Introduction

The genus Klebsiella belongs to the Enterobacteriaceae family and was named by Trevisan in 1885. The genus is heterogeneous and originally composed of three major clusters, cluster I consisting of K. granulomatis and K. pneumoniae along with its three subspecies (K. pneumoniae subsp. pneumoniae, K. pneumoniae subsp. rhinoscleromatis, and K. pneumoniae subsp. ozaenae) and cluster III consisting of K. oxytoca. The organisms originally belonging to cluster II were placed into a new genus, Raoultella containing R. terrigena, R. ornithinolytica, and R. planticola, based on biochemical, DNA–DNA hybridization and phylogenetic relationship [7]. It is also important to note that K. mobilis (=E. aerogenes) is more closely related to the genus Klebsiella than E. cloacae as evident from the phylogenetic relationship based on rpoB, gyrA, and gyrB gene sequences, shown here.

During a study investigating the microbial hot spots in the residential household, the overall levels of contamination on different surfaces, and the identification of the isolates was determined. One isolate, designated as W14T, was recovered from one of the Michigan households located in the town of Ypsilanti during the microbial hotspot study. A polyphasic approach was implemented for the characterization of isolate W14T. Evidence indicated that this isolate belonged to the genus Klebsiella within the K. oxytoca cluster [7, 12].

Materials and Methods

Bacterial Strain and Cultivation

Isolate W14T was recovered from a tooth brush holder on R2A agar medium at 30 °C. The swab utilized to collect the sample was 3 M™ Quick Swab, a rayon-tipped swab containing letheen neutralizing buffer (3 M, Minneapolis, MN, USA). The isolate was grown on Tryptic Soy Agar (TSA) (BD, Franklin Lakes, NJ) at its optimum temperature of 35 °C for further maintenance and biochemical analysis. The reference strains used in the study, K. oxytoca ATCC 13182T, K. mobilis ATCC 13048T, K. pneumoniae ATCC 13883T, and R. terrigena ATCC 33257T were grown and maintained as per ATCC’s instructions.

Biochemical and Phenotypic Characteristics

Phenotypic characteristics were observed on TSA after 24 h of incubation at 35 °C. Differential-selective media such as MacConkey, XLD, and Levine EMB (BD, Franklin Lakes, NJ) were used to confirm its position in the family Enterobacteriaceae and also to determine its growth potential on different media. Gram staining (Four-step Gram Stain Kit, BD, Franklin Lakes, NJ), motility (Motility Agar Stab), microscopic observations (Axioskop 2 Plus Imaging system, Carl Zeiss, Göttingen, Germany), and biochemical analysis were performed according to standard laboratory procedure [1]. The API20E (bioMerieux, Inc. Durham, NC), Biolog (GN2 MicroPlates, Biolog, Inc. Hayward, CA), and Enterotube II (BD, Franklin Lakes, NJ) were used to obtain the identification and biochemical profile according to the manufacturers’ instructions.

Fatty Acid Methyl Ester Analysis (FAME)

A pure culture of isolate W14T was used to perform FAME according to the protocols and software provided by Microbial ID (MIDI, Newark, DE) for the Sherlock Microbial Identification System (MIS) using automated gas chromatography. The FAME analysis was performed by growing isolate W14T on Tryptic Soy Broth Agar (TSBA) at 28 °C for 24 h and the derived FAME profile was compared against the TSBA66 database.

MALDI-TOF Analysis

Matrix-assisted laser desorption ionization-time of flight (MALDI-TOF) mass spectroscopy-based microbial identification was performed on the autoflex™ mass spectrometer according to the manufacturer’s recommendations (Bruker Daltonics, Inc., Fremont, CA). Briefly, an ethanol extract method was used to process the sample which was less than 72-h old. Ethanol extraction consisted of resuspending a 1-μL loop full of the bacterial sample in 300 μL of nuclease free water and adding 900 μL of 100 % ethanol. The cell material was then pelleted by centrifugation. The cell pellets were resuspended in formic acid and acetonitrile and 1 μL spotted onto the target plate. After samples had dried on the plate, they were overlaid with the α-cyano-4-hydroxycinnamic acid (CHCA, Protea Biosciences, Inc., Morgantown, WV) matrix and spectra produced for comparison to the reference library for identification of the isolate.

DNA Base Composition and DNA–DNA Hybridization

The DNA base composition and the DNA–DNA hybridization were performed at DSMZ. Determination of DNA base (G+C) composition was performed by the HPLC method. Cell disruption was performed by using a constant systems TS 0.75 KW instrument (IUL Instruments, Germany). The DNA was purified on hydroxyapatite according to the procedure of Cashion et al. [4]. The DNA was hydrolyzed with P1 nuclease and the nucleotides dephosphorylated with bovine alkaline phosphatase [13]. The resulting deoxyribonucleosides were analyzed by HPLC. For DNA–DNA relatedness, cells of a pure culture of isolate W14T and K. oxytoca ATCC 13182T were disrupted by using constant systems TS 0.75 KW instrument (IUL Instruments, Germany), and the DNA in the crude lysate was purified by chromatography on hydroxyapatite as described by Cashion et al. [4]. The DNA–DNA hybridization test was conducted using the spectroscopic method described by De Ley et al. [6], with modification described by Huss et al. [8] in 2× SSC + 5 % formamide at 70 °C using a model Cary100 Bio UV/Vis-spectrophotometer equipped with a Peltier-thermostatted 6 × 6 multi-cell changer and a temperature controller with in situ temperature probe (Varian, Palo Alto, CA). DNA similarity values were determined in duplicate.

Automated Ribotyping

Automated ribotyping of the isolate W14T was performed with EcoRI digested samples using a RiboPrinter® system (DuPont Qualicon, Wilmington, DE) using the recommended methods from the manufacturer.

Gene Sequencing

Sequencing of the 16S rDNA region was performed by Accugenix, Inc. using proprietary methods. Briefly, the DNA was isolated with standard alkaline lysis methods and the extract used in PCR amplification with standard M13-tailed 16S primers, 0005F and 1540R, with the primer numbering based on the Escherichia coli 16S rRNA gene [3]. Sequencing reactions were carried out using manufacturer’s recommendations and BigDye® Terminator v1.1 cycle sequencing kits (Life Technologies, Carlsbad, CA) and run on the ABI 3130XL capillary sequencer (Life Technologies, Carlsbad, CA). Protein-coding genes were also amplified using M13-tailed gene specific primers and subjected to cycle sequencing as above. Consensus sequences for rpoB and gyrB were generated using information resulting from the amplification with two sets of primers for each gene. The full length sequences were used in the analyses. Primers for gyrA: entero_gyrA6F, 5′-GTAAAACGACGGCCAGTCGACCTTGCGAGAGAAAT-3′; entero_gyrA631R, 5′-CAGGAAACAGCTATGACGTTCCATCAGCCCTTCAA-3′ were used. Primers for rpoB: entero_rpoBb_F, 5′-AACAACCCGCTGTCTGAGAT-3′; entero_rpoBb_R, 5′-CATGTTCGCACCCATCAAT-3′; entero_rpoBe_F, 5′-CAGGCGAACTCCAACCTG-3′; entero_rpoBe_R, 5′-CGCCAGTTCACCGAGGTC-3′ were used. Primers for gyrB: entero_gyrB_F, 5′-CCGATCCACCCGAATATCT-3′; entero_gyrB_R, 5′-CCCTTCCACCAGGTACAGTT-3′; entero_gyrBd_F, 5′-GTAAAACGACGGCCAGTYGCNGGNGGNAARTTYGA-3′; entero_gyrBd_R, 5′-CAGGAAACAGCTATGACARRTGNGTNCCNCC-3′ were used. Similarity matches for the sequences of the 16S rDNA, rpoB, gyrA, and gyrB genes were obtained by performing BLAST. The Ribosomal Database Project (RDP) database [5] was also searched for the 16S rDNA similarity matches. Multiple alignment of the sequences were performed using the CLUSTALX [19] software, a distance matrix was created using Kimura’s 2-parameter model [10] and neighbor-joining phylogenetic method was applied for the construction of the tree using the software MEGA5 [18] for 16S rRNA, rpoB, gyrB, and gyrA gene sequence. Bootstrapping was performed by using 1,000 replicates.

PCR Analysis of “pehX” and “gyrA” Gene

Kovtunovych et al. [11] developed a specific method to discriminate K. oxytoca from other species of the genus Klebsiella based on PCR amplification of the polygalacturonase (pehX) gene (344 bp) for clinical and ecological monitoring of K. oxytoca. In this study, the primer pair developed by Kovtunovych et al. was used to detect if the pehX gene is present in the isolate W14T. Also, the genus specific primer targeting the gyrA gene developed by Brisse and Verhoef [2] was used for the amplification and sequencing of the genus specific region (441 bp) for Klebsiella to confirm the position of the isolate W14T within this genus.

DNA was isolated using the PrepSEQ Rapid Spin reagent (Life Technologies, Foster City, CA) according to the manufacturer’s instructions. The purity and the yield of the DNA were measured using the NanoDrop 1000 spectrophotometer. The primer pair, forward 5′ GAT ACG GAG TAT GCC TTT ACG GTG 3′ and reverse 5′ TAG CCT TTA TCA AGC GGA TAC TGG 3′ was used for the amplification of the pehX gene. PCR was performed using Thermo-Start Master Mix (Thermo Scientific, Waltham, MA) and the reaction mixture consisted of 10 μL of master mix, 2.5 μL of PCR enhancer, 0.3 μL (10 μM) of forward and reverse primer, 1 μL (100 ng) of DNA template, and 4.9 μL of water to make the final volume of 20 μL. For the gyrA amplification the primer pair, forward 5′ CGC GTA CTA TAC GCC ATG AAC GTA 3′ and reverse 5′ ACC GTT GAT CAC TTC GGT CAG G 3′ was used. The reaction mixture consisted of 10 μL of master mix, 2.5 μL of PCR enhancer, 1 μL of BSA, 0.3 μL of forward and reverse primer, 0.5 μL of DNA template, and 5.4 μL of water to make the final volume of 20 μL. The reaction conditions for the PCR were initial denaturation at 95 °C for 15 min, 35 cycles of denaturation at 95 °C for 30 s, annealing at 55 °C for 30 s, extension at 72 °C for 1 min 30 s, and final extension at 72 °C for 10 min. The PCR products were visualized on 2 % agarose gel (E-Gel®, Life Technologies, Foster City, CA). A 100-bp DNA ladder was used as molecular weight marker. PCR products were purified using QIAquick PCR purification kit (Qiagen, Valencia, CA) and sequenced both directions with the same primers as used in the PCRs. The obtained sequence of the gyrA gene was matched using BLAST for sequence similarity.

Results and Discussion

Morphology

Well-isolated colonies of W14T appeared within 24 h of incubation at 35 °C on both R2A agar and TSA media. The colonies were white, circular with entire margin and convex elevation (supplementary file). They had a glistening texture due to capsule production (negative stain). The size of a single colony of an 18–24 h culture was about 0.25–0.3 cm in diameter. The isolate was capable of growing over a temperature range of 10–45 °C with an optimum of 35 °C.

Biochemical Characteristics

The isolate was also capable of growing on MacConkey Agar (bright pink colonies indicating lactose fermentation), XLD (yellow colonies, turned the complete plate from pink to yellow indicating coliform), and Levine EMB (dark colonies with green metallic sheen indicating lactose fermentation with strong acid production as seen with some Enterobacter species). The cells were gram-negative straight rods (0.5–0.8 μm × 1.0–2.0 μm) and non-motile. The NaCl and pH tolerance were up to 6 % and 10.0, respectively. Biochemical tests were performed on isolate W14T to confirm its position in the genus Klebsiella and to specifically differentiate it from K. oxytoca and K. mobilis. The isolate W14T was oxidase, arginine, ornithine, phenylalanine and H2S negative, glucose fermenter, and citrate positive indicating that it belongs to the genus Klebsiella. The key physiological and biochemical characteristics are listed in Table 1. Biolog failed to identify the isolate and the closest match was K. oxytoca with a similarity and distribution values of 0.279 and 8.97, respectively. The API20E results indicated that the isolate W14T could be K. oxytoca with a probability value of 97.7 %.

Table 1 Phenotypic profiles of isolate W14T and other related bacteria
Full size table

FAME Analysis

The major fatty acids found in isolate W14T were C16:0 (31.23 %), sum feature 8, C18:1ω6c/C18:1ω7c (21.10 %), and sum feature 3, C16:1ω7c/C16:1ω6c (19.05 %). The comparative fatty acid profile of isolate W14T with K. oxytoca and K. mobilis is shown in the supplementary file.

MALDI-TOF Analysis

Upon analysis of the spectrum produced from the MALDI-TOF system, the top match in the reference databases was K. oxytoca ATCC 13182T (Score 1.998) followed by three additional strains of K. oxytoca, then K. mobilis, and R. ornithinolytica (Score 1.788). These data indicate that the protein spectrum from W14T shares similarity to organisms in the Enterobactericeae group and is close to K. oxytoca.

DNA Base Composition and DNA–DNA Hybridization

The DNA G+C content of isolate W14T was 54.6 mol%, which falls within the range of the genus Klebsiella. DNA–DNA hybridization result indicated that the isolate W14T does not belong to the species K. oxytoca ATCC 13182T (55.7 ± 6.2 %) when the recommendations by the ad hoc committee [20] for the definition of bacterial species are considered. These recommendations include a threshold value of 70 % DNA–DNA similarity, a level clearly above the observed level of the isolate.

Ribotyping

The ribotyping pattern for W14T was visibly distinct from K. oxytoca ATCC 13182T and K. mobilis ATCC 13048T when cut with EcoRI (Table 2). The RiboPrinter system uses a statistical match at or above a similarity index of 0.85 to assign a species level match [16]. The observation that W14T held similarity values <0.46 indicated that it was only distantly related to K. oxytoca and K. mobilis.

Table 2 Comparative ribotyping patterns of isolate W14T and other related bacteria, using the restriction enzyme EcoRI
Full size table

Gene Sequence Analysis

The almost complete 16S rRNA sequence (1,395 bp) analysis and the phylogenetic tree placed isolate W14T in the cluster III of the genus Klebsiella. The most closely related species was K. oxytoca (99 %) and the range of sequence similarity among the tested strains was 96–98 %. A dendogram is presented in Fig. 1. The 16S rRNA sequence data indicated that isolate W14T belongs to the genus Klebsiella but did not permit clear differentiation between other closely related bacteria. Previous studies have shown that 16S rRNA sequence does not provide clear differentiation between the species belonging to the genus Klebsiella (K. variicola and K. pneumoniae share 99.3 % 16S rRNA gene sequence similarity) [15] Therefore, three other protein-coding housekeeping genes, gyrA, gyrB, and rpoB, were sequenced. The genes encoding gyrA and gyrB (encoding the α- and β-subunits of DNA gyrase, a type II DNA topoisomerase) and rpoB (β-subunit of the bacterial RNA polymerase) are reported to be better molecular chronometers for the study of phylogenetic relationships in bacteria [14, 21]. It was reported [9, 17] that, due to its high levels of divergence, the rpoB gene sequence serves as a useful tool for identification among members of the Enterobacteriaceae family. The rpoB gene (834 bp) for isolate W14T showed 97 % similarity with both the type strains of K. oxytoca and K. mobilis (Fig. 2). The rpoB sequence similarity between the isolate W14T and K. oxytoca was observed to be lower than the sequence similarity reported by Li et al. [12] for the comparison between K. singaporensis and K. pneumoniae (97.5 %). The similarity was 2 % lower than the 16S rRNA sequence similarity with K. oxytoca and 1 % lower than that seen with K. mobilis indicating higher resolution compared to the ribosomal sequence. The gyrB gene sequence (978 bp) of isolate W14T showed 98 % similarity with the type strains of K. oxytoca and 90 % similarity with K. mobilis (Fig. 3). The gyrA gene sequence (626 bp) of isolate W14T showed 98 % similarity with the type strains of K. oxytoca and 93 % similarity with K. mobilis (Fig. 4). Phylogenetic analysis of the gene sequences for rpoB, gyrB, and gyrA indicated that isolate W14T belongs to the cluster III of the genus Klebsiella and is related to K. oxytoca (Figs. 2, 3, 4). Based on the gene sequence analysis, the isolate W14T also deserves the status of a new species belonging to the genus Klebsiella.

Fig. 1
figure 1

Phylogenetic tree based on 16S rRNA gene sequences indicating the relationship between isolate W14T with other species of the family Enterobacteriaceae, constructed using CLUSTALX and MEGA5. Pseudomonas amygdali LMG 2123T was selected as the out group. Confidence limit for the tree was assessed using bootstrap with 1,000 replicates. The evolutionary distances were computed using the Kimura 2-parameter method. Bar 2 % sequence variation

Full size image
Fig. 2
figure 2

Phylogenetic tree based on rpoB gene sequences indicating the relationship between isolate W14T with other species of the family Enterobacteriaceae, constructed using CLUSTALX and MEGA5. Pseudomonas amygdali LMG 2123T was selected as the out group. Confidence limit for the tree was assessed using bootstrap with 1,000 replicates. The evolutionary distances were computed using the Kimura 2-parameter method. Bar 5 % sequence variation

Full size image
Fig. 3
figure 3

Phylogenetic tree based on gyrB gene sequences indicating the relationship between isolate W14T with other species of the family Enterobacteriaceae, constructed using CLUSTALX and MEGA5. Pseudomonas amygdali NCPPB 2607 was selected as the out group. Confidence limit for the tree was assessed using bootstrap with 1,000 replicates. The evolutionary distances were computed using the Kimura 2-parameter method. Bar 5 % sequence variation

Full size image
Fig. 4
figure 4

Phylogenetic tree based on gyrA gene sequences indicating the relationship between W14T with other species of the family Enterobacteriaceae, constructed using CLUSTALX and MEGA5. Bacillus subtilis KCTC 3135T was selected as the out group. Confidence limit for the tree was assessed using bootstrap with 1,000 replicates. The evolutionary distances were computed using the Kimura 2-parameter method and are in the units of the number of base substitutions per site. Bar 5 % sequence variation

Full size image

PCR Analysis of pehX and gyrA Gene

No PCR product was observed for the amplification of the pehX gene for the isolate W14T indicating that the polygalacturonase gene is absent in the isolate and it does not belong to the K. oxytoca species. The 441-bp amplicon was observed for the gyrA gene and the sequence obtained from the PCR product confirmed the position of the isolate in the genus Klebsiella.

Taxonomic Conclusion

This study was conducted to characterize and identify isolate W14T recovered from a tooth brush holder. Based on 16S rRNA, gyrB, and gyrA gene sequence analyses, the isolate W14T was related to K. oxytoca. The DNA–DNA hybridization, rpoB, FAME, ribotyping, and biochemical analysis data indicated that it deserves a status of a new species belonging to the genus Klebsiella. Additionally, the biochemical profile of isolate W14T was not only consistent with the characteristics of the genus Klebsiella but there were some key differences that might be important for distinguishing this isolate from K. oxytoca such as urease production, utilization of putrescine, and pectate degradation (a key test for K. oxytoca). Isolate W14T was negative for urease production, utilization of putrescine, and pectate degradation while K. oxytoca was positive for urease production, utilization of putrescine, and pectate degradation. W14T could be easily distinguished from K. mobilis based on motility, indole production, and ornithine decarboxylation (a key test for the genus Enterobacter). In contrast to K. mobilis, isolate W14T is non-motile, positive for indole production and negative for ornithine decarboxylation. Therefore, on the basis of physiological, chemotaxonomic, phylogenetic, and DNA–DNA hybridization characteristics of the isolate W14T, it cannot be assigned to a previously described bacterial species. We propose that isolate W14T is considered as a novel species, K. michiganensis sp. nov.

Description of K. michiganensis sp. nov

Klebsiella michiganensis (mi.chi.ga.nen.sis. N.L. masc. adj. michiganensis, of or belonging to Michigan State, USA, where the type strain was isolated).

Isolate W14T has small, circular, mucoid, convex, entire, opaque, and white colonies. The cells are gram-negative, facultatively anaerobic, straight rods, non-motile, and capsulated. Cells are 0.5–0.8 μm in diameter and 1.0–2.0 μm in length. The cells are capable of growing at 10–45 °C with an optimum temperature of 35 °C. The optimum pH is 7.0. NaCl tolerance is up to 6 %. It is oxidase, urease, and gelatinase negative, but positive for catalase, indole, adonitol, and lysine. The Voges–Proskauer test is positive and the methyl red reaction is negative. Nitrate reduction test is positive. Citrate can be used as a sole carbon source. It utilizes glucose, mannitol, inositol, lactose, sucrose, arabinose, rhamnose, melibiose, and sorbitol, but fails to produce H2S. It is also negative for arginine dihydrolase, ornithine, dulcitol, tryptophan deaminase, phenylalanine, and pectate degradation. The G+C content of the DNA is 54.6 mol%. The type strain is W14T (=ATCC BAA-2403T=DSM 25444T).