https://orcid.org/0000-0002-8646-7925
Figures
Abstract
Acetogenic bacteria are of high interest for biotechnological applications as industrial platform organisms, however, acetogenic strains from the genus Terrisporobacter have hitherto been neglected. To date, three published type strains of the genus Terrisporobacter are only covered by draft genome sequences, and the genes and pathway responsible for acetogenesis have not been analyzed. Here, we report complete genome sequences of the bacterial type strains Terrisporobacter petrolearius JCM 19845T, Terrisporobacter mayombei DSM 6539T and Terrisporobacter glycolicus DSM 1288T. Functional annotation, KEGG pathway module reconstructions and screening for virulence factors were performed. Various species-specific vitamin, cofactor and amino acid auxotrophies were identified and a model for acetogenesis of Terrisporobacter was constructed. The complete genomes harbored a gene cluster for the reductive proline-dependent branch of the Stickland reaction located on an approximately 21 kb plasmid, which is exclusively found in the Terrisporobacter genus. Phylogenomic analysis of available Terrisporobacter genomes suggested a reclassification of most isolates as T. glycolicus into T. petrolearius.
Citation: Böer T, Bengelsdorf FR, Bömeke M, Daniel R, Poehlein A (2023) Genome-based metabolic and phylogenomic analysis of three Terrisporobacter species. PLoS ONE 18(10): e0290128. https://doi.org/10.1371/journal.pone.0290128
Editor: Bashir Sajo Mienda, Federal University Dutse, NIGERIA
Received: June 26, 2023; Accepted: September 27, 2023; Published: October 10, 2023
Copyright: © 2023 Böer et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The genome sequence for T. mayombei is available at DDBJ/ENA/GenBank under the accession numbers CP101637-CP101639 and the BioProject/BioSample accession numbers are PRJNA838466 and SAMN28423240. Raw reads can be found in the NCBI sequence read archive under the accession number SRR23391581 for Illumina reads and SRR23390615 for Nanopore reads. The genome sequence for T. petrolearius is available at DDBJ/ENA/GenBank under the accession number CP102984-CP102986 and the BioProject/BioSample accession numbers are PRJNA839206 and SAMN28513175. Raw reads can be found in the NCBI sequence read archive under the accession number SRR23390574 for Illumina reads and SRR23392100 for Nanopore reads. The genome sequence for T. glycolicus is available at DDBJ/ENA/GenBank under the accession number CP117523-CP117524 and the BioProject/BioSample accession numbers are PRJNA929918 and SAMN32972286. Raw reads can be found in the NCBI sequence read archive under the accession number SRR23389261 for Illumina reads and SRR23389232 for Nanopore reads.
Funding: This work was supported by the Open Access Publication Funds of the University of Göttingen. Furthermore, we acknowledge the support of T.B. by the “Deutsche Bundesstiftung Umwelt” (DBU; 20020/640) and partial support of A.P. by the BMBF (ThermoSynCon; 031B0857C). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
Introduction
Acetogenic bacteria are of high interest to biotechnology as they enable the conversion of industrial waste gases to valuable biocommodities by synthesis gas fermentation or microbial electrosynthesis (MES). These carbon-recycling technologies can contribute to the current efforts towards achieving climate neutrality. The genus Terrisporobacter comprises isolates capable of performing MES and autotrophic growth on H2 and CO2 through acetogenesis [1–3]. Terrisporobacter consists of mesophilic, Gram-positive and strictly anaerobic members. The cells appear as slightly curved rods, form endospores and are motile by flagella [4]. The four published type strains are Terrisporobacter mayombei DSM 6539T, Terrisporobacter glycolicus DSM 1288T, Terrisporobacter petrolearius JCM 19845T and Terrisporobacter hibernicus NCTC 14625T. The ability to utilize H2 and CO2 via the Wood-Ljungdahl pathway (WLP) has only been shown for T. mayombei (Kane et al. 1991). Acetogenesis has also been described for the isolates T. glycolicus DSM 13865 and ATCC 29797, and genes required for the Wood-Ljungdahl pathway are present in all available Terrisporobacter draft genomes [2, 5]. So far, no detailed genome-based analysis is available for the genes and pathways involved in acetogenic growth of Terrisporobacter strains. The first Terrisporobacter species discovered was T. glycolicus DSM 1288T (formerly Clostridium glycolicum), which was isolated from mud of a stagnant pond near the Chesapeake and Ohio canal (USA) in 1962 [6]. T. mayombei DSM 6539T (formerly Clostridium mayombei) was isolated as the second Terrisporobacter type strain from gut contents of the soil-feeding termite Cubitermes speciosus in 1991 [3]. The genus Terrisporobacter was proposed in 2014 and T. glycolicus DSM 1288T was described as a type species [4]. The third type strain T. petrolearius JCM 19845T was isolated in 2015 using a petroleum reservoir sample from an oilfield in Shengli (China) [7]. Terrisporobacter othiniensis (08–306576) and Terrisporobacter muris (DSM 29186) were described as different strains. However, both strains appeared to be the same species, but these names have not been published [8, 9]. In 2023, Terrisporobacter hibernicus (NCTC 14625T) was published as a fourth type strain [10]. Terrisporobacter members have been found to be strongly associated to pig manure and in particular T. glycolicus DSM 1288T has been discussed as potential emerging anaerobic pathogen [11, 12]. In addition, some members show a positive correlation with the generation of bioelectricity in microbial fuel cells, indicating their electricigenic potential [13]. Correspondingly, Terrisporobacter strains have been found at the cathode of MES cells [1]. These insights support the use of strains from the genus Terrisporobacter as biocatalysts in biotechnology. The availability of high-quality genome sequences is a prerequisite for further metabolic investigations. Nevertheless, high-quality genome sequences were not available for three published type strains of the Terrisporobacter genus. We therefore sequenced and assembled the genomes of T. petrolearius JCM 19845T, T. mayombei DSM 6539T and T. glycolicus DSM 1288T to perform a genome-based metabolic and phylogenomic analysis of these Terrisporobacter type strains.
Materials and methods
Cultivation and DNA extraction
Genomic DNA was extracted from cells of a T. petrolearius JCM 19845T culture derived from the Microbe Division of the Japan Collection of Microorganisms (JCM; Tsukuba; Japan). T. glycolicus DSM 1288T and T. mayombei DSM 6539T were grown from lyophilised cultures derived from the German Collection of Microorganisms and Cell Cultures (DSMZ; Braunschweig, Germany) in an overnight culture, incubated at 30°C in DSM 311 medium supplemented with glucose (30 mM). Cell pellets were obtained by centrifugation of 10 ml culture at 13,000 g for 5 min. DNA was isolated by using the instructions for cell samples of the MasterPure™ Complete DNA and RNA purification kit (Epicentre, Madison, WI, USA).
Genome sequencing, assembly and annotation
Illumina sequencing libraries were prepared using the Nextera XT DNA sample preparation kit (Illumina, San Diego, CA, USA) and sequenced with a MiSeq system and v3 chemistry (600 cycles) following the instructions of the manufacturer (Illumina, San Diego, CA, USA). Sequencing kits and program versions for Nanopore sequencing are shown in the S1 Table. The Nanopore sequencing libraries were prepared with 1.5 μg high-molecular-weight DNA using the ligation sequencing kit 1D and the native barcode expansion kit as recommended by the manufacturer (Oxford Nanopore Technologies, Oxford, UK). Nanopore sequencing was conducted for 72 h with the MinION device Mk1B, the SpotON flow cell R9.4.1 and the MinKNOW software following the instructions of the manufacturer (Oxford Nanopore Technologies, Oxford, UK) Demultiplexing and base calling of Nanopore sequencing data were performed with the Albacore/Guppy software. Quality control and adapter trimming of paired-end Illumina sequences was performed with FastQC (v0.11.9) and Trimmomatic (v0.39; LEADING: 3, TRAILING: 3, SLIDINGWINDOW:4:15, MINLEN:50) [14]. Porechop (v0.2.4) was used for adapter trimming of Nanopore reads, which were subsequently filtered to a minimal read length of 5 kb [15]. Trimmed paired-end Illumina and filtered long Nanopore reads were used as input of Unicycler (v0.5.0) for a de novo hybrid genome assembly [16]. The resulting assembly graph was inspected with Bandage (v0.8.1) [17]. Annotation of the resulting genomes was performed with Prokka (v1.14.6) and the locus tag TEMA for T. mayombei DSM 6539T, TEPE for T. petrolearius JCM 19845T and TEGL for T. glycolicus DSM 1288T were chosen [18]. BUSCO scores were obtained using BUSCO (v5.4.4) with the lineage dataset clostridia_odb10 and the mode for prokaryotic genomes [19].
Genome-based metabolic and phylogenomic analysis
Functional annotation of genes to KEGG categories and the reconstruction of KEGG pathway modules was conducted with BlastKOALA [20]. For metabolites occurring with multiple KEGG pathway modules only the most complete ones were chosen for visualization. Assignment of genes into cluster of orthologous genes (COG) and subsequent functional annotation into COG categories was performed with the eggNOG mapper (v2.1.9) using the eggNOG 5.0 database [21, 22]. Barcharts and heatmaps were plotted in RStudio (v2022.06.0) with the ggplot2 package (v3.4.1) [23, 24]. Pan/core genome analysis was conducted with OrthoVenn2 clustering genes, identifying orthologous clusters and assigning Gene Ontology (GO) annotations [25]. The identification and classification of insertion sequence (IS) elements was performed with the ISESCAN pipeline (v1.7.2.3) [26]. Orthologous genes for the Wood-Ljundgdahl pathway and the Stickland reactions were identified by using Proteinortho (v6.0.31) running with default parameters [27]. Furthermore, the Proteinortho output was used for the verification of missing genes in KEGG modules predicted to be incomplete by BlastKOALA. Gene cluster visualizations were prepared with clinker (v1.32) [28]. Virulence factor screening was performed with the VFanalyzer pipeline using the Virulence Factor Database (VFDB) [29]. The phylogenomic analysis was performed with all Terrisporobacter genomes available at NCBI Genbank using pyani (v0.2.12) and the MUMmer alignment option [30]. ANIm percentage identity and alignment length values are shown in S2 Table. The dDDH values for T. petrolearius were calculated by the Type (Strain) Genome Server (TYGS) [31].
Results and discussion
Genome assembly and annotation
Details for the genome sequencing, assembly and annotation of the three Terrisporobacter type strains are summarized in Table 1. Each genome harbors a circular chromosome ranging from 4.03–4.09 Mb, and an additional circular plasmid ranging from 21.29–21.89 Kb (pTM1, pTP1 and pTG1). An additional circular plasmid was obtained for T. mayombei DSM 6539T (pTM2, 8.20 Kb) and T. petrolearius JCM 19845T (pTP2, 6.74 Kb). The GC content of the chromosomes varied between 28.7–28.9%. Prokka annotation resulted in 4,165, 4,208 and 4,038 potential genes for the genomes of T. mayombei, T. petrolearius and T. glycolicus, respectively. All three genomes contained 12 genes coding for 5S, 16S and 23S rRNA. One CRISPR repeat region was detected in the genomes of T. petrolearius and T. glycolicus with 67 and 28 spacers, respectively. The genome of T. mayombei contained two CRISPR repeat regions with 12 and 25 spacers. BUSCO scored a 100% completeness for all three genomes with 98.8% single copy (S) and 1.2% duplicated (D) BUSCO marker genes. No fragmented (F) or missing (M) BUSCO marker genes were identified.
Functional annotation and pan/core genome analysis
The functional annotation into COG categories of the Terrisporobacter type strain genomes is visualized in Fig 1A. The eggNOG-mapper assigned categories to 86.9% of coding sequences for T. mayombei, 88.1% of coding sequences for T. petrolearius and 89% for T. glycolicus. The genes of the three Terrisporobacter type strain genomes distributed uniformly across the different COG categories, with the exception of the category replication, recombination and repair (L). With 258, the T. petrolearius contained significantly more genes than the other two type strains of T. mayombei (161) and T. glycolicus (150). The pan genome of the three genomes comprised 3,548 gene clusters of which 2,930 gene clusters formed the core genome (Fig 1B). The genome of T. mayombei and T. glycolicus each harbored 15 exclusive gene clusters, while the genome of T. petrolearius harbored 38 exclusive gene clusters. T. mayombei contained four exclusive gene clusters, including gene clusters for acyltransferase activity (GO:0016746), phosphorelay sensor kinase activity (GO:0000155), DNA binding (GO:0003677) and metal ion binding (GO:0046872). Furthermore, T. mayombei possessed two exclusive gene clusters with GO annotations for the regulation of DNA-templated transcription (GO:0006355) and cobalamin biosynthesis (GO:0009236). T. petrolearius contained ten exclusive gene clusters with GO annotations of molecular function, comprising gene clusters for ATP-binding (GO:0005524), phosphorelay response regulator activity (GO:0000156), metal ion binding (GO:0046872), spermidine-importing ATPase activity (GO:0015595), enterobactin transmembrane transporter activity (GO:0042931), transmembrane transporter activity (GO:0022857), methionyl-tRNA formyltransferase activity (GO:0004479), sequence-specific DNA binding (GO:0043565) and two gene clusters encoding ATP hydrolysis activity (GO:0016887). Four exclusive gene clusters with over 80 predicted proteins in total were annotated for transposition (GO:0032196) and DNA-mediated transposition (GO:0006313). In addition, exclusive gene clusters were identified with GO annotations for osmosensory signaling via phosphorelay pathway (GO:0007234), shikimate metabolic process (GO:0019632), antibiotic biosynthetic process (GO:0017000), regulation of DNA-templated transcription (GO: 0006355), DNA integration (GO:0015074) and N-terminal protein amino acid acetylation (GO:0006474). T. glycolicus contained two exclusive gene clusters with annotated molecular function, including gene clusters for nickel cation transmembrane transporter activity (GO:0015099) and metal ion binding (GO:0046872). Other assigned GO annotations of exclusive gene clusters comprised regulation of systemic acquired resistance (GO:0010112), transmembrane transport (GO:005505085), ethanolamine catabolic process (GO:00463336), regulation of DNA-templated transcription (GO: 0006355), polysaccharide biosynthetic process (GO:0000271), nickel cation transport (GO:0015675) and teichoic acid biosynthetic process (GO:0019350). Remaining exclusive gene clusters for the three Terrisporobacter genomes did not receive functional annotations.
A Functional annotation of the Terrisporobacter type strain genomes into COG categories. The y-axis shows the different COG categories and the x-axis the number of genes assigned to each category by eggNOG-mapper. The bar color indicates the number of genes for the three Terrisporobacter type strains T. mayombei (green), T. petrolearius (blue) and T. glycolicus (red). B Pan/core genome analysis of three Terrisporobacter type strains. The Venn diagram visualizes the exclusive and shared gene clusters for T. mayombei (green), T. petrolearius (blue) and T. glycolicus (red) created with OrthoVenn2. The graph visualizes the total number of gene clusters in the three genomes using the same color code.
KEGG pathway analysis assigned 49.9% of genes for T. mayombei, 51.3% of genes for T. petrolearius and 52.4% of genes for T. glycolicus into functional categories (Fig 2). Between the analyzed three Terrisporobacter type strains genes distributed uniformly to the different functional categories, with the exception of the category unclassified: genetic information processing. In this category, 113 genes were assigned from the genome of T. petrolearius, while only 32 genes derived from the genome of T. mayombei and 28 genes from the genome of T. glycolicus.
The y-axis shows the different functional KEGG categories and the x-axis the number of genes assigned to each category by BlastKOALA. The bar color indicates the number of genes assigned for the three Terrisporobacter type strains T. mayombei (green), T. petrolearius (blue) and T. glycolicus (red).
The functional annotations into COG and KEGG categories match with the described Terrisporobacter substrate spectrum specializing on the fermentation of carbohydrates and amino acids [3, 6, 7]. In both functional annotation approaches the gene distribution to the different categories was very similar between the three Terrisporobacter type strain genomes with the only exception of the increased gene number in T. petrolearius assigned to the categories of unclassified: genetic information processing (KEGG) or replication, recombination and repair (COG). This difference was also identified by pan/core genome analysis, with the four exclusive gene clusters encoded by T. petrolearius containing over 80 predicted proteins for transposition. In order to investigate the difference in transposase abundance the presence of IS elements in the genomes were analyzed with the ISESCAN pipeline (Fig 3). In total 94 IS elements were identified for T. petrolearius, while only four and six IS elements were detected in the T. mayombei and T. glycolicus genomes, respectively. All three genomes harbored IS elements from the IS21 family. One element of family IS66 was present in the genomes of T. petrolearius and T. glycolicus. The family IS607 was only detected with one element in T. glycolicus. The families IS3, IS110 and IS256 were only recorded in T. petrolearius and could be identified as the major source of the detected difference in transposase abundance between the three analyzed Terrisporobacter genomes. The families IS3 and IS256 contain transposases with the common acidic triad of Asp, Asp and Glu at the active site (DDE transposases). The IS110 family characteristically comprise of IS elements with transposases containing an Asp, Glu, Asp, Asp motif at the active site (DEDD transposase). These IS elements differ from classical DDE transposases as they do not contain terminal inverted repeats and upon integration do not produce flanking target DNA repeats [32]. The identified differences indicate an increased genome plasticity of T. petrolearius in comparison to the other two type strains [33].
The y-axis shows the different IS families and the x-axis the number of IS elements assigned to each IS family by ISESCAN. The bar color indicates the number of genes assigned for the three Terrisporobacter type strains T. mayombei (green), T. petrolearius (blue) and T. glycolicus (red).
For a more detailed metabolic investigation of the three Terrisporobacter type strain genomes, we analyzed and compared regarding the pathway modules of different KEGG categories (Fig 4). Within the category carbohydrate metabolism, the modules for glycolysis, pyruvate oxidation, pentose phosphate pathway, phosphoribosyl diphosphate (PRPP) biosynthesis, glycogen biosynthesis, nucleotide sugar biosynthesis from glucose and UDP-N-acetyl-D-glucosamine biosynthesis were complete in all three genomes. The modules for gluconeogenesis, citrate cycle, semi-phosphorylative Entner-Doudoroff pathway, D-galacturonate degradation, galactose degradation, glycogen degradation, nucleotide sugar biosynthesis from galactose and propanoyl-CoA metabolism were incomplete in every genome. The module for gluconeogenesis was identified as incomplete due to the lack of a phosphoenolpyruvate (PEP) carboxykinase (both ATP-and GTP-dependent) catalyzing the conversion of oxaloacetate into PEP. Nevertheless, gluconeogenesis remains functional as acetyl-CoA is carboxylated to pyruvate by a pyruvate ferredoxin oxidoreductase. This reaction links the Wood-Ljungdahl pathway to gluconeogenesis and the incomplete citrate cycle, thereby allowing the generation of building blocks for biosynthesis during autotrophic growth [34]. With an incomplete traditional, semi- and non-phosphorylative Entner-Doudoroff pathway, hexoses are metabolized only via glycolysis. Fermentation of pentoses such as xylose can occur via the non-oxidative pentose phosphate pathway and has been described for the type strains of T. mayombei and T. glycolicus [3, 6]. All three analyzed genomes contain an incomplete citrate cycle and are not able to degrade glycogen and galactose. In the category metabolism of cofactors and vitamins every genome contained complete gene sets for the modules of thiamine salvage, NAD biosynthesis, coenzyme A biosynthesis, C1-unit interconversion and the two cobalamin biosynthesis modules. The module for the biosynthesis of pyridoxal-P was only complete in T. mayombei and T. glycolicus, while this module was identified to be incomplete in T. petrolearius lacking the PdxT subunit of the Pyridoxal 5’-phosphate synthase. Both modules for the biosynthesis of biotin were only complete in T. mayombei and incomplete in the other two type strains. The modules for the biosynthesis of thiamine, riboflavin, pantothenate, pimeloyl-ACP, lipoic acid, tetrahydrofolate, tetrahydrobiopterin, L-threo-tetrahydrobiopterin, molybdenum cofactor, heme and tocopherol/tocotorienol were incomplete in all three Terrisporobacter type strains. Siroheme biosynthesis was found to be only complete in T. mayombei and T. glycolicus. The genome of T. petrolearius is missing genes encoding the sirohydrochlorin ferrochelatase (SirB), the glutamyl-tRNA reductase (HemA) and the glutamate-1-semialdehyde 2,1-aminomutase (HemL). However, all three genomes encode sulfite reductases requiring siroheme as essential prosthetic group [35]. The incomplete module for heme biosynthesis in all three genomes imply the dependence on heme acquisition from the environment or the host during potential pathogenesis [36]. The category amino acid metabolism contained complete gene sets for the modules threonine biosynthesis, cysteine biosynthesis, lysine biosynthesis, polyamine biosynthesis, histidine biosynthesis and histidine degradation in all three genomes. The modules for tyrosine and phenylalanine biosynthesis were only complete for T. mayombei and T. glycolicus. Furthermore, the modules for leucine and tryptophan biosynthesis were only complete in T. glycolicus. The modules for the biosynthesis of serine, methionine, valine, isoleucine, arginine, proline, gamma-aminobutyric acid GABA and the shikimate pathway were incomplete in all three Terrisporobacter type strains. This also applied to the modules for the urea cycle and GABA shunt. The identified incomplete modules in the categories metabolism of cofactors and vitamins and amino acid metabolism explain the observed dependence on the addition of vitamins and amino acids for significant growth of Terrisporobacter members [3, 6, 7]. The reconstruction of KEGG pathway modules showed auxotrophies for different vitamins/cofactors and amino acids in all three genomes. Furthermore, several species-specific vitamin and amino acid auxotrophies were identified. In the category energy metabolism complete modules were detected in every Terrisporobacter type strain for reductive acetyl-CoA pathway, phosphate acetyltransferase-acetate kinase pathway and F-type ATPase. The modules reductive pentose phosphate cycle, formaldehyde assimilation and assimilatory nitrate reduction were incomplete in all three Terrisporobacter strains. No complete modules were present for the category of biosynthesis of terpenoids and polyketides. The signature modules for beta-lactam resistance and acetogen were complete in all three Terrisporobacter type strains. The prerequisites for growth on H2 and CO2 through acetogenesis are fulfilled for every Terrisporobacter type strain analyzed, as shown by the complete modules for the reductive acetyl-CoA pathway (WLP), phosphate acetyltransferase-acetate kinase pathway and the complete signature module for acetogens.
The completeness of KEGG modules is given in percentage from the total amount of required blocks and the amount of detected blocks in a pathway module for the three Terrisporobacter type strains T. mayombei (1), T. petrolearius (2) and T. glycolicus (3). For metabolites with multiple KEGG modules only the most complete modules detected are shown. The legend for the color gradient is shown in the bottom right.
Pathway reconstruction for acetogenesis
The genes necessary to employ the WLP were present in all three type strains of the Terrisporobacter genus. The detected genes for the WLP are shown in Fig 5A with the corresponding locus tags for each genome. Overall, the genomic organization of the WLP genes in the Terrisporobacter type strains resembled the structure detected in the acetogen Clostridioides difficile [37]. The methyl-branch starts with the reversible hydrogen-dependent CO2 reductase complex (HDCR) using H2 as a direct electron donor for the reduction of CO2 to formate by a formate dehydrogenase (FDH). A potential HDCR-complex was encoded by a gene cluster consisting of a selenocysteine-containing formate dehydrogenase H (fdhF), two copies of the hydrogenase-4 component A (hyfA), a NADP-reducing hydrogenase subunit (hndD), a putative formate transporter (focA) and a sulfur carrier protein (fdhD). In C. difficile the HDCR-complex gene cluster did not contain a formate transporter and fdhD was located adjacent to the fdhF gene, while in Terrisporobacter it was located at the end of the gene cluster (Fig 5B). In addition to the HDCR-complex, the oxidation of H2 in Terrisporobacter can be performed by orthologs of the electron-bifurcating hydrogenase HydABC (TEMA_26770–26790, TEPE_39410–39430, TEGL_37760–37780). The activation of formate with tetrahydrofolate (THF) is carried out under consumption of one ATP by the formyl-THF synthetase (FTS) producing N10-formyl-THF, which is subsequently transformed to N5,N10-methenyl-THF via cyclization and dehydration by the methenyl-THF cyclohydrolase (MTC). The methylene-THF dehydrogenase (MTD) reduces N5,N10-methenyl-THF to N5,N10-methylene-THF, which is then further reduced to N5-methyl-THF by the methylene-THF reductase (MTR). Finally, the methyl-group is transferred to a corrinoid-iron-sulfur protein (CoFeSP) by the methyl-transferase (MT). In the carbonyl-branch the carbon monoxide dehydrogenase (CODH) reduces CO2 to a carbonyl-group using reduced ferredoxin. This carbonyl group is fused with the methyl-group from the methyl-branch and the cofactor coenzyme A (CoA), which is catalyzed by the acetyl-CoA synthetase (ACS) and releases CoFeSP. Acetyl-CoA is converted to acetyl phosphate (Acetyl-P) by a phosphotransacetylase (PTA) or alternatively by a phosphotransbutyrylase (PTB). The transfer of the phosphate group to ADP forming ATP is catalyzed by an acetate or alternatively butyrate kinase (ACK/BUK) and leads to the formation of the final product acetate. The genus Terrisporobacter is reported to produce only acetate as a fermentation product of the WLP (homoacetogenesis) [2]. The majority of genes involved in the WLP were organized as a gene cluster of approximately 18 kb. These gene clusters showed the same organization as in C. difficile or acetogenic bacteria belonging to the genus Clostridium (Fig 5C) [38]. Energy conservation is accomplished with the RNF-complex (rnfABCDEG; TEMA_15820–15870, TEPE_33980–33930, TEGL_15820–15870) by coupling the electron transfer from reduced ferredoxin to NAD with the translocation of protons or sodium ions across the cytoplasmic membrane. A membrane-bound ATP synthase can then use the created ion gradient to produce ATP [39]. The redox pool of NADH and reduced ferredoxin is linked to the cellular redox pool (NADPH) with a Sporomusa type Nfn (STN) electron-bifurcating transhydrogenase (stnABC, TEMA_09310–09330, TEPE_23120–23140, TEGL_20000–20020) [40].
A: Genome-based reconstruction of the Wood-Ljungdahl pathway for the three type strains of the Terrisporobacter genus. Genome locus tags (black) of the catalyzing enzymes (red) and the required genes (blue) are shown for the three type strains of the Terrisporobacter genus (TEMA, T. mayombei DSM 6539T; TEPE, T. petrolearius JCM 19845T; TEGL, T. glycolicus DSM 1288T). B: Genomic organization of the HDCR-complex gene cluster in the three Terrisporobacter type strains in comparison to the type strain of C. difficile DSM 1269T. The following gene abbreviations were used: fdhF, formate dehydrogenase H; hyfA, hydrogenase-4 component A; hndD, NADP-reducing hydrogenase; focA, putative formate transporter; fdhD, sulfur carrier protein. C: Genomic organization of the Wood-Ljungdahl gene cluster in the three Terrisporobacter type strains in comparison to the type strain of C. difficile DSM 1269T. The following gene abbreviations were used: acsA, anaerobic carbon-monoxide dehydrogenase catalytic subunit; acsF, carbon monoxide dehydrogenase accessory protein; fhs, formyl THF synthetase; fchA, methenyl THF cyclohydrolase; folD, bifunctional cyclohydrolase/dehydrogenase; metV, methylene THF reductase C-terminal catalytic subunit; metF, methylene THF reductase large subunit; lpdA, dihydrolipoyl dehydrogenase; cooC, carbon monoxide dehydrogenase accessory protein; acsD, CoFeSP small subunit; acsC, CoFeSP large subunit; acsE, methyl THF CoFeSP methyltransferase; acsB, carbon monoxide dehydrogenase/acetyl-CoA synthase subunit beta; gcvH, glycine cleavage system H protein; acsV, corrinoid activation/regeneration protein. The following enzyme abbreviations were used: HDCR, hydrogen-dependent CO2 reductase; FTS, formyl-THF synthetase; MTC, methenyl-THF cyclohydrolase; MTD, methylene-THF dehydrogenase; MTR, methylene-THF reductase; MT, methyl-transferase; CODH, carbon monoxide dehydrogenase; ACS, acetyl-CoA synthetase; PTA, phosphotransacetylase, ACK, acetate kinase. The legend for size and sequence identity of translated genes are shown at the bottom of B and C in kilobases (kb) and from 0% (white) to 100% (black).
Terrisporobacter plasmids
All three Terrisporobacter type strain genomes harbored a circular plasmid with an approximate size of 21 kb (pTM1, pTP1 and pTG1). T. petrolearius and T. mayombei contained an additional smaller circular plasmid of 6.7 kb (pTP2) and 8.2 kb (pTM2) in size. The annotation of the smaller plasmids pTP2 and pTM2 yielded only genes encoding hypothetical proteins. Hence, inferring possible functional roles of these plasmids was not possible. A circular plasmid of similar size (21.7 kb) as the 21 kb plasmids was also detected in the complete genome of T. hibernicus NCTC 14625T [10]. These plasmids appeared to be a shared trait in the Terrisporobacter genus, as they were present with an almost identical genetic organization in the genomes of every Terrisporobacter type strain (Fig 6). All three larger plasmids contained a full set of genes required for the proline-dependent reductive reactions of the Stickland reaction. The Stickland reaction describes the coupled oxidation and reduction of amino acids for the conservation of energy in anaerobic bacteria [41]. The plasmids encoded the proprotein prdA, which is most likely posttranslationally cleaved resulting in one small protein containing the pyruvoyl-group for binding of D-proline and one larger scaffold protein. The proline reductase was completed by the gene products of prdB and prdC. As in C. difficile, only prdB encodes a selenocysteine residue (Sec), while the prdC gene did not encode a Sec as typically present in other amino acid-fermenting bacteria such as Acetoanaerobium sticklandii or Clostridium botulinum [42]. The use of a selenoprotein requires the cotranslational insertion of Sec guided by a complex machinery [43, 44]. The Sec insertion complex was encoded by the Terrisporobacter plasmids by the three genes selA, selB, selD and the selenium-binding tRNA gene selC. In addition to genes for proline reduction, the plasmids contained genes for the partitioning proteins ParA and ParB, a DNA-invertase (Hin), four hypothetical proteins and a putative cysteine desulfurase (Csd). The plasmid of T. mayombei additionally encoded a protein annotated as cold shock protein (Csp). Genes for the reduction of glycine/sarcosine/betaine and the oxidative branch of the Stickland reaction were all encoded chromosomally. This also applied to the proline racemase (PrdF) required for the transformation of amino acids from the L to the D stereoisomer (TEMA_04260 + 37640/TEPE_08380/ TEGL_07940) and the membrane-bound protein PrdG (TEMA_37630 + 37650/TEPE_08370 + 08390/ TEGL_07930 + 07950). Orthologs of the putative DNA-binding protein PrdX and the sigma-54-dependent transcriptional activator protein PrdR were not identified.
The plasmids encode a proline-dependent gene cluster for the reductive branch of the Stickland reaction. The legend for the size and the sequence identity of translated genes is shown at the bottom in kilobases (kb) and from 0% (white) to 100% (black). The following genes are encoded: selC, tRNA-Sec; prdD, proline reductase cluster protein PrdD; prdE; proline reductase cluster protein PrdE; csd, putative cysteine desulfurase; lysR, lysR family transcriptional regulator; selB, selenocysteine-specific elongation factor; selA, L-seryl-tRNA (Sec) selenium transferase; selD, selenide water dikinase; parA, parA family protein; hin, DNA-invertase; csp, cold shock protein; parB, DNA primase; prdB, D-proline reductase subunit gamma; prdA, D-proline reductase proprotein PrdA; prdC, proline reductase-associated electron transfer protein; hyp, hypothetical protein.
Virulence factor screening
Multiple potential virulence factors were detected in the genomes of the three Terrisporobacter type strains comprising the VF classes adherence, regulation, toxin, antiphagocytosis, cell surface components, copper uptake, enzyme, glycosylation systems, immune evasion, iron uptake, others, secretion system, serum resistance and immune evasion and stress adaptation (Table 2). Six virulence factors for cell adherence were detected. Genes encoding a fibronectin-binding protein (fbpA/fbp68) and a heat shock protein (groEL) were found in every Terrisporobacter type strain. These proteins with adhesive function are characteristically present in pathogenic Gram-positive bacteria such as C. difficile or Clostridium perfringens [45–47]. Furthermore, the Terrisporobacter type strain genomes encoded proteins for cell adherence, which are absent in genomes of other pathogenic members of the genus Clostridium. These include homologs of the adhesion protein Lap from the genus Listeria, the polar flagella from the genus Aeromonas and type IV pili from the genus Neisseria [48–50]. The genes cheA/cheY, lisR/lisK and virR/virS from the VF class regulation are virulence factors in Listeria monocytogenes and were also detected in the Terrisporobacter type strains [51–53]. Genes for the production of the toxins hemolysin and cytolysin involved in host cell lysis were detected in every Terrisporobacter strain. Furthermore, several virulence factors were detected associated to capsule formation involved in antiphagocytosis and immune evasion. For the verified Terrisporobacter pathogen T. glycolicus the ddhA gene was detected encoding an O-antigen essential for virulence in Yersinia enterocolitica [54]. Homologs of iron uptake proteins have been identified with similarity to proteins from the genus Haemophilus (HemB/L) and Vibrio (VctC), which matched with the incomplete heme biosynthesis encoded by all three genomes. Overall, the abundance and diversity of virulence factors detected in the three Terrisporobacter type strains underline the pathogenic potential harbored by this bacterial genus.
Phylogenomic analysis
The complete genome sequences were compared to available genome sequences of the Terrisporobacter genus by average nucleotide identity analysis based on a MUMmer alignment (ANIm) (Fig 7). The genomes of the type strains T. glycolicus DSM 1288T and T. mayombei DSM 6539T showed the highest identity to each other (93.5%). This value was below the species delineation threshold of 95% and supported the classification as separate Terrisporobacter species [55]. Using the same threshold, T. glycolicus DSM 1288T and T. petrolearius JCM 19845T also clustered as two distinct species with an identity of 93.1%. However, the genome T. petrolearius JCM 19845T showed the highest identity to the genome sequences of non-type T. glycolicus strains MGYG-HGUT-00005 (96.3%), KPPR-9 (96.4%), WW3900 (96.3%), UBA8115 (95.8%), FS03 (96.3%), T. petrolearius UN03-225 (96.4%) and T. hibernicus NCTC 14625T (96.4%). These identities were above the species identity threshold of 95%, suggesting that these genomes belonged to the same species, namely T. petrolearius. Mitchell et al. reported a 67.4% d4 dDDH value of the complete T. hibernicus type strain genome to the draft genome of T. petrolearius JCM 19845T calculated by the Type Strain Genome Server (TYGS) [10, 31]. TYGS submission of the complete genome of T. petrolearius JCM 19845T yielded a d4 dDDH value of 69.2% to the complete T. hibernicus genome. However, TYGS recommends the use of the dDDH values d0 and d6 for the comparison of complete genomes, as they also take the genome length into consideration in their calculations [56]. The dDDH values for d0 and d6 between the two complete genomes of T. petrolearius JCM 19845T and T. hibernicus were 72.2% and 74.1%, respectively. These dDDH values were above the generally applied species threshold of 70% and thereby further supported the classification of the isolate T. hibernicus as member of the T. petrolearius species. Our phylogenomic analysis based on complete type strain genomes suggested the existence of subspecies clusters within the T. petrolearius species rather than two distinct species. With 99.6% identity between the genome sequences of T. “othiniensis” 08–306576 and T. “muris” DSM 29186 both names describe isolates of the same species as already described by Afriza et al. [9]. Both isolates were not published under the ICNP as T. “othiniensis” was not deposited in a culture collection and T. “muris” is only deposited in one culture collection. Hence, no type strain and valid name description are currently available for these Terrisporobacter strains. In addition, the genome sequence of the isolate T. mayombei MSK4.1 showed an identity of 99.7% to the genomes of T. “othiniensis” and T. “muris” and hence represents another isolate of this species. The genome sequence of T. mayombei ban5-GNNATD2-55 isolate from human feces was below the species delineation threshold of 95% and thereby represented another undescribed Terrisporobacter species. The ANIm analysis of metagenomic assembled genomes from the Terrisporobacter genus indicated the existence of at least three more Terrisporobacter species without available isolates.
The legend for sequence identities is shown in the top left corner. Sequence identities higher than 95% are shown in red, while lower values are colored grey to blue. The three type strains of the genomes sequenced from the Terrisporobacter genus in this study are highlighted in green. Genome sequences derived from metagenomic assemblies are indicated by MAG in parentheses.
Conclusions
Complete genomes for the three Terrisporobacter type strains of T. mayombei, T. petrolearius and T. glycolicus were obtained and revealed a Terrisporobacter characteristic 21 kb plasmid encoding a gene cluster for the reductive proline-dependent branch of the Stickland reaction. Functional annotations of the genomes showed species-specific auxotrophies for amino acids and vitamins. The genetic prerequisites for acetogenesis were given in all three Terrisporobacter type strains and the WLP could successfully be reconstructed. Our metabolic analysis supports future research for biotechnological applications of acetogenic strains from the genus Terrisporobacter. The detection of multiple potential virulence factors also demonstrated a pathogenic potential harbored by the genus Terrisporobacter. Phylogenomic analysis of available Terrisporobacter genomes suggested a reclassification of most isolates now termed T. glycolicus into T. petrolearius.
Supporting information
S1 Table. Kits and program versions for Nanopore sequencing.
https://doi.org/10.1371/journal.pone.0290128.s001
(DOCX)
S2 Table. ANIm percentage identity and alignment length values.
ANIm percentage identity values (white, in %) and corresponding alignment lengths (grey, in bp) for available Terrisporobacter genomes at NCBI Genbank.
https://doi.org/10.1371/journal.pone.0290128.s002
(DOCX)
Acknowledgments
We thank Melanie Heinemann for technical assistance and Avril von Hoyningen-Huene for proofreading.
References
- 1. Wang D, Liang Q, Chu N, Zeng RJ, Jiang Y. Deciphering mixotrophic microbial electrosynthesis with shifting product spectrum by genome-centric metagenomics. Chem Eng J. 2023;451:139010.
- 2. Küsel K, Karnholz A, Trinkwalter T, Devereux R, Acker G, Drake HL. Physiological Ecology of Clostridium glycolicum RD-1, an Aerotolerant Acetogen Isolated from Sea Grass Roots. Appl Environ Microbiol. 2001;67(10):4734–41. pmid:11571179
- 3. Kane MD, Brauman A, Breznak JA. Clostridium mayombei sp. nov., an H2/CO2 acetogenic bacterium from the gut of the African soil-feeding termite, Cubitermes speciosus. Arch Microbiol. 1991;156(2):99–104.
- 4. Gerritsen J, Fuentes S, Grievink W, van Niftrik L, Tindall BJ, Timmerman HM, et al. Characterization of Romboutsia ilealis gen. nov., sp. nov., isolated from the gastro-intestinal tract of a rat, and proposal for the reclassification of five closely related members of the genus Clostridium into the genera Romboutsia gen. nov., Intestinib. Int J Syst Evol Microbiol. 2014 May 1;64(5):1600–16.
- 5. Ohwaki K, Hungate RE. Hydrogen utilization by clostridia in sewage sludge. Appl Environ Microbiol. 1977;33(6):1270–4. pmid:879782
- 6. Gaston LW, Stadtman ER. FERMENTATION OF ETHYLENE GLYCOL BY CLOSTRIDIUM GLYCOLICUM, SP. N. J Bacteriol. 1963;85(2):356–62.
- 7. Deng Y, Guo X, Wang Y, He M, Ma K, Wang H, et al. Terrisporobacter petrolearius sp. nov., isolated from an oilfield petroleum reservoir. Int J Syst Evol Microbiol. 2015;65(10):3522–6. pmid:26297478
- 8. Lund LC, Sydenham TV, Høgh SV, Skov M, Kemp M, Justesen US. Draft Genome Sequence of “Terrisporobacter othiniensis” Isolated from a Blood Culture from a Human Patient. Genome Announc. 2015;3(2):e00042–15. pmid:25744994
- 9. Afrizal A, Jennings SAV, Hitch TCA, Riedel T, Basic M, Panyot A, et al. Enhanced cultured diversity of the mouse gut microbiota enables custom-made synthetic communities. Cell Host Microbe. 2022;30(11):1630–45. pmid:36208631
- 10. Mitchell M, Nguyen S V, Connor M, Fairley DJ, Donoghue O, Marshall H, et al. Terrisporobacter hibernicus sp. nov., isolated from bovine faeces in Northern Ireland. Int J Syst Evol Microbiol. 2023;73(1):005667. pmid:36748477
- 11. Chen T, Xie G, Mi J, Wen X, Cao Z, Ma B, et al. Recovery of the Structure and Function of the Pig Manure Bacterial Community after Enrofloxacin Exposure. Microbiol Spectr. 2022;10(3):e02004–21. pmid:35604139
- 12. Cheng MP, Domingo MC, Lévesque S, Yansouni CP. A case report of a deep surgical site infection with Terrisporobacter glycolicus/T. Mayombei and review of the literature. BMC Infect Dis. 2016;16:529. pmid:27686579
- 13. Li Y, Li X, Sun Y, Zhao X, Li Y. Cathodic microbial community adaptation to the removal of chlorinated herbicide in soil microbial fuel cells. Environ Sci Pollut Res. 2018;25:16900–12.
- 14. Bolger AM, Lohse M, Usadel B. Trimmomatic: A flexible trimmer for Illumina sequence data. Bioinformatics. 2014;30(15):2114–20. pmid:24695404
- 15. Wick RR, Judd LM, Gorrie CL, Holt KE. Completing bacterial genome assemblies with multiplex MinION sequencing. Microb Genomics. 2017;3(10):e000132. pmid:29177090
- 16. Wick RR, Judd LM, Gorrie CL, Holt KE. Unicycler: Resolving bacterial genome assemblies from short and long sequencing reads. PLoS Comput Biol. 2017;13(6):e1005595. pmid:28594827
- 17. Wick RR, Schultz MB, Zobel J, Holt KE. Bandage: interactive visualization of de novo genome assemblies. Bioinformatics. 2015;31(20):3350–2. pmid:26099265
- 18. Seemann T. Prokka: rapid prokaryotic genome annotation. Bioinformatics. 2014;30(14):2068–9. pmid:24642063
- 19. Manni M, Berkeley MR, Seppey M, Simão FA, Zdobnov EM. BUSCO Update: Novel and Streamlined Workflows along with Broader and Deeper Phylogenetic Coverage for Scoring of Eukaryotic, Prokaryotic, and Viral Genomes. Mol Biol Evol. 2021;38(10):4647–54. pmid:34320186
- 20. Kanehisa M, Sato Y, Morishima K. BlastKOALA and GhostKOALA: KEGG Tools for Functional Characterization of Genome and Metagenome Sequences. J Mol Biol. 2016;428(4):726–31. pmid:26585406
- 21. Cantalapiedra CP, Hernández-Plaza A, Letunic I, Bork P, Huerta-Cepas J. eggNOG-mapper v2: Functional Annotation, Orthology Assignments, and Domain Prediction at the Metagenomic Scale. Mol Biol Evol. 2021;38(12):5825–9. pmid:34597405
- 22. Huerta-Cepas J, Szklarczyk D, Heller D, Hernández-Plaza A, Forslund SK, Cook H, et al. eggNOG 5.0: a hierarchical, functionally and phylogenetically annotated orthology resource based on 5090 organisms and 2502 viruses. Nucleic Acids Res. 2019;47(D1):D309–14. pmid:30418610
- 23.
R Core Team. R: A language and environment for statistical computing. R Found Stat Comput [Internet]. 2018; Available from: https://www.r-project.org/
- 24.
Wickham H. ggplot2: Elegant Graphics for Data Analysis. Springer-Verlag New York. 2016.
- 25. Xu L, Dong Z, Fang L, Luo Y, Wei Z, Guo H, et al. OrthoVenn2: a web server for whole-genome comparison and annotation of orthologous clusters across multiple species. Nucleic Acids Res. 2019;47(W1):W52–8. pmid:31053848
- 26. Xie Z, Tang H. ISEScan: automated identification of insertion sequence elements in prokaryotic genomes. Bioinformatics. 2017;33(21):3340–7. pmid:29077810
- 27. Lechner M, Findeiß S, Steiner L, Marz M, Stadler PF, Prohaska SJ. Proteinortho: Detection of (Co-) orthologs in large-scale analysis. BMC Bioinformatics. 2011;12:124. pmid:21526987
- 28. Gilchrist CLM, Chooi YH. clinker & clustermap.js: automatic generation of gene cluster comparison figures. Bioinformatics. 2021;37(16):2473–5. pmid:33459763
- 29. Liu B, Zheng D, Jin Q, Chen L, Yang J. VFDB 2019: a comparative pathogenomic platform with an interactive web interface. Nucleic Acids Res. 2019;47(D1):D687–92. pmid:30395255
- 30. Pritchard L, Glover RH, Humphris S, Elphinstone JG, Toth IK. Genomics and taxonomy in diagnostics for food security: soft-rotting enterobacterial plant pathogens. Anal Methods. 2016;8(1):12–24.
- 31. Meier-Kolthoff JP, Göker M. TYGS is an automated high-throughput platform for state-of-the-art genome-based taxonomy. Nat Commun. 2019;10:2182. pmid:31097708
- 32. Siguier P, Gourbeyre E, Chandler M. Bacterial insertion sequences: their genomic impact and diversity. FEMS Microbiol Rev. 2014;38(5):865–91. pmid:24499397
- 33. Bennett PM. Genome Plasticity. In: Woodford N, Johnson AP, editors. Genomics, Proteomics, and Clinical Bacteriology: Methods and Reviews. Totowa, NJ; 2004. p. 71–113.
- 34. Furdui C, Ragsdale SW. The Role of Pyruvate Ferredoxin Oxidoreductase in Pyruvate Synthesis during Autotrophic Growth by the Wood-Ljungdahl Pathway. J Biol Chem. 2000;275(37):28494–9. pmid:10878009
- 35. Crane BR, Getzoff ED. The relationship between structure and function for the sulfite reductases. Curr Opin Struct Biol. 1996;6(6):744–56. pmid:8994874
- 36. Choby JE, Skaar EP. Heme Synthesis and Acquisition in Bacterial Pathogens. J Mol Biol. 2016;428(17):3408–28. pmid:27019298
- 37. Gencic S, Grahame DA. Diverse Energy-Conserving Pathways in Clostridium difficile: Growth in the Absence of Amino Acid Stickland Acceptors and the Role of the Wood-Ljungdahl Pathway. J Bacteriol. 2020;202(20):e00233–20. pmid:32967909
- 38. Poehlein A, Cebulla M, Ilg MM, Bengelsdorf FR, Schiel-Bengelsdorf B, Whited G, et al. The Complete Genome Sequence of Clostridium aceticum: a Missing Link between Rnf- and Cytochrome-Containing Autotrophic Acetogens. MBio. 2015;6(5):e01168–15. pmid:26350967
- 39. Schuchmann K, Müller V. Autotrophy at the thermodynamic limit of life: a model for energy conservation in acetogenic bacteria. Nat Rev Microbiol. 2014;12(12):809–21. pmid:25383604
- 40. Kremp F, Roth J, Müller V. The Sporomusa type Nfn is a novel type of electron-bifurcating transhydrogenase that links the redox pools in acetogenic bacteria. Sci Rep. 2020;10:14872. pmid:32913242
- 41. Stickland LH. Studies in the metabolism of the strict anaerobes (genus Clostridium). Biochem J. 1935;29(4):889–98.
- 42. Fonknechten N, Chaussonnerie S, Tricot S, Lajus A, Andreesen JR, Perchat N, et al. Clostridium sticklandii, a specialist in amino acid degradation:revisiting its metabolism through its genome sequence. BMC Genomics. 2010;11:555. pmid:20937090
- 43. Gonzalez-Flores JN, Shetty SP, Dubey A, Copeland PR. The molecular biology of selenocysteine. Biomol Concepts. 2013;4(4):349–65. pmid:25436585
- 44. Zinoni F, Birkmann A, Stadtman TC, Böck A. Nucleotide sequence and expression of the selenocysteine-containing polypeptide of formate dehydrogenase (formate-hydrogen-lyase-linked) from Escherichia coli. Proc Natl Acad Sci. 1986;83(13):4650–4. pmid:2941757
- 45. Hennequin C, Janoir C, Barc M claude, Collignon A, Karjalainen T. Identification and characterization of a fibronectin-binding protein from Clostridium difficile. Microbiology. 2003;149(10):2779–87.
- 46. Collignon A, Bourlioux P, Waligora-Dupriet AJ, Karjalainen T, Barc MC, Hennequin C, et al. GroEL (Hsp60) of Clostridium difficile is involved in cell adherence. Microbiology. 2001;147(1):87–96. pmid:11160803
- 47. Mehdizadeh Gohari I, A. Navarro M, Li J, Shrestha A, Uzal F, A. McClane B. Pathogenicity and virulence of Clostridium perfringens. Virulence. 2021;12(1):723–53. pmid:33843463
- 48. Wampler JL, Kim KP, Jaradat Z, Bhunia AK. Heat Shock Protein 60 Acts as a Receptor for the Listeria Adhesion Protein in Caco-2 Cells. Infect Immun. 2004;72(2):931–6. pmid:14742538
- 49. Tomás JM. The Main Aeromonas Pathogenic Factors. ISRN Microbiol. 2012;2012:256261. pmid:23724321
- 50. Thanassi DG, Bliska JB, Christie PJ. Surface organelles assembled by secretion systems of Gram-negative bacteria: diversity in structure and function. FEMS Microbiol Rev. 2012;36(6):1046–82. pmid:22545799
- 51. Dons L, Eriksson E, Jin Y, Rottenberg ME, Kristensson K, Larsen CN, et al. Role of Flagellin and the Two-Component CheA/CheY System of Listeria monocytogenes in Host Cell Invasion and Virulence. Infect Immun. 2004;72(6):3237–44. pmid:15155625
- 52. Cotter PD, Guinane CM, Hill C. The LisRK Signal Transduction System Determines the Sensitivity of Listeria monocytogenes to Nisin and Cephalosporins. Antimicrob Agents Chemother. 2002;46(9):2784–90. pmid:12183229
- 53. Mandin P, Fsihi H, Dussurget O, Vergassola M, Milohanic E, Toledo-Arana A, et al. VirR, a response regulator critical for Listeria monocytogenes virulence. Mol Microbiol. 2005;57(5):1367–80. pmid:16102006
- 54. Bengoechea JA, Najdenski H, Skurnik M. Lipopolysaccharide O antigen status of Yersinia enterocolitica O:8 is essential for virulence and absence of O antigen affects the expression of other Yersinia virulence factors. Mol Microbiol. 2004;52(2):451–69.
- 55. Goris J, Konstantinidis KT, Klappenbach JA, Coenye T, Vandamme P, Tiedje JM. DNA–DNA hybridization values and their relationship to whole-genome sequence similarities. Int J Syst Evol Microbiol. 2007;57(1):81–91. pmid:17220447
- 56. Meier-Kolthoff JP, Auch AF, Klenk HP, Göker M. Genome sequence-based species delimitation with confidence intervals and improved distance functions. BMC Bioinformatics. 2013;14:60. pmid:23432962