Yersinia spp. Identification Using Copy Diversity in the Chromosomal 16S rRNA Gene Sequence.

API 20E strip test, the standard for Enterobacteriaceae identification, is not sufficient to discriminate some Yersinia species for some unstable biochemical reactions and the same biochemical profile presented in some species, e.g. Yersinia ferderiksenii and Yersinia intermedia, which need a variety of molecular biology methods as auxiliaries for identification. The 16S rRNA gene is considered a valuable tool for assigning bacterial strains to species. However, the resolution of the 16S rRNA gene may be insufficient for discrimination because of the high similarity of sequences between some species and heterogeneity within copies at the intra-genomic level. In this study, for each strain we randomly selected five 16S rRNA gene clones from 768 Yersinia strains, and collected 3,840 sequences of the 16S rRNA gene from 10 species, which were divided into 439 patterns. The similarity among the five clones of 16S rRNA gene is over 99% for most strains. Identical sequences were found in strains of different species. A phylogenetic tree was constructed using the five 16S rRNA gene sequences for each strain where the phylogenetic classifications are consistent with biochemical tests; and species that are difficult to identify by biochemical phenotype can be differentiated. Most Yersinia strains form distinct groups within each species. However Yersinia kristensenii, a heterogeneous species, clusters with some Yersinia enterocolitica and Yersinia ferderiksenii/intermedia strains, while not affecting the overall efficiency of this species classification. In conclusion, through analysis derived from integrated information from multiple 16S rRNA gene sequences, the discrimination ability of Yersinia species is improved using our method.

Introduction

The genus Yersinia is widely distributed in nature and currently has 17 species [1-3], three of which are pathogenic (Y. enterocolitica, Y. pestis, and Y. pseudotuberculosis) and have been exhaustively researched. The remaining species are referred to as Y. enterocolitica-like bacteria where their etiology in disease is not understood [4, 5]. Traditionally, bacteria were classified according to similarities and differences in phenotypes, such as morphology and biochemical reactions. The API 20E strip test is the standard for identifying Enterobacteriaceae; however this test has limitations in identifying Yersinia species. Several Yersinia species can be identified using API 20E strip test where the accuracy is influenced by passage number, culture conditions, and instability of some biochemical reactions [6]. Therefore, sensitive molecular biology methods are needed to assist the traditional approach to identify the Yersinia.

Since the 1980s, the ribosomal RNA gene has been used for phylogenetic studies, and ribosomal RNA-based approaches have been increasingly applied to bacterial classification and identification, especially using the 16S rRNA. The 16S rRNA gene is generally accepted as the best molecular sequence to use for identification because it is functionally constant and shows a mosaic of structure having conserved and variable regions and exists in all organisms; and its length is easily sequenced [7]. These properties make it uniquely well suited for systematics applications. However, the presence of multiple copies of the rRNA operons and intra-genomic heterogeneity in the 16S rRNA gene is regarded as a limiting factor in species identification [8-10]. In bacteria, three rRNA genes (16S, 23S, and 5S) are organized into a gene cluster, which is expressed as a single operon. The total number of rRNA operons in prokaryotes range from one to 15[10].With more and more bacterial genomes being completely sequenced, the heterogeneity of 16S rRNA gene at the intra-genomic level has been discussed. Pei collected 883 prokaryotic genomes from the GenBank database, representing 568 unique species, 235 of which have copy diversity in the 16S rRNA gene within a genome (0.06% to 20.38%)[9]. The diversity in 24 species are close to or over the threshold of the 16S rRNA gene-based operational definition of a species (1% to 1.3% diversity)[9], so these species maybe misclassified into a new species if a different copy is used for identification. Recently, 2,013 genomes were analyzed and 22.5% were found divergence at over 1% in the 16S rRNA gene copies [11]. In addition, the 16S rRNA gene shows high similarity comparing some different species where identical sequences were found even[12]. This limits the use of the 16S rRNA gene in identifying bacterial species. Therefore, sequencing the 16S rRNA gene from multiple operons from isolates is recommended to achieve significant phylogenetic information for species identification [13].

Given the reasons above, we used the copy diversity in the 16S rRNA gene to identify Yersinia species by analyzing multiple copies of the 16S rRNA gene from 768 strains within the genus Yersinia.

Materials and Methods

Source of Strains

We used 768 strains of Yersinia including 10 species of Yersinia. Seven hundred and eleven-one isolated strains were widely distributed within 21 provinces of China; and thirty-three reference strains were from Japan, France and the National Institutes for Food and Drug Control (NIFDC) in China. The dates of strain isolation encompass 52 years (1962–2014). Biochemical data were determined using API 20E test strips (Biomerieux, France), showing 407 strains of Y. ferderiksenii/intermedia, 119 strains of Y. kristensenii, 123 strains of Y. enterocolitica (72 nonpathogenic and 51 pathogenic), 46 strains of Y. pseudotuberculosis and 49 strains of Y. pestis. The source and hosts of strains are shown in Table 1. Twenty-four complete-genome-sequenced strains were selected from the NCBI (http://www.ncbi.nlm.nih.gov/pubmed/) with the GenBank number listed in Table 2.

Table 1

Source and host distribution of Yersinia strains used.

Source and host		Y. ferderiksenii/intermedia	Y. kristensenii	Y. enterocolitica		Y. pseudotuberculosis	Y. pestis
				Non-pathogenic	pathogenic
Strains isolated in China	Chicken	21	17	6
	Cattle	6		1
	Dogs	16	6	3
	Rats	166	31	43	8	28	2
	Ducks	2	4
	Goat	4	4	2
	Mandarin duck	5	7
	Swines	148	30	10	19	3
	Marmots						31
	Flies	2		1	1
	Ticks						2
	Fleas						6
	Diarrhea patients	12	1	4	16		4
	Food	8	9	2
	Otherse	9	4		3		4
	Total	407	119	72	51	46	49
Reference strains		8a	6b		4c	15d

aAll reference strains cited here are from NIFDC, two are Y. ferderiksenii strains (52235 and 52236) and six are Y. intermedia strains (52234, 52237, 52244,52248, 52249, and 52250).

bAmong the reference strains, four are from NIFDC (52232, 52242, 52246, and 52247), and two from Japan.

cAll reference strains are from Japan.

dAmong these, six are from NIFDC (53504, 53505, 53510, 53512, 53514, and 53518), eight (PTB3, YP1B, YB2B, YP011, YP014, YP2A, YP15, and YP6) from Japan, and one (YP010) from France.

e All bacteria strains were collected from animals, not human subjects.

Table 2

The GenBank numbers of 24 complete-genome-sequenced strains.

Strains	Genbank number	Strains	Genbank number
Y.aldovae 670–83	CP009781.1	Y. pestis Pestoides F	CP000668.1
Y. aleksiciae strain 159	CP011975.1	Y. pseudotuberculosis IP31758	CP000720.1
Y. ferderiksenii ATCC33641	KN150731.1	Y. pseudotuberculosis PB1/+	CP001048.1
Y. ferderiksenii Y225	CP009364.1	Y. pseudotuberculosis YPIII	CP000950.1
Y. intermedia strain Y228	CP009801.1	Y. rohdei strain YRA	CP009787.1
Y. kristensenii ATCC33639	CP008955.1	Y. ruckeri strain YRB	CP009539.1
Y. kristensenii Y231	CP009997.1	Y. similis strain 228	CP007230.1
Y. pestis A1122	CP002956.1	Y. enterocolitica strain 2516–87	CP009838.1
Y. pestis D106004	CP001585.1	Y. enterocolitica strain WA	CP009367.1
Y. pestis D182038	CP001589.1	Y. enterocolitica subsp. enterocolitica 8081	AM286415.1
Y. pestis KIM10+	AE009952.1	Y. enterocolitica subsp. palearctica Y11	FR729477.2
Y. pestis Nepal516	CP000305.1	Y. enterocolitica subsp. palearctica 105.5R(r)	CP002246.1

Culture and Identification of Strains

Enrichment was performed using phosphate-buffered saline with sorbitol and bile salts (PSB) at 4°C for 21 days. Then strains were inoculated onto Yersinia-selective agar (cefsulodin-irgasan-novobiocin [CIN] agar; Difco). Suspected colonies having a typical bull’s-eye appearance (deep red centers surrounded by an outer transparent zone) on CIN agar were selected and allowed to grow on brain heart infusion agar (Beijing Land Bridge Technology Co., Ltd., China) at 25°C for 24 to 48 hours to obtain pure strains[14, 15]. We used the commercial rapid identification system, API 20E strip test (bioMérieux, France), for the identification of presumptive isolates. The biotypes of Y. enterocolitica strains were identified using the scheme reviewed by Bottone [16].

T-vectors Cloning and Sequencing of the 16S rRNA Gene

The bacterial genome was extracted using a DNA nucleic acid extraction kit (Tiangen, China). The 16S rRNA gene was amplified using the universal primers 27F (5’-AGAGTTTGATCCTGGCTCAG-3’) and 1492R (5’-GGTTACCTTGTTACGACTT-3’)[17]with the Q5 Hot Start High-Fidelity 2×Master Mix(NEB, U.K.). After gel electrophoresis, specific PCR products were purified using a pEASY purification kit (Transgene, China). Connection and transformation were performed using the pEASYTM-Blunt Simple Cloning Kit (Transgene, China). The transformed bacteria were grown at 37°C for 12h on Luria-Bertani agar, containing X-gal (20 mg/ml), IPTG (24 mg/ml), and kanamycin (50 mg/ml). Five white single colonies were randomly selected for sequencing. All were sequenced with an ABI Prism BigDye Terminator cycle sequencing ready reaction kit using AmpliTaq DNA polymerase and an ABI Prism 377xl DNA sequencer (Applied Biosystems, Foster City, CA, USA) according to the instructions of the manufacturer at Tsingke BioTech Co., Ltd., sequencing the amplicons in both directions.

Sequence Analysis

Sequence alignment was performed using Seqman (Lersergene7.0). A phylogenetic tree based on a single copy of the 16S rRNA gene was constructed using Kimura’s 2-parameter distance and the neighbor-joining method (MEGA 6.0). All of the 16S rRNA gene copies from each complete-genome-sequenced strain were coded with a random number, and then five sequences were selected using a random number table. After elimination of redundancies, all the five sequences in the 16S rRNA gene of each strain were obtained corresponding to a unique pattern number. The minimum distance between species was calculated on the basis of a full array between every two strains and each of the five sequences. A phylogenetic tree based on the multi-copy of the diverse 16S rRNA gene was constructed using a distance matrix. The distance of the five sequences within each strain was also calculated.

Ethics

The sample collection and detection protocols were approved by the Ethics Review Committee [Institutional Review Board (IRB)] of National Institute for Communicable Disease Control and Prevention, Chinese Center for Disease Control and Prevention. The copy of ethical approval documents were provided comply with PLOS ONE requirements. The authors were involved in sample collections involving human subjects and authors can access to personal information before the authors had access to the data. Signed informed consent was obtained from all study participants. For all the patients under 18 years-old, a written consent form was signed by a parent or legal guardian. The field studies did not involve endangered or protected species, so the locations/activities for which specific permission was not required.

Results

16S rRNA Copies at the Intra-Genomic Level

In total, 3,840 sequences of 16S rRNA gene were obtained from 768 strains used in this study. One or more base difference between two sequences was defined as a different 16S rRNA gene type. The type number of intra-strain 16S rRNA genes is shown in Fig 1A. In general, 60% of the strains have two or three types, 17% have four, and 18% have one. Only 5% of the strains have five totally different 16S rRNA gene types. Fig 1B shows a comparison of the gene type number between the different Yersinia species. The number is primarily one or two in Y. pseudotuberculosis and Y. kristensenii, and two to four in Y. ferderiksenii/intermedia. The strains of pathogenic Y. enterocolitica usually have one to three types, whereas the non-pathogenic strains have two to three types where one strain having five identical copies is uncommon at about 8% of the strains. Y. pestis and Y. pseudotuberculosis do not have five different copy sequences.

Fig 1

A. Distribution of the type number of 16S rRNA genes in 768 Yersinia strains. The colours in different sections of the pie chart represent the type number of 16S rRNA gene in strains, one type, two types, three types, four types, five types, respectively. The number in the pie represents the number of strains that have each kind of copies of 16S rRNA gene, the percentage in parentheses represents the proportion of all strains. B. The proportion of each copy appearing in different Yersinia species. C. The identical 16S rRNA patterns that exist in different Yersinia species, except for Y. pestis and Y. pseudotuberculosis.

Numbers in the crossed circle represent the number of identical patterns in the corresponding Yersinia species. Numbers in parentheses represent the amount of total patterns in corresponding species. Specific patterns are not shown.

Pairwise Comparison of the 16S rRNA Gene at the Intra-Genomic Level

The sequence similarity comparison between pairwise copies of 16S rRNA genes in each strain is shown in Table 3 classified by Yersinia species. The sequence similarity of pairwise copies in each Yersinia species strain exceeds 99% for most strains. There are only nine Y. ferderiksenii/intermedia (accounting for 2.2%) of the strains with similarity below 98.7%, and one strain of Y. kristensenii and non-pathogenic Y. enterocolitica of relatively low similarity 96.77% and 97.94%, respectively, whose sequence variation are 47 and 30 bases, respectively.

Table 3

Pairwise comparison of 16S rRNA gene at the intra-genomic level in each Yersinia species.

Species	Similarity*			Minimum value (%)
	100%	100–99%	<98.7%
Y. ferderiksenii/intermedia	52(12.7%)	349(85.1%)	9(2.2%)	98.15
Y. kristensenii	45(37.2%)	75(62.0%)	1(0.8%)	96.77
Pathogenic Y. enterocolitica	12(22.6%)	40(75.5%)	1(1.9%)	98.22
Non-pathogenic Y. enterocolitica	6(8.0%)	68(90.7%)	1(1.3%)	97.94
Y. pseudotuberculosis	15(30.6%)	34(69.4%)		99.45
Y. pestis	5(9.1%)	50(90.9%)		99.79

* There is not one strain in our study with similarity between 99%-98.7% in different copies of 16S rRNA gene, so the group of <99%, ≥98.7% is not shown in Table 3. The group of 100–99% means the similarity <100% and ≥99%.

Cluster Analysis Based on 16S rRNA Gene Sequence

A phylogenetic tree (Fig 2), constructed on the basis of the five copies of 16S rRNA gene presented in each strain and using the minimum evolution method, can be divided into six groups: 1a, 1b, 2, 3, 4 and 5.

Fig 2

A phylogenetic tree constructed on the basis of the five copies of 16S rRNA gene in each strain using the minimum evolution method.

A. Dots with different colors represent the corresponding Yersinia species; tree branch colors are consistent with triangles in B., which represent different clustering groups.

Group 1a and group 1b comprise Y. ferderiksenii/intermedia strains, 220 and 90 strains, respectively. The similarity between strains of each group is over 90%. Group 3 primarily contains Y. enterocolitica strains; but clustered with three Y. kristenii strains and one Y. ferderiksenii strain. Y. enterocolitica strains of biotype 1b show a distinct sub-cluster (the dark blue and bold clade in Fig 1A) in group 3, three isolated strains and two complete-genome-sequenced strains WA and 8081. Two sub-clusters comprise group 4: one is a combination of Y. ferderiksenii/intermedia strains, represented by two reference strains of Y. ferderksenii, ATCC33641 and 52235; another one contains four Y. enterocolitica strains and one complete-genome-sequenced strain of Y. rohdei. Group 5 consists of all the strains of Y. pestis and Y. pseudotuberculosis involved in this study, and a complete-genome-sequenced strain of Y. similis, located outside of the group.

Group 2 contains multiple Yersinia species, such as Y. enterocolitica, Y. ferderiksenii/intermedia and Y. kristensenii, which can be divided into three clusters: a, b and c. Group 2a includes two reference strains of Y. kristensenii and one complete-genome-sequenced strain of Y. aleksiciae. Group 2b is complicated hence described in three parts. Group 2b-1 is primarily made up of Y. kristenii strains, next is two reference strains of Y. kristensenii, 52246 and 52232, and one complete-genome-sequenced strain of Y. ruckeri. Group 2b-2 has two sub-clusters, one consisting of an Y. enterocolitica strain and a whole-genome-sequenced strain of Y. aldovae, and another with 43 Y. ferderiksenii/intermedia strains of high similarity (>99.6%), two of which are reference strains of Y. intermedia. Group 2b-3 has three sub-clusters. The first one contains 13 Y. kristensenii strains, five Y. enterocolitica strains, and one complete-genome-sequenced strain of Y. ferderiksenii Y225 that is distant from the rest of the strains in this sub-cluster. The second contains 33 Y. kristensenii strains of high similarity (>99.6%), except for two Y. ferderiksenii/kristensenii strains. The third with high similarities between strains (>99.6%), presents diversity, containing 35 Y. kristensenii (two are complete-genome-sequenced) and 43 Y. ferderiksenii/intermedia strains (three reference strains of Y. intermedia and one complete-genome-sequenced strain of Y. intermedia). There is a single Y. ferderiksenii/intermedia strain alone between group 2–1 and group 2b-2. Group 2c contains nine Y. enterocolitica strains, two Y. kristensenii strains and one complete-genome-sequenced strain of Y. ferderiksenii.

The Dominant Patterns of the 16S rRNA Gene in Each Yersinia Species

Four hundred thirty-nine patterns of 16S rRNA gene sequences, coded as 0 to 438(S1 File and S1 Table), were obtained after removing the redundant sequences. Table 4 shows each Yersinia species has dominant patterns. Though non-dominant patterns are more than dominant patterns, they appear at a low frequency and usually appear only once. There exists the same dominant gene pattern in the strains of Y. kristensenii and Y. ferderiksenii/intermedia, and Y. pestis and Y. pseudotuberculosis, respectively.

Table 4

The dominant patterns of the 16S rRNA genes in each Yersinia species.

Species	No. strains	No. sequences of 16S rRNA gene	No. patterns of 16S rRNA gene	Dominant 16S rRNA gene pattern and its percentage
Pathogenic- Y. enterocolitica	53	265	48	32(51.7%)
Nonpathogenic- Y. enterocolitica	75	375	90	19(20.8%) 120(10.4%)
Y. ferderiksenii/intermedia group 1a	220	110	75	11(41.5%) 13(17.5%) 23(10.1%)
Y. ferderiksenii/intermedia group 1b	91	455	38	142(31.2%) 3(18.7%) 55(12.1%)
Y. ferderiksenii/intermedia group 2b	88	440	58	10(20.0%) 111(13.2%) 8(13.0%) 9(11.4%)
Y. ferderiksenii/intermedia group 4	7	35	20	72(14.3%) 98(11.4%)
Y. kristensenii	121	605	80	10(31.9%) 95(14.7%)
Y. pestis	55	275	45	1(34.2%)
Y.pseudotuberculosis	49	245	28	1(64.1%)

Identical 16S rRNA Gene Sequences from Different Yersinia Species

Strains of different Yersinia species have identical 16S rRNA gene sequences (Fig 1C). Non-pathogenic Y. enterocolitica strains (biotype 1A) have the same 16S rRNA gene patterns with many kinds of other Yersinia species, with nine same gene patterns with pathogenic Y. enterocolitica, four with Y. kristensenii strains and only one with the other Yersinia species. Y. kristensenii strains have three identical gene patterns with Y. ferderiksenii/intermedia strains from group 2b, one of which is dominant, accounting 20% for Y. kristensenii and 31.9% for Y. ferderiksenii/intermedia respectively. There also exist identical 16S rRNA gene patterns among non-pathogenic Y. enterocolitica, Y. kristensenii, and Y. ferderiksenii/intermedia strains from group 2b, and also among non-pathogenic Y. enterocolitica, Y. kristensenii, and pathogenic Y. enterocolitica. Only group 1a and 1b exist three same 16S rRNA gene patterns among the four groups of Y. ferderiksenii/intermedia (1a, 1b, 2 and 4). Y. pestis strains have six identical patterns with Y. pseudotuberculosis (not shown in Fig 1C), one of which is dominant in both species, accounting for 34.2% and 64.1%, respectively.

Discussion

Compared to bacteria identification methods using phenotype, the approach based on genotype stands out for its consistency. One desirable candidate is the 16S rRNA gene, highly conserved and seldom variable within species, and is becoming an important technique for phylogeny research and species classification[7]. Specifically speaking, for Yersinia, the limitation of commercial tests for identification systems such as API 20E strip test has shown a deficiency in distinguishing some species[18, 19], which can be made up through the method based on 16S rRNA gene. However, the similarity of 16S rRNA gene sequence between species is as high as 96.9%-99.8%; it is easy therefore to misclassify species of high homology[20]. Recently, the rRNA operon is shown to have multiple copies with some heterogeneity[17]. In some cases, the diversity between multiple copies of a strain is so high as to be misclassified into different species if a different copy is used to identify a bacterial strain [21, 22]. This is a limiting factor for species identification using a single 16S rRNA gene. The discrimination ability can be improved by integrating multiple 16S rRNA gene copies. Therefore, here we first report species differentiation using five copies of the 16S rRNA gene from 768 strains of Yersinia.

Previously, the copy number of Yersinia16S rRNA gene was shown to be around seven [11]. According to whole genome sequence strain from NCBI, the number of variable copies is within three. Accordingly, five colonies was randomly selected for 16S rRNA gene sequencing for each strain. Our results show 60% of the strains have two to three genotypes and only 5% have all 5 types, therefore the selection of 5 colonies is reasonable. Though 84% of the strains have more than one sequence type of 16S rRNA gene, the comparison of the intra-genomic 16S rRNA gene showed that the similarity of 98.4% of the strains is above the threshold of species identification (98.7% to 99%)[9]. In other words, for most Yersinia strains, the heterogeneity among different copies does not affect species classification. Whereas for 12 strains, the similarity of intra-genomic 16S rRNA gene is below or close to the threshold of classification. When using a single copy for identification, a different result may occur. Hence in this study, with integrated information from multiple 16S rRNA genes and a comprehensive reflection of phylogeny, species identification is more accurate.

A phylogenetic tree (Fig 2) was constructed by combining five 16S rRNA gene sequences for each strain, most clustering of species is accordant with classical identification. Y. pestis and Y. pseudotuberculosis share a large number of identical 16S rRNA gene sequences and are close to each other in the phylogenetic tree, which is consistent with the hypothesis that Y. pestis evolved from O: 1b strains of Y. pseudotuberculosis about 1,500–20,000 years ago[23, 24], indirectly supporting our method. Y. enterocolitica strains are in group 3, where Y. enterocolitica subsp. enterocolitica (biotype 1b) and Y. enterocolitica subsp. Palearctica (biotype 1A, 2, 3, 4 and 5) diverge into two clusters, coinciding with previous work[25-27].

The majority of Y. kristensenii strains are within group 2, spreading widely and having multiple branches. In group2b-3-3, Y. kristensenii and Y. ferderiksenii/intermedia strains are closely clustered with each other, sharing high similarity, due to the fact that the consensus sequence of the two species are also dominant in each species. Notably, very few Y. kristensenii strains cluster with Y. enterocolitica strains. Compared with other species, Y. kristensenii presents higher genetic variability. This may be owing to the uncertainty of biochemical identification that these ‘cross-over’ strains may be classified into other Yersinia species. Sequencing more than five random selected colonies may solve this problem and needs further investigation.

Y. ferderiksenii and Y. intermedia strains are indistinguishable using API 20E strip test where they are referred to as Y. ferderiksenii/intermedia. In this study, they are gathered within group 1a, 1b, 2b and 4. The reference strains of Y. intermedia were all in group 2b, so strains in this group should be classified to be Y. intermedia strains. Consequently, strains in group 1a, 1b and 4 belong to Y. ferderiksenii. Identified in 1980, Y. ferderiksenii included three geno-species without biochemical differences [28]. Clustering analysis showed sequence types of group 1a, 1b, and 4 clustered with geno-species 2, 3, 1 of Y. ferderiksenii strains, respectively (Fig 3). Therefore, the 16S rRNA gene sequence types can be used for unclassified strains from API 20E strip test, and also for subtyping strains.

Fig 3

Phylogenetic tree based on single 16S rRNA gene from Y. ferderiksenii/intermedia strains in group 1a, 1b, and 4 and strains of Y. ferderiksenii belonging to three geno-species.

Hollow circles represent all 16S rRNA gene types of Y. ferderiksenii /intermedia strains in group 1a, 1b, and 4; while solid circles represent Y. ferderikseniis trains of three geno-species [1]. Triangles represent identical 16S rRNA gene patterns of strains in group 1a and group 1b.

Presently there is divergence in species classification for Yersinia. In addition to traditional biochemical reactions, refined molecular biology techniques are needed to compensate for this deficiency. Y. aleksiciae and Y. kristensenii have no difference in their biochemical phenotype but differ into distinct species using 16S rRNA gene sequences[29]. In group 2a, Y. kristensenii reference strains (52242 and 52247) share 99.9% similarity with the genome sequenced strain of Y. aleksiciae, a recent phylogenetic relationship. A previous study shows Y.similis is indistinguishable from Y. pseudotuberculosis through biochemical phenotyping that it was regarded as Y. pseudotuberculosis; however, their pathogenicity are significantly different and they differed using 16S rRNA gene sequencing [30]. In this study, genome sequenced strain of Y. similis were inseparably clustered with Y. pseudotuberculosis and Y. pestis in group 5, an outlier (black dot in group 5 Fig 2A) with an outermost evolutionary distance. Further, several comparable subspecies clustered respectively in group2b-1 and group 2b-2: genome sequenced Y. ruckeri strain and four Y. kristensenii strains formed a sub-branch; Y. aldovae and one Y. enterocolitica strain formed another, far away from the rest of the Y. enterocolitica strains. We suspect this Y. enterocolitica strain may belong to another undescribed Yersinia species. Similarly, in group 4, three Y. enterocolitica strains formed another sub-branch with one genome sequenced strain of Y. rohdei. All of these strains show Yersinia is a genetically diverse genus; the strains involved here may be classified into other species of more diversity.

Based on integrated information from multiple 16S rRNA gene copies and clustering analysis, this study shows comparable species clusters of Yersinia and reduces identification chaos derived from using a single copy of the 16S rRNA gene sequence which is sometimes similar or identical between species, and remarkably increases the discrimination ability of 16S rRNA gene sequence. Further study using our classification method would allow increased differentiation of Yersinia species identification.

The 439 Patterns (Coded as 0 to 438) of 16S rRNA Gene Sequences in this study.

(PDF)

Click here for additional data file.

Patterns Code of Five 16S rRNA Gene Sequences for each Strain.

(XLSX)

Click here for additional data file.