The ability of a bacterial population to survive in different niches, as well as in stressful and rapidly changing environmental conditions, depends greatly on its genetic content. To survive such fluctuating conditions, bacteria have evolved different mechanisms to modulate phenotypic variations and related strategies to produce high levels of genetic diversity. Laboratories working in microbiological diagnosis have shown that Citrobacter freundii is very versatile in its colony morphology, as well as in its biochemical, antigenic and pathogenic behaviours. This phenotypic versatility has made C. freundii difficult to identify and it is frequently confused with both Salmonella enterica and Escherichia coli. In order to determine the genomic events and to explain the mechanisms involved in this plasticity, six C. freundii isolates were selected from a phenotypic variation study. An I-CeuI genomic cleavage map was created and eight housekeeping genes, including 16S rRNA, were sequenced. In general, the results showed a range of both phenotypes and genotypes among the isolates with some revealing a greater similarity to C. freundii and some to S. enterica, while others were identified as phenotypic and genotypic intermediary states between the two species. The occurrence of these events in natural populations may have important implications for genomic diversification in bacterial evolution, especially when considering bacterial species boundaries. In addition, such events may have a profound impact on medical science in terms of treatment, course and outcomes of infectious diseases, evading the immune response, and understanding host-pathogen interactions.
Citation: Delgado G, Souza V, Morales R, Cerritos R, González-González A, Méndez JL, et al. (2013) Genetic Characterization of Atypical Citrobacter freundii. PLoS ONE 8(9): e74120. https://doi.org/10.1371/journal.pone.0074120
Editor: Mark Alexander Webber, University of Birmingham, United Kingdom
Received: February 28, 2013; Accepted: July 29, 2013; Published: September 12, 2013
Copyright: © 2013 Delgado et al. This is an open-access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Funding: This research project was supported by grants from Dirección General de Apoyo al Personal Académico, UNAM (Pappitt project IN242602), URL:http://dgapa.unam.mx; and Consejo Nacional de Ciencia y Tecnología (project 100343), URL:http://conacyt.gob.mx. The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.
Competing interests: The authors have declared that no competing interests exist.
The correct identification and determination of bacterial species plays an important role in the diagnosis of infectious diseases, food safety, and epidemiology. Human response to bacterial infection depends on the biological properties of the infecting microorganism, as well as on the natural variation of the species and its potential to change . These variations are the adaptive responses by which bacteria undergo frequent and reversible phenotypic changes resulting from genetic modifications in their genomes . Despite the different molecular mechanisms that mediate and regulate phenotypic variation, these related gene regulatory systems are not sufficient to allow bacteria to manage environmental challenges. To survive, bacteria have evolved assorted strategies to produce high levels of genetic diversity through complex combinations and reshuffling of genetic information , . This heterogenic response enables a bacterium to utilize different niches within an ecosystem, and even to increase the overall fitness of the species, all of which are reflected in life-style differences and physiological versatility , .
Although the chromosomal organization of related bacteria is conserved, a considerable genomic variability and diversification between isolates of the same species has been observed as a result of distinct molecular events involved in the adaptive evolution of the genomes . Point mutations, homologous recombination, deletions, duplications, inversions, and horizontal gene transfer contribute to this variation, and consequently to genome diversification.
The dynamic nature of these bacterial gene pools, and in particular the ability of genes to move between non-related bacterial groups, has complicated the concept of species coherence for such microorganisms. In this respect, species boundary for bacteria continues to be controversial, especially in terms of the mechanisms of bacterial speciation and maintenance in the face of frequent genetic interchange . Some species of the genus Citrobacter present this type of behaviour  and it is this fact that has made C. freundii difficult to identify, since the species is similar to other members of the Enterobacteriaceae family. Although the majority of isolates of C. freundii are frequently identified as belonging to typical strains, atypical biotypes also exist, differing from typical biotypes in that they have variable culture, biochemical and antigenic characteristics that resemble those of the genus Salmonella and Escherichia coli . This plasticity has prevented a definitive classification of C. freundii and has created a taxonomic problem that has been under discussion for many years, with classification and nomenclature varying over time according to the author concerned –.
In addition, the Citrobacter genus has a large number of somatic (O) and flagella (H) antigens typically found in other species of the Enterobacteriaceae family, in particularly in Salmonella , . Furthermore, studies of lipopolysaccharide (LPS) structures in enteric bacteria have highlighted the structural mimicry of the somatic antigen of C. freundii –.
Other studies of C. freundii have identified virulence genes that are homologous or identical to those described in E. coli pathotypes and Salmonella –. DNA-DNA hybridization ,  used to assess divergence in species of Citrobacter and Citrobacter-like strains has also revealed substantial heterogeneity within strains identified as C. freundii. Between 1960 and 1990, various laboratories – began to isolate strains that could be treated as atypical C. freundii or as strains of lactose-positive Salmonella enterica associated with gastroenteritis. These findings highlighted the close similarity between some C. freundii strains and Salmonella enterica and the confusion that may arise in differentiating between these two genera.
The availability of a significant number of bacterial genomes has enabled the phylogenetic relationship among bacterial populations to be assessed. Such detailed research will allow a better definition of the different isolates, pathotypes and even species , .
A number of recent studies have started to shed light on the dynamic chromosomal organization, genetic content and genetic diversity of Citrobacter –. The aim of the current study was to detect the genomic changes presented in atypical C. freundii isolates during biochemical phenotype variation, and to discuss the implications of such changes with regard to their possible impact on the boundaries of the species.
Materials and Methods
Background of the Bacterial Isolates
Weekly faecal samples were obtained from each child in the study over a two-year period. Isolation, identification and classification of the enteric pathogens were carried out as described by Edwards and Ewing , . Strains identified as typical and atypical C. freundii were further used for this study. Atypical C. freundii differ mainly in their capacity to ferment lactose, decarboxylase lysine and utilize potassium cyanide (KCN). All of the isolated strains were labelled and stored on sealed Dorset egg slants, in a temperature-controlled environment.
In late 1989, V. Vázquez and A. Cravioto conducted a phenotypic variation study (data not published) with the C. freundii strains that had been isolated from the previously described longitudinal survey , . The bacteriological methods used in the study included the repeated streaking of single colonies on Salmonella-Shigella agar (SS agar) and their cultivation to avoid the possibility of both contamination and mixing of the microorganisms, all of which led to the selection of pure colonies. A strain initially identified as atypical C. freundii was selected for further studies by performing subsequent passages under identical conditions. Briefly, the atypical C. freundii isolate was grown overnight in a sterile liquid media. The resulting culture was streaked onto SS agar plates and incubated overnight at 37°C. Following incubation, all of the single colonies showing morphologic differences were selected, and then, each one were grown in different biochemical test media. The biochemical profile of each isolate were observed and scored (Figure 1, Table 1). Cells from each single colony were used to found the next selection round and so on (Figure 1). The methods of cultivation, isolation and environment conditions were exactly the same of those of the previous round. This procedure was repeated throughout all 10 selective rounds (Figure 1). This resulted in the creation of a large number of descendants, which could then be used to determine the phenotypic and genomic characteristics that made these isolates different from typical C. freundii strains (supporting information). Single colonies from these cultures were selected using a stereoscopic microscope . The selected colonies were characterized using forty standard biochemical tests as described by Ewing (Table 1).
Shows the lineages A and B derived from the Citrobacter freundii FMU108327/P, each dot represents a selected colony. The pink dots indicate a typical C. freundii biochemical phenotype; yellow dots indicate a “Salmonella-like” biochemical phenotype; the dots in gradient of pink to yellow represent the intermediates biochemical phenotypes. The solid lines indicate the number of colonies selected after each selective round. The parental strain was Citrobacter freundii FMU108327/P, and after the first selection two descendants were generated, which were used to found the next selection, which led to six descendants (in both lineages) being selected, and all of these isolates were used to found the 3th selection round. This procedure was repeated throughout the 10 selections. The numbered dots P, A1, A10, B1, B2 and B10 indicate the isolates selected for this study.
Approximately 15,500 isolates were obtained in these sequential passages, of which 23% presented the typical biochemical phenotype of C. freundii, 14% showed biochemical patterns that resembled Salmonella spp., and the remaining 63% presented variable biochemical phenotypes, making precise identification of the atypical isolates difficult using this set of tests (Figure 1, Table 1).
All of the isolates obtained from the Vázquez and Cravioto study were labelled and stored on sealed Dorset egg slants in a temperature-controlled room that housed the culture collection at the Faculty of Medicine UNAM (FMU) of the Universidad Nacional Autónoma de México (UNAM).
In the present study, six isolates derived from the Vázquez and Cravioto study were selected for further study: FMU108327 (P), and its derived isolates FMU108327/A1 (A1), FMU108327/A10 (A10), FMU108327/B1 (B1), FMU108327/B2 (B2) and FMU108327/B10 (B10) (Figure 1). These isolates were selected according to the changes found in their biochemical phenotype. To dispel any doubts that the ensuing results were due to contamination or careless mix-up when the colonies were picked from the agar plates, the purity of the selected isolates were verified using two methods: 1) streaking on Salmonella-Shigella agar (SS agar) to ensure the presence of isolated colonies followed by overnight incubation at 37°C. A colony of each isolate was then selected with the aid of a stereoscopic microscope in a sterile area ,  and standard biochemical tests were applied to confirm the original results  (Table 1); 2) the derived isolated colonies were diluted and vortexed in 0.1 ml of 0.01% Tween 40 and 0.1 M MgSO4 (V/V) and streaked onto SS agar to recheck the purity of the colony , .
Strains of Salmonella enterica serovar Typhimurium LT2 ATCC 700720 obtained from the Enteric Pathogen Reference Laboratory, Public Health Department of the Facultad de Medicina, UNAM, and C. freundii E9750 NCTC from the NCTC culture collection in London, UK, were used as reference strains. For subsequent analyses, bacteria were routinely cultured on Luria-Bertani (LB) medium and incubated overnight at 37°C. For the Pulsed Field Gel Electrophoresis (PFGE), only two isolates were used, C. freundii FMU108327/P and FMU108327/A10. All isolates deriving from this process were labelled and stored at −70°C as glycerol stocks.
I-CeuI Cleavage Map of the FMU108327/P and 108327/A10 Genome
Pulsed field gel electrophoresis (PFGE) was used to construct the ribosome skeleton of the chromosome from the FMU108327/P and FMU108327/A10 isolates. Bacterial genomic DNA from S. Typhimurium LT2 cleaved with I-CeuI was used as a molecular-weight marker. Since the size of these identifying fragments and gene markers had been determined previously , , their inclusion improved the precision of chromosome size estimation.
Preparation and Digestion of Genomic DNA in Agarose Blocks
Genomic DNA of each strain was prepared in agarose blocks using a previously described method  with some modifications. Briefly, a second incubation in proteinase K solution was carried out to increase the purity of the DNA. Agarose blocks were pre-incubated in 1× buffer for 30 min at 37°C and then incubated overnight with 100 µl of fresh 1× buffer containing 4 U of I-CeuI (New England Biolabs) at 37°C to digest DNA in the blocks.
Pulsed Field Gel Electrophoresis (PFGE)
I-CeuI (hereafter called CeuI) fragments were separated by a CHEF Mapper electrophoresis system (Bio-Rad). Electrophoresis was performed on a 1% agarose (Seakem Gold agarose, BioWhittaker Molecular Applications) gel and 0.5× TBE (90 mM Tris, 90 mM boric acid, 2 mM EDTA, pH8) buffer at 12°C. Three stage conditions were applied to the electrophoresis process: stage 1: pulse time ramped from 50 s to 2 min over 21 hrs; stage 2: pulse time ramped from 20 s to 1 min 20 s over 10 hrs; and stage 3: pulse time ramped from 3 s to 12 s over 7 hrs at 4 V/cm. The DNA fragments were then transferred onto N+nylon membrane (Amersham Biosciences) by Southern blotting as described previously .
Preparation of DNA Probe and Hybridization
The individual fragments of FMU108327/P and FMU108327/A10 corresponding to those of S. Typhimurium LT2 were determined by probing the Southern blotting membranes with polymerase chain reaction (PCR) products from S. Typhimurium LT2 genes whose location in the chromosome is already known (Figure 2). The genes metB, metC, metE, metH, fliC  and gnd were selected according to Liu et al. . These were used as marker genes to identify the individual CeuI fragments (metB on CeuI fragment E, metC on CeuI-B, metE on CeuI-D and metH on CeuI-G, fliC and gnd on CeuI-A). The primer sets used to amplify the metB, metC, metE, metH and gnd genes were designed in the laboratory (Table S1) from previously sequenced S. Typhimurium LT2 genome (AE006468)  using the Primer Select program of the DNASTAR Lasergene 7 package (DNASTAR, Inc., Madison, WI). Correspondence of Citrobacter fragments CeuI-C and CeuI-F were determined by probing the isolated DNA from S. Typhimurium LT2 CeuI fragments C and F. The PCR products and restriction fragments CeuI-C and F were purified and then 32P-labelled using the random primer method of the Megaprime DNA labelling system (Amersham Biosciences). The membrane was incubated in 10 ml hybridization solution (Rapid-hyb buffer) to which 1× Denhardt’s solution and 0.2 mg/ml of denatured sheared salmon sperm were added. Incubation was carried out at 60°C with constant, gentle shaking for 3 h. The labelled probe was then added to fresh hybridization solution and hybridization was carried out overnight at 60°C with constant and gentle shaking. The membrane was exposed to X-ray film after being washed at high astringency (64°C).
a) Pulsed-field gel shows CeuI profiles: line 1, C. freundii E9750 NCTC; lane 2, FMU108327/P; lane 3, FMU108327/A10 and lane 4; S. Typhimurium LT2 ATCC 700720. The results of Southern blotting analysis appear with capital letters. These indicate the Ceús individual fragments of C. freundii corresponding to those of S. Typhimurium LT2; b) chromosome sizes and molecular weights of individual CeuI fragments. It also shows the location of every marker gene according to its correspondence in each CeuI fragment. The CeuI-C and CeuI-F fragments of C. freundii were determined by probing the isolated DNA from S. Typhimurium LT2 CeuI-C and CeuI-F fragments; c) the gel from panel a were blotted to N+nylon membranes and probed as follows: lanes 1–4 metH (Table S1); lanes 5–8 DNA from C. freundii fragment CeuI-H and lanes 9–12, metA. CeuI fragments identified by the probes are indicated inside the images with capital letters.
PCR and Hybridization to Determine rrn Arrangement
The relative chromosomal position of each rrn operon was determined by PCR amplification of the genes adjacent to each operon. These genes are at the opposite ends of each CeuI fragment. The sequences of the used primer and their locations relative to proximal rrn operons are shown in Table S1. Primers were designed in the laboratory based on a S. Typhimurium LT2 sequence. The PCR products were purified and then 32P-labelled before being probed onto Southern blots of CeuI digested DNA.
Chromosomal DNA Isolation for PCR
Chromosomal DNA was isolated from stationary-phase bacteria taken from cultures grown overnight. DNA was purified by miniprep of bacterial genomic DNA as described previously by Murray and Thompson .
Detection of cadA and lacY Genes
The adaptive genes, cadA and lacY that encode lysine decarboxylase and lactose permease, respectively, were amplified in order to detect their presence in the C. freundii E9750 strain, the atypical FMU108327/P isolate and its derivatives, since these genes enhance bacterial fitness and confer survival advantage. The primers used to amplify these gene fragments were designed in the laboratory (Table S1). The PrimerSelect program of the DNASTAR Lasergene 7 package (DNASTAR, Inc., Madison, WI) was used to design cadA primers from previously sequenced E. coli (U00096, AE005174) ,  and S. enterica genomes (AE006468, AE014613, AL513382) , , . The lacY primer set was designed from the sequenced Citrobacter freundii lacY gene (U13675) . Thermal cycling conditions for both genes were as follows: an initial denaturation cycle at 94°C for 5 min, followed by 30 cycles at 94°C for 1 min, annealing temperature (Table S1) for 1 min and 72°C for 1 min with a final cycle of 72°C for 7 min.
Amplification and Sequencing Housekeeping Genes and 16S rRNA Gene
Eight housekeeping gene fragments and the 16S rRNA gene were amplified and sequenced for the C. freundii E9750 strain, the atypical FMU108327/P and its derivative isolates. The eight genes referred to here were: adk (U00096), aph (U00096, AE005174, AE006468, and AE014613), gnd (U00096, AE005174, AE006468, and AE014613), gyrB (U00096), icdA (U00096), mdh (U00096), purA (U00096) and recA (U00096). Some of the primers used to amplify these loci were obtained from the E. coli MLST website (http://mlst.ucc.ie/mlst/dbs/Ecoli)  while others were designed de novo to target other housekeeping genes (Table S1). The PrimerSelect program of the DNASTAR Lasergene 7 package (DNASTAR, Inc., Madison, WI) was used for primer design from previously sequenced E. coli and S. enterica genomes. PCR protocols to amplify adk, gyrB, mdh, purA and recA were performed according to the procedure specified by Wirth et al.  with minor modifications for annealing temperatures (Table S1). Thermal cycling conditions for the aph, icdA and gnd genes were as follows: an initial denaturation cycle at 94°C for 5 min, followed by 30 cycles at 94°C for 1 min, annealing temperature (according to each specific primer set Table S1) for 1 min and 72°C for 1 min with a final cycle of 72°C for 7 min. Platinum Taq DNA polymerase (Invitrogen, life technologies) was used to amplify all genes. PCR products were sequenced by High Throughput Genomics Unit at the University of Washington (http://www.htseq.org).
The sequences were obtained in forward and reverse directions, and electrophenograms for all sequences were visually inspected for consistency between strands, with any ambiguous nucleotides being resolved by re-sequencing. The SeqMan program of the DNASTAR Lasergene 7 package (DNASTAR, Inc., Madison, WI) was used to edit, trim, and assemble each sequence. The ClustalW2 program of Bioedit software version 184.108.40.206 was used to align all the sequences. The coding sequences used for the housekeeping gene fragments were read in-frame.
The GenBank IDs for the genes sequenced in this study are as follows: Accession numbers: adk (JQ606828, JX001409–JX001414), aph (JX001416–JX001422), gnd (JX001424–JX001430), gyrB (JX001432–JX001438), mdh (JX001440–JX001446), purA (JX001448–JX001454), recA (JX001456–JX001462), icd (JQ860228, JX001470–JX001474), and 16S rRNA (JX001463–JX001469).
Phylogenies based on both individual and concatenated allele sequences (haplotype) were reconstructed using PhyML software, version 3.0 . ModelTest  was used to select the optimal evolutionary model by evaluating the selected parameters using the Akaike Information Criteria (AIC). A corrected version of the AIC (AICc) was used for each locus sequence data set since the sample size (n) was small compared with the number of parameters (n/K<40). This approach suggested the following models: adk (TIM2ef+G), aph (TIM2+G), gnd (TIM3ef+I+G), gyrB (GTR+I+G), icdA (TIM2ef+I+G), recA, purA and mdh (TIM3+I+G), and the complete concatenated sequences data set (GTR+I+G). The robustness of nodes was assessed using 1000 bootstrap replicates for each locus separately and for the concatenated data set. This analysis was run using the same version of the PhyML software. Finally, trees were edited using MEGA version 5.01 .
In order to assess the role of recombination in the studied strains, the linkage model “STRUCTURE”  was run assuming a K value between 2 and 8 (20 replicates per value of K) populations with a burning of 25,000 iterations and 50,000 MCMC subsequent iterations. The optimal value was K = 4 as ascertained by comparing the posterior probabilities of the data given each value of K from 2 to 8 (Figure S1). Assignment of haplotypes to each ancestral population was performed as described by Wirth et al. .
For phylogenetic studies other enterobacteria sequences were also used. The sequences for the following strains were downloaded from the National Center of Biotechnology Information: Citrobacter freundii ballerup 7851, C. freundii 4_7_47CFAA, C. freundii GTC 09479, C. freundii GTC 09629, C. freundii MTCC1658, C. koseri, C. rodentium, Citrobacter sp. 30_2, C. youngae, Cronobacter sakazakii, Escherichia alberti, E. fergusonii, E. coli strains APEC O1, CFT073, B171, BL21-Gold(DE3), HS, O157:H7 strains, EDL933 and Sakai, K-12 substrains, MG1655, O111:H-, Klebsiella pneumoniae strains 342, NTUH-K2044, MGH 78578, rhinoscleromatis, Shigella boydii, Sh. flexneri, Salmonella enterica subsp. Arizona, S. enterica serovars: Gallinarum, Kentucky, Typhimurium LT2, Newport, Paratyphi A, Saintpaul, Schwarzengrund, Typhi strains CT18 and Ty2, Virchow and Weltevreden. Accession numbers can be found in Table 2.
DNA Polymorphism Analyses
The number of substitutions per site in terms of both synonymous and nonsynonymous changes was calculated using DnaSP v 5.10 in non-overlapping windows of 30 bp .
The isolates used in this study came from a previous study in which the aim was to determine the phenotypic changes suffered by the atypical C. freundii FMU108327 isolate systematically (Supporting information). The experimental environment consisted of a serial transfer regime in which the C. freundii was subjected to repeated subsequent cultivation followed by the selection of single colonies (Figure 1). The detail can be found in the method section above. Considering the variability in phenotypes, the original atypical C. freundii FMU108327/P isolate was selected with five derivative isolates from two different branches (Figure 1).
Biochemical Variability of Atypical FMU108327 Strain
The six selected isolates and two reference strains were streaked onto both blood agar and SS agar plates and incubated at 37°C. On inspection the following day, the original FMU108327/P (P) strain and its derived isolates, as well as the reference Citrobacter freundii E9750 and Salmonella Typhimurium LT2 strains produced unique and uniform colonies. Additionally, in SS agar, the strains of C. freundii E9750, FMU108327/P, and the isolates FMU108327/A1 and FMU108327/B1 yielded pink colonies indicating that these bacteria did ferment lactose. However, the FMU108327/A10, FMU108327/B2 and FMU108327/B10 isolates produced colourless colonies, indicating that these did not ferment lactose and/or colonies with a black dot indicative of iron sulphide production. Regarding reaction to KCN (potassium cyanide), the C. freundii strains and all derivative isolates grew, whereas S. Typhimurium did not. In terms of decarboxylation of lysine, which is the differential biochemical test between Salmonella spp. and Citrobacter spp., C. freundii E9750 and the isolates FMU108327/P and FMU108327/A1 were negative, while FMU108327/A10, FMU108327/B2, FMU108327/B10, and S. Typhimurium were positive. The FMU108327/B1 isolate decarboxylated lysine poorly resulting in a weak positive reaction. The β-galactosidase (ONPG) test helps to differentiate between Salmonella and Citrobacter cultures, particularly strains that ferment lactose slowly or fail to utilize it. The production of ONPG was positive for C. freundii E9750, FMU108327/P, FMU108327/A1 and FMU108327/B1. However, the isolates FMU108327/A10, FMU108327/B2, FMU108327/B10, and S. Typhimurium were all negative for ONPG production. These results confirm those obtained previously (Table1) (Supporting information).
I-CeuI Restriction Fragments of FMU108327/P and FMU108327/A10 Isolates
The CeuI restriction endonuclease recognizes a 26-bp sequence from position 1911 to 1936 of the 23S rRNA gene in rrn operons, with the number of CeuI fragments usually representing the number of rrn operons. The rrn skeleton of S. Typhimurium LT2 ,  had been previously mapped using chromosome digestion with CeuI endonuclease and PFGE. Data from this strain was used to construct rrn skeletons of two isolates: FMU108327/P and FMU108327A10. Initially, the CeuI chromosomal profiles of all studied isolates were carried out but the results showed only 2 different profiles among the studied isolates, one of 7 bands and the other of 8 bands. Therefore, only 2 representative isolates of each profile were selected for further analyses.
The DNA digestion of S. Typhimurium LT2 yielded a profile with 7 bands; the molecular weight of each band was the same as that of S. Typhimurium LT2 reported by Liu (Figure 2a, 2b) . C. freundii E9750 NCTC strain yielded 8 fragments, indicating that C. freundii contains 8 rrn operons in its chromosome. The isolate FMU108327/P also yielded 8 fragments while the derivative isolate FMU108327/A10 only yielded 7 fragments: A to G (Figure 2a). The calculated chromosome size was 4,739 kb for C. freundii E9750 and 4,572 kb for FMU108327/P, while the derivative isolate FMU108327/A10 had a chromosomal size of 4,520 kb (Figure 2b).
The corresponding individual fragments from the isolate FMU108327/P and the derivative isolate were determined by probing the Southern blotting membranes from PFGE that used PCR products of marker genes from S. Typhimurium LT2 (Table S1), since their location in the chromosome is known. The marker genes were: metB gene corresponding to band E, metC gene to band B, metE gene to band D, metH gene to band G, and fliC gene to band A (Figure 2).
The results obtained from these hybridizations indicated that C. freundii E9750 and FMU108327/P strains presented an extra cleavage that probably divides the 740 kb fragment (G) into smaller fragments, G (154 kb) and H (584 kb). Interestingly, the derivative isolate FMU108327/A10 did not show this additional cleavage site in fragment G, maintaining its size at 740 kb, and its rrn skeleton resembling that of S. enterica (Figure 2a, 2c).
Chromosomal Arrangement of FMU108327/P and FMU108327/A10 Isolates
PCR was used to determine the order of CeuI fragments and the rrn arrangements observed in these bacteria. To determine the chromosomal arrangement of the strains, genes adjacent to rrn operons were amplified (Table S1). In this way, the PCR assay confirmed the results obtained from hybridization on CeuI PFGE and showed the arrangement of the FMU108327/P and FMU108327/A10 isolates.
Regarding isolate FMU108327/A10, the hybridization patterns and PCR gene assays were similar to those observed for S. enterica strains. The data also indicated that the order of the rrn operons in this strain was BCDEFG; the positions of fragments B and G were adjacent to fragment A (Figure 2a) to form a circular chromosome. Meanwhile, atypical FMU108327/P showed some differences. Although the majority of the selected genes hybridized according to expected fragments (Table S1), there were a couple of differences: yrdA failed to hybridize in fragment CeuI-C, mobB did not hybridize in fragment E nor in any other fragment of the profile, although PCR assays on total genomic DNA indicated that the genes were present in the FMU108327/P isolate. These differing hybridization patterns indicate chromosomal rearrangements within the isolates concerned.
Presence of the Adaptative Genes cadA and lacY
Lactose fermentation, decarboxylation of lysine, and the utilization of potassium cyanide (KCN) are the most important criteria for differentiating Salmonella spp. (−, +, − respectively) from C. freundii (+/−, −,+respectively) (Table 1). Due to the atypical isolates in this study exhibiting changes in some of these biochemical phenotypes, the presence of the cadA gene that encodes lysine decarboxylase and the lacY gene that encodes lactose permease was investigated using PCR. The results showed the presence of cadA in FMU108327/P and in all of its derivatives, as well as in S. Typhimurium LT2, while the C. freundii strain E9750 was negative. With regards to the lacY gene, C. freundii E9750, FMU108327/P and its derivatives were positive while S. Typhimurium LT2 was negative (Figure S2).
Phylogenetic Reconstruction and Mutation Distribution
The phylogenetic analysis of the isolates showed different relationships between isolates depending on the sequenced genes. For example, when adk and recA genes were analysed (Figures 3a, 3b), the FMU108327/B2 and FMU108327/B10 isolates grouped in the Salmonella clade, while the FMU108327/P, FMU108327/A1, FMU108327/A10, FMU108327/B1 isolates and C. freundii E9750 clustered in the Citrobacter clade. For gyrB (Figure 3c), all of the isolates were located in the Citrobacter clade except for the FMU108327/B2 isolate. In the case of the icd gene (Figure 3d), FMU108327/B2, FMU108327/B10 and FMU108327/A10 isolates grouped in the Salmonella clade. Considering the purA, aph and gnd genes (Figure 3e–g), all the isolates (FMU108327/A1, FMU108327/A10, FMU108327/B1, FMU108327/B2 and FMU108327/B10) including the parental strain (FMU108327/P) grouped more closely to Salmonella, as did C. freundii E9750 when considering the purA and aph genes. Finally, with regard to the mdh gene (Figure 3h), all of the studied isolates (including the reference species C. freundii, C. youngae and Citrobacter spp.) formed a single cluster that diverged away from the Salmonella group.
All of the trees were constructed in Mega 5.01 for a) adk, b) recA, c) gyrB, d) icd, e) purA, f) aph, g) gnd, h) mdh and i) concatenated genes. The assigned names of the genus in which these alleles occur are shown. The scale bar represents the number of substitutions per site.
Although there were alleles that could be clearly assigned to one genus, in the phylogenetic reconstruction using the concatenated sequences, the haplotypes did occupy intermediate positions in the PhyML tree from the Citrobacter spp. and the progenitor strain (FMU108327/P) clade to the Salmonella spp. clade (Figure 3i). This observation of intermediate clustering in the tree of the concatenated genes is also congruent with the phenotypic variation that the derivative isolates presented.
Amazingly, the 16S rRNA sequences showed the same “undefined” nature that was observed in the tree of concatenated housekeeping genes as they did not form a clear sequence cluster . The FMU108327/A10, FMU108327/B2 and FMU108327/B10 isolates (haplotypes closer to Salmonella) showed 99% similarity to S. Typhimurium LT2 (AE006468)  and 98% similarity to both C. koseri (CP000822) and C. rodentium (FN543502) . When the sequence of the FMU108327/P isolate was compared with its derivative isolates, FMU108327/A10, FMU108327/B2 and FMU108327/B10, similarity was reduced to 97.6%. Meanwhile, the similarity among the FMU108327/P, FMU108327/A1 and FMU108327/B1 isolates (haplotypes closer to Citrobacter) and the C. freundii MRB0903 strain (GU126681) was 99.8%. Finally, the sequence similarity between the C. freundii MRB0903 strain and the FMU108327/B10 isolate was 97.4%.
In order to understand the role of genomic flux in the dynamics of the studied isolates, a more formal species assignment was performed using the linkage model of Bayesian clustering algorithm STRUCTURE (Figure 4) . The 6 isolates in the study formed two groups. In the first group, C. freundii E9750 (83%) and P (87%) strains and the A1 (87%) and B1 (69%) isolates had more than 60% identity with the Citrobacter genus while the remaining proportion identified with Salmonella. C. freundii CFAA strain was also included in this group and had 85% identity with Citrobacter and 15% with Enterobacter/Klebsiella. The second group that included the B2 and B10 isolates presented 90% identity with the Salmonella genus. It is important to highlight that the A10 derivative isolate showed 60% identity with Citrobacter ssp., 37% with Salmonella ssp and the remaining 3% with Enterobacter/Klebsiella group. In addition, C. freundii ballerup also showed this hybrid nature, this strain presented 62% identity with Citrobacter genus and the remaining 32% with Enterobacter group (Figure 4).
Results of the STRUCTURE analysis assuming 4 populations, a) every vertical line represents each of the 49 isolates and indicates the proportion of ancestry from the 4 ancestral populations, using the following colors: green for E. coli, blue for Enterobacter spp. and Klebsiella spp., purple for Citrobacter spp. and red for Salmonella spp.; b) It shows the proportion from Citrobacter and Salmonella ancestries in our isolates, as well as their biochemical phenotypes. It is interesting to highlight the A10 and B2 isolates. In spite of the fact that A10 derivative isolate has 60% identity with Citrobacter ssp., it displayed Salmonella's biochemical phenotype which has remained stable until now; in addition we have the B2 derivative isolate which had a genotype with a majority of identity to Salmonella ssp genus and its biochemical phenotype was variable, thus preventing establishing some biochemical identity.
The nested distribution analysis of the point mutations by region (30 pb sliding window) per gene, between genes and between isolates showed differences at all three levels (Figure 5). The isolates that accumulated more point mutations were FMU108327/B2 and FMU108327/B10, while the FMU108327/A10 isolate accumulated few mutations; in only 2 markers. The molecular marker that accumulated the most mutations of all isolates was the icd gene. The frequency of synonymous against nonsynonymous mutations also differed considerably, with more synonymous mutations being found in all genes and all isolates, suggesting that at the molecular level, most of the variation was neutral (Table S2).
Mutation frequencies for each of the 30 positions in 8 genes of 5 strains are shown. The isolates with the greatest number of mutations were B2 and B10, which presented the most diversified genotype, and the genes with the greatest number of mutations were adk, icd and recA.
Phenotypic variation of many phenotypic markers is a major adaptive strategy of pathogenic and commensal bacteria , . It has been shown that such variants are produced at high frequencies and in a reversible manner by bistability, bet-hedging, phase variation, large accumulation of mutations or hyper-variable methylation in specific regions of the genome , .
In the phenotypic diversification study (supporting information), atypical C. freundii FMU108327/P was shown to undergo frequent and reversible phenotypic changes (colonial morphology, biochemical characteristics) from the parental strain to its derivatives and vice-versa. This on-off molecular mechanism suggests that the changes in phenotypic expression are induced by a fluctuating environment, but surviving under these conditions poses challenges. Hypothetically, a bacterial population under such challenging conditions could produce off-spring with variable phenotypes to ensure that at least one of them, with a new phenotype, will be able to adapt to the new environment and become stable , . This has been shown previously in some bacterial species, such as C. jejuni  and Salmonella enterica ,  or more recently in Citrobacter rodentium  in which gene flux has been reported due to IS elements and prophage/phages, which are able to insert randomly throughout the chromosome and plasmids. The reshuffle of genes in both species resulted in the significant loss of functional genes associated with virulence and metabolic functions , .
Many questions about genetic, evolutionary, and ecological mechanisms behind phenotypic variation remain unresolved but it is possible that variable phenotypes aid in the survival of the bacteria under adverse conditions or, in pathogenic microorganisms might help them to increase their virulence or evade the immune responses of their hosts. In addition, these variations may serve to balance the benefits and disadvantages of a certain phenotype and in the case of competition, there may be a trade-off between the benefits of the generation of phenotypic variation and genetic diversity , .
There are four bacterial mechanisms that can shape the formation of new genotypes: a) chromosomal rearrangements; b) point mutations; c) homologous and non-homologous recombination; and d) horizontal gene transfer .
Concerning the first mechanism, chromosomal rearrangements are common in enterobacteria –. Frequently, strict host-specific serovars of S. enterica, such as serovars Typhi, Pullorum, and Gallinarum, have large-scale genomic rearrangements due to recombination between rRNA (rrn) operons –, while generalist serovars from animal and natural sources worldwide have stable genomes at the rrn level . Interestingly, similar rearrangements in rrn operons have been found in other species, such as E. coli and V. cholerae among others –.
In the current study, CeuI fragment analysis revealed that chromosome sizes and the structure of the included isolates were similar. Nevertheless, the extent of rearrangement in rrn regions was unexpected. The chromosome of C. freundii FMU108327/P was found to have eight CeuI fragments showing a very similar rrn profile to that presented by C. freundii NCTC E9750 (Figure 2). These data agree with that reported by other authors . Contrary to what had been expected, the chromosome of the derivative isolate FMU108327/A10 was found to have seven rrn operons, and appeared to have lost a CeuI site resulting in a similar profile to that observed in S. enterica serovar Typhimurium . This genetic event could have resulted from the deletion of a segment due to recombination between two rrn operons in the same replichore, thus forming a circular fragment, followed by reinsertion of a deleted segment into another location.
The second bacterial mechanism to form new genotypes, namely point mutation, can cause the loss of endonuclease cleavage sites. In the current study, this mechanism of genotype formation cannot be discarded until full genome sequences can be conducted to determine which bases might have changed in the FMU108327/P isolate and in its derivative FMU108327/A10 that led to one of the CeuI sites no longer being cleavable.
Despite the existence of a hypothesis concerning the order and structure of the chromosome being subject to selective pressure, in enterobacteria it is common to find variation in the distribution of ribosomal operons (rm). C. freundii is not an exception to these kinds of rearrangements, as shown in the above results and in the data from Kumar et al.  regarding the sequence of C. freundii MTCC 1658 that had 10 rRNA operons or in the results from Bai et al.  that found in C. freundii CF72 and CF74 8 rRNA operons. Previous studies have shown that genomic changes do occur in natural bacterial populations , . However, the majority of these are deleterious and as a result, only a very small proportion of the bacterial population undergoing genomic changes may survive under specific natural selection pressure , .
One interesting question that remains unanswered is what triggers such changes in the chromosomal synteny and what is its biological importance. The conclusions from the current study suggest that the bacteria were placed under thermal stress, which acted as a selective pressure, leading to the selection of some mutations that enhanced survival in such a specific environment, which in turn led to genomic diversification and special genomic structure that adapted better to the local environmental setting . Recently, Ciu et al.  showed the biological importance of this type of genomic rearrangement in S. aureus strains. They demonstrated that a large-scale inversion of the chromosome triggers the bacterial phenotype “on-off switch”, including colony morphology, antibiotic susceptibility, and the expression of dozens of genes.
On analysing the sequences of the eight housekeeping genes from the isolates, an astounding genetic diversity was found from a single clone. Moreover, the majority of the point mutations were synonymous, meaning that most of them did not change the enzyme function. This provided clear evidence of selective constraint on amino acid replacements. However, some of these mutated genes produced diversity in genotypes, some of which are closer (but not equal) to the genotypes of the Salmonella genus. This large quantity of mutations could only be explained by a hyper-mutant genotype induced and maintained by repair systems, such as SOS – and the DNA mismatch repair system (MMR) . The MMR system is well known for its role in maintaining genetic stability, but defects in this system lead to high mutability. Bacteria having defects in MMR are mutators or hypermutator cells, which have elevated rates of spontaneous mutation and recombination enabling a diversity of mobile DNA to be incorporated into the genome. The MMR system in particular has been widely studied in Salmonella Typhimurium –. The results show that the conversion of mutL gene between functional and defective alleles through the deletion of one of the three tandem 6 bp repeats within the gene sequence (via slipped-strand mispairing) may act as a genetic switch between genetic stability and a mutability state –. However, a hyper-mutation state may be temporary, leading to genomic diversification. The repetitive structure within the sequence of mutL enables it to function as a genetic “on-off” switch at a population level by the selection of existing cells possessing the favorable traits, which may allow return to the original wild-type phenotype , . Additionally, in Salmonella Typhimurium, a hyper-mutability state, as well as prophage/phage flux plays an important role in offspring diversity. Such a state is also related to nutritional function loss, as well as shifts to alternative biochemical pathways and colony morphology diversity , .
All of the genes involved in SOS responses (recA, lexA) and the MMR system (mutS, mutL, mutH) are found in the genomes of Citrobacter species that have been sequenced to date. The information of corresponding genes was compiled from the respective published genome sequences. These data are relevant, since the presence of such systems in C. freundii may explain the rapid and large number of accumulated mutations, as well as diversification within just a few generations.
Diversity in the16S rrn gene within a species has already been observed –. In V. vulnificus , , this heterogeneity can be related to virulence, while in Bacillus  the existence of some specific nucleotides in the 16S rRNA gene is related to the capacity to grow at both low and high temperatures. In Aeromonas spp. , a range of diversity between 96.7% and 100% has been reported, while in clinical strains of S. anginosus , micro-heterogeneity has been related to the phenotypic characteristics and the adaptation of the organism to different niches. Unfortunately, in the current study, it was not possible to relate the observed micro-heterogeneity in the sequences of 16S rrn of C. freundii and its derivative isolates to any phenotypic characteristic or function.
The phylogenetic analysis of C. freundii showed that although this species was grouped in one cluster, it seemed to be more diverse and dynamic than the Salmonella and E. coli clades. In fact in this clade, C. youngae and Citrobacter sp. clustered too, but C. rodentium and C. koseri did not and were more distantly related. Therefore, these data indicate that Citrobacter is a polyphyletic genus. In general, these results were consistent with those previously reported  with some differences in the clonal relationship between C. freundii and Salmonella sp., probably due to the analysis including more C. freundii strains in the current study. In addition, the results from the analyses of Maximum Likelihood of the FMU108327/P and its derivative isolates reveal the hybrid nature of these genotypes (Figure 3). Moreover, in the concatenated tree (Figure 3i); even though the haplotypes clustered within the C. freundii branch, they were located in an intermediate gradient between the Citrobacter and Salmonella clades. Finally, and in order to verify the presence of hybrid genotypes among the isolates, these sequences were analysed using the linkage model of Bayesian clustering algorithm “STRUCTURE”. The “STRUCTURE” (Figure 4) analysis also supported the existence of this hybrid group. In general, moving towards the interior of the Citrobacter group, each strain was very heterogeneous, showing that some of their alleles were imported from other species . Thus, this finding could explain the hybrid nature of this genetic pool as a result of convergence with other enterobacteria, a phenomenon that seems to be common in enteric bacteria, which are seen to be dynamic entities . For example, in C. rodentium, E. coli or S. typhi, and S. paratyphi, this phenomenon has been demonstrated in terms of their process of pathogenesis due to the high frequency of genetic interchange, although in all cases an immediate common ancestry does not exist , .
In studies that had analysed phenotypic variation mechanisms, it has been shown that certain bacterial species can shape their phenotype in two or more stable states depending on the environment , . The results from the current study opens the discussion further and questions if the phenotypic variation mechanisms, chromosomal rearrangements, large accumulation of mutations and genomic flux may enable a bacterial population to overcome the barriers that exists at the species level and generate new genotypes that, due to convergence, are closer to other species. Alternatively, it may be that atypical C. freundii strains exhibit different genotypic varieties of the same species co-existing in a stable way in the same time frame but expressing different phenotypes that are dependent on external environmental factors in which they are living.
Further work is needed to validate some of the hypotheses outlined in this paper. It is necessary to demonstrate the dynamics that the bacteria follow in the diversification study in order to analyse population genetics in a representative sample of the isolates that were obtained in the phenotypic study.
In addition, the genome sequence and a genomic and transcriptomics approximation of some of the lineages from the phenotypic variation study should be carried out. Furthermore, more atypical C. freundii strains should be included in order to determine if these characteristics are specific to the strain or if it is a wider-ranging genus. This would be important for providing details of genetic events to improve understanding of the dynamics and mechanisms that shape the structure of genomes, the processes related to divergence, convergence and diversification, their biological importance, and in particular, in bacterial evolution and species boundaries.
Another important consideration is the clinical implications of isolate C. freundii strains with these genetic rearrangements and genomic flux. Is it possible that this genetic plasticity induces the pathogenic behavior of these atypical strains?
Until now, definitive classification of C. freundii has been difficult creating a taxonomic problem. Published data, together with the results from this current study, highlight the great diversity and plasticity of this species and offer possibilities for separating C. freundii as a species in a more precise way.
Posterior probability of the number of genetic pools inferred by“STRUCTURE”. The optimal value was K = 4 by comparing the posterior probabilities of the data given each value of K from 2 to 8.
PCR products of the adaptive genes cadA y lacY. a) PCR products of cadA gene. Lanes: 1. S. Typhimurium LT2, 2. C. freundii E9750 NCTC, 3. FMU108327/A1, 4. FMU108327/A10, 5. DNA ladder 100 bp, 6. FMU108327/B1, 7. FMU108327/B2, 8. FMU108327/B10, 9. C. freundii FMU108327/P. b) PCR products of lacY gene. Lanes: 1 and 9. DNA ladder 100 bp, 2. C. freundii FMU108327/P, 3. FMU108327/A1, 4. FMU108327/A10, 5. FMU108327/B1, 6. FMU108327/B2, 7. FMU108327/B10, 8. C. freundii E9750 NCTC and 10. S. Typhimurium LT2.
Primer sequences used in this study.
Synonymous and Nonsynonymous substitutions.
We would like to thank R. Diaz (M.Sc) for his technical assistance in radioactive labelling. This paper is part of the doctoral research carried out by G. Delgado and constitutes partial fulfillment of the Doctorado en Ciencias Biomédicas program of the Universidad Nacional Autónoma de México (UNAM).
† DeceasedConceived and designed the experiments: GD VV AC. Performed the experiments: GD AGG JLM. Analyzed the data: GD RC AGG. Contributed reagents/materials/analysis tools: VS RM. Wrote the paper: GD RC. Critical rewiew of the article: VS RM AGG AC. Review of the article for intellectual content: VS AC. Final approval of the version to be published: VS AC.
- 1. Retchless AC, Lawrence JG (2007) Temporal fragmentation of speciation in bacteria. Science 317: 1093–1096.
- 2. Hallet B (2001) Playing Dr Jekyll and Mr Hyde: combined mechanisms of phase variation in bacteria. Curr Opin Microbiol 4: 570–581.
- 3. Veening JW, Smits WK, Kuipers OP (2008) Bistability, epigenetics, and bet-hedging in bacteria. Annu Rev Microbiol 62: 193–210.
- 4. Smits WK, Kuipers OP, Veening JW (2006) Phenotypic variation in bacteria: the role of feedback regulation. Nat Rev Microbiol 4: 259–271.
- 5. Dobrindt U, Hacker J (2001) Whole genome plasticity in pathogenic bacteria. Curr Opin Microbiol 4: 550–557.
- 6. Sheppard SK, McCarthy ND, Falush D, Maiden MCJ (2008) Convergence of Campylobacter species: implications for bacterial evolution. Science 320: 237–239.
- 7. Farmer JJ III (1981) The genus Citrobacter. In: Starr M, Stolp H, Trüper HG, Balows A, Schlegel HG editors. The prokaryotes. Springer-Verlag. 1140–1147.
- 8. Gilchrist MJR (1995) Enterobacteriaceae: opportunistic pathogens and other genera. In: Murray PR, Baron EJ, Pfaller MA, Tenover FC, Yolken RH editors. Manual of clinical microbiology 6th edition. ASM press. 457–464.
- 9. Ewing WH (1986) The genus Citrobacter. In: Edwards and Ewing’s Identification of Enterobacteriaceae. Elsevier Science Publishing Co. 341–364.
- 10. Sedlák J (1973) Present knowledge and aspects of Citrobacter.. Curr Top Microbiol Immunol 62: 41–59.
- 11. Brenner DJ, Grimont PAD, Steigerwalt AG, Fanning GR, Ageron E, et al. (1993) Classification of Citrobacteria by DNA hybridization: designation of Citrobacter farmeri sp. nov., Citrobacter youngae sp. nov., Citrobacter braakii sp. nov., Citrobacter werkmanii sp. nov., Citrobacter sedlakii sp. nov., and three unnamed Citrobacter genomospecies. Int J Syst Bacteriol. 43: 645–658.
- 12. Bettelheim KA, Evangelidis H, Pearce JL, Sowers E, Strockbine NA (1993) Isolation of a Citrobacter freundii strain which carries the Escherichia coli O157 antigen. J Clin Microbiol 31: 760–761.
- 13. Chart H, Willshaw GA, Cheasty T, Rowe B (1993) Structure and antigenic properties of Citrobacter freundii lipopolysaccharides. J Appl Bacteriol 74: 583–587.
- 14. Vinogradov E, Conlan JW, Perry MB (2000) Serological cross-reaction between the lipopolyssaccharide O-polysaccharide antigens of Escherichia coli O157: H7 and strain of Citrobacter freundii and Citrobacter sedlakii. FEMS Microbiol Lett 190: 157–161.
- 15. Nishiuchi Y, Doe M, Hotta H, Kobayashi K (2000) Structure and serologic properties of O-specific polysaccharide from Citrobacter freundii possessing cross-reactivity with Escherichia coli O157: H7. FEMS Immunol Med Microbiol 28: 163–171.
- 16. Samuel G, Hogbin JP, Wang L, Reeves PR (2004) Relationships of the Escherichia coli O157, O111 and O55 O-antigen gene clusters with those of Salmonella enterica and Citrobacter freundii, which express identical O antigens. J Bacteriol 186: 6536–6543.
- 17. Ou JT, Baron LS, Rubin FA, Kopecko DJ (1988) Specific insertion and deletion sequence 1-like DNA element causes the reversible expression of the virulence capsular antigen Vi of Citrobacter freundii in Escherichia coli. Proc Natl Acad Sci U S A 85: 4402–4405.
- 18. Guarino A, Gianella R, Thompson MR (1989) Citrobacter freundii produces an 18-amino-acid heat stable enterotoxin identical to the 18-amino-acid Escherichia coli heat stable enterotoxin (ST Ia). Infect Immun 57: 649–652.
- 19. Doran TI (1999) The role of Citrobacter in clinical disease of children: Review. Clin Infect Dis 28: 384–394.
- 20. Hess P, Altenhöfer A, Khan AS, Daryab N, Kim KS, et al. (2004) A Salmonella fim homologue in Citrobacter freundii mediates invasion in vitro and crossing of the blood-brain barrier in the rat pup model. Infect Immun 72: 5298–5307.
- 21. Pereira AL, Silva TN, Gomes A CMM, Araújo ACG, Giugliano LG (2010) Diarrhea-associated biofilm formed by enteroaggregative Escherichia coli and aggregative Citrobacter freundii: a consortium mediated by putative F pili. BMC Microbiol 10: 57–74. Available: http://www.biomedcentral.com/1471-2180/10/57. Accessed 2013 Aug 5.
- 22. Crosa JH, Steigerwalt AG, Fanning GR, Brenner DJ (1974) Polynucleotide sequence divergence in the genus Citrobacter. J Gen Microbiol 83: 271–282.
- 23. Bulmash JM, Fulton MD, Jiron J (1965) Lactose and sulfide reactions of an aberrant Salmonella strain. J Bacteriol 89: 259.
- 24. Kunz LJ, Ewing WH (1965) Laboratory Infection with a lactose-fermenting strain of Salmonella typhi.. J Bacteriol 89: 1629.
- 25. Gonzalez AB (1966) Lactose-Fermenting Salmonella. J Bacteriol 91: 1661–1662.
- 26. Le Minor L, Coynault C, Pessoa G (1974) Plasmid determinants of the atypical “lactose positive” character in “S. typhimurium” and “S. oranienburg” isolates from epidemic in Brasil, 1971–1973. Ann Microbiol (Paris) 125A: 261–285.
- 27. García MM, Vázquez V (1985) Salmonella lactosa positiva. Acta Pediatr Mex 6: 68–70.
- 28. Usera MA, Echeita A, Aladueña A, Blanco MC, Reymundo R, et al. (1996) Interregional foodborne salmonellosis outbreak due to powdered infant formula contaminated with lactose-fermenting Salmonella Virchow. Eur J Epidemiol 12: 377–381.
- 29. Petty NK, Bulgin R, Crepin VF, Cerdeño-Tárraga AM, Schroeder GN, et al. (2010) The Citrobacter rodentium sequence reveals convergent evolution with human pathogenic Escherichia coli.. J Bacteriol 192: 525–538.
- 30. Petty NK, Feltwell T, Pickard D, Clare S, Toribio AL, et al. (2011) Citrobacter rodentium is an unstable pathogen showing evidence of significant genomic flux. PLoS Pathogens 7(4): e1002018
- 31. Bai L, Xia Sh, Lan R, Liu L, Ye Ch, et al. (2012) Isolation and characterization of cytotoxic, aggregative Citrobacter freundii. PLoS ONE 7(3): e33054
- 32. Cravioto A, Reyes RE, Ortega R, Fernández G, Hernández R, et al. (1988) Prospective study of diarrhoeal disease in a cohort of rural Mexican children: incidence and isolated pathogens during the first two years of life. Epidem Inf 101: 123–134.
- 33. Cravioto A, Reyes RE, Trujillo F, Uribe F, Vázquez V, et al. (1990) Risk of diarrhea during the first year of life associated with initial and subsequent colonization by specific enteropathogens. Am J Epidemiol 131: 886–904.
- 34. Gilligan PH, Janda JM, Karmali MA, Miller JM (1992) Laboratory diagnosis of bacterial diarrhea. In: Nolte F Coordinating Editor. Cumitech 12A. ASM Press.
- 35. Flores M, González V, Brom S, Martinez E, Piñeiro D, et al. (1987) Reitered DNA sequences in Rhizobium and Agrobacterium spp. J Bacteiol 169: 5782–5788.
- 36. Flores M, González V, Pardo MA, Leija A, Martínez E, et al. (1988) Genomic Instability in Rhizobium phaseoli.. J Bacteriol 170: 1191–1196.
- 37. Liu S, Hessel A, Sanderson KE (1993) Genomic mapping with I-Ceu I, an intron-encoded endonuclease specific for genes for ribosomal RNA, in Salmonella spp., Escherichia coli, and other bacteria. Proc Natl Acad Sci U S A 90: 6874–6878.
- 38. Liu SL, Hessel A, Sanderson KE (1993) The XbaI-BlnI-CeuI genomic cleavage map of Salmonella typhimurium LT2 determined by double digestion, end labelling, and pulsed gel electrophoresis. J Bacteriol 175: 4104–4120.
- 39. Sambrook J, Fritsch EF, Maniatis T (1989) Molecular cloning: a laboratory manual. Cold Spring Harbor Laboratory Press. 9.31–9.40.
- 40. Reid SD, Selander RK, Whittam TS (1999) Sequence Diversity of flagellin (fliC) alleles in pathogenic Escherichia coli. J Bacteriol 181: 153–160.
- 41. McClelland M, Sanderson KE, Spieth J, Clifton SW, Latreille P, et al. (2001) Complete genome sequence of Salmonella enterica serovar Typhimurium LT2. Nature 413: 852–6.
- 42. Murray MG, Thompson WF (1994) Miniprep of bacterial genomic DNA. In Ausubel FM, Brent R, Kingston RE, Moore DD, Seidman JG, et al editors. Current protocols in molecular biology. John Wiley & Sons, Inc. 241–245.
- 43. Blattner FR, Plunkett G 3rd, Bloch CA, Perna NT, Burland V, et al. (1997) The complete genome sequence of Escherichia coli K-12. Science 277: 1453–62.
- 44. Perna NT, Plunkett G 3rd, Burland V, Mau B, Glasner JD, et al. (2001) Genome sequence of enterohaemorrhagic Escherichia coli O157: H7 Nature. 409: 52933.
- 45. Deng W, Liou SR, Plunkett G 3rd, Mayhew GF, Rose DJ, et al. (2003) Comparative genomic of Salmonella enterica serovar Typhi strains Ty2 and CT18. J Bacteriol 185: 2330–7.
- 46. Parkhill J, Dougan G, James KD, Thomson NR, Pickard D, et al. (2001) Complete genome sequence of a multiple drug resistant Salmonella enterica serovar Typhi CT18 Nature. 413: 848–52.
- 47. Lee JI, Okazaki N, Tsuchiya T, Wilson TH (1994) Cloning and sequencing of the gene for the lactose carrier of Citrobacter freundii.. Biochem Biophys Res Comm 203: 1882–1888.
- 48. Wirth T, Falush D, Lan R, Colles F, Mensa P, et al. (2006) Sex and virulence in Escherichia coli: an evolutionary perspective. Mol Microbiol 60: 1136–1151.
- 49. Guindon S, Gascuel O (2003) A simple, fast, and accurate algorithm to estimate large phylogenies by maximum likelihood. Syst Biol 52: 696–704.
- 50. Posada D (2008) jModelTest: Phylogenetic model averaging. Mol Biol Evol 25: 1253–1256.
- 51. Tamura K, Dudley J, Nei M, Kumar S (2007) MEGA4: Molecular evolutionary genetics análisis (MEGA) software version 4.0. Mol Biol Evol 24: 1596–1599.
- 52. Falush D, Stephens M, Pritchard JK (2003) Inference of population structure using multilocus genotype data: linked loci and correlated allele frequencies. Genetics 164: 1567–1587.
- 53. Rozas J, Sanchez-DelBarrio JC, Messeguer X, Rozas R (2003) DnaSP, DNA polymorphism analyses by the coalescent and other methods. Bioinformatics 19: 2496–2497.
- 54. Liu SL, Sanderson KE (1998) Homologous recombination between rrn operons rearranges the chromosome in host-specialized species of Salmonella. FEMS Microbiol Lett 164: 275–281.
- 55. Hanage W (2013) Fuzzy species revisited. BMC Biology 11: 41
- 56. Bayliss CD (2009) Determinants of phase variation rate and the fitness implications of differing rates for bacterial pathogens and commensals. FEMS Microbiol Rev 33: 504–520.
- 57. Van der Woude MW (2011) Phase variation: how to create and coordinate population diversity. Curr Op Microbiol 14: 205–211.
- 58. Beaumont HJE, Gallie J, Kost C, Ferguson GC, Rainey PB (2009) Experimental evolution of bet hedging. Nature 462: 90–3.
- 59. Scott AE, Timms AR, Connerton PL, Loc-Carrillo C, Radzum KA, et al. (2007) Genome dynamics of Campylobacter jejuni in response to bacteriophage predation. PLoS Pathog 3(8): e119
- 60. Thomson NR, Clayton DJ, Windhorst D, Vernikos G, Davidson S, et al. (2008) Comparative genome analysis Salmonella Enteritidis PT4 and Salmonella Gallinarum 287/91 provides insights into evolutionary and host adaptation pathways. Genome Res 18: 1624–37
- 61. Ochman H, Bergthorsson U (1995) Genome evolution in enteric bacteria. Curr Opin Genet Dev 5: 734–738.
- 62. Sanderson KE, Liu SL (1998) Chromosomal rearrangements in enteric bacteria. Electrophoresis 19: 569–572.
- 63. Liu GR, Rahn A, Liu WQ, Sanderson KE, Johnston RN, et al. (2002) The evolving genome of Salmonella enterica serovar Pullorum. J Bacteriol 184: 2626–2633.
- 64. Helm RA, Lee AG, Christman HD, Maloy S (2003) Genomic rearrangements at rrn operons in Salmonella. Genetics 165: 951–959.
- 65. Wu KY, Liu GR, Liu WQ, Wang AQ, Zhan A, et al. (2005) The genome of Salmonella enterica serovar Gallinarum: distinct insertions/deletions and rare rearrangements. J Bacteriol 187: 4720–4727.
- 66. Shu S, Setianingrum E, Zhao L, Li Z, Xu H, et al. (2000) I-CeuI fragment analysis of Shigella species: evidence for large scale chromosome rearrangement in S dysenteriae and S. flexneri. FEMS Microbiol Lett 182: 93–98.
- 67. Matthews TD, Rabsch W, Maloy S (2011) Chromosomal rearrangements in Salmonella enterica serovar Typhi strains isolated from asymptomatic human carriers. MBio 2(3): e00060–11
- 68. Liu SL, Sanderson KE (1996) Highly plastic chromosomal organization in Salmonella typhi.. Proc Natl Acad Sci U S A 93: 10303–10308.
- 69. Liu SL, Schryvers AB, Sanderson KE, Johnston RN (1999) Bacterial phylogenetic clusters revealed by genome structure. J Bacteriol 181: 6747–6755.
- 70. Helm RA, Porwollik S, Stanley AE, Maloy S, McClelland M, et al. (2004) Pigeon-associated strains of Salmonella enterica serovar Typhimurium phage type DT2 have genomic rearrangements at rRNA operons. Infect Immun 72: 7338–7341.
- 71. Kothapalli S, Nair S, Alokam S, Pang T, Khakhria R, et al. (2005) Diversity of genome structure in Salmonella enterica serovar Typhi populations. J Bacteriol 187: 2638–2650.
- 72. Liu SL, Sanderson KE (1995) I-CeuI reveals conservation of the genome of independent strains of Salmonella typhimurium. J Bacteriol 177: 3355–3357.
- 73. Hill CW, Harnish BW (1981) Inversions between ribosomal RNA genes of Escherichia coli. Proc Natl Acad Sci U S A 78: 7069–7072.
- 74. Nandi S, Khetawat G, Sengupta S, Majumder R, Kar S, et al. (1997) Rearrangements in the genome of Vibrio cholerae strains belonging to different serovars and biovars. Int J Syst Bacteriol 47: 858–862.
- 75. Khetawat G, Bhadra RK, Nandi S, Das J (1999) Resurgent Vibrio cholerae O139: rearrangements of cholera toxin genetic elements and amplification of rrn operon. Infect Immun 67: 148–154.
- 76. Hill CW (1999) Large genomic sequence repetitions in bacteria: lessons from rRNA operons and Rhs elements. Res Microbiol 150: 665–674.
- 77. Teyssier C, Marchandin H, De Buochberg MS, Ramuz M, Jumas-Bilak E (2003) Atypical 16S rRNA gene copies in Ochrobactrum intermedium strains reveal a large genomic rearrangements by recombination between rrn copies. J Bacteriol 185: 2901–2909.
- 78. Yap MN, Barak JD, Charkowski AO (2004) Genomic diversity of Erwinia carotovora subsp. carotovora and its correlation with virulence. App Environ Microbiol 70: 3013–3023.
- 79. Kumar S, Kaur C, Kimura K, Takeo M, Mayilraj S, et al. (2013) Draft genome sequence of the type species of the genus Citrobacter, Citrobacter freundii MTCC 1658. Genome Announc 1(1): e00120–12
- 80. Liu GR, Edwards K, Eisenstark A, Fu YM, Liu WQ, et al. (2003) Genomic diversification among archival strains of Salmonella enterica serovar Typhimurium LT7. J Bacteriol 185: 2131–2142.
- 81. Liu SL, Sanderson KE (1995) Rearrangements in the genome of the bacterium Salmonella typhi. Proc Natl Acad Sci U S A 92: 1018–1022.
- 82. Cui L, Neoh Hm, Iwamoto A, Hiramatsu K (2012) Coordinated phenotype switching with large-scale chromosome flip-flop inversion observed in bacteria. Proc Natl Acad Sci U S A 109: E1647–56.
- 83. Fernandez de Henestrosa AR, Calero S, Barbé J (1991) Expression of the recA gene of Escherichia coli in several species of gram negative bacteria. Mol Gen Genet 226: 503–506.
- 84. Fernandez de Henestrosa AR, Ogi T, Aoyagi S, Chafin D, Hayes JJ, et al. (2000) Identification of additional genes belonging to the LexA regulon in Escherichia coli. Mol Microbiol 35: 1560–1572.
- 85. Janion C (2008) Inducible SOS response system of DNA repair and mutagenesis in Escherichia coli. Int J Biol Sc 4: 338–344.
- 86. Gong J, Liu WQ, Liu GR, Chen F, Li JQ, et al. (2007) Spontaneous conversion between mutL and 6 bpΔmutL in Salmonella typhimurium LT7: association with genome diversification and possible roles in bacterial adaptation. Genomics 90: 542–549.
- 87. Chen F, Liu WQ, Eisenstark A, Johnston RN, Liu GR, et al. (2010) Multiple genetic switches spontaneously modulating bacterial mutability. BMC Evol Biol 10: 277.
- 88. Chen F, Liu WQ, Liu ZH, Zou QH, Wang Y, et al. (2010) mutL as a genetic switch of bacterial mutability: turned on or off through repeat copy number changes. FEMS Microbiol Lett 312: 126–132.
- 89. Eisenstark A (2010) Genetic diversity among offspring from archived Salmonella enterica ssp. enterica serovar Typhimurium (Demerec collection): in search of survival strategies. Annu Rev Microbiol 64: 277–292.
- 90. Sacchi CT, Whitney AM, Reeves MW, Mayer LW, Popovic T (2002) Sequence diversity of Neisseria meningitides 16S rRNA genes and use of 16S rRNA gene sequencing as a molecular subtyping tool. J Clin Microbiol 40: 4520–4527.
- 91. Vickery MC, Nilsson WB, Strom MS, Nordstrom JL, DePaola A (2007) A real-time PCR assay for rapid determination of 16S rRNA genotype in Vibrio vulnificus. J Microbiol Methods 68: 376–384.
- 92. Aznar R, Ludwig W, Amann RI, Schleifer KH (1994) Sequence determination of rRNA genes of pathogenic Vibrio species and whole-cell identification of Vibrio vulnificus with rRNA-targeted oligonucleotide probes. Int J Syst Bacteriol 44: 330–337.
- 93. Prüβ BM, Francis KP, von Stetten F, Scherer S (1999) Correlation of 16S ribosomal DNA signature sequences with temperature-dependent growth rates of mesophilic and psychrotolerant strains of Bacillus cereus group. J Bacteriol 181: 2624–2630.
- 94. Alperi A, Figueras MJ, Inza I, Martinez-Murcia AJ (2008) Analysis of 16S rRNA gene mutations in subset of Aeromonas strains and their impact in species delineation. Int Microbiol 11: 185–194.
- 95. Clarridge JE 3rd (2004) Impact of 16S rRNA gene sequence analysis for identification of bacteria on clinical microbiology and infectious diseases. Clin Microbiol Rev 17: 840–862.
- 96. den Bakker HC, Didelot X, Fortes ED, Nightingale KK, Wiedmann M (2008) Lineage specific recombination rates and microevolution in Listeria monocytogenes.. BMC Evol Biol 8: 277.
- 97. Retchless AC, Lawrence JG (2010) Phylogenetic incongruence arising from fragmented speciation in enteric bacteria. Proc Natl Acad Sci U S A 107: 11453–11458.
- 98. Liu WQ, Feng Y, Wang Y, Zou QH, Chen F, et al. (2009) Salmonella paratyphi C: Genetic divergence from Salmonella choleraesuis and pathogenic convergence with Salmonella typhi. PLoS One 4(2): e4510
- 99. Kucerova E, Clifton SW, Xia XQ, Long F (2010) Genome sequence of Cronobacter sakazakii BAA-894 and comparative genomic hybridization analysis with other Cronobacter species. PLoS One 5(3): e9556
- 100. Johnson TJ, Kariyawasam S, Wannemuehler Y, Mangiamele P, Johnson SJ, et al. (2007) The genome sequence of avian pathogenic Escherichia coli strain O1: K1: H7 shares strong similarities with human extraintestinal pathogenic E. coli genomes. J Bacteriol 189: 3228–36.
- 101. Welch RA, Burland V, Plunkett G 3rd, Redford P, Roesch P, et al. (2002) Extensive mosaic structure revealed by the complete genome sequence of uropathogenic Escherichia coli. Proc Natl Acad Sci U S A 99: 17020–4.
- 102. Makino K, Yokoyama K, Kubota Y, Yutsudo CH, Kimura S, et al. (1999) Complete nucleotide sequence of the prophage VT2-Sakai carrying the verotoxin 2 genes of the enterohemorrhagic Escherichia coli O157: H7 derived from the Sakai outbreak. Genes Genet Syst 74(5): 227–39.
- 103. Rasko DA, Rosovitz MJ, Myers GS, Mongodin EF, Fricke WF, et al. (2008) The pangenome structure of Escherichia coli: comparative genomic analysis of E. coli commensal and pathogenic isolates J Bacteriol. 190(20): 6881–93
- 104. Ogura Y, Ooka T, Iguchi A, Toh H, Asadulghani M, et al. (2009) Comparative genomics reveal the mechanism of the parallel evolution of O157 and non-O157 enterohemorrhagic Escherichia coli. Proc Natl Acad Sci U S A 106: 17939–44
- 105. Fouts DE, Tyler HL, DeBoy RT, Daugherty S, Ren Q, et al. (2008) Complete genome sequence of the N2-fixing broad host range endophyte Klebsiella pneumoniae 342 and virulence predictions verified in mice. PLoS Genet 4(7): e1000141
- 106. Wu KM, Li LH, Yan JJ, Tsao N, Liao TL, et al. (2009) Genome sequencing and comparative analysis of Klebsiella pneumoniae NTUH-K2044, a strain causing liver abscess and meningitis. J Bacteriol 191: 4492–501
- 107. Yang F, Yang J, Zhang X, Chen L, Jiang Y, et al. (2005) Genome dynamics and diversity of Shigella species, the etiologic agents of bacillary dysentery. Nucleic Acids Res 33: 6445–58.
- 108. Jin Q, Yuan Z, Xu J, Wang Y, Shen Y, et al. (2002) Genome sequence of Shigella flexneri 2a: insights into pathogenicity through comparison with genomes of Escherichia coli K12 and O157. Nucleic Acids Res 30: 4432–41.
- 109. Fricke WF, Mammel MK, McDermott PF, Tartera C, White DG, et al. (2011) Comparative genomics of 28 Salmonella enterica isolates: evidence for CRISPR-mediated adaptive sublineage evolution. J Bacteriol 193: 3556–68
- 110. Holt KE, Thomson NR, Wain J, Langridge GC, Hasan R, et al. (2009) Pseudogene accumulation in the evolutionary histories of Salmonella enterica serovars Paratyphi A and Typhi. BMC Genomics 10: 36
- 111. Cañedo I (1989) Estudios de patogenicidad en Citrobacter ssp. Bachelor in science thesis. Universidad Nacional Autónoma de México.
- 112. Selander RK, Caugant DA, Ochman H, Musser JM, Gilmour MN, et al. (1986) Methods of multilocus enzyme electrophoresis for bacterial population genetics and systemic. Appl Environ Microbiol 51: 873–884.
- 113. Lane DJ (1991) 16S/23S rRNA sequencing. In: Stackebrandt E, Goodfellow M, editors. Nucleic acid techniques in bacterial systematic. Chichester Wiley pp. 115–175.