The role of sheep ked (Melophagus ovinus) as potential vector of protozoa and bacterial pathogens

The sheep ked (Melophagus ovinus) hematophagous insect may act as a potential vector of vector-borne pathogens. The aim of this study was to detect the presence of Trypanosoma spp., Bartonella spp., Anaplasma phagocytophilum and Borrelia burgdorferi sensu lato in sheep ked collected from sheep in Poland. In total, Trypanosoma spp. was detected in 58.91% of M. ovinus, whereas Bartonella spp. and B. burgdorferi s.l. were found in 86.82% and 1.55% of the studied insects, respectively. A. phagocytophilum was not detected in the studied material. In turn, co-infection by Trypanosoma spp. and Bartonella spp. was detected in 50.39%, while co-infection with Trypanosoma spp. and Bartonella spp. and B. burgdorferi s.l. was found in 1.55% of the studied insects. The conducted study showed for the first time the presence of B. burgdorferi s. l. in M. ovinus, as well as for the first time in Poland the presence of Trypanosoma spp. and Bartonella spp. The obtained results suggest that these insects may be a potential vector for these pathogens, but further-more detailed studies are required.

The family Hippoboscidae (Diptera) is a group of blood-sucking flies of veterinary importance that parasitize on mammals and birds. Worldwide, the fauna of Hippoboscinae consists of more than 213 species, 21 genera and three subfamilies: (Ornithomyinae, Hippoboscinae and Lipopteinae) 1,2 . In Europe, about 30 species of hippoboscid have been described, of which 10 species are found in Poland 2,3 . The sheep ked (Melophagus ovinus) is a blood-sucking wingless ectoparasite of sheep. Their life-cycle consists of three stages: larva, pupa and adult, and occurs in the fleece of the sheep host and can be carried from one sheep to another by direct contact 4 . Although the sheep ked parasitizes mainly sheep, their incidental occurrence has also been reported in red fox, rabbit and the European bison 4,5 . This species of flies occurs in North America, Oceania, Asia, China, Africa and Europe 4,6 . The sheep keds have an economic impact by reducing the production of meat, milk and the wool of sheep, and the infestation of sheep also has harmful effects, such as: weight loss, anemia, anxiety and reduction in wool growth. In addition, skin damage due to abrasion and scratching can lead to secondary microbiological infections 4 . The sheep ked has been reported to be responsible for the transmission of zoonotic pathogens, such as: Bartonella spp., Anaplasma spp., bluetongue virus, border disease virus (BDV), Rickettsia spp., Trypanosoma spp. [6][7][8][9][10][11][12][13] . The occurrence of arthropod-borne pathogens in M. ovinus has not yet been studied in Poland.
Trypanosomes (genus Trypanosoma, family Trypanosomatidae, order Kinetoplastea), are flagellated protozoa with a worldwide distribution and have been isolated from the gut of sheep ked and the blood of sheep 14 . Due to the developmental mode, it belong to the Stercorarian group. The infection of trypanosomes in the mammalian host takes place through damaged skin or mucous membranes when the trypanosomes leave the insect organism with the faeces 14 . Trypanosomiasis in animals are usually subclinical, although anaemia, leucocytosis, weight loss, neonate death and decreased milk production have been noted 15 .
Anaplasma phagocytophilum causes granulocytic anaplasmosis in humans (HGE) as well tick-borne fever in ruminants, equine anaplasmosis in horses, and severe febrile diseases in dogs and cats 16 . It is Gram-negative intracellular bacteria of the family Anaplasmataceae that grow and multiply in the membrane-bound vacuoles of vertebrate and invertebrate host cells 17  www.nature.com/scientificreports/ appetite, and weak or painful limbs. The cells most commonly infected are neutrophilic granulocytes. Humans can be the host, and the animal reservoir for A. phagocytophilum depends on the geographic region and includes many animal species, including wild animals such as rodents, carnivores, deer and domestic animals, such as cattle, goats, and sheep [18][19][20] . Molecular studies based on the 16S rRNA gene, two different genetic variants of A. phagocytophilum have been described. The AP-ha strain is pathogenic to both humans and animals, while the AP -variant 1 strain is infectious to animals but not to humans 21,22 . In Europe, zoonotic reservoirs of human A. phagocytophilum strains are wild boar, the hedgehog, and possibly carnivores 23,24 . The Ixodes ricinus tick is a primary vector of this bacteria, but their presence has been also confirmed in deer keds (Lipoptena cervi) nor in some species of blood-sucking flies from the Tabanidae family 25,26 . Bartonella spp. are small, intracellular, hemotropic Gram-negative bacteria that cause long-lasting infections in their mammalian hosts and are mainly transmitted by arthropod vectors 27 . In the last 20 years, the number of Bartonella species descriped has increased rapidly, with at least 26 species now designated and with some species containing more than one subspecies 28 . The clinical manifestations of human bartonellosis depend on the species of the infecting Bartonella, and most often manifest as various cardiovascular, neurological and rheumatological conditions 28 . Tsai and co-authors 29 showed that in recent decades a variety of insect vectors and mammal hosts have been associated with Bartonella sp. infections. Bartonella species were identified mainly based on PCR amplification in a wide range of hematophagous arthropods. These bacterial pathogens were detected in human lice (Pediculus humanus), cat fleas (Ctenocephalides felis), sand flies (Lutzomyia verrucarum), and various hard tick species, such as Rhipicephalus sanguineus and varoius species from genera Ixodes spp., Dermacentor spp., Haemaphysalis spp. 30 .
Lyme disease, a mul-systemic, chronic, and often clinically diverse human disease is caused by the spirochetes of Borrelia burgdorferi sensu lato (Spirochaetia, order Spirochaetales, family Spirochaetaceae). The B. burgdorferi s.l. complex consist of 22 genospecies, of which 11 are circulating in Europe and five of them, namely B. afzelii, B. garini, B. burgdorferii sensu stricto., B. spielmanii, B. bavariensis, are pathogenic to humans and associated with human Lyme disease 31,32 . Blood-sucking flies, such as the deer fly, horse fly and flea, have been found to be infected with B. burgdorferi s.l. in Europe and North America 33,34 . These pathogens have been described recently in sheep ked in China 9 .
In the current study, a molecular screening was conducted of sheep keds for the presence of Trypanosoma spp. Anaplasma spp. Bartonella spp. and Borrelia burgdorferi s.l. pathogens in sheep ked collected from a sheep farm near the Białowieża Primeval Forest in north-eastern Poland. To the best of the authors' knowledge, this is the first report to confirm the presence of these pathogens in sheep ked in Poland. Evidence is provided that two (Trypanosoma spp. and Bartonella spp.) and three pathogens (Trypanosoma spp. and Bartonella spp. and B. burgdorferi s.l.) can co-infect sheep ked (Melophagus ovinus).

Results
In total, 129 M. ovinus (69 ♀; 60 ♂) were collected from sheep ( Fig. 1)The presence Trypanosoma spp. was detected in 76 of the 129 (58.91%) studied insects. In turn, Bartonella spp. was found in 122 of the 129 (86.82%) flies. The presence of B. burgdorferi s.l. was detected in 2 of the 129 (1.55%) sheep keds, whereas no A. phagocytophilum was detected in the tested group of insects. The positivity to protozoan and bacterial pathogens DNA among female and male M. ovinus from different locations is given in Table 1. Co-infection with the two pathogens Trypanosoma spp. and Bartonella spp was detected in 65 of the 129 (50.39%) M. ovinus, whereas co-infection with Trypanosoma spp. and Bartonella spp. and B. burgdorferi s.l was detected in only 2 of the 129 (1.55%) sheep keds (Fig. 2). In the current study, A. phagocytophilum was not detected in the tested group of flies.
Sequence analysis showed that four partial rpoB gene sequences of Bartonella spp. (Gen Bank Accession No., MW929188, MW929189, MW929190, MW92919) were identical to each other and showed 100% identity with Bartonella melophagi from sheep ked from the USA (EF605288). In total, eight partial 18S rDNA sequences of Trypanosoma spp. were obtained in the current study, three of which (Acc. No., MZ014571, MZ014572, MZ014565) shared 100% similarity with Trypanosoma melophagium from M. ovinus in the United Kingdom and Croatia (FN666409, HQ664912). While the other five sequences (Acc. No., MZ014566, MZ014567, MZ014568, MZ014569, MZ014570) showed 100% identity with Trypanosoma sp. from Lipoptena fortisetosa in Poland (MT393991). Analysis of two partial fla B gene nucleotide sequences of B. burgdorferi s.l. (GenBank Acc. No., MW929186, MW929187) obtained in this study showed that they were identical to each other, and shared 100% similarity with B. burgdorferi strains from Ixodes ricinus ticks in Poland (MK604273, MK604272, MF150052, KX64620, KF422802, HM345911), Switzerland (KF422803) and Serbia (AB189460) and with B. burgdorferi isolates obtained from the human serum of a patient with Lyme in the Czech Republic (FJ231333-FJ231335). This result confirm the presence of strain typical for Middle Europe. Moreover, they were also identical with B. burgdorferi from cotton mouse (Peromyscus gossypinus) from the USA (EU220782).
Statistical comparisons (Table 2) revealed that there were three significant associations between variables, i.e., for prevalence of Trypanosoma spp. on M. ovinus and sampling location (X 2 = 51.8768, p ≤ 0.01) and for prevalence of any of examined pathogens on M. ovinus and sampling location (X 2 = 15.6845, p ≤ 0.01) as well as for prevalence of Bartonella spp. on M. ovinus and subject gender (X 2 = 7.064, p ≤ 0.01). All of the other associations tested were not statistically significant.

Discussion
The   13 .
The results obtained in the current study show for the first time in Europe evidence of B. burgdorferi s.l. in M. ovinus collected from sheep. However, this bacteria was evident in only 1.5% of the studied insects. The study conducted in China by Chu et al. 9 showed a twice as high percentage in M. ovinus infected with B. garinii and B. valaisiana (related group of B. burgdorferi s.l).
In the presented study, DNA of A. phagocytophilum was not detected in the tested group of flies, while Hornok et al. 8 detected the presence DNA of Anaplasma ovis in all (81 specimens) of sheep ked collected from sheep in Hungary. In Algieria, 25.88% of the M. ovinus were infected by bacteria from the Anaplasmataceae family 13 . In a study conducted by Zhang et al. 36 the infection rates were 39.1%, 17.4%, and 9.8% for A. ovis, A. bovis, and A. phagocytophilum in M. ovinus, respectively.
The mixed infection of sheep keds with two bacterial genus, Bartonella and Anaplasma, were noted, with frequency co-infection at 18.64% 13 . The obtained results also demonstrated that two pathogens, Trypanosoma spp. and Bartonella spp., can co-infect the same sheep ked.  Table 2. Statistics of the chi-square tests performed to examine associations between prevalences of (i) Trypanosoma spp., (ii) Bartonella spp., (iii) Trypanosoma spp., Bartonella spp., Borrelia burgdorferi s.l. or Anaplasma phagocytophilum on Melophagus ovinus and (i) sampling location, (ii) sex of M. ovinus. *The result is significant at p ≤ 0.01.

Melophagus ovinus infected/non-infected by
Sampling location (I-Waliły, II-Gródek, III-Rzepiska, IV-Nowoberezowo) Trypanosoma spp. The presence of Trypanosoma spp., Bartonella spp., and B. burgdorferi s.l. pathogens in M. ovinus collected from sheep may be related to their biology. The sheep ked spends its entire life cycle on the host; however, adults may circulate among the animals of the same herd, and are commonly transferred from ewes to their offspring 4 . Hippoboscids are also likely to be mechanical vectors of infectious agents due to their blood feeding behavior 37 . Some pathogens may be transferred from an infected host to non-infected individuals through their mouthparts 38 , although the keds were probably infected through receiving the pathogens from sheep. However, it is interesting that keds and sheep are generally thought to be reservoirs incompetent for B. burgdorferi s.l. 9,39 . In the current study, the keds probably acquired the B. burgdorferi s.l infections via co-feeding transmission when one infected Ixodes spp. tick infected a ked feeding nearby. At this stage of the study, it is difficult to explain this speculation. This is the first report in Europe on the detection of B. burgdorferi s.l. DNA in sheep keds.
Recently, vertical transmission of A. ovis in sheep keds has been described 13,40 . The presence of Bartonella DNA in all of the M. ovinus samples, even at the pupal stage, has been described 30 . This provides new evidence for the potential of these flies as mechanical or biological vectors 13 .
Due to the results of statistically significant comparisons (chi-square tests) between prevalence of Trypanosoma spp. and Bartonella spp. on M. ovinus versus sampling location and sex of the flie further research on this subject should be carried out.
In conclusion, this study provides the first molecular evidence of the presence of DNA of Trypanosoma spp, Bartonella spp., B. burgdoreferi s.l. in a tested group of sheep ked. Detection of the pathogen in M. ovinus is important evidence that other blood-sucking flies, including hippoboscid flies, are important candidates as potential vectors of infectious diseases.
The DNA was isolated from single individuals by using a Genomic Mini AX Tissue kit (A&A Biotechnology, Gdynia, Poland), according to the manufacturer's recommended protocol. Concentration was measured spectrophotometrically at 260/280 nm wave length. Isolated material was stored at − 20 °C until further molecular analysis. Pathogens in M. ovinus were detected by PCR and nested PCR methos.
The detection of Trypanosoma spp. was based on PCR amplification of a 18S rDNA gene fragment of approximately 650 bp. Two oligonucleotide primers, TrypF 150 and TrypR 800, were used, as previously described 41 . For amplification, a 200 ng DNA template was used, and for the reactions Allegro Taq DNA Polymerase (Novazym, Poznań, Poland), was used, and to detect Bartonella spp., a pair of primers-1400F and 2300R-were used to amplify an 850 bp fragment of the rpoB gene 42 . PCR reactions were conducted according to Paziewska et al. 43 . For the reaction mixture, RUN Taq polymerase (A&A Biotechnology, Gdynia, Poland) was used.
In turn, B. burgdorferi s.l., was detected in insects with the use of the two pairs of primers specific to the fla B gene fragment, as previously described 44 . For amplification, a 200 ng DNA template was used. In turn, for the re-amplification, 1 µl of the amplification product was used. DFS-Plus DNA Taq Polymerase (GeneOn, Germany) was used for both reactions. The presence of 824 bp and 605 bp reaction products were considered positive.
Anaplasma phagocytophilum was detected in M. ovinus with the use of two pairs of primers specific to the 16S rDNA gene fragment, as previously described 45 . A 200 ng DNA template and 1 µl of amplification product were used for amplification and re-amplification, respectively. For both reactions, Taq DNA Polymerase (EURx, Poland) was used. The presence of 932 bp and 546 bp reaction products were considered positive.
PCR and nested PCR products were visualized on 1% and 2% ethidium bromide stained agarose gels. Next, gels were visualized using ChemiDoc, MP Lab software (Imagine, BioRad, Hercules, USA) or Omega 10 (UltraLum, USA) and TotalLab software (TotalLab, UK). The positive products of PCR and nested PCR were purified using the QIAEX II Gel extraction kit (Qiagen, Hilden, Germany) or Agarose-Out DNA Purification Kit (EURx, Poland), and sequenced by Genomed (Warsaw, Poland). Next, the sequences were assembled into contigs using ContigExpress, Vector NTI Advance 11.0 (Invitrogen Life Technologies, New York, USA). The sequences were then aligned with reference sequences available in GenBank by BLAST (BasicLocal Alignment Search Tool) and analyzed using MEGA 5.0 software. X 2 statistical analyses were performed to examine the relation between the prevalence data and: (i) sampling locations as well as (ii) sex of M. ovinus, and the 5% level of probability was set for rejection of the null hypothesis.
Ethics approval and consent to participate. Not applicable.