Identifying blood meals in cat fleas (Ctenocephalides felis) from a plague-endemic region of Uganda using a SYBR Green real-time polymerase chain reaction-based assay

Abstract

A zoonotic disease that has killed millions over the course of at least three pandemics, plague remains a threat in regions where the etiologic agent, Yersinia pestis, persists in natural cycles involving small mammals and their fleas. Numerous flea species have been implicated as Y. pestis vectors, and some provide a "bridge" from zoonotic hosts to humans, particularly during the epizootics that decimate susceptible small mammal populations. In order to serve as a bridging vector, a flea species must be able to transmit Y. pestis, it must feed on infectious zoonotic hosts, and it must feed on humans. Identifying bridging vector species in plague-endemic regions can aid in the development of vector-control activities aimed at reducing the incidence of human plague. The West Nile region is an established plague focus in northwest Uganda. Since 1999, more than 2400 suspect human plague cases have been reported from Vurra and Okoro counties. The most likely source of infection for humans in this region is the black rat, Rattus rattus, which commonly infests human habitations and is highly susceptible to Y. pestis infection. Other potential zoonotic hosts include other rodent and shrew species that predominate in the peridomestic environment and occasionally enter huts. Two rat flea species, Xenopsylla cheopis and X. brasiliensis, both among the most efficient flea vectors of Y. pestis, are very likely to serve as bridging vectors to humans in Vurra and Okoro counties. Recent investigations, however, have found that the cat flea, Ctenocephalides felis, comprises more than 88% of host-seeking (off-host) fleas captured in huts in this region. Though an inefficient vector, this species is capable of transmitting Y. pestis. Given its dominance in human habitations and its catholic feeding habits in other regions, we hypothesized that C. felis might serve as a secondary bridging vector in Vurra and Okoro counties. In order to address this hypothesis, we sought to determine what proportion of blood meals in off-host cat fleas collected in huts in this region come from humans, and what proportion come from potentially-infectious small mammal species. Blood meal assays have long been used to examine the feeding behavior of a wide variety of disease vectors, but existing blood meal assays were deemed inadequate for our purposes because they were either not sensitive enough to detect the very small amounts of host DNA in field-collected fleas, or they were unable to capture the wide range of potential cat flea hosts in the West Nile region. Therefore, we developed a blood meal assay that takes advantage of the exquisite sensitivity of SYBR Green I-based real-time polymerase chain reaction (PCR) and combines it with the specificity and flexibility afforded by sequencing. We found that this highly-sensitive assay was subject to human DNA contamination, so we analyzed vertebrate DNA detection in artificially-fed and unfed fleas to establish a threshold cycle (Ct) cutoff that would optimize specificity without completely sacrificing sensitivity. Specifically, we identified a Ct cutoff that maximized positive predictive value. Using the established cutoff, our assay was 94 percent specific, detecting contaminating human DNA in 3 of 50 unfed fleas, and it detected and correctly identified the source of human and rat blood meals in 100 percent of artificially fed fleas held alive for up to 4 hours post feeding. Assay sensitivity declined as the time between feeding and collection increased, but we were able to detect and identify human and rat DNA in a proportion of artificially-fed fleas held alive for up to 72 hours post feeding. Using this assay, we detected and identified vertebrate DNA in 148 off-host C. felis collected in human habitations in Vurra and Okoro counties, none of it from wild rodents or shrews. Our findings indicate that cat fleas infesting huts in the West Nile region probably feed on humans, but the majority of off-host C. felis blood meals came from domesticated species that are unlikely to play a significant role in perpetuating transmission of Y. pestis. We concluded that C. felis is unlikely to serve as a bridging vector for Y. pestis in the West Nile region.