Relationship among bats, parasitic bat flies, and associated pathogens in Korea

Bats are hosts for many ectoparasites and act as reservoirs for several infectious agents, some of which exhibit zoonotic potential. Here, species of bats and bat flies were identified and screened for microorganisms that could be mediated by bat flies. Bat species were identified on the basis of their morphological characteristics. Bat flies associated with bat species were initially morphologically identified and further identified at the genus level by analyzing the cytochrome c oxidase subunit I gene. Different vector-borne pathogens and endosymbionts were screened using PCR to assess all possible relationships among bats, parasitic bat flies, and their associated organisms. Seventy-four bat flies were collected from 198 bats; 66 of these belonged to Nycteribiidae and eight to Streblidae families. All Streblidae bat flies were hosted by Rhinolophus ferrumequinum, known as the most common Korean bat. Among the 74 tested bat flies, PCR and nucleotide sequencing data showed that 35 (47.3%) and 20 (27.0%) carried Wolbachia and Bartonella bacteria, respectively, whereas tests for Anaplasma, Borrelia, Hepatozoon, Babesia, Theileria, and Coxiella were negative. Phylogenetic analysis revealed that Wolbachia endosymbionts belonged to two different supergroups, A and F. One sequence of Bartonella was identical to that of Bartonella isolated from Taiwanese bats. The vectorial role of bat flies should be checked by testing the same pathogen and bacterial organisms by collecting blood from host bats. This study is of great interest in the fields of disease ecology and public health owing to the bats’ potential to transmit pathogens to humans and/or livestock.

that Anaplasma ovis, Bartonella spp., Rickettsia spp., and Trypanosoma spp. are present in these insects [8][9][10]. Bat flies are also considered vectors. In recent studies, bat ectoparasite burden was found to be proportional to Bartonella infection; moreover, Bartonella spp. were also detected in bat flies and host bats, underscoring the parasite vector potential. However, more research is warranted [11,12]. Furthermore, it has not been demonstrated that bat flies transmit Bartonella bacteria. The vector potential of bat flies was demonstrated only in Polychromophilus spp. [13].
Bat flies are obligate ectoparasites for bats and include endosymbiotic prokaryotes that are not yet well understood; however, it is assumed that they establish a symbiotic relationship with mutualistic bacteria [14]. Members of the superfamily Hippoboscoidea require milk secretion for larval development, and certain bacteria such as Bartonella and Wolbachia can be vertically transmitted during this process. These bacteria can also be horizontally transmitted through parasitoids or via contact with contaminated saliva [15,16]. However, horizontal transmission has not been proven in bat flies or any other hippoboscids.
Wolbachia is a bacterium belonging to the order Rickettsiales, which includes the genera Anaplasma, Ehrlichia, and Rickettsia. Wolbachia influences host reproduction through extensive symbiotic interactions in some species and is estimated to be present in up to 66% of insect species [17]. Wolbachia has become an integral component of vector-mediated disease control due to its ability to spread through insect populations and influence vector competence through pathogen protection [18]. Bartonella spp. are parasitic bacteria that infect the red blood cells of vertebrates. Several different bacterial species [19], including Bartonella mayotimonensis, are associated with bats, some of which are zoonotic and can cause disease in humans [5].
However, to date, few studies have examined the pathogenic relationships among bat flies, although previous studies reported the possibility of Bartonella and Wolbachia bacteria occurring transiently [20,21]. In general, the high degree of host specificity in bat flies reduces the likelihood of interspecies transmission of pathogens, but bat flies are still likely to carry transmissible pathogens within the host population [7]. Most previous studies on microorganisms, including those on Wolbachia spp., in bat flies have focused on endosymbiotic characteristics or distribution [2,20].
In Korea, many vector-mediated diseases and causative agents, including Anaplasma, Bartonella, and Borrelia, occur in humans [22][23][24]. However, there are few data regarding bats and bat flies in Korea. Furthermore, a recent Korea-focused study assessed the local distribution of bat flies without considering the pathogens mediated by these flies [25]. Therefore, the purpose of this study was to investigate the relationships among bats, parasitic bat flies, and their associated bacteria in Korea.

Study area, sample collection, and species identification of bats and bat flies
Bats and bat flies were collected from caves, forests, and abandoned mines in 12 cities across seven provinces [Gangwon (Inje), Chungbuk (Danyang), Gyeongbuk (Yeongju, Andong, Yeongcheon, and Gyeongju), Ulsan, Jeonbuk (Sunchang), Gwangju, and Jeonnam (Muan, Jindo, and Wando)] of Korea from February 2016 to August 2017 (Fig. 1). A total of 198 dead bats were found, and 74 bat flies were collected and immersed in 70% ethyl alcohol solution. The bat flies were collected by ecologists licensed from the National Institute of Environmental Research, Korea. The bat species were identified based on their morphological characteristics as previously described [26,27], and all collection-related information was provided from the ecologists.
Bat fly species were initially identified using key morphological characteristics, such as the presence or absence of wings, using a dissecting microscope (Fig. 2) [28]. The species were further identified at least at the genus level by analyzing the cytochrome c oxidase subunit I (COI) gene (approximately 658 bp length) [29,30], which was amplified through PCR using invertebratespecific primers [31,32].

DNA extraction and PCR assay
Bat fly DNA was extracted using the DNeasy ® Blood & Tissue Kit (Qiagen, Hilden, Germany) as per the manufacturer's instructions. An Infinite ® 200 PRO NanoQuant (Tecan, Männedorf, Switzerland) plate reader was used to assess the quality and quantity of bat fly DNA by calculating the ratio of the absorbance at 260 nm and 280 nm. DNA samples were stored at − 20 °C until further use.
The commercially available AccuPower ® HotStart PCR Premix kit (Bioneer, Daejon, Korea) was used for PCR. This premix product includes most of the elements required for PCR, including DNA polymerase, dNTPs, reaction buffer, and metal ions, lyophilized in a single tube. For PCR amplification, primers, DNA template, and distilled water are added until the total volume of the mixture reaches 20 μl. Bat fly-mediated pathogens and bacterial organisms were detected by amplification using primers specific to a target gene in each microorganisms. All reactions were performed using 20 μl reaction mixture containing 16 μl distilled water, 1 μl of 10 μM of each primer pair, and 2 μl template DNA.
We amplified the 16S rRNA regions of Rickettsiales (Anaplasma, Ehlrichia, Rickettsia, and Wolbachia species) and Coxiella spp.; 5S-23S rRNA regions of Borrelia spp.; 18S rRNA regions of Babesia, Theileria, and Hepatozoon species; and internal transcribed spacer I (ITS-1) regions of Bartonella spp. [23,[33][34][35][36]. Positive DNA samples were confirmed using a second set of PCR primers to amplify other regions of the gene, including citrate synthase gene (gltA) of Bartonella spp. [37] and cell division protein FtsZ (ftsZ) of Wolbachia spp. [36]. These primers are listed in Additional file 1: Table S1, along with their expected amplicon sizes. All PCR amplifications were performed using the Mastercycler ® nexus GSX1 (Eppendorf, Hamburg, Germany) under conditions outlined in Additional file 2: Table S2. The PCR products were electrophoresed on a 1% agarose gel stained with ethidium bromide. All amplicons were visualized and photographed using a UV transilluminator, and PCR-positive products were sent

Sequencing and phylogenetic analysis
The obtained sequence data were aligned and edited using BioEdit 7.2.5 [38]. MEGA 7 was used to construct phylogenetic trees for each species using the maximum likelihood method with 1000 replicates based on the fragments of COI, ftsZ, and gltA of bat flies, Wolbachia spp., and Bartonella spp., respectively [39]. Reference sequence data were obtained using NCBI Web BLAST (http:// www. ncbi. nlm. nih. gov/ blast). The phylogenetic tree of Wolbachia spp. based on ftsZ was constructed using 27 GenBank database entries and Ehrlichia sp. as the outgroup. The phylogenetic tree of the Bartonella spp. based on gltA was constructed using 15 GenBank database entries and Brucella sp. as the outgroup.

Identification of bat fly species
A total of 74 bat flies were collected from 198 bats. Ectoparasites other than bat flies were not detected. Organisms belonging to the families Nycteribiidae (89.2%, n = 66) and Streblidae (10.8%, n = 8) were identified in three species of host bat ( Table 2). Among the Nycteribiidae specimens, five species from three genera were identified, and the % similarities of the bat fly species based on their GenBank match were as follows: Nycteribia allotopa (100% with LC522000), Nycteribia parvula (96.7% with KF021501), Nycteribia pleuralis (94.3% with AB632553), Penicillidia jenynsii (98.6% with AB632562), and Phthiridium hindlei (99.9% with AB632569). Among the Streblidae specimens, only one species of the genus Brachytarsina was identified, and % similarity of the bat fly species based on GenBank match was Brachytarsina kanoi (93.0% with AB632571) ( Fig. 3; Table 2). Although it is the closest homology to this species, there is the probability of another closely related species apart from this. Most bat fly specimens were found on M. fuliginosus (51.4%, n = 38). All P. hindlei (n = 15) and Brachytarsina sp. (n = 8) were parasitic on the bat species R. ferrumequinum. Most P. jenynsii (family Nycteribiidae) flies were collected from the specimens obtained from the Jeonbuk Province (24/26), whereas only two such individuals were identified in specimens from the Chungbuk Province. Brachytarsina sp. (family Streblidae) was identified in three specimens from   Gyeongbuk, three specimens from Jeonnam, and two specimens from Ulsan (Table 3).

Screening for pathogens and endosymbionts mediated by bat flies
Of the identified pathogens and endosymbionts, 35  Through phylogenetic analysis of the Wolbachia spp., the three sequences from Phthiridium spp. bat flies were clustered as supergroup A, and the other two sequences from Nycteribia spp. and Penicillidia spp. bat flies were clustered as supergroup F (Fig. 4) [17].

Discussion
According to the morphological characteristics of bats identified in Korea, 24 species have been reported from 11 genera [26,27]. According to the key to the order Chiroptera in Korea, four bat families (Miniopteridae, Molossidae, Rhinolophidae, and Vespertilionidae) have been reported in Korea. In this study, 11 species belonging to seven genera in three families (Miniopteridae, Rhinolophidae, and Vespertilionidae) were identified. The collection places of bats were recorded for all except 31 specimens (these specimen data on tubes were erased because of leakage of ethanol during transfer), most of which were found in mines (79 specimens), caves (76 specimens), and the rest forests (12 specimens). Rhinolophus ferrumequinum was found in all regions except Gangwon; however, this does not necessarily mean that it does not occur in Gangwon. Only two bat specimens were collected from Gangwon. Rhinolophus ferrumequinum is known to be the most widely occurring bat in Korea [26]. A total of 74 bat flies were collected. One possible reason for this small number could be that the flies were collected from dead bats. Therefore, several bat flies would have left after the host bat died.
Molecular identification and phylogenies of bat flies have been widely utilized over the last decade to characterize different fly species [40,41]. COI, in particular, has been proven a useful marker in the documentation of invertebrates and insects [41][42][43]. Previous studies on Korean bat fly were limited to morphological characteristics [25]. Moreover, traditional morphological identification methods are extremely time-consuming because bat flies are complex, extremely small, and diverse in species. Although this does not justify the lack of morphology-based identification, our results showed that a molecular approach would be useful in the quick identification of the different species of bat flies, at least at the genus level. There was a limitation that COI was not sufficient to reliably identify bat flies by itself at the species level because sequence data of many species are still missing from the GenBank. Therefore, in this study, COI sequencing enabled genus-level identification and indicated the closest GenBank match species. In some species, the most similar GenBank sequences showed a similarity of < 95% (N. pleuralis, 94.3%; B. kanoi, 93.0%), whereas others showed a similarity of > 98% (N. allotopa, 100%; P. jenynsii, 98.6%; P. hindlei, 99.9%). Therefore, more comparable data on COI will allow a clear distinction of bat flies at the species level. Previously, COI was also used for species identification in other families of Diptera [44,45].
Phthiridium hindlei is an ectoparasite of R. ferrumequinum, which has not been reported in Korea previously. Most Nycteribia spp. were detected in M. fuliginosus (8/12, 66.7%), which is consistent with the findings of a previous study that collected Nycteribia spp. from Miniopterus sp. in Korea [25].
Penicillidia jenynsii (family Nycteribiidae) was mostly collected from the Jeonbuk Province (24/26), with only two individuals identified in the Chungbuk Province. Brachytarsina sp. (family Streblidae) was confirmed only  14:503 in Jeonnam, Gyeongbuk, and Ulsan Provinces, but not in Jeonbuk and Chungbuk Provinces. However, because there is a history of discovery of Brachytarsina sp. in Jeju Island and Gangwon Province [25] in the southernmost and northernmost regions of Korea, respectively, more population studies on ectoparasites and identification are required.
This study confirmed the high prevalence of Wolbachia and Bartonella bacteria and that P. hindlei was highly coinfected with Wolbachia and Bartonella spp. (9/10). According to the identification results of these pathogens and endosymbionts in this study, Wolbachia spp. was identified at a rate of 53.0% (35/66) in the family Nycteribiidae [Penicillidia sp. (22/26, 84.6%), Nycteribia spp. (2/12, 16.7%), Phthiridium sp. (10/15, 66.7%) and an unidentified Nycteribiidae species (1/13, 7.7%)] but not in the family Streblidae (Brachytarsina sp.). This is believed to be associated with the vertical transfer of the endosymbiont from mother to offspring through the mammary glands. Endosymbiont localization is consistently observed in all nycteribiid bat flies. In particular, P. jenynsii females exhibit a different pattern from that of males. In the abdominal cavity of females, larvae were found around the mammary glands, which supplied secretions and exhibited endosymbiont signals [20]. Bartonella spp. were identified at a rate of 27.3% (18/66)  A previous study reported that the most common microparasites in bat flies were bacteria (n = 149), with high numbers of Bartonella spp. (n = 91, 61.0%) but few Wolbachia spp. (n = 8, 5.4%) [2]. However, in this study, Wolbachia spp. were detected at a higher rate (35/74, 47.3%), whereas Bartonella bacteria detection was less frequent (20/74, 27.0%). This could be due to the differences in sample collection time and areas.
As per the phylogeny results of Wolbachia and Bartonella bacteria, we confirmed that each species of bat flies clustered as separated roots of each type. The Wolbachia endosymbionts detected in our study were clustered into two supergroups, A and F. Supergroup A was found in arthropods, whereas supergroup F was found in both filariae and arthropods [17]. All supergroup A Fig. 4 A phylogenetic tree was constructed with Wolbachia ftsZ gene-amplifying sequences generated in this study using the maximum likelihood method based on the Tamura-Nei model (1000 replicates). Sequences identified in this study are marked with black circles (•) with isolated ID and host species scientific name. The Ehrlichia sequences were used as outgroup endosymbionts were detected in P. hindlei. Supergroup F endosymbionts were detected in other bat flies (Penicillidia sp. and Nycteribia spp.). The presence of Wolbachia spp. confirmed that various arthropods could be vectors, and the bat fly Wolbachia spp. could have sequences similar to both filariae and arthropods. In addition, the possibility of Wolbachia bacteria transmission through bat blood should be studied to determine the connection between bat flies and bats.
Bartonella endosymbionts were also grouped according to their bat fly host except P. hindlei. However, these sequences clustered into distinct Bartonella groups and were considered a separate Bartonella species because of its < 96% identity [46,47]. Interestingly, in this study, with two separate branch groups, Bartonella host specificity in bats and bat flies was confirmed; one is P. jenynsii group and the other N. allotopa group, all of which had Miniopterus sp. bat hosts. In a previous study conducted in northern China, bat and Bartonella host specificity was recorded [48], and it was suggested that N. allotopa is the vector for transmitting Bartonella in bats [49]. However, the correlation between Bartonella species originating from bats and other bat fly species (another Nycteribia sp. and P. hindlei) requires further study.
Bartonella bacterial infection has been reported in humans and animals in African countries, where it was observed that a large number of local bat flies were positive for Bartonella spp. [50]. Bat ectoparasites generally exhibit high host specificity; therefore, their impact on other animal species and humans may be low, but the spread of bat-borne Bartonella spp. poses a global risk [29,51]. Furthermore, it must be considered that endosymbionts of bat flies may come from their bat hosts. Recent research suggests that bat flies transfer viruses to host bats as well as humans [7]. Nycteribiids are known to host several Bartonella spp., and bat and Bartonella bacteria associations have been studied in several parts of the world, including Asia [30]. In this study, our data indicated that Wolbachia and Bartonella bacteria are associated with bat fly species. In

Conclusions
This study employed morphological and molecular techniques to identify bat fly species in Korea. We also determined the distribution of P. hindlei and its endosymbionts. Using molecular methods, we identified several microorganisms, such as the endosymbiotic Wolbachia and possibly pathogenic or endosymbiotic Bartonella of bat flies, that are parasites for Korean bats. This is the first study to use such methods to identify Korean bat flies. Although the possibility of pathogen transmission through direct contact with a bat fly is low, subsequent studies on bat blood are required to confirm the potential for direct infection between bats and bat flies. This is an important public health concern owing to its potential for transmission to other species through bats.
Abbreviations COI: Cytochrome c oxidase subunit I; ITS-1: Internal transcribed spacer I.