and a polyphosphate kinase (PKK2). In this study, we investigated the relationship of PorA and PorH to one another. The results suggested that the major cell wall channels of Corynebacterium glutamicum, Corynebacterium efficiens, and Corynebacterium diphtheriae need the obligatory presence of two distinct polypeptides, one of class PorA and one of class PorH, to form an active cell wall channel. Identification of genes coding for homologous proteins in the chromosome of Corynebacterium callunae suggested a similar result for this strain. Contrary to our previous reports on channel-forming proteins in these strains, a heterooligomeric structure composed of PorA and PorH is needed in all of them to form the major cell wall channel. This was concluded from complementation experiments using a porH- and porA-deficient C. glutamicum strain. The stringent necessity of proteins of either class to recover the wild-type channels was demonstrated by black lipid bilayer experiments using detergent or organic solvent extracts of the complemented porH- and porA-deficient C. glutamicum strain. The channel-forming capability of recombinant expressed, affinity-purified PorA and PorH proteins of C. glutamicum revealed that the channels consisted solely of these two components. This agreed with results obtained from a transcript coding for both channel-forming components identified in C. glutamicum by Northern blot analysis and reverse transcription-PCR analysis. The transcription start point of the genes was determined by the rapid amplification of cDNA ends approach, allowing the prediction of the -35 and -10 regions of the promoter. The results demonstrate that the cell wall channels within the genus Corynebacterium may be formed by two-component oligomers.

FIG. 1.

ORF analyses of a plasmid-encoded C. diphtheriae fragment. Depicted is the fragment of strain C. diphtheriae ATCC 11913 which is an integral part of plasmid pX_Cd-WT. The fragment complements the channel deficiency of the C. glutamicum ΔporH ΔporA mutant (53). Gene silencing disclosed that the 59-amino-acid small protein CdPorH, together with CdPorA, accounted for the channel activity. The polypeptide sequence of CdPorH shows 20 and 24% identify to PorH of C. efficiens and C. glutamicum, respectively. The genes cdporH and cdporA of C. diphtheriae are highlighted in black. Dots below the sequence of the cloned C. diphtheriae fragment indicate identical nucleotides, whereas mutations are specified in boldface type. Boldface italics show a putative ribosome binding site (RBS) in front of each porin gene. Underlined italics mark the XbaI and EcoRI sites used for cloning. A transcriptional terminator is specified by underlined letters (53). The specific size of the fragment in the plasmid is displayed by the numbers behind the DNA sequence. The asterisks specify plasmids with a functional C. diphtheriae channel domain.

FIG. 2.

(A) Arrangement of cgporA and cgporH in the genome of C. glutamicum and primers used for Northern blot analysis. (B) Transcript analysis of cgporH and cgporA of C. glutamicum. Total RNA was isolated from the wild-type strain and a mutant strain deficient in cgporA (9). Northern blot assays of a single membrane in which 5 μg of total RNA of C. glutamicum cells grown in BHI medium was probed with either a cgporH or a cgporA probe. A cotranscript of approximately 600 bases was detected.

FIG. 3.

Comparison of the genotypes of wild-type and ΔporH ΔporA mutant C. glutamicum strains. The sequence of the mutant strain used for the alignment with the wild-type (WT) sequence is deduced from a sequenced PCR product obtained with primers FP Cg_KO_HA_CHECK2/RP Cg_OPERON_1_5 (Table 2) and template DNA of the C. glutamicum ΔporH ΔporA mutant (data not shown). The deleted region comprises 472 bp, including both the cgporH and cgporA genes (black highlighted) coding for the components of the main cell wall channel of C. glutamicum. Boldface type indicates CDS adjacent to the porins. The TSP of the cgporH-cgporA cotranscript is marked by a broken arrow. Putative −10 and −35 boxes are shown by underlined italics. Boldface italics indicate putative ribosome binding sites (RBS), whereas double-underlined letters indicate rho-independent hairpin terminators (26). The numbers to the right of the nucleotide sequence correspond to positions in the genome of C. glutamicum (GenBank accession number NC_003450).

FIG. 4.

Single-channel recordings for a PC/n-decane membrane in the presence of channel-forming proteins of C. glutamicum. The aqueous phase contained 1 M KCl. The applied membrane potential was 20 mV, and the temperature was 20°C. Arrows indicate fluctuations caused by sample addition. (A to F) Crude organic solvent extracts of C. glutamicum rporHrporA (A), C. glutamicum rporHrporA expressing CgPorA (B), C. glutamicum rporHrporA expressing CgPorH (C), wild-type C. glutamicum (D), C. glutamicum rporHrporA expressing CgPorH and CgPorA (E), and a mixture of C. glutamicum rporHrporA expressing CgPorH and wild-type C. glutamicum (F). (G) A 1:1 protein mixture of purified CgPorH_CHis and factor Xa-treated CgPorA_NHis (20 ng/ml).

FIG. 5.

Silver staining of Ni-NTA-purified, histidine-tagged PorA and PorH proteins of C. glutamicum separated by 12% Tricine SDS-PAGE. The proteins were expressed independently of each other in a ΔporH ΔporA mutant strain of C. glutamicum. Note that the forms of the bands are caused by some overloading of the gels and the highly hydrophobic nature of PorA and PorH.

FIG. 6.

Histogram of the probability of the occurrence of ascertained conductivity units observed in the experiment of Fig. 4G. The most frequent single-channel conductances were 2.5 nS (left-side maximum) and 5.5 nS (right-side maximum), for a total of 164 single-channel events. The data were collected from more than three individual membranes.

FIG. 7.

Single-channel recordings and single-channel analysis of porin proteins of C. efficiens expressed in the C. glutamicum ΔporH ΔporA mutant. The aqueous phase contained 1 M KCl. The applied membrane potential was 20 mV, and the temperature was 20°C. (A) A combination of organic solvent cell wall preparations which contained either the CePorH protein (using plasmid pX_Ce-H) or the CePorA protein (using plasmid pX_Ce-A). Each sample itself was non-channel forming. (B) PorH and PorA of C. efficiens were expressed together using plasmid pX_Ce-HA. About 20 ng/ml ether-precipitated organic solvent extract was added to the aqueous phase on both sides of the membrane. (C) Statistical analysis of the channel conductance of 143 single-channel events of panel B derived from more than three individual membranes.

FIG. 8.

Single-channel recording of a PC/n-decane membrane in the presence of channel-forming proteins of C. diphtheriae expressed in the C. glutamicum ΔporH ΔporA mutant. About 20 ng/ml ether-precipitated organic solvent extract (see text) was added to the aqueous phase on both sides of the membrane. The aqueous phase contained 1 M KCl. The applied membrane potential was 20 mV. The temperature was 20°C.

FIG. 9.

Agarose (0.8%) gel electrophoresis of PCR-amplified homologous porin regions of C. glutamicum and C. callunae. Amplicons were obtained with primers FP Cc_HA1 and RP Cc_HA3 using chromosomal DNA of both strains as the template. The sequence of the C. callunae PCR product is shown in Fig. 10.

FIG. 10.

Analysis of the C. callunae porin domain. The amplicon of Fig. 9 was cloned into the TOPO 2.1 vector and sequenced. The primers used for cloning of the C. callunae region are marked by underlined italics. The boldface type represents the C. callunae homologues of, e.g., NCgl2620 and NCgl2621 of C. glutamicum ATCC 13032. The identified C. callunae ccporH and ccporA genes (black background) thus have the same chromosomal localization as the porH and porA genes of other corynebacteria investigated in this study. As with C. glutamicum, a putative RBS shown in boldface italics attends each porin gene. Both genes are surrounded by distinct hairpin structures marked as double-underlined letters. Numbers behind the DNA sequence display the specific size of the PCR product inserted into the TOPO 2.1 vector.

FIG. 11.

Ethidium bromide-stained agarose (0.8%) gel electrophoresis of PCR amplification products obtained with primers FP Cg_KO_HA_CHECK2 and RP Cg_OPERON_1_5 (Table 2) binding in the flanking regions of the porH-porA genes of C. glutamicum. Clearly visible are the absence of cgporA in lane 3 and the absence of cgporA and cgporH in lane 4 when the data are compared to PCR in the presence of wild-type (WT) DNA (lane 2).