Emergence of Babesia canis in southern England

The United Kingdom is considered free of autochthonous transmission of canine babesiosis although cases are reported in dogs associated with recent travel abroad. During the winter months of 2015/16, a cluster of cases of disease in dogs with signs suggestive of canine babesiosis were reported in Harlow, Essex. Babesia species were detected in dog blood samples by Giemsa staining of blood smears and by pan-piroplasm PCRs. Babesia species were also detected in extracts of tick DNA using pan-piroplasm PCRs. DNA sequencing and phylogenetic analysis was used to confirm the species of Babesia present in dog blood and tick samples. Tick species were identified by PCR-sequencing based on amplification of the cytochrome c oxidase subunit one (cox1) gene. Dermacentor reticulatus ticks were sampled from field sites in England and Wales. Blood smear analysis on samples taken from some of the affected dogs confirmed the presence of a large Babesia species within erythrocytes. A tick recovered from one of these cases was identified as Dermacentor reticulatus, a species with a limited distribution in England and Wales, but a known vector of canine babesiosis in continental Europe. Babesia canis was subsequently identified in blood samples obtained from three clinical cases (all dogs) within the area and from ticks associated with these dogs. A field survey detected 17 adult D. reticulatus ticks from one area visited by the affected dogs. Fourteen of these ticks were shown to be positive for the B. canis parasite, implicating them as a potential source for babesiosis in Harlow. In order to assess whether the parasite is present in more than one tick population, D. reticulatus ticks from across England and Wales were screened for the presence of Babesia species. In addition to the Harlow site, a further five locations where D. reticulatus is present were screened for Babesia species. Babesia was not detected from most sites tested but one tick from a single location in Wales was positive for B. canis. Infection with B. canis was confirmed in a number of dogs in Harlow, Essex, with no history of travel outside of the country. The same pathogen was identified in field-caught D. reticulatus ticks in the same area and is considered the likely source of infection. This highlights the need for vigilance by veterinary surgeons for future outbreaks of tick-borne disease in dogs.


Background
Babesiosis is one of the major tick-borne infections of dogs around the world [1]. It is caused by haemoprotozoan parasites, or piroplasms, belonging to the genus Babesia that infect erythrocytes in the vertebrate host. In extreme cases, the dog exhibits fever, jaundice, and anaemia that if left untreated can lead to death. However, the disease is treatable if recognised [1], although there are no treatments currently licensed for use in dogs in the UK. The underlying disease mechanisms are poorly understood and can present as a range of disease signs that if not treated develop into a number of immune-mediated syndromes [2]. One potential cause for variation in disease severity is the species of Babesia. In Europe, four species of piroplasm have been reported to cause clinical disease in dogs including B. canis, B. vogeli, B. gibsoni and B. microti-like [3]. In addition, "Theileria annae" has been reported to cause severe regenerative anaemia, renal failure and thrombocytopenia in dogs [4]. This species has tentatively been renamed "Babesia vulpes sp. nov." [5], although this is disputed. The most commonly encountered species are B. canis, associated with transmission by Dermacentor reticulatus [6] and B. vogeli associated with transmission by the brown dog tick, Rhipicephalus sanguineus [7]. Although other piroplasms including B. caballi [8] and Theileria equi [9] have been detected in the blood of dogs in Europe using molecular methods; their significance in the epidemiology of disease is unknown at this time.
Canine babesiosis caused by B. canis is highly prevalent in countries of southern and central Europe including France [10], Portugal [11] and Hungary [12]. The disease is rarely encountered in the UK and is usually associated with dogs returning to the UK after travel abroad to countries where the disease is endemic [13][14][15]. However, occasional cases have been reported where autochthonous transmission has been suspected [16]. The absence of canine babesiosis was believed in part due to the absence of key tick vectors such as R. sanguineus and the restricted distribution of D. reticulatus, being limited to coastal locations in Devon, West Wales and Essex [17,18] or to poor reporting of clinical cases. In recent years, there has been a significant increase in dogs travelling with owners from the UK to continental Europe as a result of the introduction and modification of the Pet travel scheme (PeTS) [19,20], enabling dogs vaccinated against rabies and treated for tapeworm, to enter the UK without the need for a period of quarantine. This meant the removal for the legal requirement to treat dogs for ticks prior to entering the UK. As a consequence there are concerns that travelling dogs could return from continental Europe infested with ticks and/or infected with the diseases ticks carry [21,22].
In February 2016, a cluster of cases of canine babesiosis were reported from the town of Harlow in Essex [23]. A number of dogs, examined at the same veterinary surgery, had no history of foreign travel. Subsequent testing of blood samples from dogs, ticks attached to the dogs and ticks collected from an area within Harlow where the dogs had been exercised, tested positive for B. canis indicating a potential source of the disease in dogs [24,25]. The aims of this study are to report the testing conducted on the original outbreak of canine babesiosis in Harlow and a subsequent investigation of known D. reticulatus populations in England and Wales to assess the presence and distribution of B. canis infected D. reticulatus ticks. Sequence data obtained from this study suggested a single introduction of disease in this case rather than multiple introductions. All D. reticulatus specimens were collected by dragging a 1 × 1 m cloth over vegetation at the various locations across England and Wales. All ticks found attached to the cloth were collected, counted and stored in plastic vials during transit, frozen and stored at -80°C. All ticks were identified morphologically using published keys [26,27] and the species confirmed by sequence derived from the cox1 gene (see below).

Detection of Babesia in thin blood smears
Samples of EDTA treated blood taken from the cephalic vein of affected dogs were used to prepare thin blood smears which were fixed in methanol, stained in 10% Gurrs improved R66 Giemsa [28] and examined under light microscopy at 1,000× magnification using oil immersion. A sample was considered positive if intraerythrocytic, single trophozoites and/or paired piriform merozoites of a large Babesia sp. were detected.

Detection of piroplasms by polymerase chain reaction
For the dog blood samples and the D. reticulatus specimens from Essex, Devon and Wales, DNA extraction was carried out from EDTA-treated blood or manually homogenised tick leg, removed with haemolymph, using the DNeasy Blood and Tissue Kit (Qiagen, Manchester, UK), following the manufacturer's instructions. DNA extractions were stored at 4°C until tested by polymerase chain reaction (PCR). DNA extractions were performed from individual ticks. Samples were treated individually at each stage to reduce the risk of PCR contamination.
DNA samples were initially screened by real-time PCR using primers Bbmit2 and Bbmit3 ( Table 1) that amplify an 150 base pair (bp) fragment of the cytochrome c oxidase subunit 1 gene (cox1), present within the mitochondrial genome. Subsequently, piroplasm genome was detected by real-time PCR using primers PIRO-A and PIRO-B ( Table 1) that amplify a 423 bp piroplasm sequence [29]. The final PCR reaction mix for both PCRs (final volume 40 μl) consisted of: 13 μl H 2 O, 20 μl SYBR®Green JumpStart TM Taq Ready Mix TM (Sigma-Aldrich, Poole, UK), 1 μl of each primer (at 20 μM for primers PIRO-A and PIRO-B; 10 μM for primers Bbmit2 and Bbmit3) and 5 μl of the same DNA extract used above. Amplification was performed as follows: an initial denaturation step of 94°C for 10 s followed by 45 cycles of 94°C for 30 s, 58°C for 30 s (when using the primers PIRO-A and PIRO-B; or 48°C for 30 s when using primers Bbmit2 and Bbmit3), and 72°C for 1 min. Samples were then sequenced using flanking primers at 1 pmol/μl using the ABI PRISM® BigDye® Terminator v3.1 Cycle Sequencing Kit (Applied Biosystems, Warringtom, UK), following the manufacturer's protocols.
For the D. reticulatus specimens from Devon, ticks were pooled prior to RNA extraction (5 ticks/pool), which were screened by real-time PCR using primer PIRO-A and PIRO-B. The final PCR reaction mix (final volume 25 μl) consisted of: 6.25 μl H 2 O, 12.5 μl SYBR®Green Jump-Start TM Taq Ready Mix TM (Sigma-Aldrich), 2 μl of each primer (at 20 μM primers PIRO-A and PIRO-B), 0.25 μl QuantiTect RT Enzyme (Qiagen) and 2 μl of RNA. Amplification was performed as described above.

Sequencing and sequence analysis
Piroplasm and tick amplicons generated by PCR were separated by electrophoresis in a 1% agarose gel impregnated with SYBR safe and visualised by UV illumination. Automated DNA sequencing using flanking primers was achieved as previously described [31]. Individual forward and reverse traces were oriented, edited, and aligned using the Sequencer (v.4.5; Genes Codes Corporation, Ann Harbour, MI, USA), GenDoc (v. 2.6.02) and Clus-talX sequence analysis programs. Sequences were edited in Lasergene version 12.1 (DNASTAR) and assigned to a particular species, by BLAST (NCBI) search, when agreement was ≥ 98% to sequences of known taxa in GenBank (for Babesia spp.). In the case of tick cox1 barcode sequences, BLAST searches were performed at NCBI and BOLD databases for molecular species validation.
Once identification was achieved for Babesia spp., sequences were further analysed in MEGAv.6 [32]. Neighborjoining (NJ) trees were exported as JPG files in Adobe Acrobat 8 Professional, and edited using Adobe Photoshop CS3 (v.10.0.1). The K2P distance metric was used to determine the intra-and inter-specific genetic distances, and a NJ tree analysis was carried out to represent the specimen's clustering pattern. Bootstrap values were calculated to test the robustness of the cluster, which was obtained by conducting 1,000 pseudoreplicates.

Results
A summary of the clinical presentation of the dogs associated with the outbreak is provided in Table 2. The first dog presented with non-specific signs including lethargy  on the 5th of October 2015 but babesiosis was not suspected at the time and only diagnosed retrospectively on a second visit the following year based on a positive PCR test. A second case presenting on the 26th of October 2015 raised suspicions of babesiosis, particularly as the dog had no history of travel outside the country, and a commercial test confirmed the presence of Babesia in a blood sample. The remaining dogs presented in the first three months of 2016. All of the dogs involved were resident in Harlow (Fig. 1a).
The first blood sample and an associated tick were received at APHA on the 8th February, 2016, and belonged to a dog which was found to be positive for Babesia by blood smear analysis by the attending veterinary surgeon and which was successfully treated for babesiosis. Both the tick removed from this dog and the blood sample, were tested for piroplasms using PCR and sequencing, which confirmed the presence of Babesia canis in the tick but not in the dog blood. Subsequent blood samples from additional dogs with disease suggestive of canine babesiosis were submitted for testing from Harlow. Of these, 3 of 5 blood samples were positive for B. canis either by blood-smear analysis (Fig. 1b) and/or the PCR, and the parasite was associated with either the blood-meal or the tick removed from four of the dogs affected. One engorged female tick was submitted live and maintained in a humid chamber for three weeks until oviposition. A leg taken from this tick and eggs derived from it were positive for B. canis.
The detection of a cluster of cases of canine babesiosis and associated B. canis-infected D. reticulatus ticks prompted a field survey of a site that most of the dogs had visited in the Harlow area. This survey collected 17 ticks (eight females, nine males; Fig. 2a) from one area close to a car park located next to Third Avenue and Abercrombie Way (51.762683, 0.088548). The area surveyed is a common dog-walking route, with dogs likely accessing this area as soon as they leave the car park with their owners as previously described by Hansford et al. [33]. The collected ticks were all identified as adult D. reticulatus based on morphology. Neighbor-joining (NJ) analysis using the cox1 gene sequence derived from each tick confirmed the morphological identification of the dog-associated and field collected ticks as D. reticulatus showing 100% identity to this species from continental Europe (data not shown). Pan-piroplasm PCR detected fourteen ticks (six females and eight males) positive for the presence of a piroplasm using assays that amplify either 18S sequence (data not shown) or a sequence of the mitochondrial genome (Fig. 2b). Sequencing of the amplicons derived from these samples established the presence of B. canis in these ticks. NJ analysis confirmed that the sequences derived from infected dogs, associated ticks and field-collected ticks were identical (Fig. 3), sharing 100% sequence identity with B. canis from Lithuania (GenBank KM111283) and France (GenBank KC902833). A representative sequence for this outbreak has been submitted to GenBank with accession number KY694436. To further investigate the distribution of B. canis infection in UK ticks, a retrospective study of D. reticulatus collected between 2010 and 2016 were tested. DNA was extracted from ticks originating from southern England and Wales (Table 3). With the exception of the ticks sampled from Harlow, the majority of ticks tested (n = 95) were negative for piroplasm DNA (Table 3). However, one male D. reticulatus was found positive for the presence of piroplasm DNA from the population sampled from Harlech, Wales. Further sequencing analysis confirmed it was 100% identical to Babesia canis sequence available on GenBank (accession number KC207822) and to that found in samples from Harlow.

Discussion
This study describes the first outbreak of B. canis in the UK with evidence for autochthonous transmission by indigenous ticks. There was no history of foreign travel by any of the dogs affected and ticks surveyed at a site visited by the dogs were positive for B. canis. A further case of B. canis infection has been reported from Essex [34]. This dog presented in November of 2015 with no history of travel outside of the county. The presence of the parasite in blood samples, ticks associated with the affected dogs and field collected ticks from a site known to have been visited by the infected dogs all point to Harlow, Essex as being the source of infection. The species of tick detected in the area, D. reticulatus, is a known vector of B. canis and has been associated with transmission of B. canis in northern Europe in Germany [35], the Netherlands [36] and Belgium [37]. In recent years, cases of canine babesiosis in the UK have been associated with travel to endemic countries, although the species of infecting Babesia is rarely reported. A single case of autochthonous transmission was suspected in a dog with no travel history [16] but no further investigations were conducted to support or refute this claim. Sequence analysis also suggested this case was more closely related to B. vogeli, than B. canis.
Dermacentor reticulatus is indigenous to the UK, although populations of this tick species are rare and prior to this study historical records suggested that this species was localised to a number of coastal sites [18]. A retrospective survey of D. reticulatus samples from known populations were tested for Babesia infection. Most populations tested were negative, but one location in Harlech, Wales, yielded a single positive specimen. No cases of canine babesiosis have been reported from this area [38] so the significance of this observation is unclear, and further surveys and testing would be required to assess the prevalence of infection in this tick population. Other pathogens have been detected in D. reticulatus from Wales [39].
The response to this outbreak illustrates a number of features. All of the cases detected were treated for babesiosis and recovered from infection. Information was disseminated rapidly through both the local, national and veterinary media outlets to raise awareness amongst dog owners and veterinary surgeons. Acaricide treatment was recommended as the current best practice guidance to prevent transmission while the importance of this for dogs returning to the UK from countries where canine babesiosis is indigenous was emphasised. Tick repellents could also be used to prevent tick infestation. Local measures to restrict access for dogs to the infested site were implemented by the local council and based on an understanding of vector behaviour, modification of the vegetation of the area was considered to reduce tick activity [40].
The source of infected ticks and canine babesiosis cases in Essex is presumed to be continental Europe. Although the mechanism for introduction remains speculative, it could have been introduced by an asymptomatically infected dog, as a recent arrival from abroad, providing a source of transmission to a tick (s) from a previously undetected population of D. reticulatus in the locality of the outbreak. Alternatively, a returning dog could have been infested with an infected tick (s) that dropped off at the site, causing the current infestation and subsequent outbreak. Either event is likely to have occurred in 2015 or earlier to produce an adult tick population that was Babesia-infected and sampled in 2016 [25]. Recent studies have shown that new populations of D. reticulatus are establishing in The Netherlands and Belgium, and that areas affected by this tick are increasing [41]. Alternative pathways for introduction include importation on livestock, horses or birds. Reports of autochthonous transmission of B. canis in previously disease-free areas have been reported from Norway [42] and Latvia [43].

Conclusions
Currently, there is no evidence to confirm the route of introduction of B. canis into the UK in this case, but the underlying cause is likely to have been the movement of an infected or infested dog into the country. In addition to reducing the risk of further outbreaks of canine babesiosis, tick treatments that are effective and applied correctly, can considerably reduce the risk of introducing other non-endemic species. This could provide additional protection from the introduction of pathogens such as B. caballi, Rickettsia conorii and tick-borne encephalitis virus. Currently, acaricide treatment is recommended. The responsibility for tick control is entirely with pet owners. With over 100,000 dogs and cats entering the UK annually [20], the risk of further outbreaks of canine babesiosis caused by B. canis and the introduction of non-endemic ticks is heightened if owners do not treat their pets with an acaricide.