New Records of Bartonella Spp. And Rickettsia Spp. In Lice Collected From Small Rodents

Background: Lice are blood-sucking insects that are of medical and veterinary signi�cance as parasites and vectors for various infectious agents. More than half of described blood-sucking lice species are found on rodents. Rodents are important hosts of several Bartonella and Rickettsia species and some of these pathogens are characterised as human pathogens in Europe. Rodent ectoparasites, such as �eas and ticks, are important vectors of Bartonella spp. and Rickettsia spp., but knowledge about the presence of these bacteria in lice is limited. The aim of this study was to determine the prevalence of Bartonella and Rickettsia bacteria in lice collected from rodents in Slovakia. Methods: The ectoparasites were collected from small rodents captured from 2010 to 2015 at four different sites in eastern Slovakia. The presence of Bartonella and Rickettsia pathogens in lice samples was screened by real-time PCR, targeting ssrA and gltA genes respectively. The molecular characterisation of the Bartonella strains was based on sequence analysis of partial rpoB and ITS genes, and of the Rickettsia species on sequence analysis of the gltA gene. Results: A total of 1074 lice of seven species were collected from six rodent species in Slovakia from 2010 to 2015. Bartonella DNA was detected in three species of lice Hoplopleura a�nis (collected from Apodemus agrarius, A. �avicollis and Myodes glareolus), Polyplax serrata (from A. agrarius) and Hoplopleura sp. (from A. �avicollis). Sequence analysis revealed that the Bartonella strains belonged to the B. coopersplainsensis, B. tribocorum and B. taylorii genogroups. Rickettsia DNR was detected in H. a�nis and P. serrata lice collected from A. agrarius. Sequence analysis revealed the presence of two Rickettsia species: R. helvetica and Rickettsia sp. Conclusions: To the best of the authors’ knowledge, this is the �rst report on the occurrence and diversity of Bartonella spp. and Rickettsia spp. in lice collected from small rodents in Europe. This study is also the �rst to detect B. coopersplainsensis in Slovakia.


Background
Small rodents are important hosts of ectoparasites such as eas, ticks, mites and lice and are reservoir hosts or carriers of medically important pathogens [1].
Bartonella spp.and Rickettsia spp.are gram-negative bacteria that can cause severe disease in humans and animals [4,5].Currently, 37 Bartonella species and three subspecies have been identi ed [6].Small rodents represent an important group of potential reservoirs for many Bartonella infections.More than 20 Bartonella species have been detected in different rodent species.At least six Bartonella species found in rodents in Europe have been implicated in human illnesses [2].Bartonella spp.have been reported in rodents and their ectoparasites (ticks, mites, and eas) in Sweden [7], Denmark [8], Poland [9], Spain [10], Lithuania [11,12], Slovakia [13,14] and Germany [15].
Sucking lice (Phthiraptera: Anoplura) are obligate blood-feeding insects and permanent ectoparasites of eutherian mammals.All their life cycle stages are closely related with their vertebrate hosts and they cannot survive without them.More than 540 species of blood-sucking lice have been described that parasitise over 840 mammal species belonging to 12 mammalian orders [26,27].Small rodents are the most common hosts of sucking lice: about 67% of the described sucking lice species are found on rodents [28].The sucking lice are of medical and veterinary signi cance as vectors of louse-borne pathogens (viruses, bacteria, fungi and protozoa) to vertebrate hosts [29].Blood-sucking lice are highly host-speci c and each species of sucking louse parasitises a single host species or only a few closely related host species [26,30].In Europe, reports of pathogens in lice are scarce.Hornok et al. [29] were the rst to report Rickettsia spp. in lice from livestock animals and proved that lice could be potential vectors of arthropod-borne pathogens.However, there is limited information available with regard to Bartonella spp.and Rickettsia in lice collected from rodents in Europe.The role of lice in the life cycles of Rickettsia and Bartonella is still not clear.The aim of this study was to determine the prevalence of Bartonella and Rickettsia species in lice collected from rodents in Slovakia.
At each site, 50 traps were placed 5 m apart in transects (approximately 250 m in length) for two consecutive nights.Captured animals were transported to the laboratory where they were determined to species level and euthanised under licenses from the Ministry of Environment of the Slovak Republic No.
The ectoparasites (ticks, eas, mites and lice) were collected and placed in 70 % ethanol until determination.Lice were then determined by species and sex using light microscopy according to Smetana [33] and Wegner [34].

Molecular analyses
Lice from each rodent host were grouped in pools by species, life stage and sex.A total of 275 sample pools (between one and ten lice per pool): 38 pools of larvae, 151 pools of females and 86 pools of males were analysed.DNA from lice was extracted using 2.5 % ammonium hydroxide solution [35].Bartonella and Rickettsia DNA in samples was detected using a duplex TaqMan real-time PCR targeting 124 bp fragment of ssrA and 103 bp fragment of citrate synthase (gltA) genes respectively.The qPCR ampli cations were carried out in a 15-μl nal volume consisting of 1 μl of extracted DNA, (1x) SensiMix™ II Probe No-ROX Kit (Bioline Reagents Ltd, UK), 1 μM of each primer and 0.5 μM of each probe.The reaction was carried out in a real-time thermocycler Rotor-Gene Q 5plex model with software version 1.7 (Qiagen GmbH, Germany).The optimised thermal cycler programme was 95 °C for 10 minutes (1 cycle), followed by 50 cycles of denaturation at 95 °C for 20 seconds, annealing at 50 °C for 1 minute, and extension at 72 °C for 10 seconds.Results that satis ed the ampli cation cut-offs below 40 Ct (cycle threshold) when the threshold was 0.10101 indicated positive samples.Bartonella-positive samples were tested further in two PCRs using a set of genus-speci c primers targeting the 795 bp fragment of the RNA polymerase β-subunit (rpoB) gene [36] and primers targeting the 16S-23S rRNA gene intergenic species region (ITS) (0.9-1.6 kb) [37,38].A nested PCR that targeted the partial gltA gene (338 bp fragment) [19] was used for ampli cation of Rickettsia spp.The primer sequences and target genes used in this study are presented in Table 1.Negative (dH2O) and positive controls (DNA of Bartonella-infected rodents and the DNA of Rickettsia-infected ticks, con rmed by sequencing) were included in real-time PCR, conventional and nested PCRs runs.Products of ampli cation were identi ed in 1.5 % agarose gel after undergoing electrophoresis under standard conditions and staining with ethidium bromide solution (2 μg/ml), and then visualised using the UV transilluminator (EASY Win32, Herolab, Germany).
Representative positive PCR products were extracted from the agarose gel and puri ed using the GeneJET Gel Extraction Kit (ThermoFisher Scienti c, Lithuania) according to the manufacturer's instructions (Macrogen Europe, Netherlands).The obtained sequences were edited, aligned with one other and compared with the sequence data available from NCBI GenBank, using the Mega X program and the NCBI BLAST® blastn suite applet.The most appropriate model of nucleotide substitution for each alignment dataset was determined according to the Bayesian information criterion (BIC).Phylogenetic trees were constructed using the maximum-likelihood (ML) method with the Tamura-Nei model.Bootstrap support was calculated by means of 1,000 replicates.Bartonella and Rickettsia sequences obtained in this study were deposited in the GenBank database under the accession numbers MT840662 -MT840520 (Bartonella ITS region), MT876371 -MT876377, MT833866 (BartonellarpoB gene) and MT876378 -MT876382 (RickettsiagltA gene).

Statistical analysis
The prevalence of pathogens in lice was calculated as a minimum infection rate (MIR) with 95 % con dence intervals (CI).MIR was calculated as the ratio of the number of positive pools to the total number of lice tested.The underlying MIR assumption was that only one infected individual exists in a positive pool [39].
Bartonella-positive PCR products of good quality were subjected to sequence analysis.A total of 14 good-quality sequences of BartonellarpoB (n=8) gene and ITS region (n=6) were obtained and analysed.The ITS region sequences of Bartonella derived from lice were 100 % identical to each other and 98-100 % identical to Bartonella coopersplainsensis and Bartonella tribocorum sequences deposited in GenBank (Fig. 1).Sequences (samples MT840662, MT840663, MT840664) derived from H. a nis (two pools of females and one pool of larvae collected from two A. agrarius rodents) were 100 % identical to each other, 100 % identical to B. coopersplainsensis sequences detected in A. agrarius from Lithuania (GenBank: MH547343) and 98 % identical to B. coopersplainsensis sequences detected in rats from Italy (GenBank: MK562489) and Australia (GenBank: EU111770) (Fig. 1).Sequences (samples MT840518, MT840519, MT840520) derived from H. a nis (two different pools of females collected from a single A. avicollis and a single M. glareolus) and P. serrata (one pool of males collected from a single A. agrarius) were 100 % identical to each other and with B. tribocorum sequences detected in A. agrarius rodents from Lithuania (GenBank: MH687379) and South Korea (GenBank JN810856) (Fig. 1).

Rickettsia infection in lice
Six lice pools of H. a nis (four pools) and P. serrata (two pools) collected from A. agrarius (n=6) were found to be positive for Rickettsia spp.Rickettsia pathogens were detected in males (2.3 % positive pools out of 86) and females (2.7 % out of 151).A total of ve good-quality sequences of Rickettsia gltA gene were obtained and analysed.Sequence analysis of the partial gltA gene revealed the presence of two Rickettsia species: Rickettsia helvetica (n=4) and unrecognised Rickettsia sp.(n=1).Rickettsia sequences (samples MT876379, MT876380, MT876381 and MT876382) derived from H. a nis (three pools of females) and P. serrata (one pool of males) shared 99 % identity (with one nucleotide difference) and were 100 % identical to the gltA sequence of R. helvetica detected in A. avicollis from Lithuania (GenBank: MF491764) and R. helvetica sequences detected in eas from Slovakia (GenBank: MN276064, MK85717) and in Ixodes ricinus ticks from Slovakia (GenBank: MK85717), Poland (GenBank: EY779822) and Italy (GenBank: MN226407).The Rickettsia sequence (sample MT876378) isolated from H. a nis (one pool of males) was 100 % identical to the closely phylogenetically related sequences deposited in GenBank for R. raoultii (GenBank: MN550895; MH064450; MK875750; MK792599), R. aeschlimann (GenBank: JF803905) and R. heilongjiangensis (GenBank: JX945522).

Discussion
In Europe, lice collected from rodents have never been examined before for the presence of these bacteria.This study is the rst report on the prevalence and diversity of Bartonella and Rickettsia species in lice collected from rodents in Slovakia.Phylogenetic analysis based on the Bartonella rpoB gene and ITS region and the Rickettsia gltA gene revealed the presence of B. tribocorum, B. coopersplainsensis, B. taylorii, R. helvetica and Rickettsia sp. in rodent lice.
In this study, B. taylorii was detected in H. a nis lice collected from A. avicollis.In previous studies, B. taylorii has been con rmed in the small mammals A. agrarius, A. avicollis, M. glareolus, M. arvalis and Talpa europaea in Slovakia [13,14].B. taylorii strains in small mammals and their ectoparasites have also been reported in several studies conducted in Europe, including in Germany [15], England [40], Lithuania [11,12], Slovenia [41], Poland [9] and Spain [10].B. taylorii can infect several sympatric woodland rodents at a given site.A high diversity of B. taylorii strains is frequently found in Apodemus mice and in Myodes and Microtus voles [2].The pathogenic potential of B. taylorii is as yet unknown [11,40].
In this study, the B. tribocorum infection was detected in P. serrata and H. a nis lice collected from A. avicollis, A. agrarius and M. glareolus.This Bartonella species is pathogenic to humans [2].Previous studies have strongly supported the association of B. tribocorum with rats of the genus Rattus.B. tribocorum has been detected in rats and their eas in Thailand [42] and Bartonella strain closely related to B. tribocorum has been detected in louse (adult P. spinulosa) collected from rats in Egypt [43].In the striped eld mouse A. agrarius, B. tribocorum was detected for the rst time in South Korea [44] and closely related strains were later con rmed in A. agrarius from Slovakia [13] and Lithuania [12].
The present study is the rst to detect the B. coopersplainsensis infection in Slovakia in H. a nis lice collected from A. agrarius.Previously, B. coopersplainsensis has been isolated in rats from Australia [45] and New Zealand [46] and in one louse pool (Hoplopleura spp.) collected from rats in Thailand [42].B. coopersplainsensis has also been reported in A. agrarius in Lithuania [12].There is a lack of information on B. coopersplainsensis, therefore the public health impact of this bacteria is unknown [46].The present study is also the rst to demonstrate the presence of R. helvetica and Rickettsia sp. in lice collected from rodents in Slovakia.Two R. helvetica strains were detected in H. a nis and P. serrata lice collected from A. agrarius.R. helvetica are considered to be agents of human rickettsioses [17].In recent studies conducted in Slovakia, R. helvetica has been identi ed in rodents and in eas, mites and ticks collected from rodents [3,19,20,21].R. helvetica has also been reported in rodents and their ectoparasites in other European countries, such as the Netherlands [47], Hungary [22], Germany [18], Poland [17] and Lithuania [24,25].
Based only on sequence analysis of the gltA gene, the Rickettsia sp.detected in this study in H. a nis lice pool collected from A. agrarius was not identi ed to species level.The obtained gltA sequence showed 100% identity with the corresponding sequences of R. aeschlimann, R. heilongjiangensis and R. raoultii in the GenBank database.
The presence of Bartonella spp.and Rickettsia spp. in lice may result from the acquisition of these bacteria via blood meals from infected rodents.Bartonella spp.are transmitted via horizontal transmission: arthropod vectors become infected with Bartonella bacteria while feeding on infected hosts, including rodents, and can then transfer the bacteria to another host [42].Worldwide, the prevalence of Bartonella spp. in rodents ranges from 25 to 80%, which suggests a reciprocal adaptation between the bacteria and their reservoirs [1].As a result of their blood-feeding habits, lice could transfer disease agents between closely-related host species [26,29] and physical contact between individual rodents may promote the transmission of different Bartonella species [1].Some SFG rickettsiae are thought to circulate in enzootic or epizootic cycles between wild vertebrates and arthropod vectors.The high prevalence of R. helvetica previously obtained in small rodents suggests that they may play an important role as potential natural reservoir hosts for this pathogen [18,25].
The rodents from which the lice were collected have previously been tested for the presence of Bartonella and Rickettsia pathogens [3,13,14].However, almost all the Bartonella-infected and Rickettsia-infected lice were derived from non-infected rodent hosts (except for two specimens of A. agrarius; data not shown).In this case, lice could become infected by parasitising on other infected hosts.Examined small rodents infested with lice also harbour other ectoparasites species such as mites, eas and I. ricinus ticks [3,20].The presence of Bartonella pathogens in lice may also result from acquisition pathogens by co-feeding with Bartonella-infected eas.Fleas are the main vectors for the maintenance and transmission of B. grahamii, B. taylorii and B. rochalimae among populations of small mammals [2].B. tribocorum has been detected in eas and B. coopersplainsensis in ticks and lice [42].Lice that infest rodents could acquire Rickettsia pathogens by co-feeding with infected I. ricinus ticks, eas and mites.Horizontal transmission through a shared blood meal has been demonstrated for some rickettsial pathogens [48].In a previous study conducted in Slovakia, Rickettsia spp. was detected in four species of mites, I. ricinus ticks and four ea species, with an overall prevalence of 9.3%, 17.2% and 3.5% respectively [3].R. helvetica has been identi ed in eas, ticks and mites [3,21].
Although, the results of the present study con rm the circulation of Bartonella spp.and Rickettsia spp. in lice, the role of lice in the transmission of Bartonella and Rickettsia species remains unknown.Thus, future studies should be performed to determine the speci c roles of different species of lice parasitising small rodents in the transmission of Bartonella spp.and Rickettsia spp.bacteria in order to estimate the potential risks for other mammals (e.g.cats) and humans.

Conclusions
To the best of the authors' knowledge, this is the rst report on the occurrence and diversity of Bartonella spp.and Rickettsia spp. in lice collected from small rodents in Europe.The data presented in this paper add to knowledge about the distribution of Bartonella spp.and Rickettsia spp. in rodent ectoparasites, and demonstrate the presence of Bartonella pathogens in three species of lice -H. a nis, P. serrata and Hoplopleura sp.-and of Rickettsia pathogens in two lice species -H. a nis and P. serrata.This study is also the rst to detect B. coopersplainsensis in Slovakia.
Tables Table 1.Primers and probes used for real-time PCR, conventional and nested-PCRs.
Maximum-likelihood phylogenetic tree for the partial gltA gene of Rickettsia spp.The phylogenetic tree was created using the Tamura-Nei model and bootstrap analysis of 1000 replicates.Samples sequenced in the present study are marked.Abbreviations: A. agr -Apodemus agrarius, A. a -Apodemus avicollis, F -female, M -male.
b External primers, c Internal primers.

Figures Figure 1
Figures

Figure 2 Maximum
Figure 2 Maximum-likelihood phylogenetic tree for the partial rpoB gene of Bartonella spp.The phylogenetic tree was created using the Tamura-Nei model and bootstrap analysis of 1000 replicates.Samples sequenced in the present study are marked.Abbreviations: A. agr -Apodemus agrarius, M. agr -Microtus agrestis, A. a -A.avicollis, F -female, M -male, L -larva.

Table 2
Presence of Bartonella spp.and Rickettsia spp. in lice collected from different species of small rodents.