Presence of Coxiella burnetii in Fleas in Cyprus

Introduction

Qfever is a zoonosis caused by Coxiella burnetii, a small, obligate, intracellular, Gram-negative bacterium that is prevalent throughout the world (Angelakis and Raoult 2010). It can infect and persist in a wide range of domestic and free-living vertebrates, namely ungulates (although the pathogen is not exclusively associated to them), and ticks. A number of studies have been published on the presence and dispersal of C. burnetii infection in wildlife, whereas over 40 tick species have been found to be naturally infected by C. burnetii. Nevertheless, little is known about the presence of C. burnetii in other ectoparasites, such as fleas. In fact, a single study has been published in the past demonstrating the presence of the pathogen in fleas (Loftis et al. 2006). In Cyprus, a high prevalence of C. burnetii in humans, goats, sheep, and bovines was recorded in past studies (Psaroulaki et al. 2006). Recently, 12.8% of rat collected in different areas of the island revealed antibodies against the pathogen (Psaroulaki et al. 2010). The pathogen has been detected both in endemic and migratory birds, as well as in mouflons and their ticks (Ioannou et al. 2009, 2011). Recent still unpublished data of a cross-sectional study have verified the presence of C. burnetii in ticks collected from wildlife. The aim of the present study was to record the presence of C. burnetii in fleas collected from wild animals that may take part in the pathogen's cycle in Cyprus.

Materials and Methods

During a 6-year (2000–2006) study, fleas were collected from rats (Rattus norvegicus and R. rattus), foxes (Vulpes vulpes indutus), and hares (Lepus europaeus) captured from 48 different areas of Cyprus. Captures were carried out throughout the year, but mainly in the spring and autumn months. Collected ectoparasites were identified using standard taxonomic keys (Ioannou et al. 2011). Fleas were preserved in sterile conditions at room temperature before being sent to the Laboratory of Clinical Bacteriology (Q fever reference center) in Heraklion, Crete, Greece. Prior to DNA isolation, each flea was rinsed twice in 70% ethanol and sterile water for 15 min. Fleas were placed in separate eppendorf tubes into 200 μL of distilled water, smashed using surgical scissors, and then pooled, with each pool containing three to nine fleas, according to species, host, and region of collection, to a final volume of 200 μL. In some cases, fleas from different different host individuals, but always belonging to the same species, were pooled together. DNA was extracted using a QIAamp Tissue Kit (Qiagen, Hilden, Germany). The extracted DNA was handled under sterile conditions to avoid cross-contamination and stored at −20^°C until PCR assays. Alongside the samples, extraction from sterile water also took place; the extracted product was used as a negative marker. For the detection of C. burnetii, a real-time PCR targeting the IS1111 insertion sequence was used (Denison et al. 2007). Two sets of negative controls (DNA from noninfected specimens and sterile water) were used.

Results

Of the 2072 wild animals tested, namely 622 rats (983 fleas), 324 foxes (118), and 1126 hares (46), 652 were infested by ticks and/or fleas, and 1147 fleas were removed from them. Three flea species were identified: Xenopsylla cheopis (729), Ctenocephalides felis (304), and C. canis (114). A total of 153 pools were prepared. C. burneti was identified in 25 (16.3%) pools. None of the fleas parasitizing hares was positive for C. burnetii, as opposed to fleas collected from rats (12%) and foxes (47.6%). The highest presence was recorded in C. canis (38.1%) compared to C. felis (16.6%) and X. cheopis (10.8%). All positive C. canis were removed from foxes, and all positive X. cheopis were removed from rats. The infection rate in flea pools was estimated using the formula maximum likelihood estimation (MLE)=1− (1 − Y/X), where Y=number of positive pools, X=number of pools (Walter et al. 1980). All results are summarized in Table 1.

Discussion

Fleas are primarily associated with the transmission of pathogens other than C. burnetii, such as Rickettsia typhi, R. felis, Bartonella henselae, Yersinia pestis, etc. (Eisen and Gage 2012). Very few studies have been conducted aiming to detect C. burnetii in fleas; in fact, in the past, the pathogen has been detected in a single study in two (one X. cheopis and one C. felis) of the 987 fleas collected in a study conducted in Egypt (Loftis et al. 2006). This low infection rate was not unexpected, because C. burnetii is transmitted primarily by air, milk products, or ticks (McQuiston and Childs 2002). Relatively higher recordings were revealed in the current study, implying that fleas may play a role in the transmission of C. burnetii among wild vertebrates. Of course it should always be at the back of our minds that the presence of the pathogen in fleas could be due to an infected meal from its host. In that case, the flea could well be infected by C. burnetii, but this does not imply that it could be able to transmit the pathogen in other hosts and/or in humans. To prove which of two hypotheses is valid, or even both, since we are talking about different flea species, more studies are required that will test the ability of fleas to digest blood from infected hosts and transmit it to noninfected ones.

Concerning hares, few studies have been conducted in the past studying the relationship between this animal species and C. burnetii; nevertheless, the presence of the pathogen has been demonstrated either by molecular (Astobiza et al. 2011) or by serological means (Hubalek et al. 1993, Marrie et al. 1993). In our case, the absence of the pathogen in fleas collected from hares may be attributed to their low number collected from this host. Rats have been proven to carry the pathogen and can serve as vertebrate amplifying hosts, aiding the maintenance of the transmission cycles of the pathogen in urban areas (Comer et al. 2001). In fact, a number of studies have demonstrated the presence of C. burnetii in this animal species either by serology (Psaroulaki et al. 2010, McCaughey et al. 2010) or by genotyping (de Bruin et al. 2012, Jado et al. 2012), enhancing the opinion that rats may act as potential reservoirs of the pathogen (Reusken et al. 2011).

Regarding foxes, a single bibliographical study has recorded an attempt to detect C. burnetii in this animal species without any success (Almeida et al. 2013). Most of the wild animals from which fleas, tested in the current study, were collected may carry the pathogen and may potentially transmit it to fleas and vice versa. Of course, it should not be underestimated that study of vectors by collection only provides a frame of a specific time of the wild cycle. Thereafter, the role of fleas in the maintenance and transmission of C. burnetii remains to be verified by further studies.