This study represents the first isolation of R. felis in cell culture from cat fleas in Australia. This pathogenic agent is an obligate intracellular bacteria which requires nucleated eukaryotic cells to grow , and grows best at temperatures under 32°C . XTC-2 cell lines are derived from Xenopus laevis, a South African clawed toad, which grows at 28°C and is suited to support the growth of R. felis at optimum levels. In contrast, the optimal growing temperatures for typhus group (35°C) and spotted fever group (32°C) rickettsiae are higher [29, 30].
Previous isolation of R. felis in XTC-2 cell lines was attempted using the shell vial centrifugation technique . This technique is sensitive and frequently utilised for isolation of agents from clinical specimens [31, 32] that contain a low burden of microorganisms. However, it is laborious, requires expertise and is not suitable for downstream production of antigen for serological assays. In this study, conventional cell culture was carried out utilising cell culture flasks to enable production of R. felis antigens in large amounts. Rickettsia felis has also been reportedly successfully cultivated in vertebrate and arthropod cell lines, including Vero cells, L929, ISE6 and C6/36 [22, 33–35].
Our study represents the first to provide serological evidence for R. felis exposure in dogs in Australia. The high seroprevalence (50.7%) is in agreement with a study conducted in Spain, where 51.1% of dogs were reported as exposed to this agent . The high seroprevalence of R. felis in dogs in the present study was not unexpected. The cat flea, C. felis, is known to be the most common ectoparasite and dominant flea infesting dogs in Australia and its wide geographical distribution across the country [11, 36] suggests that the seroprevalence of R. felis reported in this study could be a representation of most populated areas of Australia. We found no significant difference in seropositivity between dogs located in SE QLD and NT despite the variation in climate. This suggests that dogs from these two regions have been equally exposed to R. felis, which is in turn likely attributable to frequent exposure to cat fleas. However, flea infestation in dogs in the current study was not evaluated, hence an association with the presence of R. felis antibodies could not be confirmed.
Besides fleas, DNA of R. felis has also been isolated from the brown dog tick, Rhipicephalus sanguineus. This tick is highly prevalent in dogs in the NT due to its preference for the humid warm tropics with relative humidity of 60%-90% and temperatures of 20°C −30°C . Whether this tick species acts as a true biological vector as opposed to simply being an incidental mechanical vector remains uncertain at this time.
Serological cross-reactivity among Rickettsia spp is common. R. felis antibodies have been known to be more reactive to R. typhi from the typhus group, than to the spotted fever group [2, 7, 39, 40]. Moreover, a recent serosurvey study in Spain showed dogs that were positive for R. felis antibodies did not necessarily cross-react with R. typhi, with prevalences of 9.7% and 51.1% respectively . A seroepidemiological study of R. felis, R. typhi and R. conorii infection in humans in Spain also demonstrated low levels of cross-reaction between R. felis and R. typhi or R. conorii. These findings might suggest the possibility of high specificity of R. felis serological tests.
The current study highlights the importance of flea control in pets by demonstrating a significant association between active ectoparasite control and the absence of R. felis exposure. Although it is not statistically significant in the multivariable model, intact female animals in the current study showed higher seroprevalence of R. felis compared to neutered dogs, suggestive of possible association with gonadal hormonal factors that might influence the outcome of an infection . Sex-associated behaviour such as roaming in intact males may predispose them to wider exposure to fleas and the pathogens they carry. This phenomenon has been observed in a number of studies whereby neutering decreased the prevalence of both endoparasites and tick-borne diseases in dogs [42–45].
The high seroprevalence in dogs in the present study, the detection of R. felis DNA in dog blood  and high infection rates in cat fleas sourced from dogs  support the role of dogs as potential reservoir hosts for this zoonosis . Previous studies have demonstrated infection with rickettsial spotted fever in humans positively associated with owning or contacting dogs [47, 48]. In Spain, seropositivity was associated with humans who had contact with domestic animals compared to farm and wild animals . A dog whose owners were infected with FSF was also found to be infected by the same agent . This study further provides evidence of the risks this emerging zoonosis poses, especially to companion animal owners and their families.