Morphological identification and characterization of ticks
Between 1959 and 2019 more than 20,000 ticks were collected from a wide range of hosts (e.g. cattle, goats, camels, hedgehogs) across China (Figure 1, Table 1, Additional file 1: Table S1). Species identification of these ticks were carried out based on morphological characters, such as the number of punctuations, size and shape of mouthparts, and genital aperture [27–29]. This revealed a total of at least 46 species comprising two families (Ixodidae and Argasidae) and eight genera: Haemaphysalis (19 species), Ixodes (10 species), Dermacentor (6 species), Rhipicephalus (four species), Hyalomma (two species), Amblyomma (three species), Argas (one species) and Carios (one species). Among these were the six most common tick species in China [6]: Rh. microplus, Rh. sanguineus, Ix. persulcatus, Ha. longicornis, De. silvarum and Hy. asiaticum. In addition, other common species were collected and analyzed, including Ix. sinensis (the vector of B. burgdorferi [30]), Ix. ovatus, De. steini, Ha. yeni, Ha. concinna, Hy. scupense, Rh. turanicus, Rh. haemaphysaloides and Ar. persicus, all of which were sampled from at least two provinces. In comparison, other species were more locally defined, such as Ha. tibetensis from Tibet, Ha. qinghaiensis and Ha. danieli from Qinghai, Hy. asiaticum, Ha. punctata, De. marginatus, and De. niveus from Xinjiang, Ix. acutitarsus from Hubei, and Ha. lagrangei and Ha. mageshimaensis from Hainan. In addition, we obtained a number of every rare species, particularly Am. javanense, Ix. simplex, Ix. nuttallianus, Ix. crenulatus, Ix. kuntzi, Ha. kitaokai, and Ca. vespertilionis, some of which were collected from wildlife animals including bats, pangolins and flying squirrels (Pteromyini). Other samples were collected by drag-flagging methods or were historical samples preserved in ethanol for more than 60 years. For example, the oldest sample from our data set (Am. javanense) was collected in the 1960s from a wild Chinese pangolin (M. pentadactyla). In the case of two samples (i.e. C20, A29) we could only identify them to the genus level, in Amblyomma and Ixodes, respectively. Since we were unable to identify them at the species level using morphological characteristics, they were tentatively assigned as potential new species.
Table 1
Sampling locations, animal hosts and bacterial pathogens of tick species in China at the genu level. The number of libraries denotes that how many libraries are included in each tick genus while bracketed numbers denote the numbers of mt genomes successfully obtained in each group.
Species(genus)
|
Location
|
Host
|
No. of libraries
|
Pathogen
|
Rhipicephalus
|
Hubei, Jiangxi, Yunnan, Beijing, Zhejiang, Xinjiang, Hainan
|
Cattle, dog, hedgehog
|
12 (11)
|
Rickettsia, Coxiella,
|
Hyalomma
|
Xinjiang, Hubei
|
Cattle, goat, camel
|
4 (3)
|
—
|
Dermacentor
|
Hubei, Jiangxi, Inner Mongolia, Beijing, Hebei, Zhejiang, Xinjiang
|
Cattle, goat, rabbit, hedgehog, wild boar
|
18 (15)
|
Rickettsia, Coxiella, Rickesiales unclassified
|
Amblyomma
|
Hubei, Zhejiang, Hainan
|
Wild boar, Malaysia pangolin, Chinese pangolin
|
6 (2)
|
Rickettsia
|
Haemaphysalis
|
Hubei, Jiangxi, Fujian, Zhejiang, Hainan, Yunnan, Beijing, Shaanxi, Heilongjiang, Liaoning, Qinghai, Tibet, Xinjiang, Inner Mongolia
|
Cattle, yak, pheasant, dog, hedgehog, muntjac, goat, wild boar, hog badger
|
30 (23)
|
Rickettisa, Coxiella,
|
Ixodes
|
Hubei, Jiangxi, Jiangsu, Zhejiang, Hainan, Inner Mongolia, Jilin, Tibet, Heilongjiang
|
Cattle, goat, Eurasian badger, hog badger, rabbit, bat, pteromyini, tupaia
|
22 (16)
|
Rickettisa, Coxiella, Wolbachia
|
Carios
|
Hubei, Henan
|
—
|
2 (2)
|
Coxiella, Rickettisa, Borrelia
|
|
Heilongjiang, Xinjiang
|
Chicken
|
2 (2)
|
Rickettsia, Coxiella
|
Total
|
—
|
—
|
96 (74)
|
—
|
Mitochondrial Genomes Of 74 Ticks Of 46 Species
We sequenced the total DNA of 96 individual or mixed tick samples, which generated an average of 7.87 Gb of clean reads for de novo assembly and annotation. Complete or near-complete mt genome sequences were successfully obtained from 74 of the 96 libraries, including 30 species whose mt genomes were reported for the first time. The length of the newly identified mt genomes ranged from 14,428 bp to 15,307 bp, and the AT content varied from 72.29% (Ixodes sp. A29) to 81.06% (Ha. danieli Z14), similar to previously identified mt genomes of ticks [8]. Furthermore, the structure, composition, as well as the arrangement of genes largely followed their closest relatives within the same genus [8, 31]. The only differences were observed in the length and composition of non-coding regions, some of which contain more than one tandem repeat region (Additional files 2, 3, 4: Figure S1-S2, Table S2). For example, the mt genome of De. marginatus E48 has an extra copy of the non-coding region so that its length (i.e. 15,307 bp) has surpassed that of Ix. tasmani (NC 041086.1, 15227 bp) [8] to become the longest tick mt genome identified to date (Additional file 2: Figure S1). Furthermore, we found inconsistencies in the control region within some individual samples. For example, cloning of sequencing of PCR products spanning the control region between trnQ and trnF reveals various copy numbers of short repeat sequences within the same (single-tick, De. marginatus E1) sample (Additional file 3: Figure S2C).
Molecular Identification And Genetic Diversity Of Ticks
Both maximum likelihood (ML) and Bayesian phylogenetic trees were estimated based on sequences of 13 protein coding genes and two rRNA genes derived from 136 tick mt genomes, including 74 generated in this study and 62 reference mt genomes from GenBank. The ML and Bayesian methods resulted in highly similar tree topologies that placed the diversity of Chinese ticks within a global context with high resolution (Figure 2, Additional file 5, 6, 7, 8, 9, 10, 11, 12: Figure S3-S10). Importantly, the newly added genomes greatly expanded the diversity of many groups, particularly the genera Haemaphysalis, Ixodes and Dermacentor (Figure 2). In addition to the new species identified in this study, 28 tick species previously only known through morphological characteristics or incomplete mt genomes (e.g. Ix. kuntzi, Ix. acutitarsus, Ha. mageshimaensis and Ha. colasbelcouri) were also included (Figure 2, Additional file 5, 6: Figure S3-4). Furthermore, at least five potential cryptic species were identified – Rh. sanguineus, De. steini, De. marginatus, and Ix. ovatus – adding to the previously reported cryptic species identified in Rh. microplus [6, 32]. Each contained at least two divergent (70.93% ~ 94.21% identity) phylogenetic clusters while sharing the same morphological characteristics (Figure 2). Conversely, De. sinicus, De. nuttalli, and De. silvarum shared a very close relationship (> 98.46% identity) even though these were separate species based on morphological characteristics. Interestingly, De. nuttalli, and De. silvarum cannot be distinguished based on mt genome phylogeny, although they had quite distinctive Trochanter I dorsal spur (Additional file 13: Figure S11)
Discovery and characterization of bacterial endosymbionts and pathogens in ticks
Based on DNA sequencing, we searched for members of the tick-borne bacterial groups that are known to harbor human pathogens: the order Rickettsiales, genus Coxiella and genus Borrelia. Overall, 56% (54/96) of tick libraries were positive for these bacterial groups, among which Coxiella had the highest prevalence (40/96, 42%), followed by Rickettsia (26/96, 27%), Wolbachia (1/96, 1%), and Borrelia (1/96, 1%) (Figure 3A, Additional file 14: Table S3).
Most of the species in the order Rickettsiales belonged to the genus Rickettsia, within which 14 bacterial species were identified from all tick genera included in this study (with the exception of Hyalomma; Figure 3B), including a number of human pathogens. For example, R. raoultii, which causes human tick-borne lymphadenitis [33, 34], was identified from De. marginatus and De. niveus in Jinghe, Xinjiang province, a region where R. raoultii have previously been reported [35–37]. Within Xinjiang (Bole and Zhaosu) we identified R. sibirica and R. africa circulating in De. sinicus and Ix. vespertilionis, which are responsible for a range of tick-borne diseases, including Siberian tick typhus (STT) in Asia and African tick bite fever (ATBF) in Africa [19]. In addition, we discovered R. heilongjiangensis, the newly reported agent of Far-Eastern spotted fever (FESF) [38]. This bacterium was previously found in De. sylvarum ticks from Heilongjiang, and herein it was associated with the tick species Ha. campanulata and Ha. cornigera in Hubei and Jiangxi provinces located in central China. Other pathogenic Rickettsia species included R. tamurae, R. monacensis and R. helvetica identified from Am. testudinarium, Ix. sinensis and Ix. kuntzi.
Notably, we found a potentially novel species within Spotted Fever Group that was relatively divergent (< 99.46% genetic identity in six genes) from the rest of the bacteria in this group (Additional file 15: Table S4). Since the species were identified from Ha. megaspinasa, we tentatively named it Rickettsia endosymbiont of Haemaphysalis megaspinasa. Furthermore, we identified four genetically divergent Rickettsia species that occupied basal phylogenetic positions. Among these, Rickettsia endosymbiont of Ixodes persulcatus H5 and N2 clustered with R. canadensis (97.78% identity), Rickettsia endosymbiont of Argas persicus H1 fell with R. bellii and Rickettsia sp. MEAM1 (Bemisia tabaci) (92.16% and 90.96% identity), and Rickettsia endosymbiont of Ixodes vespertilionis A54b, Rickettsia endosymbiont of Carios vespertilionis X1 formed a monophyletic group with Rickettsia endosymbiont of Culicoides newsteadi despite a high level of divergence (86.08% and 85.97% identity, respectively) (Additional file 15, 16, 17, 18: Table S4, Figure S12-S14).
In addition to Rickettsia, we identified two potential new species within the order Rickettsiales. One, Wolbachia endosymbiont of Ixodes vespertilionis A54, clustered with W. pipientis strain FL2016 (95.37%) and Wolbahcia endosymbiont of Drosophila melanogaster (95.16%) within the genus Wolbachia. The identification of Wolbachia in ticks has been reported in recent years [39, 40]. The other - Rickettsiales endosymbiont of Dermacentor - clustered with an unclassified Rickettsiales bacterium Ac37b identified from Am. cajennense in Brazil (86.62% identity). Together they may represent a new genus or even family within the Rickettsiales (Additional file 15, 16, 18: Table S4, Figure S12-S13).
Bacteria of the genus Coxiella had the highest prevalence among the tick species examined (74%, 40/54). The newly discovered Coxiella species in the present study are highly diverse and greatly expand the genetic diversity within this group (Figure 3C). Indeed, new genetic lineages were defined based on our phylogenetic analysis, including Coxiella endosymbiont of Dermacentor marginatus, Coxiella endosymbiont of Haemaphysalis concinna and Coxiella endosymbiont of Ixodes ovatus, most of which were divergent from exiting members of Coxiella and generally associated with specific tick genera (Figure 3C, Additional file 19: Figure S15). In contrast, we discovered a C. burnetii, the causative agent of Q fever [41], from a De. sinicus tick sampled from Xinjiang province, which had high abundance (29.92 RPM) and was closely related with the “Dugway 5J108-111” strain sampled from the United States and transmitted by De. andersoni [42] (Figure 3C, Additional file 19: Figure S15). In addition to C. burnetii, we identified a single species of Borrelia from a soft tick Ca. vesperitilionis in Henan province. Based on the phylogenetic analyses, the newly identified bacteria, named Borrelia henanensis X1, fell within a clade “RF” that contains pathogens causing tick-borne relapsing fever (Figure 3D, Additional file 20: Figure S16) [43, 44].
Ecological and evolutionary patterns in ticks and their associated bacterial symbionts
We used Mantel tests to examine whether the tick host and/or geographic factors shape the genetic diversity of the bacteria they carry. For both Rickettsia and Coxiella, our results revealed positive and significant (P < 0.0005) correlations between tick and bacteria genetic distance matrices. However, no such significant correlation was found between bacterial genetic distance and geographic distance. Similar results were obtained using (i) partial Mantel analyses, in which we tested the effect between two factors while controlling for the third, and (ii) multiple linear regression analyses in which we tested the effect between three matrices (Table 2, Additional file 21: Table S5). These results suggested that bacterial genetic diversity was primary shaped by host genetic distance, with geographic distribution having little or no impact. The strong impact of host on bacterial genetic diversity were also reflected in the phylogenetic analysis in which we observed a significant clustering of bacterial genetic diversity at the host general level (Rickettsia: association index (AI) = 2.760, P < 0.001; Coxiella: AI=0.969, P < 0.001).
Table 2
Results of the Mantel test and Partial Mantel test comparing two factors (host genetic distance and geographic distance) that predict the structure of genetic diversity in bacterial pathogens.
|
Model
|
r value (P value)
|
Rickettsia
|
Hosta
|
0.5868 (0.0004)
|
Host | geographyb
|
0.5952 (0.0009)
|
Geographya
|
0.0028 (0.3916)
|
Geography | hostb
|
-0.1225 (0.8633)
|
Coxiella
|
Hosta
|
0.4939 (0.0001)
|
Host | geographyb
|
0.4939 (0.0001)
|
Geographya
|
0.0013 (0.4499)
|
Geography | hostb
|
0.0053 (0.4378)
|
aMantel test. |
bPartial Mantel test. |
| Indicate that the first factor excludes the effect of the second. |
We next examined whether the phylogeny of the ticks and their bacterial symbionts exhibited a pattern of bacterial-host co-divergence over evolutionary time. We first tested hypothesis of co-divergence using an event-based framework, based on which we reconciled the phylogenies of ticks and their associated bacteria (i.e. Rickettsia and Coxiella, respectively) by accounting for four processes: co-divergence, duplication, host switching and loss [45]. This revealed significantly fewer non-co-divergence events (i.e. duplication, host switching and loss) than expected by chance alone. We similarly examined the co-divergence hypothesis using a distance method, in which we evaluated the overall phylogenetic congruence by comparing the tick and bacterial symbionts patristic distance [46]. This confirmed the significant overall similarity (ParafitGlobal, P = 0.0021 and 0.0003, respectively, for Rickettsia and Coxiella, at 9999 permutations) between the tick and bacterial symbionts phylogenies (Figure 4). Collectively, these results suggest that the symbiotic bacteria from genera Rickettsia and Coxiella have co-diverged with their tick hosts for many millions of years.