Diversity unearthed by the estimated molecular phylogeny and ecologically quantitative characteristics of uncultured Ehrlichia bacteria in Haemaphysalis ticks, Japan

Ehrlichia species are obligatory intracellular bacteria transmitted by arthropods, and some of these species cause febrile diseases in humans and livestock. Genome sequencing has only been performed with cultured Ehrlichia species, and the taxonomic status of such ehrlichiae has been estimated by core genome-based phylogenetic analysis. However, many uncultured ehrlichiae exist in nature throughout the world, including Japan. This study aimed to conduct a molecular-based taxonomic and ecological characterization of uncultured Ehrlichia species or genotypes from ticks in Japan. We first surveyed 616 Haemaphysalis ticks by p28-PCR screening and analyzed five additional housekeeping genes (16S rRNA, groEL, gltA, ftsZ, and rpoB) from 11 p28-PCR-positive ticks. Phylogenetic analyses of the respective genes showed similar trees but with some differences. Furthermore, we found that V1 in the V1–V9 regions of Ehrlichia 16S rRNA exhibited the greatest variability. From an ecological viewpoint, the amounts of ehrlichiae in a single tick were found to equal approx. 6.3E+3 to 2.0E+6. Subsequently, core-partial-RGGFR-based phylogenetic analysis based on the concatenated sequences of the five housekeeping loci revealed six Ehrlichia genotypes, which included potentially new Ehrlichia species. Thus, our approach contributes to the taxonomic profiling and ecological quantitative analysis of uncultured or unidentified Ehrlichia species or genotypes worldwide.


Scientific Reports
| (2021) 11:687 | https://doi.org/10.1038/s41598-020-80690-7 www.nature.com/scientificreports/ Ehrlichia species or genotypes from ticks as well as wild mammals worldwide, including Japan. Representatives of the published reports are summarized in Supplementary Table S2. Because these previous studies used the individual PCR primers for respective target genes (e.g., 16S rRNA, groEL, gltA, rpoB, and dsb) [8][9][10][11][12][13][14][15][16][17][18][19] , all Ehrlichia species or genotypes detected cannot be directly compared or classified. Moreover, although the amount of ehrlichiae in a single tick is thought to be significant information from an ecological viewpoint, such as tick transmission and surveillance, these data remain elusive. Specific quantification data for some identified Ehrlichia species obtained by quantitative real-time PCR (qPCR) have previously been reported 18,[20][21][22] , but these assays are not available for other or unknown Ehrlichia bacteria in ticks or wild mammals. This study aimed to (1) identify uncultured Ehrlichia bacteria from ticks in Japan by specific p28-PCR screening, (2) characterize the detailed diversity unearthed by estimated molecular phylogeny based on five additional housekeeping genes, and (3) quantify the copy number of ehrlichiae in a single tick by a newly developed qPCR assay. Therefore, this study is expected to provide an improved understanding of the taxonomic status and ecologically quantitative properties of uncultured Ehrlichia bacteria in a tick, and our approach will be applicable and contribute to the comprehensive taxonomic and ecological profiling of unidentified Ehrlichia species or genotypes worldwide

Results
Ehrlichia detection from Haemaphysalis ticks by p28-PCR screening. We collected 616 unfed ticks in endemic areas for Japanese spotted fever as well as Anaplasma phagocytophilum-infected ticks inhabiting these areas in Japan 23 (Table 1). For PCR screening, we selected p28 multigenes of Ehrlichia spp. as target genes because p28 paralogous genes are highly specific for Ehrlichia spp. 9,24 . As determined by p28-PCR screening, 11 out of 616 ticks (1.8%; nine H. longicornis individuals [six males and three females), one H. hystricis individual (female) and one H. flava individual (nymph)] were found to be positive for ehrlichial infection by PCR (Table 1). Statistical analysis using Fisher's extract test showed that the ratio of positive adult ticks (males + females) was significantly higher than that of positive nymphs (p = 0.003). Three tick species, H. cornigera, H. kitaokai, and H. formosensis, yielded negative PCR results. The amplicons of p28 multigenes from eight positive ticks were successfully cloned, and the recombinant clones were randomly selected and sequenced. However, three amplicons from a female H. longicornis individual, a female H. hystricis individual, and a H. flava nymph could not be cloned because the amount of gel-purified amplicons was extremely low. A phylogenetic tree was constructed based on the deduced amino acid sequences from 21 p28-different clones obtained in this study and eight closely related clones from the other Ehrlichia spp. (similarities of 61.4-98.9% among all p28 clones within the tree, Fig. 1 S1). In the V1-V9 regions, the V1 region was found to be exhibit the most diversity among 33 Ehrlichia 16S rRNA sequences (Fig. 3). The calculated numbers of polymorphic sites in the V1-V9 regions confirmed showed that the V1 region had the highest number of these sites ( Table 2).   www.nature.com/scientificreports/ Estimation of the copy numbers of uncultured Ehrlichia bacteria in a single tick. We first performed real-time PCR targeting groEL, gltA, ftsZ and rpoB using serially diluted genomic DNA extracted from E. chaffeensis-infected DH82 cells. As a result, we selected gltA as the target that appears to be the most adequate in terms of specificity and sensitivity for our purpose. Calibration curves were then prepared by qPCR using dilutions of the standard gltA amplicons as the DNA templates in each experiment (a representative curve is shown in Supplementary Fig. S2). Based on this analysis, the limit of detection of the qPCR assay was estimated to equal three copies. Subsequently, the gltA copy numbers corresponding to Ehrlichia numbers in 1 µL of Ehrlichiapositive tick samples were determined to equal 2.1E+2 to 6.6E+4 (Table 3). Based on the DNA elution volumes, the number of ehrlichiae in a single tick was ultimately estimated to be in the approx. range of 6.3E+3 to 2.0E+6.
Taxonomic characteristics of uncultured Ehrlichia bacteria unearthed through estimated molecular phylogeny. The taxonomic status of the uncultured ehrlichiae without genome sequence data was unearthed via estimated molecular phylogeny through a core-partial-RGGFR-based phylogenetic analysis based on the concatenated sequences of five housekeeping loci in the order 16S rRNA-groEL-gltA-ftsZ-rpoB (total length of 2845-2846 bp) using the closely related Anaplasmataceae bacteria with genome sequence data available in the database (Fig. 4). The tree revealed that the uncultured ehrlichiae were located separately into    Fig. 3 and Supplementary Fig. S1 is shown here. www.nature.com/scientificreports/   . Core-partial-RGGFR-based phylogenetic analysis of uncultured Ehrlichia bacteria and cultured Ehrlichia species with genome sequence data. Tree based on five housekeeping genes (16S rRNA-groEL-gltA-ftsZ-rpoB) of uncultured Ehrlichia bacteria in this study and cultured Ehrlichia species for which complete genome sequence data have previously been obtained were constructed using the maximum likelihood method with the Kimura two-parameter model and "complete deletion" for gap/missing data treatment. The bootstrap values from 1000 replications are shown on the branch nodes. The scale bar indicates the evolutionary distance. The uncultured Ehrlichia bacteria in this study are shown by the colored characters, and their "genotypes" that were estimated based on the core-partial-RGGFR alignment and summarized in Table 4 are shown on the right side. The accession numbers and the location of isolated or identified Ehrlichia species or genotypes are shown in parentheses. Welgevonden. Based on these taxonomic analyses, six Ehrlichia genotypes, including potentially new Ehrlichia species, were found to exist in Japan and are summarized in Table 4.

Discussion
The present study demonstrated the molecular-based diversity among 11 newly identified Ehrlichia members from Haemaphysalis ticks (H. longicornis, H. hystricis, and H. flava) through our developed taxonomic profiling approach. The genotypes of these 11 members based on the current phylogenetic analyses are summarized in Table 4 and discussed here. Based on the 16S rRNA and groEL analyses, five Ehrlichia sp., MieHl1, MieHl25, MieHl31, MieHl145, and MieHl192 (termed genotype 1), are likely members of the Candidatus E. shimanensis group. On the core-partial-RGGFR-based tree, two Ehrlichia sp., MieHl92 and MieHl94 (genotype 2), were mainly associated with E. muris, Ehrlichia sp. HF followed by E. chaffeensis. The analyses of groEL and gltA revealed that Ehrlichia sp., MieHfN113 and MieHl182 (genotypes 3 and 5), might be related to E. ewingii. It should be considered that E. ewingii infects granulocytes 1,2 , but most Ehrlichia members infect monocytes/ macrophages in mammalian cells. Biological information regarding such E. ewingii-related members will be investigated in future studies. Ehrlichia sp. MieHl173 (genotype 4) was located in a large clade in the rpoB-based tree but was classified into different clades in the other gene-based trees. Ehrlichia sp. MieHh24 (genotype 6) was found to likely be independently located in all trees. A previous study suggested that genomes from bacteria with the same species epithet consistently exhibit more than 96.8% identity in their core genome alignments 7 (Table 4).
In general, most Ehrlichia species or genotypes are horizontally transmitted between mammals as natural reservoirs and ticks as arthropod vectors through tick bites. Hence, many tick species have become vector candidates for several Ehrlichia species or genotypes. Indeed, previous studies conducted in East Asia have identified various uncultured Ehrlichia species or genotypes detected from Haemaphysalis ticks, such as H. longicornis 8,9,[26][27][28] , H. flava 8,27 , and H. megaspinosa 27 in Japan, H. longicornis 29,30 , H. flava 29 , and H. hystricis 31 in China, and H. longicornis 32,33 in South Korea. Furthermore, different Ehrlichia species or genotypes have been detected from Ixodes persulcatus 11,12 , I. ovatus 11,24,27,34 , I. granulatus 35 , and I. turdus 27 in Japan. Among these, E. muris subsp. muris and Ehrlichia sp. HFs, which have been isolated and maintained in culture (the genome information is shown in Supplemental Table S1) are potentially transmitted by I. persulcatus and I. ovatus, respectively 11,12,24,27,28,34,36 . Thus, the current and previous studies suggest that a single tick species carries several Ehrlichia species or genotypes Table 4. Genotypes of uncultured Ehrlichia bacteria in the current study based on the similarities of the core-partial-RGGFR alignment. Under consideration based on the bacterial species criteria (96.8%) due to core genome alignments as previously described 7 , the 11 uncultured Ehrlichia members could be divided into six genotypes, probably including novel Ehrlichia species. *H: Haemaphysalis. **Five Ehrlichia members in genotype 1 may belong to "Candidatus E. shimanensis" that has previously been proposed based on 16S rRNA and groEL 28 . www.nature.com/scientificreports/ through horizontal transmission. In terms of transmission ability, the variety of Ehrlichia species or genotypes is unlikely limitless and dependent on the tick species (probably "regulated due to adaptation in tick species"). Therefore, the global classification of unidentified Ehrlichia bacteria in ticks and mammalian hosts will be highly significant and demanded, e.g., as in the current study aiming to perform comprehensive taxonomic profiling without genome sequencing. The field survey of ehrlichiae in the current study was conducted in a narrow area of the Mie prefecture in Japan, which is a high-risk area for Japanese spotted fever and anaplasmosis 23,37 , because we have confirmed the presence of antibody against E. chaffeensis in the sera of some tick-borne-suspected patients with fever of unknown origin in this area. Accordingly, the members of the uncultured Ehrlichia genotypes in the current study have become candidate human pathogens. In Taiwan, two cases of human infections with E. chaffeensis have been reported 38,39 . The short 16S rRNA sequence (182 bp) detected from the first patient 38 was identical to that from all 11 members of the uncultured Ehrlichia genotypes as well as E. chaffeensis, and the Ehrlichia 16S rRNA sequence (345 bp) from the second patient 39 was only identical to that of Ehrlichia sp. MieHl92 and MieHl94 (genotype 2) as well as E. chaffeensis. Taken together, the results indicate that the members of genotype 2 that are mainly related to E. muris and Ehrlichia sp. HF followed by E. chaffeensis are the most likely candidates to act as human ehrlichiosis agents in Japan.
We performed p28-PCR screening and phylogenetic analysis based on the amino acid sequences of p28 clones. The p28 multigenes encode a major outer membrane protein (OMP) family consisting of 22 protein species 5,[40][41][42] . In general, the spontaneous mutation of bacterial OMP genes, including p28 multigenes, occurs frequently due to the pressure of environmental bias, e.g., the repeated host changes needed for Ehrlichia survival due to horizontal transmission between mammals and ticks. From this perspective, we constructed a phylogenetic tree of p28 (OMP) clones using the sequences of amino acids rather than nucleotides to confirm the phenotypic variation. Additionally, p28 clone data were not used for the multiple loci-based construction of trees for taxonomic profiling of the uncultured Ehrlichia bacteria in Fig. 4 because OMP gene mutations are thought to be undergo more rapid accumulation than those of housekeeping genes due to environmental bias.
16S rRNA sequences are available and frequently used for bacterial taxonomy, and the bacterial 16S rRNA sequences contain V1-V9 regions 25 . In 16S rRNA-based metagenomics, the V3-V4 regions have frequently been used as a target in bacterial identification at the family level and even the genus level 43 . However, the V1 region of Ehrlichia members is found to exhibit the highest variability, which suggests that the V1 region benefits the identification of Ehrlichia species or genotypes.
From an ecological perspective, the tick stage associated with Ehrlichia transmission is likely the "adult stage" based on the statistical significance between the positive ratios of adults and nymphs. It is possible that nymphs acquire Ehrlichia organisms from natural host mammals through horizontal transmission and become infected adult ticks through molting. In contrast, the nymph stage plays an important role in A. phagocytophilum transmission in Japan as described previously 23 . The number of ehrlichiae in a single infected tick that was estimated by the newly developed qPCR method was found to be in the wide range of 6.3E+3 to 2.0E+6. Some infected ticks can unexpectedly carry a large number of ehrlichiae. To the best of our knowledge, this study provides the first estimate of the number of ehrlichiae in an individual tick in nature by qPCR.
As mentioned above, the current study provides significant information regarding the taxonomic and ecological characteristics of uncultured Ehrlichia bacteria from Haemaphysalis ticks in Japan. Our approach is applicable and contributes to the detailed molecular-based characterization and the surveillance of unidentified or uncultured Ehrlichia species or genotypes worldwide in future studies.

Materials and methods
Tick collection and DNA preparation. A total of 616 unfed ticks were collected by flagging at 13 sites (within a square area located from 34°29′33.6″ to 34°36′59.9″N and 136°53′88.8″ to 136°62′48.6″E) of the forests or weedy regions in Shima Peninsula at the Mie prefecture, Japan, during June and July of 2018. These areas are known to be endemic for Japanese spotted fever 37 and at high risk for anaplasmosis 23 . The ticks obtained were morphologically identified as shown in Table 1 and maintained in sterile tubes under continuous humidity at 16 °C. For DNA preparation, these live ticks were washed with 0.12% sodium hypochlorite solution followed by 70% ethanol supplemented with 1% povidone-iodine solution for 10 min during each disinfection cycle to avoid contamination by soil bacteria on the body surface of the ticks and rinsed with phosphate-buffered saline (PBS, pH 7.4). The ticks were then individually dissected using a sterile and disposable blade, and total DNA was extracted from the whole tissues of each tick using the InstaGene Matrix Kit (Bio-Rad Laboratories, Hercules, CA, USA) or QIAamp DNA Mini Kit (QIAGEN, Hilden, Germany) according to the manufacturer's instructions. The final elution volume for DNA extraction was 30 µL, and the DNA samples were stored at − 30 °C until use.
Ehrlichia detection by p28-PCR screening from ticks. The DNA samples from all the ticks were individually screened by conventional PCR targeting p28-paralougaous genes, which are known as a p28-multigene family specific for Ehrlichia members 9,24 . The primers are shown in Supplementary Table S3, and nested PCR was conducted in a 25-µL reaction mixture containing 12.5 µL of 2× GoTaq (Promega, USA), 400 nM of each primer, and 1 or 2 µL of the DNA template under previously-described conditions 24 . The p28 amplicons obtained were subjected to gel-purification using a Wizard SV Gel and PCR Clean-Up System Kit (Promega, USA) and cloned into the pCR2.1 vector using the TA Cloning Kit (Thermo Fisher Scientific, Waltham, MA, USA). The recombinant plasmids were then introduced into Escherichia coli DH5α cells (Toyobo Co., Ltd., Osaka, Japan). The randomly selected recombinant p28 clones were sequenced and phylogenetically analyzed as described below. . Based on the results, we selected gltA as the target for the quantification of uncultured ehrlichiae in ticks because this gene appears to be the most adequate target in terms of specificity and sensitivity. For quantification, the DNA template prepared using genomic DNA extracted from E. chaffeensis-infected DH82 cells was subjected to PCR using external primers of gltA (Eh_gltA_112_F1 and Eh_gltA_686_R1 in Supplementary Table S3). The amplicon was subjected to gel purification, and the gltA copy number was calculated based on the concentrations of the amplicon using Qubit dsDNA HS Assay Kits (Thermo Fisher Scientific, Waltham, MA, USA). The purified amplicon was serially diluted (3 × 10 7 , 3 × 10 6 , 3 × 10 5 , 3 × 10 4 , 3 × 10 3 , 3 × 10 2 , 3 × 10 1 , 3 × 10 0 , and 3 × 10 -1 copies/µL) and used as standard DNA. qPCR with the serially diluted standards and the DNA samples was performed using internal primers of gltA (Eh_gltA_137_F2 and Eh_gltA_614_R2 in Supplementary Table S3). The gltA copy number corresponding to the Ehrlichia number per 1 µL of each sample was determined based on the calibration curve of the standard DNAs ( Supplementary  Fig. S2). The copy number of Ehrlichia in a single tick was estimated from a DNA elution volume of 30 µL.
In silico analysis and statistical analysis. The similarity of the nucleotide and amino acid sequences among uncultured Ehrlichia members investigated in this study and the related bacteria existing in GenBank was calculated using MegAlign of DNASTAR software in Lasergene version 14 (DNASTAR, Madison, WI, USA). For phylogenetic analysis, the uncultured Ehrlichia sequences were aligned with reference sequences using ClustalW (default parameters) within the MEGA program (version 7.0.26). Based on the sequence alignment, phylogenetic trees were constructed using the maximum likelihood method with the Jones-Taylor-Thornton model (using the default parameters except "all sites" for gaps/missing data treatment) for the amino acid sequences of p28 clones and using the Kimura two-parameter model (using the default parameters except "complete deletion" for gaps/missing data treatment) for the nucleotide sequences of respective genes. The core-partial-RGGFR-based phylogenetic tree was constructed using the concatenated sequences in the order 16S rRNA-groEL-gltA-ftsZ-rpoB (length of 2845-2846 bp) using the same procedure. Bootstrap values were obtained with 1000 replicates, and values higher than 70 were considered to indicate good confidence 44 . A statistical analysis between positive adult ticks and positive nymphs was performed using Fisher's extract test in R (R Core Team 2020, Version 4.0.2, https ://www.R-proje ct.org/). www.nature.com/scientificreports/