- Research article
- Open Access
Complete genome sequence of Corynebacterium variabile DSM 44702 isolated from the surface of smear-ripened cheeses and insights into cheese ripening and flavor generation
BMC Genomicsvolume 12, Article number: 545 (2011)
Corynebacterium variabile is part of the complex microflora on the surface of smear-ripened cheeses and contributes to the development of flavor and textural properties during cheese ripening. Still little is known about the metabolic processes and microbial interactions during the production of smear-ripened cheeses. Therefore, the gene repertoire contributing to the lifestyle of the cheese isolate C. variabile DSM 44702 was deduced from the complete genome sequence to get a better understanding of this industrial process.
The chromosome of C. variabile DSM 44702 is composed of 3, 433, 007 bp and contains 3, 071 protein-coding regions. A comparative analysis of this gene repertoire with that of other corynebacteria detected 1, 534 predicted genes to be specific for the cheese isolate. These genes might contribute to distinct metabolic capabilities of C. variabile, as several of them are associated with metabolic functions in cheese habitats by playing roles in the utilization of alternative carbon and sulphur sources, in amino acid metabolism, and fatty acid degradation. Relevant C. variabile genes confer the capability to catabolize gluconate, lactate, propionate, taurine, and gamma-aminobutyric acid and to utilize external caseins. In addition, C. variabile is equipped with several siderophore biosynthesis gene clusters for iron acquisition and an exceptional repertoire of AraC-regulated iron uptake systems. Moreover, C. variabile can produce acetoin, butanediol, and methanethiol, which are important flavor compounds in smear-ripened cheeses.
The genome sequence of C. variabile provides detailed insights into the distinct metabolic features of this bacterium, implying a strong adaption to the iron-depleted cheese surface habitat. By combining in silico data obtained from the genome annotation with previous experimental knowledge, occasional observations on genes that are involved in the complex metabolic capacity of C. variabile were integrated into a global view on the lifestyle of this species.
Cheese is one of the oldest dairy products and generally based on the addition of rennet and lactic acid bacteria to milk of cows, sheep, goats, or buffalos. Nowadays, cheese is also the most diverse group of dairy products with hundreds of varieties that are capable of being differentiated mostly by their type of ripening . Bacterial smear-ripened cheeses, such as Appenzeller, Gubbeen, Limburger, Livarot, Munster, and Tilsit, are characterized by the development of a viscous, red-orange smear on the surface of the cheese body during ripening. This smear is a microbial mat composed of various species of yeasts and bacteria. Their combined metabolic activities are mainly responsible for the typical flavor and textural properties of this type of cheese [2, 3]. The microbiology of the red-orange smear is poorly understood so far. Generally, acid-tolerant yeasts grow during the first days of ripening of smear-ripened cheeses and metabolize the lactic acid produced by the lactic starter cultures to CO2 and H2O, thereby increasing the pH and producing growth factors, such as pantothenic acid, that permit the growth of Gram-positive and Gram-negative bacteria [4, 5]. Yeast species can also deaminate amino acids to ketoacids and NH3, which leads to deacidification of the cheese curd in such a way that bacteria are able to grow [6, 7]. The bacterial microflora dominating the later stages of ripening of smear-ripened cheeses is composed of salt-tolerant micrococci, staphylococci, and corynebacteria [2, 8]. In the past, it was supposed that the coryneform species Brevibacterium linens represents the major organism on the surface of smear-ripened cheeses , but recent investigations emphasized the importance of other coryneform bacteria to the ripening process and identified Arthrobacter nicotianae, Arthrobacter arilaitensis, Brevibacterium ammoniagenes, Microbacterium gubbeenense, Rhodococcus fascians, and members of the genus Corynebacterium (Corynebacterium ammoniagenes, Corynebacterium casei, and Corynebacterium variabile) dominating the microflora of Gubbeen, Tilsit, and other smear-ripened cheeses [10–14].
The genus Corynebacterium includes (besides numerous human and animal pathogens) a diverse collection of non-pathogenic species that have been detected in a wide variety of habitats, such as soil, plant material, waste water, and dairy products . C. casei, for instance, belongs to the main lineage of the genus Corynebacterium [4, 16] and was identified as the dominant species on the surface of a Gubbeen-type Irish farmhouse smear-ripened cheese by taxonomic analyses [12, 17]. The abundance of C. casei on the cheese surface is closely followed by C. variabile, which is taxonomically located in a distinct subline (cluster 3) of the genus Corynebacterium, with Corynebacterium jeikeium and Corynebacterium urealyticum as the most prominent phylogenetic relatives [18, 19]. C. casei and C. variabile are salt-tolerant and able to grow in the presence of 8.0% NaCl and at pH values below 4.9 . Both corynebacteria also metabolize the lactate produced by the starter bacteria, implying that they can grow from the beginning of cheese ripening and may therefore not depend on significant yeast growth during the initial stages of the ripening process .
In the present study, we characterize the complete genome sequence of C. variabile DSM 44702 (formerly Corynebacterium mooreparkense) that was originally isolated from the smear-ripened cheese Gubbeen and identified by 16S rDNA sequencing and DNA-DNA hybridization studies . It was initially described as the type strain of a separate species named C. mooreparkense because the organism was isolated at the Dairy Products Research Centre in Moorepark, Ireland . However, additional molecular taxonomic studies demonstrated considerable similarities of all C. mooreparkense isolates with the species C. variabile at the levels of 16S rDNA gene sequence and DNA-DNA relatedness . It was therefore concluded that the name C. mooreparkense is a later heterotypic synonym of C. variabile . In the present study, the annotation of the complete genome sequence and the deduced genetic repertoire of C. variabile DSM 44702 provide detailed insights into the lifestyle and metabolic features of this species, which is strongly adapted to the cheese surface habitat, where it contributes to flavor and texture of the final product.
Pyrosequencing and annotation of the C. variabile DSM 44702 genome
The DNA sequence of the C. variabile DSM 44702 chromosome was determined by a whole-genome shotgun approach using pyrosequencing. A quarter of a sequencing run with the Genome Sequencer FLX Instrument and Titanium chemistry yielded 253, 845 reads and 97, 765, 747 bases that were assembled into 93 large contigs (≥500 bases) and 41 small contigs, indicating numerous repetitive elements and insertion sequences in the C. variabile DSM 44702 genome. The remaining gaps in the chromosomal sequence were closed by PCR strategies, supported by the r2cat tool  and the Consed program . The final assembly of the DNA sequences yielded a circular chromosome with a size of 3, 433, 007 bp (Figure 1). A 28-fold coverage was thus obtained by pyrosequencing when considering the final size of the C. variabile DSM 44702 chromosome. The mean G+C content of the C. variabile DSM 44702 genome is 67.15%, which is significantly higher than the value (60%) initially determined by a reverse-phase HPLC method . The annotation of the C. variabile genome sequence was performed with the GenDB software system  and resulted in the detection of 3, 071 protein-coding regions. Furthermore, 59 tRNA genes were predicted by the tRNAscan-SE program  and six rrn operons were detected on the leading strands of the C. variabile chromosome (Figure 2).
General architecture of the C. variabile DSM 44702 genome
The plot of the calculated G/C skew [(G-C)/(G+C)] indicated a typical bi-directional replication mechanism of the C. variabile chromosome (Figure 1). According to the presence and distribution of six conserved DnaA boxes, the origin of chromosomal replication (oriC) is located downstream of the dnaA coding region . Moreover, the biased distribution of architecture imparting sequences (AIMS) on the leading strands of the C. variabile chromosome indicated the presence of a dif region  at position 1, 603 kb on the chromosomal map, dividing the chromosome of C. variabile into two replichores of nearly similar size (Figure 2). Due to the highly biased presence of AIMS on the leading strands of the C. variabile chromosome (Figure 2), we also examined the gene-strand bias in this strain. In total, 59.6% of the predicted gene repertoire is located on the leading strands of the C. variabile chromosome, revealing a moderate gene-strand bias in this species. However, it has been suggested previously that essentiality is the driving force for gene-strand bias in bacterial genomes . We therefore calculated also the gene-strand bias for candidate essential genes in C. variabile. For this purpose, the set of 658 candidate essential genes detected in the genome of Corynebacterium glutamicum R by high-density transposon mutagenesis [28, 29] was compared to the predicted gene repertoire of C. variabile by reciprocal BLASTP matches with the EDGAR software . This comparative content analysis revealed 373 of the candidate essential genes from C. glutamicum R having orthologs in the C. variabile genome. The majority of these genes (80.4%) are located on the leading strands of the C. variabile chromosome, with 68% of all candidate essential genes being located on the left replichore.
Synteny analysis by reciprocal best BLASTP hits  revealed a highly conserved order of orthologous genes between the chromosomes of C. variabile DSM 44702 and C. jeikeium K411, with the exception of a translocated DNA region comprising a segment of approximately 125 kb in size (Figure 3). This result is consistent with the very close phylogenetic relationship of both species that belong to the same subline (cluster 3) of the genus Corynebacterium [18, 19], and the general observation that genetic rearrangements are rare in corynebacterial genomes [15, 31]. The genome of C. variabile DSM 44702 thus contains the previously predicted inversions around dif and oriC, which are both characteristic genomic features of cluster 3 species in the genus Corynebacterium . However, the synteny between both genomes is interrupted due to the presence of additional genes in the C. variabile chromosome that were functionally assigned to a phage island, several transposon hotspots, and species-specific gene islands, the latter containing for instance large gene clusters involved in iron acquisition and siderophore biosynthesis (Figure 3). Some of these species-specific gene islands show a significant deviation from the mean G+C content of the C. variabile chromosome and are flanked by several insertion sequences, suggesting that horizontal gene transfer contributed to the evolution of the current gene repertoire of strain DSM 44702 (Figure 1).
The phage island of the C. variabile chromosome has a size of about 48.3 kb and comprises 60 genes (CVAR_0826-CVAR_0886). This genomic region is locted between two tRNAGly genes that may represent the integration site of a putative phage. This view is supported by the presence of 50-bp direct repeats, which are part of the tRNA genes and flank the prophage-like region. Most genes (39 out of 60) of the phage island encode hypothetical proteins of unknown function, whereas others encode enzymes involved in phage DNA replication, recombination, and repair. The structural proteins of the putative phage revealed similarities to the tail structure of Rhodococcus phage ReqiPine5  and to the capsid structure of Listeria phage A006 .
The calculated reciprocal best BLASTP hits  were moreover used to compare the predicted proteome of C. variabile DSM 44702 with the complete set of proteins encoded in the genomes of C. jeikeium K411  and C. urealyticum DSM 7109 , which are close taxonomic relatives with known genome sequences. This comparative content analysis at proteome level revealed that 1, 120 proteins (36.5%) of C. variabile share homologs in the genomes of C. jeikeium and C. urealyticum (Figure 4A). According to this comparative data, C. variabile contains 1, 699 proteins with no homologous counterparts in the proteomes of the taxonomic relatives C. jeikeium and C. urealyticum. However, it has to be considered that the three genomes of the taxonomically closely related species differ considerably in their sizes as well as in the numbers of protein-coding regions [31, 34]. Thus, the 3.31 Mb sequence data of the soil bacterium C. glutamicum R  was incorporated into this comparative analysis, revealing a comparably extensive set of 1, 534 genes that are specific for C. variabile when compared to the selected corynebacteria (Figure 4B). This calculation indicates the presence of an extensive and still unexplored variability in the gene equipment of non-pathogenic corynebacteria. This set of genes might contribute to the characteristic features of C. variabile that define its distinct metabolic capabilities.
In the following sections, we present the detailed analysis of these metabolic features of C. variabile DSM 44702 and combine them with data deduced from the predicted transcriptional regulatory repertoire. This approach in evaluating the complete genome sequence of C. variabile DSM 44702 revealed a collection of relevant genes contributing to the lifestyle of this species and their integration into a transcriptional gene-regulatory network.
General metabolic features of C. variabile DSM 44702 deduced from the complete genome sequence
The experimental work on the taxonomic description of C. variabile DSM 44702 as the type strain of the former species C. mooreparkense revealed the ability of this cheese isolate to utilize glucose, fructose, mannose, and ribose as carbon and energy sources . Bioinformatic analysis of the gene repertoire assigned to the central carbohydrate metabolism detected a complete set of genes involved in glycolysis, gluconeogenesis, and the pentose phosphate pathway, as well as the presence of glucose- and fructose-specific components (ptsG and ptsF) of the bacterial phosphoenolpyruvate:carbohydrate phosphotransferase system (ptsH and ptsI) and at least one ABC-type transport system (sugABCD) for sugar uptake (Additional file 1). The stcRS genes of a two-component signal transduction system are located directly upstream of the putative sugABCD operon and might be involved in the transcriptional control of this sugar importer. Glucokinase and ribokinase genes (glk and rbsK) are present in C. variabile DSM 44702, allowing the conversion of "free" sugars into phosphorylated central pathway intermediates (Additional file 1).
According to the genome annotation, C. variabile DSM 44702 can utilize gluconate that is imported by gluconate permease (gntP) and converted to 6-phosphogluconate by gluconate kinase (gntK) (Additional file 1). C. variabile DSM 44702 can moreover channel propionate via its methylcitrate cycle genes  into the tricarboxylic acid cycle (Additional file 1). Calcium propionate occurs naturally in many dairy products including cheese, and some types of cheese contain as much as 1% of natural propionic acid . Propionate is imported into C. variabile DSM 44702 by a monocarboxylic acid transporter (mctC) . This secondary transport system can probably also import pyruvate that exists naturally in cheese in small amounts [37, 38].
The tricarboxylic acid cycle of C. variabile DSM 44702 and the glyoxylate bypass are complete, with the exception of the sucCD genes encoding two subunits of succinyl-CoA synthetase. The absence of these genes was also observed in other Corynebacterium species, such as C. diphtheriae , C. jeikeium , and C. urealyticum . It has been suggested that a succinyl-CoA:CoA transferase (cat1) catalyzes the conversion of succinyl-CoA to succinate instead [39, 40]. Candidate succinyl-CoA:CoA transferases are encoded by the neighboring paralogs cat1 (CVAR_0586) and cat2 (CVAR_0587) in the C. variabile DSM 44702 chromosome. Anaplerotic reactions in C. variabile DSM 44702 are accomplished by pyruvate carboxylase (pyc) and phosphoenolpyruvate carboxylase (ppc).
All components constituting a non-branched respiratory chain are present in C. variabile DSM 44702, along with complete menaquinone and heme biosynthesis pathways (Additional file 2). The terminal oxidase of the respiratory chain most likely consists of both cytochrome bc 1 oxidase (qcrCAB) and cytochrome aa 3 oxidase (ctaCDEF) that might constitute a cytochrome bc 1-aa 3 supercomplex, as previuosly described for C. glutamicum . Moreover, the genome of C. variable DSM 44702 contains the cydAB genes encoding the subunits of cytochrome bd oxidase. In mycobacteria, the cytochrome bd oxidase is induced under microaerobic conditions because of the high oxygen affinity of this enzyme . This suggests that the cytochrome bc 1-aa 3 supercomplex is of basic importance for C. variabile DSM 44702 under aerobic growth conditions. The genes encoding subunits of the F1F0-ATP synthase complex that is essential for ATP generation by oxidative phosphorylation are present and organized in a putative operon (atpBEFHAGDC).
The genome sequence of C. variabile DSM 44702 was moreover screened for genes encoding enzymes, which degrade, modify, or create glycosidic bonds according to data stored in the CAZy database . This seed information of CAZy includes glycoside hydrolases, glycosyl transferases, and carbohydrate esterases. The content of the CAZy database also covers the associated carbohydrate-binding module, comprising carbohydrate-active enzymes that often display a modular structure with the aforementioned proteins. Candidate carbohydrate-active enzymes encoded in C. variabile DSM 44702 were identified by a combined approach using BLASTp analysis of all corynebacterial enzymes stored in the CAZy database and hidden Markov models  created from amino acid sequence alignments downloaded from the Pfam database . Glycoside hydrolases (GH) hydrolyze glycosidic bonds and are a widespread group of enzymes with at least 128 families [43, 46]. C. variabile DSM 44702 encodes 15 candidate glycoside hydrolases associated to eight families. The protein families GH13, GH23, and GH25 are represented by three proteins each, whereas the families GH32, GH65, GH76, and GH19 contain only a single protein member. Two proteins of C. variabile DSM 44702 were associated to the glycoside hydrolase family GH3. Glycosyl transferases (GT) catalyze the transfer of sugar moieties during polysaccharide biosynthesis and are classified into 94 families [43, 47]. A total number of 28 candidate glycosyl transferases was detected in the genome of C. variabile DSM 44702 and grouped into 13 protein families. The largest families are GT2 and GT4, each containing six proteins, followed by family GT87 with three enzymes, and GT1, GT3, and GT51 with two members. One protein was identified in each of the families GT5, GT20, GT28, GT35, GT39, GT53, and GT85. Carbohydrate esterases (CE) catalyze the de-O-acylation or de-N-acylation of substituted saccharides and are grouped into 16 families [43, 48]. C. variabile DSM 44702 encodes 14 candidate carbohydrate esterases, six of them grouped into family CE1 and six enzymes classified into family CE9. Family CE14 contains two proteins. A carbohydrate-binding module (CBM) is defined as a contiguous amino acid sequence within a carbohydrate-active enzyme with a discreet fold having carbohydrate-binding activity . In C. variabile DSM 44702, two proteins were assigned to family CBM48, but the respective proteins are simultaneously associated to family GH13. Altogether, 57 candidate proteins constitute the set of carbohydrate-active enzymes of C. variabile DSM 44702.
Moreover, the hitherto known biosynthesis pathways for vitamins and cofactors in corynebacteria were reconstructed and shown to be complete in C. variabile DSM 44702, including the synthesis of molybdenum cofactor (Additional file 2). The anabolic pathways for purine and pyrimidine nucleotides are also complete (data not shown), as are all pathways for the de novo biosynthesis of standard proteinogenic amino acids (Additional file 3). The C. variabile DSM 44702 genome is equipped with several transport systems allowing the import of "free" amino acids (Additional file 3). The content of amino acids varies between different types of cheese mainly due to the duration and intensity of proteolysis during the ripening process . Seven amino acids are often concentrated in cheese and contribute to the taste of the final product: leucine, lysine, and phenylalanine (bitter); proline and valine (bitter sweet); aspartate and glutamate (salty umami) . According to the detailed knowledge of amino acid transporters in C. glutamicum, putative importers for these amino acids are encoded in C. variabile DSM 44702 by brnQ (leucine, valine, and isoleucine ), lysI (lysine ), pheP (phenylalanine ), aroP (phenylalanine, tyrosine, and tryptophan ), proP and putP (proline [55, 56]), and gluABCD (glutamate ) (Additional file 3). In addition, the sulphur-containing amino acid methionine, which is the precursor of the flavor compound methanethiol, is imported into C. variabile DSM 44702 by two transport systems, an ABC-type transporter encoded by the metQNI genes and a two-subunit member of the neurotransmitter:sodium symporter family encoded by the metP-metPS gene pair  (Additional file 3).
The secretome of C. variabile DSM 44702 deduced from the complete genome sequence
Like other members of the genus Corynebacterium, C. variabile DSM 44702 encodes two types of protein secretion systems, the general secretory (Sec) pathway and the twin-arginine translocation (Tat) pathway. The Sec translocation pathway has two distinct functions in bacterial protein trafficking, the cotranslational integration of proteins into the cytoplasmic membrane and the posttranslational translocation of proteins across the membrane . The posttranslational export of proteins is facilitated by the SecYEG translocon in conjunction with the multifunctional protein SecA. The latter protein binds to cytosolic precursor proteins destined for export and delivers them to the SecYEG translocon. As the SecA protein is also an ATPase, it provides energy for the stepwise translocation of the precursor protein through the channel of the SecYEG translocon. The precursor proteins have a typical signal sequence at the amino-terminal end, which is cleaved from the precursor by either signal peptidase I or signal peptidase II .
The genome sequence of C. variabile DSM 44702 encodes all protein components of the basic Sec machinery necessary for protein secretion. The lepB1 and lepB2 genes encode homologs of signal petidase I (EC 220.127.116.11), whereas the lspA gene codes for signal petidase II, also named lipoprotein signal peptidase (EC 18.104.22.168). In addition, the secA2 gene of C. variabile DSM 44702 provides a homolog of the preprotein translocase subunit SecA. The presence of the accessory protein SecA2 was reported previously for other Gram-positive bacteria. In M. tuberculosis and M. smegmatis, the SecA2 protein is required for the export of a small subset of exoproteins, whereas SecA1 remains the essential housekeeping translocase . On the other hand, both SecA homologs are probably essential for viability in C. glutamicum . C. variabile DSM 44702 moreover encodes the signal recognition particle SRP, which is essential for targeting of almost all inner membrane proteins in E. coli , as well as its receptor FtsY .
To estimate the number of secreted proteins encoded by C. variabile DSM 44702, the first 70 amino acid residues of each protein were used to search for amino-terminal signal peptides with SignalP 4.0 . In this way, a total number of 363 proteins were identified to be potentially secreted by C. variabile. To furtheron deduce the number of putative lipoproteins, all secreted proteins were screened for the presence of a lipobox motif using the DOLOP database tools . Altogether, 66 lipobox-containing proteins were identified in the proteome of C. variabile DSM 44702. The signal peptides of these proteins are therefore most likely recognized by signal peptidase II. The signal peptides of the remaining 297 proteins are recognized by signal peptidase I and the respective proteins are exported by the machinery of the Sec pathway of C. variabile DSM 44702. However, it has to be considered that proteins with membrane-spanning domains might also contain signal peptides that are recognized by type I signal peptidases . Out of the 363 proteins predicted to be secreted by C. variabile DSM 44702, a set of 100 proteins with membrane-spanning domains were detected by the TMHMM tool , which therefore may be destined for the integration into the cytoplasmic membrane. This sub-set of proteins comprises eleven membrane-associated enzymes with functions in the respiratory chain or in cell wall turnover of C. variabile DSM 44702 and 25 putative permeases with transport functions for diverse substrates. Two-thirds of the proteins were annotated as putative membrane proteins with unknown functions. After exclusion of proteins with membrane-spanning domains, the remaining 263 predicted extracellular proteins of C. variabile DSM 44702 were grouped into functional categories according to the classification scheme of the Clusters of Orthologous Groups of proteins (COG) . A large portion of the predicted extracellular proteins was assigned to COG classes R (16%), S (6.4%), and X (39.8%), containing proteins that are currently not grouped into COGs or whose precise functions are hitherto unknown. Prominent functional classes of the extracellular proteins of C. variabile DSM 44702 comprise inorganic ion transport and metabolism (COG class P; 16%), cell wall and membrane biogenesis (COG class M; 7.6%), and signal transduction mechanisms (COG class T; 3.4%).
The Tat pathway of C. variabile DSM 44702 is represented by two members of the TatC protein family encoded by tatC1 and tatC2, as well as four members of the TatA family encoded by paralogous versions of tatA and tatB . Proteins are targeted to the Tat pathway by N-terminal signal peptides containing the twin-arginine motif that is followed by a hydrophobic stretch of amino acids . In C. variabile DSM 44702, a total number of ten proteins containing signal peptides meeting these criteria were detected by the prediction tools TATP  and TATFIND . This small number of proteins secreted by the Tat pathway is comparable to those predicted previously in C. diphtheriae , C. glutamicum, and C. efficiens [65, 73]. The set of proteins secreted by the Tat pathway of C. variabile DSM 44702 includes for instance an alkaline phosphatase (EC 22.214.171.124), a ceramidase (EC 126.96.36.199), an esterase of the SGNH type, and a Dyp-type peroxidase.
The transcriptional regulatory repertoire of C. variabile DSM 44702 deduced from the complete genome sequence
The repertoire of candidate transcription regulators encoded in the C. variabile DSM 44702 genome was deduced from the functional genome annotation, taking into account the comprehensive knowledge of the gene-regulatory network of C. glutamicum ATCC 13032 . A collection of 156 genes encoding DNA-binding transcription regulators (132 genes), sigma factors (8 genes), and response regulators of two-component signal transduction systems (16 genes) can be regarded as the minimal regulatory repertoire of C. variabile DSM 44702 (Figure 5). It is noteworthy that among the genes for response regulators, four are not associated with a corresponding sensor histidine kinase gene (CVAR_0325, CVAR_0577, CVAR_1009, and CVAR_2072). The deduced set of candidate transcription regulators represents 5.1% of the predicted protein-coding genes of the C. variabile genome. This value is in the range known from the 3.28 Mb genome of C. glutamicum ATCC 13032 (5.3%)  and is in agreement with previous observations that less than 10% of the total number of predicted proteins is associated with transcriptional regulatory processes in bacteria .
The collection of potential transcription regulators was grouped into regulatory protein families according to their amino acid sequence similarities and domain organizations . This bioinformatic classification assigned the predicted transcription regulators of C. variabile to 37 regulatory protein families that vary significantly in their number of representatives (Figure 5). The largest family of DNA-binding transcription regulators is TetR with 30 members, followed by AraC with 13 proteins. The TetR family of transcription regulators is widely distributed among bacterial species  and is also the most prevalent group of regulatory proteins in other corynebacteria . The AraC family of transcription regulators comprises a diverse group of proteins that are involved in the regulation of various biological processes, such as carbon and nitrogen metabolism, adaptive responses, stress responses, and virulence . The number of AraC-type transcriptional regulators is remarkably high in C. variabile, as other non-pathogenic corynebacteria encode only three or four members of this protein family, whereas genes encoding AraC-type regulators are less prevalent or even absent in the genomes of pathogenic corynebacteria [78, 80, 81].
C. variabile encodes only 22 proteins of the previously detected collection of 24 transcription regulators that constitute the core set of DNA-binding transcription regulators in corynebacteria. The genome of C. variabile DSM 44702 lacks orthologs of the conserved regulators RamB and SugR that are involved in the transcriptional control of acetate metabolism and central sugar metabolism, respectively [74, 78]. To detect regulatory interactions contributing to metabolic pathways relevant for cheese ripening and flavor generation by C. variabile, the deduced knowledge of the transcriptional regulatory repertoire was combined with bioinformatic motif searches for DNA-binding sites of prominent regulatory proteins . In this way, relevant metabolic pathways of C. variabile DSM 44702 described below were linked with the transcriptional regulatory network of this cheese isolate.
Specific features of C. variabile DSM 44702 related to carbohydrate metabolism and its regulation in the cheese habitat
The characteristic features of the central carbohydrate metabolism in C. variabile DSM 44702 with respect to the ripening of smear-ripened cheeses are in part specified by genes providing the ability to metabolize citrate, L-lactate, taurine, and γ-amino butyric acid (Figure 6). These metabolites are well-known as prominent components on the surface of smear-ripened cheeses from several previous studies [1, 83–86].
Citrate uptake in C. variabile DSM 44702
The metabolic pathway reconstruction of C. variabile DSM 44702 revealed the presence of the citP gene encoding a citrate transporter of the CitMHS protein family (Figure 6). Members of this transporter family import citrate in symport with Mg2+ or Ca2+ ions. Citrate is naturally occuring in milk, containing 8 to 9 mM of citrate during fermentation . Therefore, C. variabile DSM 44702 can benefit from the natural compositon of the cheese habitat by importing citrate as an additional carbon and energy source. Citrate metabolism is widespread among lactic starter bacteria and can contribute to the synthesis of flavor compounds, such as acetoin and butanediol [88–91].
The expression of the citrate transporter gene citP is most likely regulated by the two-component signal transduction system CitAB. The corresponding citAB genes are located upstream of citP in the C. variabile DSM 44702 genome. The orthologous two-component system of C. glutamicum ATCC 13032 was shown to regulate genes involved in citrate uptake and metabolism . Target genes of the CitAB system are apparently coregulated by the cAMP-sensing transcription regulator GlxR in this species . The GlxR protein is currently the only known global transcription regulator in corynebacteria, connecting different functional modules of the gene-regulatory network . The glxR gene is conserved among all hitherto sequenced corynebacterial species and also present in C. variabile DSM 44702. A GlxR binding site was detected upstream of the citP coding region, indicating that the expression of this gene is under local control by CitAB and under global control by GlxR in C. variabile DSM 44702.
Lactate metabolism in C. variabile DSM 44702
Two main sources of carbon available to microorganisms at the beginning of cheese ripening are lactose and lactate . Lactose is rapidly converted to lactate in the early days of cheese ripening by the large number of lactic starter bacteria present in the cheese. Lactate is thus the most important carbon source after a few days of cheese ripening [1, 94]. According to the taxonomic description and genome sequence data, C. variabile DSM 44702 is unable to utilize lactose . On the other hand, the growth rate of C. variabile and lactate dehydrogenase activity were significantly increased in the presence of lactate . The C. variabile DSM 44702 genome contains the lldT gene coding for L-lactate permease involved in L-lactate uptake, which is located upstream of the lldD1 gene. A paralog of lldD1, named lldD2, is located elsewhere on the chromosome. Both genes encode L-lactate dehydrogenase (EC 188.8.131.52) that facilitates the conversion of L-lactate to pyruvate. In addition, L-lactate can be converted to acetate by lactate 2-monooxygenase (EC 184.108.40.206) encoded by the lldA gene of C. variabile DSM 44702. D-Lactate is also used as substrate by three dld paralogs (dld1-dld3) coding for D-lactate dehydrogenases (EC 220.127.116.11). D-Lactate is a natural product in smear-ripened cheeses formed by lactic acid bacteria that can produce both stereoisomers of lactic acid . Due to efficient metabolism of L-lactate and D-lactate, C. variabile can probably grow from the beginning of the ripening process and is independent of significant yeast growth during the initial stages of cheese ripening . GlxR binding sites were detected upstream of the putative lldT-lldD1 operon, upstream of the lldD2 coding region as well as in front of the paralogous dld genes (data not shown). Lactate metabolism of C. variabile DSM 44702, with the exception of lldA expression, is thus under global control by GlxR, as it was demonstrated previously in C. glutamicum ATCC 13032 [74, 96].
Taurine utilization by C. variabile DSM 44702
Taurine (2-aminoethanesulfonic acid) is a sulfonated amino acid that is not incorporated into polypeptides. It derives from the metabolism of sulphur-containing amino acids and is the most abundant free amino acid in goat's milk, with a mean value of 66.2 mg/l . The amount of taurine in cow's milk is considerably lower (1.6-10 mg/l) . The C. variabile DSM 44702 genome contains a gene cluster involved in the assimilation of taurine (Figure 6). The respective genomic region comprises the tauCBA genes encoding a predicted taurine transporter of the ABC superfamily, the tauD1 gene encoding a putative taurine dioxygenase, and the tauR gene coding for a ROK-type transcription regulator. Moreover, C. variabile DSM 44702 contains a paralog of tauD1 not located in this gene cluster, tauD2, encoding a second taurine dioxygenase. This enzyme (EC 18.104.22.168) catalyzes the key step in taurine assimilation by generating sulfite, aminoacetaldehyde, and succinate (Figure 6). As the resulting sulfite can be converted to cysteine, the cheese component taurine can serve as a combined sulphur and carbon source for C. variabile DSM 44702.
The expression of the tauCBA and tauD1 genes is apparently regulated by TauR, a ROK-type transcription regulator, whose corresponding gene is located upstream of the tauD1CBA operon. ROK-type regulators are known from C. glutamicum ATCC 13032 to contribute to the transcriptional control of sulphur metabolism that is under hierarchical control by the TetR-type regulator McbR [99–101]. McbR belongs to the core set of corynebacterial transcription regulators  and acts as the master regulator of all aspects of sulphur metabolism in C. glutamicum and C. jeikeium [101, 102]. The McbR protein is also encoded in C. variabile DSM 44702 by the mcbR gene that is located at the boundary of the small duplicated genomic region CVAR_2865-CVAR_2874 (mcbR). The second copy of this DNA segment is part of the iron acquisition island II (Figure 3), comprises the coding regions CVAR_2821-CVAR_2830, and ends with a 5' truncated version of the mcbR gene, named mcbR2. A genome-wide search for McbR binding sites in C. variabile DSM 44702 detected 23 operators that are probably controlling 41 genes involved in (S-adenosyl)methionine biosynthesis, methionine import, cysteine biosynthesis, and sulfate reduction (data not shown). In addition, McbR binding sites were detected in the tauR-tauD1 intergenic region and in front of the tauD2 coding region. This data strongly indicates that all genes involved in taurine assimilation are embedded in the sulphur regulatory network of C. variabile DSM 44702.
GABA utilization by C. variabile DSM 44702
GABA (γ-amino butyric acid) is a four-carbon, non-proteinogenic amino acid known to occur in cheese with a varying concentration that is strongly affected by the protocol of cheese manufacture, the type of primary starter cultures, the cheese microbiota, and the ripening conditions [85, 86]. Several species of lactic starter bacteria are able to synthesize GABA from L-glutamate by glutamate decarboxylase (EC 22.214.171.124)  that is widely distributed among eukaryotes and prokaryotes  and also encoded in the C. variabile DSM 44702 genome by the gadB gene. In addition, GABA is degraded to succinate by the enzymes γ-aminobutyrate aminotransferase (EC 126.96.36.199) and succinate semialdehyde dehydrogenase (EC 188.8.131.52) in C. variable DSM 44702 (Figure 6). These three enzymatic reactions belong to a metabolic pathway known as GABA shunt . Two genes encoding GABA transporters (gabP1 and gabP2) were detected in C. variabile DSM 44702. The first one is located in a cluster comprising the genes gabD1 encoding succinate semialdehyde dehydrogenase, gabT1 coding for γ-aminobutyrate aminotransferase, and gabR1 specifying a transcription regulator. The second gene cluster attributed to GABA utilization is composed of gabP2, gabT2, and gabR2, and is thus lacking a gabD component. In total, six potential gabD genes (gabD1-gabD6) are distributed across the C. variabile DSM 44702 chromosome.
According to this genome annotation, the uptake and catabolism of GABA in C. variabile DSM 44702 is represented by at least three gene regions (Figure 6). The first gene cluster (gabT1-gabD1-gabP1) is preceded by the regulatory gene gabR1 coding for a transcription regulator of the COG2508 class. Likewise, the second gene cluster (gabT2-gabP2) is preceded by the regulatory gene gabR2 also coding for a transcription regulator assigned to the COG2508 class. The presence of this type of regulatory genes upstream of gab gene clusters is conserved among GABA-utilizing corynebacteria , suggesting that the respective regulators of the COG2508 class are involved in the transcriptional control of the divergently arranged gabT(D)P genes. The succinate semialdehyde dehydrogenase gene gabD2 is located elsewhere on the chromosome and organized in an operon with the regulatory gene gabR3, which encodes a transcription regulator of the LysR protein family . Moreover, GlxR binding sites were detected in the intergenic region of gabR1/gabT1 and gabR2/gabT2, as well as upstream of the gabR3 coding region, indicating that the respective GABA-related genes of C. variabile DSM 44702 are under hierarchical regulation by local GabR regulators and the global cAMP-sensing regulator GlxR. Integration of GABA utilization into the global GlxR network of C. variabile DSM 44702 is reasonable, because the central pathway intermediate succinate is finally generated by the GABA shunt (Figure 6). Furthermore, it is remarkable that GABA utilization as well as taurine and propionate metabolism channel succinate into the tricarboxylic acid cycle, as the typical generation of this compound by succinyl-CoA synthetase is not encoded in C. variabile DSM 44702.
Nitrate, nitrite, and oxaloacetate metabolism of C. variabile DSM 44702 in the cheese habitat
Low amounts of nitrates and nitrites are natural constituents of many types of cheese . Both compounds may exert important effects on human health when present in higher concentrations, as they may react with naturally present amines to form potent carcinogens. The content of nitrates and nitrites in cheese is therefore strictly regulated by law in many European countries. The annotation of the C. variabile DSM 44702 genome sequence revealed the presence of the nrtABC genes encoding an ABC-type transport system involved in nitrate uptake (Figure 6). Nitrate can be reduced by a putative nitrate reductase (EC 184.108.40.206) encoded by the narB gene of C. variabile DSM 44702, which is located upstream of nirB and nirD, both coding for subunits of nitrite reductase (EC 220.127.116.11). This enzyme converts nitrite to ammonium hydroxide, resulting in the detoxification of nitrite and the generation of a suitable nitrogen source for C. variabile DSM 44702. Alternatively, nitrite can be exportet by the putative nitrite extrusion protein NarT, which is encoced in the same gene cluster by the narT gene.
Moreover, the oxalate:formate antiporter gene oxlT was identified in C. variabile DSM 44702 (Figure 6). The OxlT protein is a member of the major facilitator superfamily and exchanges formate for oxalate across the cytoplasmic membrane . Oxalate is present in milk and cheese in relatively small amounts. Nevertheless, oxalic acid can react with calcium ions in the milk producing calcium oxalate, thereby preventing this vital nutrient from being absorbed in the human intestine . In principle, oxalate can be converted to formate by oxalate decarboxylase (EC 18.104.22.168), but the respective oxdD gene is absent in the genome of C. variabile DSM 44702. However, formate can be oxidized in C. variabile DSM 44702 by formate dehydrogenase (EC 22.214.171.124) encoded by the fdhFD genes (Additional file 1). Due to import of oxalic acid by C. variabile DSM 44702, the binding of calcium ions probably occurs to a lower degree in smear-ripened cheeses in the presence of this strain. Therefore, the gene equipment of C. variabile DSM 44702 provides a substantial benefit for the quality of smear-ripened cheeses.
Lipolysis and fatty acid metabolism by C. variabile DSM 44702 in the cheese habitat
Cow's milk contains high amounts of lipids and triacylglycerides (about 35 g/l), which may serve as carbon and energy sources for the microflora of smear-ripened cheeses . Free fatty acids derive from breakdown of triacylglycerides, and their further degradation can lead to volatile compounds having low olfactory thresholds in cheese flavor perception . In smear-ripened cheeses, more than 4 g/kg of free fatty acids are available for the surface microbiota . The genome of C. variabile DSM 44702 is equipped with a large set of genes encoding enzymes involved in lipolysis and the degradation of fatty acids (Figure 6). It contains five genes (estA1-estA5) encoding secreted esterases of the SGNH-hydrolase superfamily and the estB gene coding for an esterase of the AB hydrolase superfamily . Secreted lipases are encoded by five genes in the C. variabile DSM 44702 genome. The paralogous lipA genes (lipA1-lipA3) encode class 2 lipases that hydrolyse ester bonds in triacylglycerol giving diacylglycerol, monoacylglycerol, glycerol, and free fatty acids . The genes lipS1 and lipS2 belong to the LIP superfamily containing lipases with broad lipolytic activities . All these enzymes may thus contribute to the generation of free fatty acids from precursor molecules present in the cheese habitat.
Furthermore, 13 fadD genes encoding acyl-CoA synthetases were identified in the genome of C. variabile DSM 44702, including the fadD1 gene that is involved in mycolic acid biosynthesis . The amino acid sequences of the FadD enzymes in C. variabile DSM 44702 revealed low levels of overall similarity, but the characteristic AMP-binding motif of acyl-CoA synthetase is highly conserved among the twelve paralogs (FadD2-FadD13). In principle, acyl-CoA synthetases are involved in activating free fatty acids to form acyl-CoA of various chain lengths concomitant with the transport into the bacterial cell . The amino acid sequence diversity of the predicted acyl-CoA synthetases might indicate different substrate specificities of these enzymes, which would enable C. variabile to utilize a broader range of fatty acid substrates in its natural cheese habitat, probably leading to a growth advantage as a well-adapted species.
In addition to the set of fadD genes, a complete β-oxidation pathway for fatty acid degradation is encoded in the C. variabile DSM 44702 genome (Figure 6). In this pathway, the acyl-CoA is converted into a product with a shortened acyl chain (n-2) and acetyl-CoA that can be used in the tricarboxylic acid cycle to generate energy, whereas the shortened acyl-CoA is used in a new round of β-oxidation . In principle, C. variabile DSM 44702 can utilize exogenous fatty acids also as carbon sources, since the glyoxylate bypass genes aceA and aceB are present and the gluconeogenesis pathway is complete (Additional file 1). The predicted enzymes involved in β-oxidation of C. variabile DSM 44702 are encoded by ten paralogs of fadE (encoding acyl-CoA dehydrogenase), the bifunctional fadB1 gene (enoyl-CoA hydratase/hydroxyacyl-CoA dehydrogenase), three monofunctional fadB genes (hydroxyacyl-CoA dehydrogenase), three echA genes (enoyl-CoA hydratase), and five paralogs of fadA (ketoacyl-CoA thiolase). The amino acid sequences of these paralogous gene families vary substantially, again suggesting diverse substrate specificities of the respective enzymes, at least in the case of the ten acyl-CoA dehydrogenases. Interestingly, all genes coding for secreted esterases or lipases and almost all fad and echA genes of C. variabile DSM 44702 contain at least one GlxR binding site in their upstream regions (data not shown), indicating that the complete metabolic process of lipolysis and β-oxidation is under global control by the GlxR protein.
In contrast to the lipid-auxotrophic relatives C. jeikeium and C. urealyticum, the metabolic pathway analysis of the C. variabile DSM 44702 genome revealed the presence of the fatty acid synthase gene fas, showing that growth of this strain is not dependent on the availability of exogenous fatty acids [31, 34]. These compounds are, moreover, building blocks for the synthesis of corynomycolic acids, which are major constituents of the corynebacterial cell envelope . It was shown previously that a polyketide synthase encoded by the pks13 gene represents a condensase that catalyzes the key step in the biosynthesis of corynomycolic acids, in conjunction with an acyl-CoA carboxylase and a distinct acyl-AMP ligase [117–119]. The pks13 gene is located in a conserved cluster in the C. variabile DSM 44702 genome, including genes coding for an acyl-CoA carboxylase (accD3), an acyl-CoA synthetase/acyl-AMP ligase (fadD1), the envelope lipids regulation factor ElrF (elrF), and trehalose corynomycol transferases (cmt). It is noteworthy that eight genes encoding trehalose corynomycolyl transferases (cmtA-cmtH) were identified in the C. variabile DSM 44702 genome, catalyzing the transfer of mycolic acids from trehalose monocorynomycolate to another molecule of trehalose monocorynomycolate or on the cell wall arabinogalactan . Five of these genes (cmtC-cmtG) are clustered with the pks13 gene.
A second gene cluster involved in fatty acid metabolism of C. variabile DSM 44702 includes genes coding for putative acyl-CoA carboxylase subunits (CVAR_2092 and CVAR_2091), an acyl-CoA dehydrogenase domain-containing protein (CVAR_2090), an enoyl-CoA hydratase domain-containing protein (CVAR_2089), a citrate lyase β-subunit (citE), an acyl-CoA synthetase (fadD6), aycl-CoA:3-ketoacid-CoA transferase subunits (scoA and scoB), and a ketoacyl-CoA thiolase (fadA4). A similar gene cluster is present only in the genome of the lipophilic isolate C. jeikeium K411, whereas some genes were found in a similar genetic arrangement in the genome of Mycobacterium tuberculosis H37Rv . As most of the genes present in these conserved clusters are linked to fatty acid degradation, they might be involved in the activation and subsequent degradation of a distinct type of fatty acid substrate, whose chemical composition is currently unknown.
Proteolysis and utilization of caseins by C. variabile DSM 44702 in the cheese habitat
Protein degradation contributes to cheese texture and the generation of cheese flavor in all varieties of smear-ripened cheeses, as many relevant flavor compounds derive directly from amino acids [1, 87, 91, 121]. Milk and especially cheese have a high protein content, which is represented by a mixture of αS1-, αS2-, and β-caseins . Extracellular caseins can be degraded by C. variabile DSM 44702 by means of secreted proteolytic enzymes, such as a secretory serine protease encoded by the sepP gene and a secretory aminopeptidase encoded by the pepC gene (Figure 6). In addition, amino-terminal prolines are released from peptides by the enzymatic action of proline iminopeptidase (EC 126.96.36.199) encoded by the pepI gene. The enzymatic activity of an extracellular proline iminopeptidase from C. variabile NCDO 2101 was demonstrated previously . Since caseins are very rich in proline residues and proline is the most abundant amino acid in cheese, it may constitute a prominent substrate for C. variabile. The import of "free" or released proline is facilitated in C. variabile DSM 44702 by the amino acid transporters ProP and PutP (Additional file 3). The subsequent oxidation of proline to glutamate is catalyzed as a two-step reaction by a bifunctional proline dehydrogenase (EC 188.8.131.52 and EC 184.108.40.206) encoded by the putA gene .
Interestingly, McbR binding sites were detected in the upstream regions of sepP and pepC, suggesting the integration of these genes into the sulphur regulatory network of C. variabile DSM 44702. Therefore, proteolysis might be used by C. variabile DSM 44702 to cover the demand for sulphur-containing amino acids and is probably enhanced under sulphate starvation conditions. Other genes assigned to proteolysis by C. variabile DSM 44702 are also linked to unfavorable environmental conditions by regulatory features. The proline importer gene putP, for instance, is specified by the presence of a typical SOS box and seems to be under the direct transcriptional control by the SOS response regulator LexA, as previously demonstrated in C. glutamicum . Moreover, the proline iminopeptidase gene pepI is organized in an operon and located downstream of sodC in the C. variabile DSM 44702 genome. The sodC gene encodes a secreted copper, zinc-dependent superoxide dismutase (EC 220.127.116.11) that is characterized by a lipobox motif to be anchored in the cell membrane . The extracellular location of SodC indicates that it helps to protect the cell surface of C. variabile against superoxide generated externally by the microbiota of smear-ripened cheeses. The gene arrangement of pepI and sodC implies that the proline iminopeptidase gene plays a role in the responses to external stresses, which is reasonable since proline can be used as compatible solute in corynebacteria under osmotic stress conditions .
Osmotic and oxidative stress responses of C. variabile DSM 44702
Due to permanently changing environmental conditions during the ripening process of smear-ripened cheeses, the bacteria of the cheese microbiota are constantly exposed to several external stresses, such as pH stress and osmotic stress [110, 126]. As the pH of the cheese curd can be as low as 4.5 during the early stage of cheese processing, C. variabile has to cope with this external stress condition. It was demonstrated in C. glutamicum that potassium transport is essential for corynebacterial pH homeostasis and growth at acidic pH . The potassium channel CglK was found to be relevant for the maintenance of the internal pH and the adjustment of the membrane potential. The genome sequence of C. variabile DSM 44702 was therefore screened for the presence of genes coding for actinobacterial potassium transporters, of which seven different types were described very recently . C. variabile DSM 44702 encodes a homolog of the potassium channel CglK (CVAR_2147) and a Kef-type potassium transport system (CVAR_0647). Moreover, a sodium or potassium/H+ antiporter of the NhaP-type (CVAR_1814) might contribute to pH homeostasis in C. variabile DSM 44702.
C. variabile DSM 44702 is also well equipped with genes offering protection from osmotic stress that derives from high salt concentrations on the cheese surface during the ripening process (Figure 6). One mechanism to overcome this osmotic stress is the accumulation of osmoprotectants, such as ectoine, proline, and glycine betaine in the cytoplasm . The genome of C. variabile DSM 44702 contains the ectP gene encoding an ectoine transporter, proP coding for an osmoregulated proline transporter, and six genes (betP1-betP3 and betT1-betT3) encoding putative betaine/carnitine/choline transporters of the BCCT family. Betaine can be synthesized in two steps from the precursor choline by choline oxidase (EC 18.104.22.168) and betaine-aldehyde dehydrogenase (EC 22.214.171.124) mediated reactions. Milk and cheese contain high amounts of choline , making it available as precursor for the synthesis of betaine by C. variabile. The C. variabile DSM 44702 genome contains two betA genes (betA1 and betA2) encoding choline oxidases, of which the betA1 gene is clustered with the betaine-aldehyde dehydrogenase gene betB and the putative choline transporter gene betT1. The homologous uptake systems for osmoprotectants were extensively studied in C. glutamicum ATCC 13032 , as was the two-component system MtrAB that is involved in the transcriptional control of the genes encoding osmoregulated compatible solute carriers . The mtrAB gene pair was also identified in the genome of C. variabile DSM 44702, suggesting the presence of a similiar MtrAB-controlled network of osmoprotection in this species.
The C. variabile DSM 44702 genome is moreover equipped with several genes involved in oxidative stress responses, such as two genes (katA and katG) encoding catalases (EC 126.96.36.199), duplicated genes (sodA1 and sodA2) coding for intracellular manganese-dependent superoxide dismutases (EC 188.8.131.52), and the aforementioned secreted copper, zinc-dependent superoxide dismutase gene sodC. Additional functions involved in oxidative stress responses are provided by bcp coding for a peroxiredoxin, tpx encoding a thiol peroxidase, and the ahpCD gene pair encoding the subunits of alkyl hydroperoxide reductase (EC 184.108.40.206). The latter protein is an antioxidant enzyme responsible for directly reducing organic hyperoxides , whereas thiol peroxidase represents an oxidative stress defense system that uses reduction equivalents from thioredoxin and thioredoxin reductase to reduce alkyl hydroperoxides . Thioredoxin (EC 220.127.116.11) and thioredoxin reductase (EC 18.104.22.168) are also encoded by the trxA and trxB genes in the genome of C. variabile DSM 44702. The Bcp protein belongs to the bacterioferritin comigratory protein subfamily of the peroxiredoxin superfamily and represents also a thioredoxin-dependent thiol peroxidase .
Furthermore, the C. variabile DSM 44702 genome contains the regulatory gene oxyR encoding a homolog of the central regulator of oxidative stress responses in bacteria [134, 135]. The oxyR gene belongs to the core set of corynebacterial transcription regulators  and is located adjacent to the ahpCD gene pair of C. variabile DSM 44702. This gene order is also conserved in several corynebacterial genomes and in genomes of slow-growing mycobacteria, where oxyR is involved in the activation of ahpCD gene expression [136, 137]. According to this conserved gene arrangement in actinobacterial species and experimental data from mycobacteria, it is likely that OxyR controls at least the expression of the ahpCD genes in C. variabile DSM 44702. The katG gene of C. variabile DSM 44702 is probably under transcriptional control by LexA, as a DNA-binding site (SOS box) of this regulator was detected in front of the katG coding region (data not shown). The LexA protein is the key component of the bacterial SOS response , and the lexA gene is conserved in all corynebacterial genomes .
Iron acquisition by C. variabile DSM 44702 in the iron-depleted cheese habitat
For most bacteria, iron is essential as a cofactor for proteins involved in important cellular functions, such as DNA biosynthesis and aerobic respiration . Thus, iron acquisition is a vital function for bacterial survival. However, milk and cheese contain the iron-binding protein lactoferrin that can deny iron to microorganisms in such a way that smear-ripened cheeses contain only small amounts of readily available iron . To circumvent this limitation of iron in the cheese habitat, bacteria of the cheese microflora have evolved several molecular strategies, including the synthesis of siderophores during the ripening process . Siderophores are high-affinity iron-chelating compounds, often synthesized by non-ribosomal peptide synthetases in response to low-iron availability in the environment . The genome of C. variabile DSM 44702 contains five genes (nrpS1-nrpS5) encoding proteins with sequence similarities to multi-domain peptide synthetases (Figure 6). The peptide synthetase genes nrpS2 to nrpS5 are clustered in iron-acquisition island II of the C. variabile genome (Figure 3). This genomic island contains also the entC2 and entE2 genes, coding for isochorismate synthase (EC 22.214.171.124) and 2, 3-dihydroxybenzoate-AMP ligase (EC 126.96.36.199), respectively. These gene products revealed similarities to proteins involved in the biosynthesis of the catechol siderophores bacillibactin from Bacillus subtilis  and enterobactin from Escherichia coli . Additional enzymatic functions involved in the biosynthesis of catechol siderophores are encoded in the entA-entC1-entB-fes gene cluster of C. variabile DSM 44702. The entA gene encodes 2, 3-dihydro-2, 3-dihydroxybenzoate dehydrogenase (EC 188.8.131.52) and the entB gene codes for isochorismatase (EC 184.108.40.206). The product of the fes gene is similar to enterobactin esterase that is probably required for the removal of iron from the siderophore. In conjunction with a non-ribosomal peptide synthetase, these genes provide a complete pathway for the synthesis of catechol siderophores with chorismate as the precursor . Among the cluster 3 species of the genus Corynebacterium, similar siderophore biosynthesis genes are only present in the genome of C. jeikeium K411 .
Further analysis of the C. variabile DSM 44702 genome sequence revealed the presence of 29 genes with predicted functions in the utilization of iron-siderophores, named either siu (siderophore-iron utilization protein) or sii (siderophore-iron interacting protein) (Figure 7). Siu gene clusters are represented by siuS genes encoding substrate-binding proteins (siuS1-siuS17), siuU genes encoding siderophore utilization proteins (siuU1 and siuU2), and sii genes coding for siderophore-interacting proteins (siiA-siiE), which might be necessary for the removal of iron from iron-siderophore complexes . In addition, five regulatory genes (siuR1-siuR5) encoding transcription regulators of the AraC protein family were assigned to these gene clusters. Interestingly, numerous genes involved in iron utilization are located in the iron acquisition islands I (CVAR_230-CVAR_0239) and II (CVAR_2806-CVAR_2820) of the C. variabile genome (Figure 3).
A further aspect of iron aquisition is the transport of this trace element into the bacterial cell. The C. variabile DSM 44702 genome contains 23 genes with assigned functions in ferric iron uptake (fiu) (Figure 7). The fiu gene clusters are represented by genes coding for substrate-binding proteins (fiuS), permease components (fiuP and fiuQ), ATP-binding proteins (fiuA), and an iron-utilization protein (fiuU1). In addition, three fiu gene clusters contain regulatory genes (fiuR1-fiuR3) encoding transcription regulators of the AraC family (Figure 7). The accumulation of genes coding for AraC-like proteins in the siu and fiu gene clusters explains the large number of transcription regulators assigned to this protein family in C. variabile DSM 44702 (Figure 5), when compared with the regulatory repertoires in other corynebacterial genomes [78, 80, 81].
As iron limitation poses a problem for the microflora of the cheese surface, the microorganisms have to compete for iron, but nonetheless, they must regulate iron metabolism to prevent excess iron that can initiate the generation of toxic oxygen radicals from normal products of metabolism by the Fenton reaction . Bacteria have solved the problem of iron acquisition and homeostasis by encoding uptake systems that are tightly regulated at the transcriptional level . DtxR is the key transcription regulator controlling complex gene-regulatory networks involved in iron homeostasis in corynebacteria [74, 147]. The complete genome sequence of C. variabile DSM 44702 was screened for the presence of DtxR binding sites, using the experimentally validated binding motif from C. glutamicum ATCC 13032 as input . This genome-wide motif search detected 26 operators that are probably controlling the expression of 62 target genes, including for instance dps (DNA protection during starvation protein), ftn (ferritin), hmuO (heme oxygenase), sdhCAB (succinate dehydrogenase), and the suf operon involved in Fe-S cluster assembly. Among the detected operators, 17 are located in the siu and fiu gene clusters of C. variabile DSM 44702, regulating in total 41 genes that are organized in 18 transcription units (Figure 7). This result indicates that almost all siu and fiu gene clusters are integrated into the iron-regulatory network of C. variabile DSM 44702 that obviously exerts a hierarchical control of gene expression by DtxR in concert with the AraC-type regulators of these clusters. Such a complex transcriptional regulation of iron homeostasis has not been detected in any sequenced corynebacterial genome before .
Contribution of C. variabile DSM 44702 to flavor generation during cheese ripening
The flavor compounds commonly found in smear-ripened cheeses are produced by the cheese microbiota as a diverse group of volatile substances, including volatile sulphur-containing compounds, esters, aldehydes, and ketones . Half of these compounds derive from lactose fermentation, citrate degradation, and a few from lipolysis. The second half derives from amino acid catabolism by complex metabolic pathways . Methanethiol, for instance, derives from methionine and is a prominent aroma compound in smear-ripened cheeses. Several cheese-related strains of B. linens produce high amounts of methanethiol due to the presence of L-methionine-γ-lyase (EC 220.127.116.11) . This enzyme converts methionine to methanethiol via a γ-elimination step. A corresponding mgl gene was not detectable in the genome of C. variabile DSM 44702, although this strain was shown previously to produce methanethiol . This observation indicates that other enzymes must be responsible for the synthesis of methanethiol in C. variabile DSM 44702. Cystathionine β-lyase (EC 18.104.22.168) and cystathionine γ-lyase (EC 22.214.171.124) catalyze similar reactions in lactococci [150, 151], but only the former enzyme is encoded by the aecD gene in C. variabile DSM 44702.
Acetoin (3-hydroxybutanone) is another aroma compound of smear-ripened cheeses and gives butter its characteristic flavor [87, 152]. C. variabile DSM 44702 produces acetoin from pyruvate, with acetolactate being an intermediate. The conversion of acetolactate to acetoin is mediated by acetolactate decarboxylase (EC 126.96.36.199) encoded by the alsD gene (Figure 6). Moreover, acetoin can be converted to 2, 3-butanediol by the enzymatic activity of acetoin reductase (EC 188.8.131.52) encoded by the butA gene in C. variabile DSM 44702. Butanediol is also an important flavor compound in some varieties of smear-ripened cheeses .
Cheese is one of the oldest dairy products produced by humanity and its origin predates recorded history . Nowadays, hundreds of cheese varieties are produced throughout the world in wide-ranging textures, forms, and flavors. The production processes of all cheese varieties share one common characteristic: the microflora present in each cheese contributes to the typical properties of the final product, such as taste, aroma, and color [1, 94]. The complex interactions of these microorganisms and the various biochemical processes occuring during cheese ripening are the result of centuries of adaptation of yeasts, moulds, and bacteria to the cheese habitat . During the traditional ripening of smear-ripened cheeses, the surface flora of old, ripenend cheeses was transfered to young, unripened cheeses by a brining step in the early stage of ripening, a process named "old-young smearing" . Therefore, the combined metabolic activities of a highly adapted bacterial community on the surface of smear-ripened cheeses have evolved over many years to determine all aspects of the mature product . However, the process of "old-young smearing" involves the transfer of a complex, undefined, and variable microflora. Therefore, a significant improvement in the industrial production of smear-ripened cheeses is nowadays achieved by using defined surface starter cultures. C. variabile was shown previously to be an important bacterial component on the surface of the smear-ripened cheeses Brick and Gubbeen [4, 10, 17, 20]. Likewise, C. casei is also a dominant component on the surface of Gubbeen, but is moreover found to be dominant on the surface of other smear-ripened cheeses, such as Ardrahan, Durrus, and Milleens . Therefore, we reconstructed in this study the metabolic capabilities of C. variabile DSM 44702 from the complete genome sequence to understand its adaptation to the cheese surface habitat and its physiological role in cheese ripening and flavor generation.
The C. variabile genome (3.43 Mb) contains large regions with high levels of synteny with the chromosome of the close phylogenetic relative C. jeikeium (2.46 Mb), a multi-drug resistant nosocomial pathogen . However, the size of the C. variabile genome is rather characteristic of environmental corynebacteria , such as C. glutamicum (3.28-3.31 Mb), Corynebacterium efficiens (3.14 Mb), and Corynebacterium nuruki (3.11 Mb), than of pathogenic relatives of the cluster 3 subline [31, 34, 80]. In contrast to the genetic repertoires of the cluster 3 species C. jeikeium and C. urealyticum that both show a clear reduction of genes associated with catabolic activities of carbon sources, the metabolic reconstruction of C. variabile revealed the presence of a complete set of enzymes required for central carbohydrate metabolism, vitamin and cofactor biosynthesis, and the synthesis of amino acids. C. variabile is equipped for instance with several additional enzymes and proteins required for the utilization of substrates typically present on the cheese surface, including citrate, lactate, propionate, taurine, and γ-amino butyric acid (GABA). The utilization of the sulfonated amino acid taurine as a combined carbon and sulphur source by C. variabile has not been reported so far in conjunction with other bacterial species associated with smear-ripened cheeses. GABA was shown to be synthesized by lactic starter bacteria within the body of 22 Italian cheese varieties, such as Gorgonzola, Mozzarella, and Pecorino , and is also utilized as carbon source by Arthrobacter arilaitensis , which is one of the major bacterial species detected on the surface of smear-ripened cheeses. This bacterium is also able to utilize the stereoisomers L-lactate and D-lactate, but not citrate, presumably as a result of a mutation in the gene cluster encoding the two-component system CitAB necessary for activating the expression of the cognate citrate import system .
C. variabile is also equipped with an extensive set of enzymes participating in β-oxidation or the degradation of lipids and with enzymes involved in proteolysis. Cheese has a high content of lipids and triacylglycerides  that are probably made available for the metabolism of C. variabile by the enzymatic actions of secreted lipases and esterases. Likewise, caseins are the major protein sources in milk and cheese  and are thus prominent substrates for proteolysis by C. variabile, presumably carried out by secreted proteases. Moreover, proline iminopeptidase can contribute to proteolysis , as prolines are the major amino acid constituents of caseins . It is obvious that these metabolic specificities of C. variabile are closely linked to its niche adaption. Another specific feature of C. variabile is its high salt tolerance . The bodies of smear-ripened cheeses are brined with salt solutions prior to the start of the ripening period to control microbial growth and to extend shelf-life . The efficient accumulation of osmoprotectants by C. variabile may contribute to its salt tolerance, as a large number of genes accounting for the accumulation of compatible solutes were detected in the genome sequence.
Moreover, the C. variabile genome is well equipped with genes involved in iron acquisition, presumably as a result of adaptation to the iron-depleted cheese habitat. At least one complete siderophore biosynthesis cluster was identified in this species, which shows similarity to gene clusters involved in the synthesis of catechol siderophores in B. subtilis  and E. coli . The biosynthesis of siderophores was also observed in cheese-related strains of B. linens and A. arilaitensis that are frequently found on the surface of smear-ripened cheeses [110, 156]. Moreover, about 50 additional genes were assigned to iron-siderophore and ferric iron uptake components, which have few orthologs in other corynebacteria, indicating that the acquisition of iron is a key factor for growth of C. variabile on the surface of smear-ripened cheeses. It would be interesting to investigate whether C. variable is able to use siderophores produced by other species sharing its habitat, as it has been assumed for A. arilaitensis .
In principle, flavor compounds in cheese can derive from pathways of lipolysis or proteolysis and the metabolism of lactate and citrate . Volatile sulphur compounds, such as methanethiol, are particularly important in the flavor of smear-ripened cheeses and produced by the bacterial microflora . There is experimental evidence that C. variabile produces methanethiol , although no methionine-γ-lyase encoding gene (mgl) was detectable in this study. Another flavor compound produced by C. variabile from pyruvate is acetoin, which is also generated by lactococci in fresh unripened cheese and can be reduced to 2, 3-butanediol by the action of the enzyme acetoin reductase . The corresponding butA gene was detected in the C. variabile genome, indicating that the synthesis of 2, 3-butanediol might contribute to flavor generation of smear-ripened cheeses. Moreover, lipolysis plays an important role in the formation of cheese flavor in several cheese varieties, such as Swiss cheese , Hispánico cheese , Domiati cheese , and smear-ripened cheeses . The quantification of the production of flavour compounds from fat during cheese ripening by the action of lipases and esterases indicated that lipolysis can reach up 20% in mould-ripened cheeses, such as Camembert . Due to the presence of several lipases and esterases encoded by C. variabile, it is likely that this species may also contribute to cheese flavor by producing free fatty acids from lipids and triacylglycerides and by generating volatile compounds with low olfactory threshold values in cheese flavor perception.
Furthermore, the detected regulatory pattern for relevant metabolic pathways led to the conclusion that almost all major processes occuring in C. variabile are regulated in a hierarchical manner. We identified local regulators, whose corresponding genes are clustered with their proposed target genes, for instance the tauR gene that is located adjacent to the tauDCBA cluster involved in uptake and assimilation of taurine, gabR genes located upstream of gabT(D)P genes responsible for GABA utilization, as well as siuR and fiuR genes that are embedded in iron-siderophore and ferric iron uptake gene clusters and are all encoding members of the AraC family of transcription regulators. In addition, DNA-binding sites of master regulators known from C. glutamicum were detected in the C. variabile genome sequence. McbR binding sites, for instance, indicate the connection of the tauABCD genes to the sulphur regulon of C. variabile, whereas DtxR binding sites within the siu and fiu gene clusters indicate the integration of the respective genes into an iron-dependent gene-regulatory network. The presence of GlxR binding sites support the view of a hierarchical regulation of several prominent aspects of the C. variabile metabolism, including the utilization of γ-amino butyric acid, lipolysis, and β-oxidation.
The annotation of the C. variabile DSM 44702 genome in conjunction with comparative analysis provides evidence that a strong selective pressure exists on the surface of smear-ripened cheeses, concerning the ability to catabolize substrates, such as γ-amino butyric acid, lactate, lipids, and amino acids (mainly proline), and the presence of efficient iron acquisition and salt-tolerance systems. Our analysis of the genome data revealed that C. variabile is strongly adapted to this habitat and contributes to the main aspects of cheese quality, such as flavor formation. Future studies will certainly benefit from the comprehensive genomic information on C. variabile by integrating this knowledge into models describing the complex cheese ecosystem, which is composed of various bacterial species and yeasts and their microbial interactions.
Bacterial strain and growth conditions
C. variabile DSM 44702 (CIP 107183; JCM 12073; LMG S-19265; NCIMB 30131) was originally isolated from the surface of an Irish farmhouse smear-ripened cheese . The strain used in this study was obtained as a lyophilized culture from the German Culture Collection DSMZ (Braunschweig, Germany). C. variabile DSM 44702 was formerly described as the type strain of the species C. mooreparkense, which turned out to be a later heterotypic synonym for C. variabile [4, 20]. C. variabile DSM 44702 was routinely cultivated in BYT complex medium  at 30°C.
Sequencing of the C. variabile DSM 44702 genome
Genomic DNA was purified from C. variabile cells by an alkaline lysis procedure , using 20-ml aliquots of an overnight culture grown in liquid BYT complex medium that was supplemented with 1.25% (w/v) glycine. The alkaline lysis protocol applied in this study includes two minor modifications: (i) The C. variabile cells were incubated in a 30 mg/ml lysozyme solution at 37°C for 1 h. (ii) The harvested C. variabile cells were lysed in 0.7 ml 10% (w/v) SDS solution at 37°C for 15 min. A total amount of 5 μg of purified genomic DNA was used for constructing a single-stranded template DNA library. The DNA concentration of the library was measured by using the Agilent RNA 6000 Nano Kit. The preparation and sequencing of the DNA library were performed according to standard protocols from Roche Applied Science. Pyrosequencing of the genomic DNA sample was carried out with the Genome Sequencer FLX System and Titanium chemistry (Roche Applied Science). The sequence reads were assembled with the GS Assembler Software (version 2.3).
Gap closure of the C. variabile DSM 44702 genome
The gap closure was supported by the related reference contig arrangement tool r2cat, using the complete C. jeikeium K411 genome sequence as a reference . The remaining gaps in the genome sequence of C. variabile were closed by PCR with Phusion hot start high-fidelity DNA polymerase (Finnzymes) and genomic template DNA. All primers used in this study were synthesized by Metabion. The PCR assays were carried out with a TProfessional PCR thermocylcer (Biometra) according to standard protocols (Finnzymes). Amplified DNA fragments linking individual contigs were sequenced by IIT Biotech. All DNA sequences were uploaded into the Consed program  and manually inspected to generate the complete genome sequence of C. variabile DSM 44702.
Annotation and bioinformatic analysis of the C. variabile DSM 44702 genome
The assembled genome sequence of C. variabile was uploaded into the bacterial genome annotation system GenDB . The annotation of the C. variabile genome sequence was performed as described previously , followed by manual curation. A combined gene prediction strategy was executed by using GLIMMER 2.1 and the CRITICA program suite  in conjuction with postprocessing by the RBSfinder tool . Additional regional information on protein-coding regions was obtained with REGANOR  and integrated into the final genome annotation. The deduced proteins of C. variabile were functionally characterized by automated searches in public databases according to default search parameters of the GenDB system . The origin of chromosomal replication of C. variabile was predicted with the web version of the Ori-Finder . The genome plot of C. variabile was generated with the DNAPlotter tool . Comparative analyses of the predicted gene content and reconstructions of metabolic properties of C. variabile were accomplished by the computer programs EDGAR  and CARMEN . The software tool CARMEN supports the in silico reconstruction and visualization of automatically derived metabolic networks based on pathway information from the KEGG database and user-defined SBML templates. The visualization and manual curation of pathway maps was performed with the CellDesigner software (version 3.2) . The synteny between the genomes of C. variabile DSM 44702 and C. jeikeium K411 was calculated with reciprocal best BLASTP matches by the EDGAR software . The detection and classification of the transcriptional regulatory repertoire of C. variabile was performed as described previously . The DNA motif discovery for regulon predictions followed a combined workflow, using both position weight matrices and hidden Markov models . The transcriptional regulatory network transfer between corynebacteria on the genome-scale was described in detail previously .
The annotated genome sequence of C. variabile DSM 44702 has been deposited in the GenBank database with accession number CP002917 and is also available from the RefSeq database with accession number NC_015859.
Fox PF, Guinee T, Cogan T, McSweeney P: Fundamentals of Cheese Science. 2000, Berlin: Springer
Beresford T, Fitzsimons N, Brennan N, Cogan T: Recent advances in cheese microbiology. Int Dairy J. 2001, 11: 259-274. 10.1016/S0958-6946(01)00056-5.
Reps A: Bacteria surface-ripened cheeses. Cheese: chemistry, physics and microbiology. Edited by: Fox P. 1993, London: Aspen Publishers, 2: 137-172.
Brennan NM, Brown R, Goodfellow M, Ward AC, Beresford TP, Simpson PJ, Fox PF, Cogan TM: Corynebacterium mooreparkense sp. nov. and Corynebacterium casei sp. nov., isolated from the surface of a smear-ripened cheese. Int J Syst Evol Microbiol. 2001, 51: 843-852. 10.1099/00207713-51-3-843.
Larpin-Laborde S, Imran M, Bonaiti C, Bora N, Gelsomino R, Goerges S, Irlinger F, Goodfellow M, Ward AC, Vancanneyt M, et al: Surface microbial consortia from Livarot, a French smear-ripened cheese. Can J Microbiol. 2011, 57: 651-660. 10.1139/w11-050.
Viljoen BC: The interaction between yeasts and bacteria in dairy environments. Int J Food Microbiol. 2001, 69: 37-44. 10.1016/S0168-1605(01)00570-0.
Corsetti A, Rossi J, Gobbetti M: Interactions between yeasts and bacteria in the smear surface-ripened cheeses. Int J Food Microbiol. 2001, 69: 1-10. 10.1016/S0168-1605(01)00567-0.
Bockelmann W, Hoppe-Seyler T: The surface flora of bacterial smear-ripened cheeses from cow's and goat's milk. Int Dairy J. 2001, 11: 307-314. 10.1016/S0958-6946(01)00060-7.
Rattray FP, Fox PF: Aspects of enzymology and biochemical properties of Brevibacterium linens relevant to cheese ripening: a review. J Dairy Sci. 1999, 82: 891-909. 10.3168/jds.S0022-0302(99)75308-7.
Valdes-Stauber N, Scherer S, Seiler H: Identification of yeasts and coryneform bacteria from the surface microflora of brick cheeses. Int J Food Microbiol. 1997, 34: 115-129. 10.1016/S0168-1605(96)01171-3.
Mounier J, Gelsomino R, Goerges S, Vancanneyt M, Vandemeulebroecke K, Hoste B, Scherer S, Swings J, Fitzgerald GF, Cogan TM: Surface microflora of four smear-ripened cheeses. Appl Environ Microbiol. 2005, 71: 6489-6500. 10.1128/AEM.71.11.6489-6500.2005.
Monnet C, Correia K, Sarthou AS, Irlinger F: Quantitative detection of Corynebacterium casei in cheese by real-time PCR. Appl Environ Microbiol. 2006, 72: 6972-6979. 10.1128/AEM.01303-06.
Rea MC, Gorges S, Gelsomino R, Brennan NM, Mounier J, Vancanneyt M, Scherer S, Swings J, Cogan TM: Stability of the biodiversity of the surface consortia of Gubbeen, a red-smear cheese. J Dairy Sci. 2007, 90: 2200-2210. 10.3168/jds.2006-377.
Irlinger F, Bimet F, Delettre J, Lefevre M, Grimont PA: Arthrobacter bergerei sp. nov. and Arthrobacter arilaitensis sp. nov., novel coryneform species isolated from the surfaces of cheeses. Int J Syst Evol Microbiol. 2005, 55: 457-462. 10.1099/ijs.0.63125-0.
Ventura M, Canchaya C, Tauch A, Chandra G, Fitzgerald GF, Chater KF, van Sinderen D: Genomics of Actinobacteria: tracing the evolutionary history of an ancient phylum. Microbiol Mol Biol Rev. 2007, 71: 495-548. 10.1128/MMBR.00005-07.
Otsuka Y, Kawamura Y, Koyama T, Iihara H, Ohkusu K, Ezaki T: Corynebacterium resistens sp. nov., a new multidrug-resistant coryneform bacterium isolated from human infections. J Clin Microbiol. 2005, 43: 3713-3717. 10.1128/JCM.43.8.3713-3717.2005.
Brennan NM, Ward AC, Beresford TP, Fox PF, Goodfellow M, Cogan TM: Biodiversity of the bacterial flora on the surface of a smear cheese. Appl Environ Microbiol. 2002, 68: 820-830. 10.1128/AEM.68.2.820-830.2002.
Khamis A, Raoult D, La Scola B: rpoB gene sequencing for identification of Corynebacterium species. J Clin Microbiol. 2004, 42: 3925-3931. 10.1128/JCM.42.9.3925-3931.2004.
Khamis A, Raoult D, La Scola B: Comparison between rpoB and 16S rRNA gene sequencing for molecular identification of 168 clinical isolates of Corynebacterium. J Clin Microbiol. 2005, 43: 1934-1936. 10.1128/JCM.43.4.1934-1936.2005.
Gelsomino R, Vancanneyt M, Snauwaert C, Vandemeulebroecke K, Hoste B, Cogan TM, Swings J: Corynebacterium mooreparkense, a later heterotypic synonym of Corynebacterium variabile. Int J Syst Evol Microbiol. 2005, 55: 1129-1131. 10.1099/ijs.0.63420-0.
Husemann P, Stoye J: r2cat: synteny plots and comparative assembly. Bioinformatics. 2011, 26: 570-571.
Gordon D, Abajian C, Green P: Consed: a graphical tool for sequence finishing. Genome Res. 1998, 8: 195-202.
Meyer F, Goesmann A, McHardy AC, Bartels D, Bekel T, Clausen J, Kalinowski J, Linke B, Rupp O, Giegerich R, Pühler A: GenDB--an open source genome annotation system for prokaryote genomes. Nucleic Acids Res. 2003, 31: 2187-2195. 10.1093/nar/gkg312.
Lowe TM, Eddy SR: tRNAscan-SE: a program for improved detection of transfer RNA genes in genomic sequence. Nucleic Acids Res. 1997, 25: 955-964. 10.1093/nar/25.5.955.
Gao F, Zhang CT: Ori-Finder: a web-based system for finding oriC s in unannotated bacterial genomes. BMC Bioinformatics. 2008, 9: 79-10.1186/1471-2105-9-79.
Hendrickson H, Lawrence JG: Selection for chromosome architecture in bacteria. J Mol Evol. 2006, 62: 615-629. 10.1007/s00239-005-0192-2.
Rocha EP, Danchin A: Essentiality, not expressiveness, drives gene-strand bias in bacteria. Nat Genet. 2003, 34: 377-378. 10.1038/ng1209.
Yukawa H, Omumasaba CA, Nonaka H, Kos P, Okai N, Suzuki N, Suda M, Tsuge Y, Watanabe J, Ikeda Y, et al: Comparative analysis of the Corynebacterium glutamicum group and complete genome sequence of strain R. Microbiology. 2007, 153: 1042-1058. 10.1099/mic.0.2006/003657-0.
Suzuki N, Okai N, Nonaka H, Tsuge Y, Inui M, Yukawa H: High-throughput transposon mutagenesis of Corynebacterium glutamicum and construction of a single-gene disruptant mutant library. Appl Environ Microbiol. 2006, 72: 3750-3755. 10.1128/AEM.72.5.3750-3755.2006.
Blom J, Albaum SP, Doppmeier D, Pühler A, Vorhölter FJ, Zakrzewski M, Goesmann A: EDGAR: a software framework for the comparative analysis of prokaryotic genomes. BMC Bioinformatics. 2009, 10: 154-10.1186/1471-2105-10-154.
Tauch A, Trost E, Tilker A, Ludewig U, Schneiker S, Goesmann A, Arnold W, Bekel T, Brinkrolf K, Brune I, et al: The lifestyle of Corynebacterium urealyticum derived from its complete genome sequence established by pyrosequencing. J Biotechnol. 2008, 136: 11-21. 10.1016/j.jbiotec.2008.02.009.
Summer EJ, Liu M, Gill JJ, Grant M, Chan-Cortes TN, Ferguson L, Janes C, Lange K, Bertoli M, Moore C, et al: Genomic and functional analyses of Rhodococcus equi phages ReqiPepy6, ReqiPoco6, ReqiPine5, and ReqiDocB7. Appl Environ Microbiol. 2011, 77: 669-683. 10.1128/AEM.01952-10.
Dorscht J, Klumpp J, Bielmann R, Schmelcher M, Born Y, Zimmer M, Calendar R, Loessner MJ: Comparative genome analysis of Listeria bacteriophages reveals extensive mosaicism, programmed translational frameshifting, and a novel prophage insertion site. J Bacteriol. 2009, 191: 7206-7215. 10.1128/JB.01041-09.
Tauch A, Kaiser O, Hain T, Goesmann A, Weisshaar B, Albersmeier A, Bekel T, Bischoff N, Brune I, Chakraborty T, et al: Complete genome sequence and analysis of the multiresistant nosocomial pathogen Corynebacterium jeikeium K411, a lipid-requiring bacterium of the human skin flora. J Bacteriol. 2005, 187: 4671-4682. 10.1128/JB.187.13.4671-4682.2005.
Bott M: Offering surprises: TCA cycle regulation in Corynebacterium glutamicum. Trends Microbiol. 2007, 15: 417-425. 10.1016/j.tim.2007.08.004.
Mann DA, Beuchat LR: Combinations of antimycotics to inhibit the growth of molds capable of producing 1, 3-pentadiene. Food Microbiol. 2008, 25: 144-153. 10.1016/j.fm.2007.06.005.
Jolkver E, Emer D, Ballan S, Krämer R, Eikmanns BJ, Marin K: Identification and characterization of a bacterial transport system for the uptake of pyruvate, propionate, and acetate in Corynebacterium glutamicum. J Bacteriol. 2009, 191: 940-948. 10.1128/JB.01155-08.
Akyol I, Serdaroglu K, Gezginc Y, Dayisoylu K, Ekinci M, Ozkose E: Redirection of pyruvate pathway of lactic acid bacteria to improve cheese quality. Food Biotechnology. 2009, 23: 200-213. 10.1080/08905430903102562.
Cerdeno-Tárraga AM, Efstratiou A, Dover LG, Holden MT, Pallen M, Bentley SD, Besra GS, Churcher C, James KD, De Zoysa A, et al: The complete genome sequence and analysis of Corynebacterium diphtheriae NCTC13129. Nucleic Acids Res. 2003, 31: 6516-6523. 10.1093/nar/gkg874.
Söhling B, Gottschalk G: Molecular analysis of the anaerobic succinate degradation pathway in Clostridium kluyveri. J Bacteriol. 1996, 178: 871-880.
Bott M, Niebisch A: The respiratory chain of Corynebacterium glutamicum. J Biotechnol. 2003, 104: 129-153. 10.1016/S0168-1656(03)00144-5.
Kana BD, Weinstein EA, Avarbock D, Dawes SS, Rubin H, Mizrahi V: Characterization of the cydAB-encoded cytochrome bd oxidase from Mycobacterium smegmatis. J Bacteriol. 2001, 183: 7076-7086. 10.1128/JB.183.24.7076-7086.2001.
Cantarel BL, Coutinho PM, Rancurel C, Bernard T, Lombard V, Henrissat B: The Carbohydrate-Active EnZymes database (CAZy): an expert resource for glycogenomics. Nucleic Acids Res. 2009, 37: D233-238. 10.1093/nar/gkn663.
Eddy SR: Profile hidden Markov models. Bioinformatics. 1998, 14: 755-763. 10.1093/bioinformatics/14.9.755.
Finn RD, Mistry J, Tate J, Coggill P, Heger A, Pollington JE, Gavin OL, Gunasekaran P, Ceric G, Forslund K, et al: The Pfam protein families database. Nucleic Acids Res. 2010, 38: D211-222. 10.1093/nar/gkp985.
Henrissat B, Davies G: Structural and sequence-based classification of glycoside hydrolases. Curr Opin Struct Biol. 1997, 7: 637-644. 10.1016/S0959-440X(97)80072-3.
Coutinho PM, Deleury E, Davies GJ, Henrissat B: An evolving hierarchical family classification for glycosyltransferases. J Mol Biol. 2003, 328: 307-317. 10.1016/S0022-2836(03)00307-3.
Lombard V, Bernard T, Rancurel C, Brumer H, Coutinho PM, Henrissat B: A hierarchical classification of polysaccharide lyases for glycogenomics. Biochem J. 2010, 432: 437-444. 10.1042/BJ20101185.
Boraston AB, Bolam DN, Gilbert HJ, Davies GJ: Carbohydrate-binding modules: fine-tuning polysaccharide recognition. Biochem J. 2004, 382: 769-781. 10.1042/BJ20040892.
Kabelova I, Dvorakova M, Cizkova H, Dostalek P, Melzoch K: Determination of free amino acids in cheeses from the Czech market. Czech J Food Sci. 2009, 27: 143-150.
Tauch A, Hermann T, Burkovski A, Krämer R, Pühler A, Kalinowski J: Isoleucine uptake in Corynebacterium glutamicum ATCC 13032 is directed by the brnQ gene product. Arch Microbiol. 1998, 169: 303-312. 10.1007/s002030050576.
Seep-Feldhaus AH, Kalinowski J, Pühler A: Molecular analysis of the Corynebacterium glutamicum lysl gene involved in lysine uptake. Mol Microbiol. 1991, 5: 2995-3005. 10.1111/j.1365-2958.1991.tb01859.x.
Zhao Z, Ding JY, Li T, Zhou NY, Liu SJ: The ncgl1108 (PheP (Cg)) gene encodes a new L-Phe transporter in Corynebacterium glutamicum. Appl Microbiol Biotechnol. 2011, 90: 2005-2013. 10.1007/s00253-011-3245-z.
Wehrmann A, Morakkabati S, Krämer R, Sahm H, Eggeling L: Functional analysis of sequences adjacent to dapE of Corynebacterium glutamicum reveals the presence of aroP, which encodes the aromatic amino acid transporter. J Bacteriol. 1995, 177: 5991-5993.
Peter H, Weil B, Burkovski A, Krämer R, Morbach S: Corynebacterium glutamicum is equipped with four secondary carriers for compatible solutes: identification, sequencing, and characterization of the proline/ectoine uptake system, ProP, and the ectoine/proline/glycine betaine carrier, EctP. J Bacteriol. 1998, 180: 6005-6012.
Peter H, Bader A, Burkovski A, Lambert C, Krämer R: Isolation of the putP gene of Corynebacterium glutamicum and characterization of a low-affinity uptake system for compatible solutes. Arch Microbiol. 1997, 168: 143-151. 10.1007/s002030050480.
Kronemeyer W, Peekhaus N, Krämer R, Sahm H, Eggeling L: Structure of the gluABCD cluster encoding the glutamate uptake system of Corynebacterium glutamicum. J Bacteriol. 1995, 177: 1152-1158.
Trötschel C, Follmann M, Nettekoven JA, Mohrbach T, Forrest LR, Burkovski A, Marin K, Krämer R: Methionine uptake in Corynebacterium glutamicum by MetQNI and by MetPS, a novel methionine and alanine importer of the NSS neurotransmitter transporter family. Biochemistry. 2008
du Plessis DJ, Nouwen N, Driessen AJ: The Sec translocase. Biochim Biophys Acta. 2011, 1808: 851-865. 10.1016/j.bbamem.2010.08.016.
Rigel NW, Braunstein M: A new twist on an old pathway--accessory Sec [corrected] systems. Mol Microbiol. 2008, 69: 291-302. 10.1111/j.1365-2958.2008.06294.x.
Caspers M, Freudl R: Corynebacterium glutamicum possesses two secA homologous genes that are essential for viability. Arch Microbiol. 2008, 189: 605-610. 10.1007/s00203-008-0351-0.
Luirink J, von Heijne G, Houben E, de Gier JW: Biogenesis of inner membrane proteins in Escherichia coli. Annu Rev Microbiol. 2005, 59: 329-355. 10.1146/annurev.micro.59.030804.121246.
Nielsen H, Engelbrecht J, Brunak S, von Heijne G: A neural network method for identification of prokaryotic and eukaryotic signal peptides and prediction of their cleavage sites. Int J Neural Syst. 1997, 8: 581-599. 10.1142/S0129065797000537.
Madan Babu M, Sankaran K: DOLOP--database of bacterial lipoproteins. Bioinformatics. 2002, 18: 641-643. 10.1093/bioinformatics/18.4.641.
Hansmeier N, Chao T-C, Pühler A, Tauch A, Kalinowski J: The cytosolic, cell surface and extracellular proteomes of the biotechnologically important soil bacterium Corynebacterium efficiens YS-314 in comparison to those of Corynebacterium glutamicum ATCC 13032. Proteomics. 2006, 6: 233-250. 10.1002/pmic.200500144.
Krogh A, Larsson B, von Heijne G, Sonnhammer EL: Predicting transmembrane protein topology with a hidden Markov model: application to complete genomes. J Mol Biol. 2001, 305: 567-580. 10.1006/jmbi.2000.4315.
Tatusov RL, Koonin EV, Lipman DJ: A genomic perspective on protein families. Science. 1997, 278: 631-637. 10.1126/science.278.5338.631.
Yen MR, Tseng YH, Nguyen EH, Wu LF, Saier MH: Sequence and phylogenetic analyses of the twin-arginine targeting (Tat) protein export system. Arch Microbiol. 2002, 177: 441-450. 10.1007/s00203-002-0408-4.
Berks BC: A common export pathway for proteins binding complex redox cofactors?. Mol Microbiol. 1996, 22: 393-404. 10.1046/j.1365-2958.1996.00114.x.
Bendtsen JD, Nielsen H, Widdick D, Palmer T, Brunak S: Prediction of twin-arginine signal peptides. BMC Bioinformatics. 2005, 6: 167-10.1186/1471-2105-6-167.
Rose RW, Brüser T, Kissinger JC, Pohlschröder M: Adaptation of protein secretion to extremely high-salt conditions by extensive use of the twin-arginine translocation pathway. Mol Microbiol. 2002, 45: 943-950. 10.1046/j.1365-2958.2002.03090.x.
Hansmeier N, Chao T-C, Kalinowski J, Pühler A, Tauch A: Mapping and comprehensive analysis of the extracellular and cell surface proteome of the human pathogen Corynebacterium diphtheriae. Proteomics. 2006, 6: 2465-2476. 10.1002/pmic.200500360.
Dilks K, Rose RW, Hartmann E, Pohlschröder M: Prokaryotic utilization of the twin-arginine translocation pathway: a genomic survey. J Bacteriol. 2003, 185: 1478-1483. 10.1128/JB.185.4.1478-1483.2003.
Schröder J, Tauch A: Transcriptional regulation of gene expression in Corynebacterium glutamicum: the role of global, master and local regulators in the modular and hierarchical gene regulatory network. FEMS Microbiol Rev. 2010, 34: 685-737.
Rodionov DA: Comparative genomic reconstruction of transcriptional regulatory networks in bacteria. Chem Rev. 2007, 107: 3467-3497. 10.1021/cr068309+.
Brune I, Brinkrolf K, Kalinowski J, Pühler A, Tauch A: The individual and common repertoire of DNA-binding transcriptional regulators of Corynebacterium glutamicum, Corynebacterium efficiens, Corynebacterium diphtheriae and Corynebacterium jeikeium deduced from the complete genome sequences. BMC Genomics. 2005, 6: 86-10.1186/1471-2164-6-86.
Ramos JL, Martínez-Bueno M, Molina-Henares AJ, Terán W, Watanabe K, Zhang X, Gallegos MT, Brennan R, Tobes R: The TetR family of transcriptional repressors. Microbiol Mol Biol Rev. 2005, 69: 326-356. 10.1128/MMBR.69.2.326-356.2005.
Brinkrolf K, Brune I, Tauch A: The transcriptional regulatory network of the amino acid producer Corynebacterium glutamicum. J Biotechnol. 2007, 129: 191-211. 10.1016/j.jbiotec.2006.12.013.
Ibarra JA, Perez-Rueda E, Segovia L, Puente JL: The DNA-binding domain as a functional indicator: the case of the AraC/XylS family of transcription factors. Genetica. 2008, 133: 65-76. 10.1007/s10709-007-9185-y.
Tauch A, Schneider J, Szczepanowski R, Tilker A, Viehoever P, Gartemann KH, Arnold W, Blom J, Brinkrolf K, Brune I, et al: Ultrafast pyrosequencing of Corynebacterium kroppenstedtii DSM44385 revealed insights into the physiology of a lipophilic corynebacterium that lacks mycolic acids. J Biotechnol. 2008, 136: 22-30. 10.1016/j.jbiotec.2008.03.004.
Trost E, Ott L, Schneider J, Schröder J, Jaenicke S, Goesmann A, Husemann P, Stoye J, Dorella FA, Rocha FS, et al: The complete genome sequence of Corynebacterium pseudotuberculosis FRC41 isolated from a 12-year-old girl with necrotizing lymphadenitis reveals insights into gene-regulatory networks contributing to virulence. BMC Genomics. 2010, 11: 728-10.1186/1471-2164-11-728.
Baumbach J, Wittkop T, Kleindt CK, Tauch A: Integrated analysis and reconstruction of microbial transcriptional gene regulatory networks using CoryneRegNet. Nat Protoc. 2009, 4: 992-1005. 10.1038/nprot.2009.81.
Izco JM, Irigoyen A, Torre P, Barcina Y: Effect of the activity levels of the added proteolytic enzyme mixture on free amino acids in ripening Ossau-Iraty cheese. J Chromatogr A. 2000, 881: 69-79. 10.1016/S0021-9673(00)00285-5.
Pasqualone A, Caponio F, Alloggio V, Gomes T: Content of taurine in Apulian Cacioricotta goat's cheese. Eur Food Res Technol. 2000, 211: 158-160. 10.1007/s002179900121.
Siragusa S, De Angelis M, Di Cagno R, Rizzello CG, Coda R, Gobbetti M: Synthesis of gamma-aminobutyric acid by lactic acid bacteria isolated from a variety of Italian cheeses. Appl Environ Microbiol. 2007, 73: 7283-7290. 10.1128/AEM.01064-07.
Nomura M, Kimoto H, Someya Y, Furukawa S, Suzuki I: Production of gamma-aminobutyric acid by cheese starters during cheese ripening. J Dairy Sci. 1998, 81: 1486-1491. 10.3168/jds.S0022-0302(98)75714-5.
McSweeney P, Sousa M: Biochemical pathways for the production of flavour compounds in cheeses during ripening: A review. Lait. 2000, 80: 293-324. 10.1051/lait:2000127.
De Figueroa R, Cerutti de Guglielmone G, Betino de Cardenas I, Oliver G: Flavour compound production and citrate metabolism in Lactococcus rhamnosus ATCC 7469. Milchwissenschaft. 1998, 53: 617-619.
Drinan DF, Robin S, Cogan TM: Citric acid metabolism in hetero- and homofermentative lactic acid bacteria. Appl Environ Microbiol. 1976, 31: 481-486.
Hugenholtz J: Citrate metabolism in lactic acid bacteria. FEMS Microbiol Rev. 1993, 12: 165-178. 10.1111/j.1574-6976.1993.tb00017.x.
Marilley L, Casey MG: Flavours of cheese products: metabolic pathways, analytical tools and identification of producing strains. Int J Food Microbiol. 2004, 90: 139-159. 10.1016/S0168-1605(03)00304-0.
Brocker M, Schaffer S, Mack C, Bott M: Citrate utilization by Corynebacterium glutamicum is controlled by the CitAB two-component system through positive regulation of the citrate transport genes citH and tctCBA. J Bacteriol. 2009, 191: 3869-3880. 10.1128/JB.00113-09.
Kohl TA, Tauch A: The GlxR regulon of the amino acid producer Corynebacterium glutamicum: Detection of the corynebacterial core regulon and integration into the transcriptional regulatory network model. J Biotechnol. 2009, 143: 239-246. 10.1016/j.jbiotec.2009.08.005.
Brennan NM, Cogan TM, Loessner M, Scherer S: Bacterial surface-ripened cheeses. Cheese: Chemistry, Physics, and Microbiology. Edited by: Fox PF, McSweeney PL, Cogan TM, Guinee T. 2004, Amsterdam: Elsevier Academic Press, 2: 199-222.
Mounier J, Rea MC, O'Connor PM, Fitzgerald GF, Cogan TM: Growth characteristics of Brevibacterium, Corynebacterium, Microbacterium, and Staphylococcus spp. isolated from surface-ripened cheese. Appl Environ Microbiol. 2007, 73: 7732-7739. 10.1128/AEM.01260-07.
Kohl TA, Baumbach J, Jungwirth B, Pühler A, Tauch A: The GlxR regulon of the amino acid producer Corynebacterium glutamicum: in silico and in vitro detection of DNA binding sites of a global transcription regulator. J Biotechnol. 2008, 135: 340-350. 10.1016/j.jbiotec.2008.05.011.
Tripaldi C, Martillotti F, Terramoccia S: Content of taurine and other free amino acids in milk of goats bred in Italy. Small Ruminant Res. 1998, 30: 127-136. 10.1016/S0921-4488(98)00095-9.
Erbersdobler HF, Trautwein E, Greulich HG: Determinations of taurine in milk and infant formula diets. Eur J Pediatr. 1984, 142: 133-134. 10.1007/BF00445594.
Rückert C, Milse J, Albersmeier A, Koch DJ, Pühler A, Kalinowski J: The dual transcriptional regulator CysR in Corynebacterium glutamicum ATCC13032 controls a subset of genes of the McbR regulon in response to the availability of sulphide acceptor molecules. BMC Genomics. 2008, 9: 483-10.1186/1471-2164-9-483.
Koch DJ, Rückert C, Albersmeier A, Hüser AT, Tauch A, Pühler A, Kalinowski J: The transcriptional regulator SsuR activates expression of the Corynebacterium glutamicum sulphonate utilization genes in the absence of sulphate. Mol Microbiol. 2005, 58: 480-494. 10.1111/j.1365-2958.2005.04836.x.
Rey DA, Nentwich SS, Koch DJ, Rückert C, Pühler A, Tauch A, Kalinowski J: The McbR repressor modulated by the effector substance S-adenosylhomocysteine controls directly the transcription of a regulon involved in sulphur metabolism of Corynebacterium glutamicum ATCC 13032. Mol Microbiol. 2005, 56: 871-887. 10.1111/j.1365-2958.2005.04586.x.
Brune I, Barzantny H, Klötzel M, Jones J, James G, Tauch A: Identification of McbR as transcription regulator of aecD and genes involved in methionine and cysteine biosynthesis in Corynebacterium jeikeium K411. J Biotechnol. 2011, 151: 22-29. 10.1016/j.jbiotec.2010.11.005.
Ueno H: Enzymatic and structural aspects on glutamate decarboxylase. J Mol Catal. 2000, 10: 67-79. 10.1016/S1381-1177(00)00114-4.
Bown AW, Shelp BJ: The Metabolism and Functions of [gamma]-Aminobutyric Acid. Plant Physiol. 1997, 115: 1-5.
Mavromatis K, Chu K, Ivanova N, Hooper SD, Markowitz VM, Kyrpides NC: Gene context analysis in the Integrated Microbial Genomes (IMG) data management system. PLoS One. 2009, 4: e7979-10.1371/journal.pone.0007979.
Maddocks SE, Oyston PC: Structure and function of the LysR-type transcriptional regulator (LTTR) family proteins. Microbiology. 2008, 154: 3609-3623. 10.1099/mic.0.2008/022772-0.
Kyriakidis N, Tarantili-Georgiou K, Tsani-Batzaka E: Nitrate and nitrite content of greek cheeses. Journal of Food Composition and Analysis. 1997, 10: 343-349. 10.1006/jfca.1997.0549.
Hirai T, Subramaniam S: Structure and transport mechanism of the bacterial oxalate transporter OxlT. Biophys J. 2004, 87: 3600-3607. 10.1529/biophysj.104.049320.
Center of Disease Control and Prevention: Effects of switching from whole to low-fat/fat-free milk in public schools - New York city, 2004-2009. Morb Mortal Wkly Rep. 2010, 59: 70-73.
Monnet C, Loux V, Gibrat JF, Spinnler E, Barbe V, Vacherie B, Gavory F, Gourbeyre E, Siguier P, Chandler M, et al: The Arthrobacter arilaitensis Re117 genome sequence reveals its genetic adaptation to the surface of cheese. PLoS One. 2010, 5: e15489-10.1371/journal.pone.0015489.
Wolf IV, Meinardi CA, Zalazar CA: Production of flavour compounds from fat during cheese ripening by action of lipases and esterases. Protein Pept Lett. 2009, 16: 1235-1243. 10.2174/092986609789071289.
Marchler-Bauer A, Lu S, Anderson JB, Chitsaz F, Derbyshire MK, DeWeese-Scott C, Fong JH, Geer LY, Geer RC, Gonzales NR, et al: CDD: a Conserved Domain Database for the functional annotation of proteins. Nucleic Acids Res. 2011, 39: D225-229. 10.1093/nar/gkq1189.
Gavalda S, Leger M, van der Rest B, Stella A, Bardou F, Montrozier H, Chalut C, Burlet-Schiltz O, Marrakchi H, Daffe M, Quemard A: The Pks13/FadD32 crosstalk for the biosynthesis of mycolic acids in Mycobacterium tuberculosis. J Biol Chem. 2009, 284: 19255-19264. 10.1074/jbc.M109.006940.
Black PN, DiRusso CC: Transmembrane movement of exogenous long-chain fatty acids: proteins, enzymes, and vectorial esterification. Microbiol Mol Biol Rev. 2003, 67: 454-472. 10.1128/MMBR.67.3.454-472.2003.
Hiltunen JK, Qin Y: Beta-oxidation - strategies for the metabolism of a wide variety of acyl-CoA esters. Biochim Biophys Acta. 2000, 1484: 117-128.
Bayan N, Houssin C, Chami M, Leblon G: Mycomembrane and S-layer: two important structures of Corynebacterium glutamicum cell envelope with promising biotechnology applications. J Biotechnol. 2003, 104: 55-67. 10.1016/S0168-1656(03)00163-9.
Portevin D, De Sousa-D'Auria C, Houssin C, Grimaldi C, Chami M, Daffe M, Guilhot C: A polyketide synthase catalyzes the last condensation step of mycolic acid biosynthesis in mycobacteria and related organisms. Proc Natl Acad Sci USA. 2004, 101: 314-319. 10.1073/pnas.0305439101.
Gande R, Gibson KJ, Brown AK, Krumbach K, Dover LG, Sahm H, Shioyama S, Oikawa T, Besra GS, Eggeling L: Acyl-CoA carboxylases (accD2 and accD3), together with a unique polyketide synthase (Cg-pks), are key to mycolic acid biosynthesis in Corynebacterianeae such as Corynebacterium glutamicum and Mycobacterium tuberculosis. J Biol Chem. 2004, 279: 44847-44857. 10.1074/jbc.M408648200.
Portevin D, de Sousa-D'Auria C, Montrozier H, Houssin C, Stella A, Laneelle MA, Bardou F, Guilhot C, Daffe M: The acyl-AMP ligase FadD32 and AccD4-containing acyl-CoA carboxylase are required for the synthesis of mycolic acids and essential for mycobacterial growth: identification of the carboxylation product and determination of the acyl-CoA carboxylase components. J Biol Chem. 2005, 280: 8862-8874.
Barzantny H, Brune I, Tauch A: Molecular basis of human body odour formation: insights deduced from corynebacterial genome sequences. Int J Cosmet Sci. 2011
Deetae P, Bonnarme P, Spinnler HE, Helinck S: Production of volatile aroma compounds by bacterial strains isolated from different surface-ripened French cheeses. Appl Microbiol Biotechnol. 2007, 76: 1161-1171. 10.1007/s00253-007-1095-5.
Gobbetti M, Smacchi E, Semeraro M, Fox PF, Lanciotti R, Cogan T: Purification and characterization of an extracellular proline iminopeptidase from Corynebacterium variabilis NCDO 2101. J Appl Microbiol. 2001, 90: 449-456. 10.1046/j.1365-2672.2001.01264.x.
Ling M, Allen SW, Wood JM: Sequence analysis identifies the proline dehydrogenase and delta 1-pyrroline-5-carboxylate dehydrogenase domains of the multifunctional Escherichia coli PutA protein. J Mol Biol. 1994, 243: 950-956. 10.1006/jmbi.1994.1696.
Jochmann N, Kurze AK, Czaja LF, Brinkrolf K, Brune I, Hüser AT, Hansmeier N, Pühler A, Borovok I, Tauch A: Genetic makeup of the Corynebacterium glutamicum LexA regulon deduced from comparative transcriptomics and in vitro DNA band shift assays. Microbiology. 2009, 155: 1459-1477. 10.1099/mic.0.025841-0.
Krämer R: Osmosensing and osmosignaling in Corynebacterium glutamicum. Amino Acids. 2009, 37: 487-497. 10.1007/s00726-009-0271-6.
Klein J, Schwarz T, Lentzen G: Ectoine as a natural component of food: detection in red smear cheeses. J Dairy Res. 2007, 74: 446-451.
Follmann M, Becker M, Ochrombel I, Ott V, Krämer R, Marin K: Potassium transport in Corynebacterium glutamicum is facilitated by the putative channel protein CglK, which is essential for pH homeostasis and growth at acidic pH. J Bacteriol. 2009, 191: 2944-2952. 10.1128/JB.00074-09.
Ochrombel I, Ott L, Krämer R, Burkovski A, Marin K: Impact of improved potassium accumulation on pH homeostasis, membrane potential adjustment and survival of Corynebacterium glutamicum. Biochim Biophys Acta. 2011, 1807: 444-450. 10.1016/j.bbabio.2011.01.008.
Patterson K, Bhagwat S, Williams J, Howe J, Holden J: USDA Database for the choline content of common foods. Book USDA Database for the choline content of common foods. 2008, City, 2:
Morbach S, Krämer R: Osmoregulation in Corynebacterium glutamicum. Corynebacterium glutamicum. Edited by: Eggeling L, Bott M. 2005, Boca Raton: CRC Press LLC, 417-431.
Möker N, Brocker M, Schaffer S, Krämer R, Morbach S, Bott M: Deletion of the genes encoding the MtrA-MtrB two-component system of Corynebacterium glutamicum has a strong influence on cell morphology, antibiotics susceptibility and expression of genes involved in osmoprotection. Mol Microbiol. 2004, 54: 420-438. 10.1111/j.1365-2958.2004.04249.x.
Hunter S, Apweiler R, Attwood TK, Bairoch A, Bateman A, Binns D, Bork P, Das U, Daugherty L, Duquenne L, et al: InterPro: the integrative protein signature database. Nucleic Acids Res. 2009, 37: D211-215. 10.1093/nar/gkn785.
Baker LM, Poole LB: Catalytic mechanism of thiol peroxidase from Escherichia coli. Sulfenic acid formation and overoxidation of essential CYS61. J Biol Chem. 2003, 278: 9203-9211. 10.1074/jbc.M209888200.
Pomposiello PJ, Demple B: Redox-operated genetic switches: the SoxR and OxyR transcription factors. Trends Biotechnol. 2001, 19: 109-114. 10.1016/S0167-7799(00)01542-0.
Lushchak VI: Adaptive response to oxidative stress: Bacteria, fungi, plants and animals. Comp Biochem Physiol C Toxicol Pharmacol. 2011, 153: 175-190. 10.1016/j.cbpc.2010.10.004.
Pagan-Ramos E, Song J, McFalone M, Mudd MH, Deretic V: Oxidative stress response and characterization of the oxyR-ahpC and furA-katG loci in Mycobacterium marinum. J Bacteriol. 1998, 180: 4856-4864.
Pagan-Ramos E, Master SS, Pritchett CL, Reimschuessel R, Trucksis M, Timmins GS, Deretic V: Molecular and physiological effects of mycobacterial oxyR inactivation. J Bacteriol. 2006, 188: 2674-2680. 10.1128/JB.188.7.2674-2680.2006.
Butala M, Zgur-Bertok D, Busby SJ: The bacterial LexA transcriptional repressor. Cell Mol Life Sci. 2009, 66: 82-93. 10.1007/s00018-008-8378-6.
Andrews SC, Robinson AK, Rodriguez-Quinones F: Bacterial iron homeostasis. FEMS Microbiol Rev. 2003, 27: 215-237. 10.1016/S0168-6445(03)00055-X.
Oram JD, Reiter B: Inhibition of bacteria by lactoferrin and other iron-chelating agents. Biochim Biophys Acta. 1968, 170: 351-365. 10.1016/0304-4165(68)90015-9.
Ong SA, Neilands JB: Siderophores in microbially processed cheese. J Agric Food Chem. 1979, 27: 990-995. 10.1021/jf60225a033.
Brown JS, Holden DW: Iron acquisition by Gram-positive bacterial pathogens. Microbes Infect. 2002, 4: 1149-1156. 10.1016/S1286-4579(02)01640-4.
Dertz EA, Stintzi A, Raymond KN: Siderophore-mediated iron transport in Bacillus subtilis and Corynebacterium glutamicum. J Biol Inorg Chem. 2006, 11: 1087-1097. 10.1007/s00775-006-0151-4.
Braun V: Iron uptake by Escherichia coli. Front Biosci. 2003, 8: s1409-1421. 10.2741/1232.
May JJ, Wendrich TM, Marahiel MA: The dhb operon of Bacillus subtilis encodes the biosynthetic template for the catecholic siderophore 2, 3-dihydroxybenzoate-glycine-threonine trimeric ester bacillibactin. J Biol Chem. 2001, 276: 7209-7217. 10.1074/jbc.M009140200.
Touati D: Iron and oxidative stress in bacteria. Arch Biochem Biophys. 2000, 373: 1-6. 10.1006/abbi.1999.1518.
Brune I, Werner H, Hüser AT, Kalinowski J, Pühler A, Tauch A: The DtxR protein acting as dual transcriptional regulator directs a global regulatory network involved in iron metabolism of Corynebacterium glutamicum. BMC Genomics. 2006, 7: 21-10.1186/1471-2164-7-21.
Yvon M, Rijnen L: Cheese flavour formation by amino acid catabolism. Int Dairy J. 2001, 11: 185-201. 10.1016/S0958-6946(01)00049-8.
Amarita F, Yvon M, Nardi M, Chambellon E, Delettre J, Bonnarme P: Identification and functional analysis of the gene encoding methionine-gamma-lyase in Brevibacterium linens. Appl Environ Microbiol. 2004, 70: 7348-7354. 10.1128/AEM.70.12.7348-7354.2004.
Dias B, Weimer B: Conversion of methionine to thiols by lactococci, lactobacilli, and brevibacteria. Appl Environ Microbiol. 1998, 64: 3320-3326.
Smit G, Smit BA, Engels WJ: Flavour formation by lactic acid bacteria and biochemical flavour profiling of cheese products. FEMS Microbiol Rev. 2005, 29: 591-610. 10.1016/j.fmrre.2005.04.002.
Le Bars D, Yvon M: Formation of diacetyl and acetoin by Lactococcus lactis via aspartate catabolism. J Appl Microbiol. 2008, 104: 171-177.
Dimos A, Urbach G, Miller A: Changes in flavour and volatiles of full-fat and low-fat cheeses during maturation. Int Dairy J. 1996, 6: 981-995. 10.1016/S0958-6946(97)84214-8.
Bockelmann W, Hoppe-Seyler T, Krusch U, Hoffmann W, Heller K: The microflora of Tilsit cheese. Part 2. Development of a surface smear starter culture. Nahrung. 1997, 41: 208-212. 10.1002/food.19970410405.
Mookadam F, Cikes M, Baddour LM, Tleyjeh IM, Mookadam M: Corynebacterium jeikeium endocarditis: a systematic overview spanning four decades. Eur J Clin Microbiol Infect Dis. 2006, 25: 349-353. 10.1007/s10096-006-0145-8.
Noordman WH, Reissbrodt R, Bongers RS, Rademaker JL, Bockelmann W, Smit G: Growth stimulation of Brevibacterium sp. by siderophores. J Appl Microbiol. 2006, 101: 637-646. 10.1111/j.1365-2672.2006.02928.x.
Bassit N, Boquien CY, Picque D, Corrieu G: Effect of initial oxygen concentration on diacetyl and acetoin production by Lactococcus lactis subsp. lactis biovar diacetylactis. Appl Environ Microbiol. 1993, 59: 1893-1897.
Dherbecourt J, Bourlieu C, Maillard MB, Aubert-Frogerais L, Richoux R, Thierry A: Time course and specificity of lipolysis in Swiss cheese. J Agric Food Chem. 2010, 58: 11732-11739. 10.1021/jf102572z.
Picon A, Gaya P, Fernandez-Garcia E, Rivas-Canedo A, Avila M, Nunez M: Proteolysis, lipolysis, volatile compounds, texture, and flavor of Hispanico cheese made using frozen ewe milk curds pressed for different times. J Dairy Sci. 2010, 93: 2896-2905. 10.3168/jds.2009-3041.
Awad S, Ahmed N, El Soda M: Influence of microfiltration and adjunct culture on quality of Domiati cheese. J Dairy Sci. 2010, 93: 1807-1814. 10.3168/jds.2009-2649.
Tauch A, Bischoff N, Pühler A, Kalinowski J: Comparative genomics identified two conserved DNA modules in a corynebacterial plasmid family present in clinical isolates of the opportunistic human pathogen Corynebacterium jeikeium. Plasmid. 2004, 52: 102-118. 10.1016/j.plasmid.2004.05.003.
Hopwood D: Genetic Manipulation of Streptomyces: A Laboratory Manual. 1985, New York: Cold Spring Harbor Laboratory Press
McHardy AC, Goesmann A, Pühler A, Meyer F: Development of joint application strategies for two microbial gene finders. Bioinformatics. 2004, 20: 1622-1631. 10.1093/bioinformatics/bth137.
Suzek BE, Ermolaeva MD, Schreiber M, Salzberg SL: A probabilistic method for identifying start codons in bacterial genomes. Bioinformatics. 2001, 17: 1123-1130. 10.1093/bioinformatics/17.12.1123.
Linke B, McHardy AC, Neuweger H, Krause L, Meyer F: REGANOR: a gene prediction server for prokaryotic genomes and a database of high quality gene predictions for prokaryotes. Appl Bioinformatics. 2006, 5: 193-198. 10.2165/00822942-200605030-00008.
Carver T, Thomson N, Bleasby A, Berriman M, Parkhill J: DNAPlotter: circular and linear interactive genome visualization. Bioinformatics. 2009, 25: 119-120. 10.1093/bioinformatics/btn578.
Schneider J, Vorhölter FJ, Trost E, Blom J, Musa YR, Neuweger H, Niehaus K, Schatschneider S, Tauch A, Goesmann A: CARMEN - Comparative analysis and in silico reconstruction of organism-specific metabolic networks. Genet Mol Res. 2010, 9: 1660-1672. 10.4238/vol9-3gmr901.
Funahashi A, Jouraku A, Matsuoka Y, Kitano H: Integration of CellDesigner and SABIO-RK. In Silico Biol. 2007, 7: S81-90.
The valuable contribution of the CeBiTec genome sequencing team (A. Albersmeier, K. Brinkrolf, and R. Szczepanowski) to this genome project is greatly acknowledged. The authors thank S. Jaenicke and J. Schneider (BRF/CeBiTec) for implementing and maintaining the C. variabile GenDB project and for additional bioinformatic support. The authors acknowledge support of the publication fee by Deutsche Forschungsgemeinschaft and the Open Access Publication Funds of Bielefeld University.
JS sequenced and annotated the C. variabile genome and prepared the manuscript. IM participated in the gap closure process. ET and AT participated in data evaluation. AT supervised the project. All authors read and approved the final version of the manuscript.
Electronic supplementary material
Authors’ original submitted files for images
Below are the links to the authors’ original submitted files for images.