Vector-Borne Pathogens in Ticks and Fleas of Client-Owned Dogs in Metro Manila, Philippines

Rhipicephalus sanguineus s.l. and Ctenocephalides felis are considered the most prevalent ectoparasites of dogs in the Philippines. Vector-borne pathogens (VBPs) in these ectoparasites pose health risks to humans and animals. This study aimed to confirm the morphological and molecular identity of tick and flea species parasitising dogs in Metro Manila (Philippines) and molecularly investigate the possible presence of Bartonella spp., Rickettsia spp., Ehrlichia canis, and Anaplasma platys DNA. Ticks (n = 58) and fleas (n = 52) on dogs from three veterinary clinics in Metro Manila were collected and identified morphologically and molecularly via amplification and sequencing of cytochrome c oxidase I (cox1). Aliquots of ectoparasite DNA underwent real-time polymerase chain reaction (qPCR) screening for VBPs. All ticks were R. linnaei (formerly R. sanguineus s.l. “tropical lineage”), while all fleas were C. felis from clade 6 of the tropical II cluster/“Cairns” clade known from Australia. DNA of B. clarridgeiae was detected in 10% of fleas. DNA of R. felis was detected in 10% of fleas and in 3.8% of ticks. DNA of E. canis and A. platys was not detected. This study confirmed the presence of ticks and fleas as frequent ectoparasites on dogs and VBP presence emphasises the importance of preventative actions for animal health and welfare.


Introduction
Ticks and fleas are significant ectoparasites in companion animals as they carry vectorborne pathogens (VBPs) that pose a health risk to dogs and humans within the dogs' households [1,2]. Both ticks and fleas thrive in the Philippines because of the warm and humid climate. These environmental and species-specific drivers facilitate broad dog tick and flea distribution, further enabled by an increasing population of urban dogs [3,4]. As of 2020, there are approximately 10.8 million dogs in the Philippines, many of which are unowned and free-roaming street dogs [5].
The brown dog tick (Rhipicephalus sanguineus sensu lato) is considered the most common tick infesting dogs in the Philippines [3,6]. Genetic and biological differences divide R. sanguineus into tropical and temperate lineages [7]. The tropical lineage, prevalent in Asia and Australia, is now recognised as a distinct species, Rhipicephalus linnaei, and is suspected to be the only lineage in the Philippines [8]. The brown dog tick, R. sanguineus s.l., is the principal vector of Ehrlichia canis and Anaplasma platys [3,6,9]. The canine monocytic ehrlichiosis caused by E. canis is a potentially fatal disease of dogs with a seroprevalence as high as 33% in client-owned dogs in the Philippines [9,10]. Comparably, A. platys has a seroprevalence of 17% in client-owned dogs in the Philippines and causes canine infectious cyclic thrombocytopenia [3,10]. Both E. canis and A. platys are zoonotic, the former causing human ehrlichiosis and the latter potentially causing human anaplasmosis [11,12]. Knowledge regarding tick-borne pathogens in stray unowned and/or free-roaming street dogs and their role in disease transmission is incompletely understood in the Philippines [3].
The cat flea (Ctenocephalides felis) is the most common flea infesting dogs in the Philippines [1,6,13]. Mitochondrial genetic differences divide C. felis into temperate, tropical I, tropical II, and African clusters, but to date no specimens from the Philippines have been genetically characterised [14]. Models predicting the cat flea global distribution expect the Philippines to harbor the tropical I and tropical II cluster [14]. The tropical II cat flea cluster was shown to be the dominant "Cairns" clade in Australia thriving in the tropical climates [15]. Previous studies have noted the presence of the dog flea (Ctenocephalides canis) and the increasing prevalence of Ctenocephalides orientis in East and Southeast Asia, although its' presence in the Philippines remains unconfirmed [6,14]. The cat flea is the principal vector of Rickettsia spp. and Bartonella spp. [1,6,13]. A recent survey demonstrated a R. felis prevalence of 27.7% in C. felis samples collected from dogs in the Philippines [6]. Dogs are asymptomatic reservoir hosts for Rickettsia species, several of which are zoonotic. For instance, R. felis causes flea-borne spotted fever [16]. In contrast, Bartonella spp. cause a broad range of clinical signs including fever and endocarditis in dogs [17]. In humans, some Bartonella species-such as B. henselae and B. clarridgeiae-cause cat-scratch disease [1,4]. Others such as B. quintana causes bacillary angiomatosis, and B. koehlerae causes endocarditis [17]. The seroprevalence of Bartonella spp. antibodies in dogs from the Philippines is 2.6% indicating a low exposure to the pathogen [1].
This study aimed to (i) confirm the identity of the most common tick and flea species parasitising dogs in Metro Manila, Philippines; and (ii) investigate the possible presence of canine VBP DNA. Ticks and fleas were collected from three veterinary clinics and morphological and molecular techniques were used to unequivocally identify their species. Canine VBPs (Rickettsia spp., Bartonella spp., E. canis and A. platys) were screened using real-time PCR (qPCR) assays.

Results
Ticks (n = 58) and fleas (n = 52) were collected from a total of 42 client-owned dogs presenting to one of three veterinary clinics in Metro Manila, Philippines. The majority of dogs (92.9%, n = 39) were from clinics 1 (n = 20) and 2 (n = 19), located in San Juan City, whereas clinic 3 in Quezon City provided 7.1% (n = 3) of dogs ( Figure 1). Among the 42 dogs in the study, 50% had ticks only (n = 21), 42.9% had fleas only (n = 18), and 7.1% had both (n = 3). Female dogs accounted for 57.1% of the dogs sampled (n = 24). The age of the sampled dogs ranged from 0.2 to 13 years old with a median age of four years, however, four dogs had unknown ages. Of the sampled dogs, 45.2% were exclusively housed indoors (n = 19), while 40.5% were housed both indoors and outdoors (n = 17) and 14.3% were exclusively housed outdoors (n = 6) ( Table 1).
All ticks were morphologically identified as unambiguous R. sanguineus s.l. From the total number of ticks collected, 26 specimens from 24 dogs (at least 1 tick per dog) underwent cox1 amplification and DNA sequencing, all of which revealed high similarity (>99%) to R. linnaei (formerly R. sanguineus "tropical lineage"). In total, there were four R. linnaei cox1 haplotypes. The most numerous (17/26) R. linnaei cox1 haplotype was 100% identical while the remaining three were >99% identical with the reference mtDNA of R. linnaei (MW429381) from Australia [8].
All fleas were morphologically identified as unambiguous C. felis. In total, 20 flea specimens from 20 dogs (1 flea per dog) were subject to cox1 amplification and DNA sequencing, confirming C. felis identity. All but one C. felis specimen belonged to the M_h1 haplotype (one belonged to M_h2), which is identical to haplotype h3 sensu Lawrence et al. [14]. There was only a single nucleotide difference between M_h1 and M_h2. Both haplotypes belonged to the C. felis "Cairns" clade [15].
VBPs were detected in the DNA of ticks and fleas from 5 dogs. Bartonella and Rickettsia multiplex qPCR testing of 20 C. felis and 26 R. linnaei DNA samples was performed.
DNA sequencing of the three Bartonella-positive qPCR samples demonstrated that two 100% matched B. clarridgeiae ssrA (JN982716), while one had insufficient DNA quantity to be sequenced. All (n = 5) Rickettsia positive and Rickettsia suspect positive samples were subjected to conventional nested PCR to amplify the ompA and gltA genes. Four were successfully amplified and DNA sequence comparison to reference R. felis (CP000053) revealed all samples to be 100% R. felis at both loci [21]. The Rickettsia suspect positive tick sample failed to amplify using the ompA and gltA nested PCR and was therefore considered negative for Rickettsia spp.
Real-time PCR testing of 20 C. felis and 26 R. sanguineus DNA samples did not detect any E. canis or A. platys DNA ( Table 2). All ticks were morphologically identified as unambiguous R. sanguineus s.l. From the total number of ticks collected, 26 specimens from 24 dogs (at least 1 tick per dog) underwent cox1 amplification and DNA sequencing, all of which revealed high similarity (>99%) to R. linnaei (formerly R. sanguineus "tropical lineage"). In total, there were four R. linnaei cox1 haplotypes. The most numerous (17/26) R. linnaei cox1 haplotype was 100% identical while the remaining three were >99% identical with the reference mtDNA of R. linnaei (MW429381) from Australia [8].
All fleas were morphologically identified as unambiguous C. felis. In total, 20 flea specimens from 20 dogs (1 flea per dog) were subject to cox1 amplification and DNA sequencing, confirming C. felis identity. All but one C. felis specimen belonged to the M_h1 haplotype (one belonged to M_h2), which is identical to haplotype h3 sensu Lawrence et al. [14]. There was only a single nucleotide difference between M_h1 and M_h2. Both haplotypes belonged to the C. felis "Cairns" clade [15].
VBPs were detected in the DNA of ticks and fleas from 5 dogs. Bartonella and Rickettsia multiplex qPCR testing of 20 C. felis and 26 R. linnaei DNA samples was performed.  (Table A1). Sample locations for this study are in San Juan City represented in yellow (clinic 1) and green (clinic 2), and Quezon City in red (clinic 3). Google Maps, Philippines (2021).

Discussion
The majority of tick and flea samples came from the two clinics located in San Juan City. The third clinic that donated least ectoparasites is located in a more affluent area and receives more clients with a greater ability to purchase antiparasitics and other preventatives [22]. Tick infestation was reported to be 2.6% in dogs from Metro Manila, whereas tick and flea infestations in non-Metro Manila dogs were as high as 67.5% and 80%, respectively [10,23]. Factors influencing tick and flea infestation include the availability and ability to purchase antiparasitics which are likely contributing to the disparity between urban and rural dog infestation [10,23,24]. There are conflicting reports regarding age predilection to VBPs. For instance, one study reported VBPs to be more prevalent in <1-year-old dogs, whereas another reported an age range of 1-3 years [3,25]. Regardless, age predisposition has not been definitively established [26].
In this study, only the tropical brown dog tick (R. linnaei, formerly R. sanguineus s.l. "tropical lineage") was detected on dogs in Metro Manila, consistent with previous studies (Table A1) [6,13]. Molecular analysis based on cox1 typing confirmed R. linnaei. This is similar to previous studies based on 16S rDNA typing of 35 R. sanguineus s.l samples throughout East and Southeast Asia, including nine samples from the Philippines (MN685295-MN685303) [6,10]. The tropical brown dog tick (R. linnaei) appears to be the most prevalent species within R. sanguineus s.l. distributed across East Asia, Southeast Asia, and Australia, and it is the main vector for the transmission of VBPs such as Ehrlichia spp., Rickettsia spp., and Babesia spp. [6,8,10]. The temperate lineage of R. sanguineus s.l. and R. sanguineus sensu stricto are found in Beijing, which has cooler climates as compared to the Philippines [6,10]. Other ticks collected from dogs in East and Southeast Asia include Haemophysalis lognicornis and Haemophysalis hysterics, however, previous tick surveys found only R. sanguineus s.l. on dogs in the Philippines (Table A1) [6,8,10,13,18]. The cat flea (C. felis) was the only flea species found on dogs in Metro Manila in this study. All C. felis samples belonged to M_h1 haplotype except one (M_h2). A previous survey sequenced cox1 from one C. felis from the Philippines (MT027207) which was identical to the M_h1 haplotype in this study [10]. Both haplotypes (M_h1 and M_h2) belong to clade 6 of the tropical II cluster and the "Cairns" clade that was predicted to be distributed around Southeast Asia and North Australia [14,15,27]. This was to be expected as the ecological niche of the "Cairns" clade is similar to the climatic conditions of Metro Manila [15,28]. Aside from C. felis, previous surveys noted the presence of C. orientis and C. canis, which were not found in this study (Table A1) [6,10]. It is known that C. orientis is distributed across tropical Asia, including Southeast Asia [6,10,27]. In the previous studies, C. orientis could have been mistaken for C. canis as they have remarkably similar morphology [10,29].
The DNA of B. clarridgeiae was detected in 10% of C. felis samples (95% CI 1.57-31.3%), which is relatively similar to previous reports of Bartonella spp. prevalence in C. felis within East and Southeast Asia (16.5%, 95% CI 12.4-21.7%) and the Philippines (0%, 95% CI 0-9.0%) [6]. Cats are considered the main reservoir for Bartonella spp. as 31% of domestic cats in Metro Manila and Cebu City were seropositive for B. clarridgeiae and 68% were seropositive for B. henselae [4]. Dogs in Metro Manila were reported to have a low seroprevalence of B. henselae (2.6%), suggesting they are a spill-over host rather than a true host [1]. The low detection frequency of Bartonella spp. in this study could also support this notion. This study failed to detect any B. henselae, which may be due to the low seroprevalence in dogs and the small sample size. Nevertheless, the presence of B. clarridgeiae in fleas is important to note as it is considered a causative agent of cat-scratch disease [30].
This survey demonstrated a low prevalence of R. felis DNA in ticks (3.8%, 95% CI 0-20.5%) and fleas (10%, 95% CI 1.57-31.3%), which is roughly consistent with previous reports of a 1.6% (95% CI 0-9.3%) prevalence in ticks and 27.7% in fleas (95% CI 15.6-42.6%) [6]. Despite the low prevalence, the presence of R. felis DNA in ticks and fleas poses a risk to humans. Rickettsia spp. are a group of Gram-negative, obligate intracellular bacteria whose asymptomatic reservoir hosts are dogs [16]. Specifically, R. felis causes symptoms similar to murine typhus and dengue in humans bitten by infected fleas or ticks [31]. Because of the presence of R. felis in ticks and fleas from dogs, rickettsia infections may be considered by physicians for patients with dengue-like symptoms [32]. This study reports potential co-infection of R. felis and B. clarridgeiae in one flea, which has only been previously reported in Taiwan and France [33]. Detection of B. clarridgeiae DNA needs to be further investigated because the positive result could have been the result of co-feeding of the two ectoparasites and thus passive transfer of the DNA from flea to tick.
No E. canis and A. platys DNA in ticks were detected, implying 0-15% prevalence assuming a 95% confidence interval. This result is expected because of a detection frequency range of 3.2% to 8.3% for E. canis and 0.6-0.8% for A. platys in ticks, as reported by previous studies (Table A1) [3,6,19,25]. Seroprevalence of A. platys (17%, 95% CI 11.1-25.0%) and E. canis in dogs are high, indicating high exposures to the pathogens [10,34]. Dogs from Metro Manila were reported to have a higher seroprevalence of E. canis (95.3%, 95% CI 90.9-97.9%) as compared to dogs from non-Metro Manila cities (33%, 95% CI 25.0-42.2%) [10,34]. This suggests that dogs from Metro Manila have a higher exposure to E. canis, despite having a low tick infestation prevalence, supporting the notion that E. canis infection risk is independent of tick infestation levels [23,24]. The Philippines has many free-roaming stray dogs or the street dogs whose exact numbers are not known. These dogs are neglected, likely representing a reservoir for diseases, especially tick-borne pathogens as was demonstrated for Babesia gibsoni [35]. These dogs may facilitate the spread and maintenance of pathogens, especially since they frequently come into close contact with outdoor owned dogs in low-income households, which increases the risk of infection and infestation of client-owned dogs [10,36].

Collection and Morphological Identification of Ticks and Fleas from Dogs in Metro Manila, Philippines
Metro Manila is a metropolitan area in the Philippines that has a mean temperature of 25.5 • C in the coldest months, a mean temperature of 28.3 • C in the warmest months, and annual rainfall of 965 mm to 4064 mm [28]. All ticks and fleas were collected from Metro Manila between January 2021 and March 2021 and donated by three veterinary practices: Aso, Pusa atbp. Animal Shelter and Veterinary Services Clinic in San Juan City (Clinic 1), The Pet Project Veterinary Clinic in San Juan City (Clinic 2), and Vets in Practice in Quezon City (Clinic 3). Client-owned dogs visiting the practices were examined by a veterinarian for the presence of ticks and fleas. We had no control over effort and whether all fleas and ticks were found on individual dogs or any recent application of tick and flea preventatives. Ticks were removed from the dogs via tweezers and the fleas were removed using a flea comb according to veterinary practice routine procedure. Ticks and fleas are removed and discarded routinely from animals by all veterinary practices as part of veterinary care. For our purpose, removed ticks and fleas were donated to us rather than discarded, hence animal ethics approval was not required. Samples were stored in 1.5 mL tubes with 70% ethanol at room temperature. Samples were de-identified and submitted to the Veterinary Parasitology Laboratory (VPL) at the University of Sydney along with dog age, sex, indoor and/or outdoor status, and collection date as recorded by the veterinarian; the data were summarised descriptively, no further statistical analysis was attempted due to low sample sizes. Upon arrival in Sydney, all specimens were morphologically identified by the primary authors to the genus and species level via a stereomicroscope and morphological keys [37,38].

Molecular Characterisation of Ticks and Gleas at Cytochrome C Oxidase Subunit I (cox1)
Between one to three ticks and/or fleas representing at least one of each identified species per dog were selected for DNA isolation at VPL. A small incision was made to the body of each tick and flea using single-use sterile scalpel blades and dried in a heat block at 60 • C for 1 h [39]. Total genomic DNA was isolated using Monarch ® Genomic DNA Purification Kit (New England Biolabs, Australia). A blank isolation with no flea/tick DNA was included to control for cross-contamination (negative extraction control, NEC). DNA was eluted into 75 µL of elution buffer and stored at −20 • C.

Molecular Detection of Vector-Borne Pathogens in Ticks and Fleas
An aliquot of extracted tick and flea DNA was submitted to The Elizabeth Macarthur Agricultural Institute (EMAI) Laboratory (NSW Department of Primary Industries and Environment), Menangle, New South Wales) for Ehrlichia canis DNA and Anaplasma platys DNA diagnostic evaluation using real-time PCR following OIE protocols and assays [41,42].
Flea DNA underwent further screening at VPL at the University of Sydney using a multiplex TaqMan qPCR targeting the Rickettsia spp. and Bartonella spp. genes gltA (citrate synthase) and ssrA (transfer-messenger RNA), respectively [21,43,44]. The reactions were performed in duplicate using the CFX96 Touch TM Real-Time PCR Detection System (BioRad, Australia) and contained Luna ® Universal Probe qPCR Master Mix (New England BioLabs, Omnico, Australia) as described [21]. Results were considered positive if duplicates yielded C t values < 36. Results were considered suspect positive if one or more duplicates yielded C t values ≥ 36 and samples were considered negative if neither duplicate crossed the threshold (C t > 40). Positive Bartonella spp. results were sent to Macrogen for sequencing (Macrogen Ltd., Seoul, South Korea) and compared to reference Bartonella spp. sequences. Samples considered either positive or suspect positive for Rickettsia spp. (Ct value < 38) were further characterised using a pair of conventional nested PCRs targeting the outer membrane protein A (ompA) gene and gltA [21,45]. PCR products were sequenced at Macrogen Inc. (Seoul, Korea), assembled using CLC Main Workbench 21 (CLC bio, Qiagen, Australia), inspected manually by eye and compared to reference Rickettsia spp. sequences, i.e., R. felis (CP000053) [21].

Conclusions
This study confirms that the tropical brown dog tick (R. linnaei) and the cat flea (C.   n/a n/a Cebu