Two New Haplotypes of Bartonella sp. Isolated from Lipoptena fortisetosa (Diptera: Hippoboscidae) in SE Poland

Simple Summary Lipoptena fortisetosa is a hematophagous ectoparasite of game animals feeding accidentally on companion animals and humans. Since the presence of numerous pathogenic microorganisms has been described in this species, monitoring its geographic distribution is of great epidemiological importance. To the best of our knowledge, we present two new haplotypes of Bartonella sp. isolated from L. fortisetosa in south-eastern Poland and confirm the presence of this invasive species in Lublin Voivodeship since 2013. Abstract Insects of the genus Lipoptena are parasitic arthropods with a broad host range. Due to the type of parasitism (hematophagy), their potential role as vectors of pathogens, i.e., Bartonella sp., Anaplasma phagocytophilum, Rickettsia spp., and Borrelia burgdorferi is considered. As the range of their occurrence has been changing dynamically in recent years and infestations of humans have increasingly been reported, these organisms are now the subject of numerous studies. Our research aimed to present the molecular characteristics of Bartonella sp. detected in Lipoptena fortisetosa parasitizing wild cervids in south-eastern Poland. Adults of Lipoptena spp. were collected from carcasses of roe deer and red deer between spring and autumn in 2013. The PCR method was used to detect Bartonella sp. in the insects. We report two new haplotypes of the rpoB gene of Bartonella sp. isolated from L. fortisetosa feeding on wild cervids in south-eastern Poland and the presence of this invasive ectoparasitic species in the studied area since 2013. Phylogenetic analyses of newly obtained Bartonella sp. haplotypes confirmed their unique position on the constructed tree and network topology. The rpoB gene sequences found belonging to lineage B support the view that this phylogenetic lineage represents a novel Bartonella species.

In Poland, L. fortisetosa was first found in Lower Silesia at the end of the 1980s [18]. This deer ked species was found again in 2007-2014 on red deer (Cervus elaphus Linnaeus, 1758) and roe deer (Capreolus capreolus Linnaeus, 1758) in the north [19,20], in environments in north-eastern and southern Poland, including the Polish part of the Tatra Mountains [21,22], and recently in northern and western Poland [23].
Depending on the geographical region as well as climate and ecological conditions, the level of prevalence and severity of invasion of specific ectoparasites varies significantly, e.g., in the case of L. cervi, it mainly depends on the host species [4,24] and exhibits seasonal differences: the highest prevalence is most often noted in autumn and winter [25]. L. cervi parasitize domestic and wild animals, primarily representatives of Cervidae-red deer, roe deer, and moose (Alces alces Linnaeus, 1758) [15,26], own observations. The L. fortisetosa host species have not been clearly defined, but they are probably the same animals as the hosts of L. cervi [13,16]. Human infestations by Lipoptena adults in their habitats are increasingly being reported. Their bites cause dermatitis in humans [3,[27][28][29]. In animals, the parasitism of these flies induces clinical symptoms related to anemia and skin mechanical damage [30,31].
Our study presents the molecular characteristics of Bartonella sp. detected in L. fortisetosa parasitizing wild cervids in south-eastern Poland.

Species Identification
Identification of the species and sex of the adult insects was carried out in the laboratory using an OLYMPUS SZX16 (Olympus, Tokyo, Japan) stereoscopic microscope and the key for identification of arthropod species compiled by Borowiec [45].

DNA Extraction and Polymerase Chain Reaction
The molecular analysis included 24 specimens of Lipoptena spp., each blood-fed adult from a different animal host. The DNA from 15 L. cervi (6 females and 9 males) and 9 L. fortisetosa (7 females and 2 males) randomly selected for the pilot study was isolated with the ammonia method [46]. Next, its concentration was measured spectrophotometrically using a nanospectrophotometer Pearl (Implen, Germany) at a 260/280 wavelength. Then, the samples were frozen at −20 • C and stored until further analysis. The PCR method and a pair of primers (1400F and 2300R) specific to the rpoB gene were used to detect Bartonella sp. in the insects [47]. The amplification product was separated electrophoretically in 2% ethidium bromide-stained agarose gel. Then, the gel was visualized in ultraviolet light in an Omega 10 device (Ultra Lum, Claremont, CA, USA). Next, the samples were analyzed with the use of Total Lab software (TotalLab, Newcastle-Upon-Tyne, UK). The presence of an 825 base pair PCR product was treated as positive. Next, this product was isolated from the agarose gel with the use of an Agarose Out kit (EURx, Gdansk, Poland) according to the manufacture's protocol and sequenced (Genomed, Warsaw, Poland).

Sequencing and Phylogenetic Analysis
The resulting sequences of the rpoB gene for RNA polymerase beta subunit were aligned and revised manually in BioEdit v7.0.4 [48]. The obtained sequences were submitted to GenBank. To test the phylogenetic relationships among our newly obtained rpoB gene haplotypes and sequences downloaded from GenBank, we constructed a phylogenetic tree using a maximum-likelihood (ML) algorithm in Mega v5.05 [49] with 1000 bootstrap replicates. The GTR+I+G model of substitution was selected as the best-fitting model by the AIC test (Akaike Information Criterion) with jModelTest [50] for the ML tree. We also calculated and visualized the relationships among founders in our study and downloaded rpoB gene haplotypes from GenBank by constructing a haplotype network using the medianjoining method available in Network version 10.2.0.0 (http://www.fluxus-engineering.com (accessed on 10 February 2021).

Results
Two species, i.e., L. cervi and L. fortisetosa, were identified among the Lipoptena adults collected from C. elaphus and C. capreolus. The preliminary analyses of the presence of Bartonella sp. in the deer keds involved 15 L. cervi specimens (3 females and 6 males from C. elaphus and 3 females and 3 males from C. capreolus) and 9 L. fortisetosa specimens (3 females and 2 males from C. elaphus and 4 females from C. capreolus). In total, Bartonella sp. were detected in 3/24 (12.5%) of the studied insects. The presence of the bacteria was shown in only 3/9 (33.3%) L. fortisetosa adults (2/7 of the studied females and 1/2 of studied males). No Bartonella sp. were detected in the L. cervi adults. The derived sequences of Bartonella sp. were submitted to the GenBank database under the accession numbers: MZ061868, MZ061869. The sequences obtained in this study share from 96.6 to 98.3% similarity with Bartonella sp. Honshu isolated from sika deer blood in Japan (GenBank accession no. AB703145).
The analysis of a rpoB gene fragment yielded two new haplotypes of Bartonella sp.: haplotype H1 (MZ061868) and haplotype H2 (MZ061869), as defined by three polymorphic sites, all being transitions. The maximum-likelihood phylogenetic reconstructions produced a strong topology (Figure 2). The ML tree revealed that our two rpoB haplotypes belong to lineage B described by Sato et al. [51]. The median-joining network based on sequences from this study and haplotypes representing different species of Bartonella obtained from GenBank (Table 1) suggested the presence of a distinct phylogenetic branch created by the discovered haplotypes inside lineage B. It also showed that they are grouped closely with haplotypes H4 (AB703145) and H7 (AB703147) described for new species of Bartonella obtained from Japanese sika deer in Japan (Figure 3). Table 1. List of species and GenBank accession numbers of their RNA polymerase beta subunit (rpoB) gene sequences used in the network phylogenetic analysis (Figure 3).  Numbers listed at the nodes represent the percent support for the node from 1000 bootstrap replicates. The ML tree has been rooted with sequences of Brucella melitenis, a microorganism closely related with Bartonella sp., as they together belong to the same order, Hyphomicrobiales. The haplotypes of Bartonella sp. found in this study are marked in blue. Lineage B, according to Sato et al. [51].  Table 1). Missing haplotypes are indicated by a grey dot.

Discussion
The zoonotic pathogen Bartonella sp. is a Gram-negative hemotropic bacterium, which is an etiological agent of bartonellosis. The disease usually manifests as an acute or sub-acute febrile illness in humans and animals [59]; however, a long-term symptomless infection with bacteremia in mammalian reservoir hosts (e.g., dogs and cats) was also noted [55,60,61]. The role of this bacterium as a causative agent or cofactor in endocarditis has been reported [62,63]. Lipoptena spp. may serve as a potential vector of this bacterium [34,35,39,40].
The prevalence of Bartonella sp. in Lipoptena is high. For instance, Bartonella DNA was detected in 85% of wingless adults of L. cervi collected from free-ranging cervids in Norway [64], and even in 94% of these deer keds collected from roe deer in France [35]. In Mazury forests (northern part of Poland), Szewczyk et al. showed the prevalence of Bartonella sp. in these insects at the level of 75.12% [44]. The latest data from northern and western Poland indicate the presence of Bartonella sp. in 49.4% of L. fortisetosa adults [23]. In turn, the Bartonella sp. infection rate in L. fortisetosa collected in Japan was estimated at 87.9% by real-time PCR and 51.5% in culture [40].
In this study, this bacterium was not detected in L. cervi. However, the absence of Bartonella sp. in the studied deer keds may be related to the smaller number of analyzed samples. Five sequences of Bartonella sp. obtained by Szewczyk et al. showed 94.4% similarity with Bartonella sp. from Japanese sika deer (GenBank accession no. AB703149) [44]. In turn, the two other sequences showed 99.7% similarity with Bartonella sp. isolated from Japanese sika deer in Wakayama Prefecture, Japan (GenBank accession no. AB703149) and with Bartonella sp. isolated from Japanese sika deer in Nara Prefecture, Japan (GenBank accession no. AB703146). The sequences obtained in this study showed high similarity with Bartonella sp. Honshu isolated from sika deer blood in Japan (GenBank accession no. AB703145) but did not show similarity with the sequences obtained by Szewczyk et al. from L. cervi [44].
In the maximum-likelihood (ML) algorithm based on the rpoB gene sequences, our two haplotypes formed a distinct branch with high bootstrap support within lineage B described by Sato et al. [51]. The ML phylogenetic analyses corroborated the result obtained from the nucleotide network and confirmed that the two haplotypes obtained in this study created a separate branch within the different species of Bartonella. Our newly discovered haplotypes differed by at least nine substitutions from haplotype 4 (GenBank accession no. AB703145, HonshuWD-9.3) and by at least 10 mutations from haplotype 7 (GenBank accession no. AB703147, Bartonella sp. HonshuWD-18.5), both described by Sato et al. [51]. Interestingly, these two GenBank haplotypes of Bartonella were isolated from Japanese sika deer in Japan. As shown by the network analysis, our two haplotypes and haplotypes 4 and 7 created a distinct group together, which additionally supports the view proposed by Sato et al. that lineage B represents a novel Bartonella species [51]. The presence of these two new Bartonella sp. haplotypes in L. fortisetosa and the haplotypes obtained by Szewczyk et al. in L. cervi may suggest that red deer in Poland seem to harbor the novel Bartonella species discovered in Japanese sika deer [44,51]. Moreover, it seems to be necessary to obtain and analyze more sequences of Bartonella directly from red deer blood to resolve the relationships of Bartonella species in deer from Japan and Poland. In turn, the role of this deer ked species as a potential vector of this bacterium needs further study.