Prevalence and infection load dynamics of Rickettsia felis in actively feeding cat fleas

Abstract

Background: Rickettsia felis is a flea-associated rickettsial pathogen recurrently identified in both colonized and wild-caught cat fleas, Ctenocephalides felis. We hypothesized that within colonized fleas, the intimate relationship between R. felis and C. felis allows for the coordination of rickettsial replication and metabolically active periods during flea bloodmeal acquisition and oogenesis.

Methodology/principal findings: A quantitative real-time PCR assay was developed to quantify R. felis in actively feeding R. felis-infected fleas. In three separate trials, fleas were allowed to feed on cats, and a mean of 3.9x10(6) R. felis 17-kDa gene copies was detected for each flea. A distinct R. felis infection pattern was not observed in fleas during nine consecutive days of bloodfeeding. However, an inverse correlation between the prevalence of R. felis-infection, which ranged from 96% in Trial 1 to 35% in Trial 3, and the R. felis-infection load in individual fleas was identified. Expression of R. felis-infection load as a ratio of R. felis/C. felis genes confirmed that fleas in Trial 3 had significantly greater rickettsial loads than those in Trial 1.

Conclusion/significance: Examining rickettsial infection dynamics in the flea vector will further elucidate the intimate relationship between R. felis and C. felis, and facilitate a more accurate understanding of the ecology and epidemiology of R. felis transmission in nature.

Figure 1. R. felis-infection load in individual C. felis lysates.

For nine consecutive days, 10 fleas were randomly selected and removed from the feeding capsule situated on a cat host. In Rickettsia positive fleas, qPCR was used to quantify the mean (±SEM) R. felis-infection load. Results are presented as the mean daily R. felis-infection load (17-kDa antigen gene copy number) of individual, whole flea lysates during flea bloodmeal acquisition. Within each trial significant differences from Day 0 (unfed fleas) are marked with an asterisk. ^ΨIn Trial 3, Day 3 only one flea was positive for R. felis infection.

Figure 2. Estimation of C. felis 18S rDNA copy number using Southern blot.

C. felis genomic DNA (6 µg/enzyme) was digested with Eag I, EcoR I, Pst I, Xba I and Xho I. Uncut gDNA served as a negative control and PCR product of a portion of C. felis 18S rDNA served as a positive control. Genomic DNA was hybridized with Cf18S bp probe to estimate the number of 18S rDNA gene copies in C. felis. Digestion with each enzyme results in a single digestion product. (*DNA was not completely digested by Pst I. Top band is uncut DNA.)

Figure 3. R. felis-infection load in C. felis expressed as a ratio of R. felis and C. felis genes.

For nine consecutive days 10 fleas were randomly selected and removed from the feeding capsule situated on a cat host. In Rickettsia positive fleas, R. felis infection was determined by quantifying the R. felis 17-kDa copy number. For R. felis-infected fleas, C. felis 18S rDNA copy number was quantified to serve as a comparison point for R. felis infection. R. felis-infection density was determined by logarithmically transforming and taking the ratio of R. felis 17-kDa and C. felis 18S rDNA (Rf17kDa/Cf18S) copy numbers and the mean (±SEM) daily Rf17kDa/Cf18S ratios for each trial were calculated. Within each trial significant differences from Day 0 (unfed fleas) are marked with an asterisk. ^ΨIn Trial 3, Day 3 only one flea was positive for R. felis infection. All trials have significantly higher ratios than the previous trial, indicating that rickettsial burdens are increasing in the LSU C. felis colony.

Figure 4. Experimental model depicting the relationship between R. felis and C. felis.

Supported by the results of this study, R. felis prevalence and individual flea infection load in the LSU C. felis colony are inversely correlated. The ratio of Rf17kDa/Cf18SrDNA also increased significantly between Trial 1 and Trial 3 indicating fleas are infected at a greater density. Trials are situated within the model according to their individual results. As a population, fleas in Trial 1 had the highest prevalence of R. felis infection and the lowest mean individual R. felis-infection load. Conversely, fleas in Trial 3 had the lowest prevalence of R. felis-infection and the greatest mean individual R. felis-infection load. Trial 2 fleas represent a median demonstrating the progression of decreasing colony prevalence and increasing infection load from Trial 1 to Trial 3. The ratio of R. felis and C. felis genes increases significantly from Trial 1 to Trial 3, demonstrating that fleas in Trial 3 have a greater R. felis burden (higher Rf17kDa/Cf18S ratio) than fleas in Trial 1. These results indicate that at increased infection loads, R. felis may influence flea fitness to facilitate their own successful transmission to the next generation of fleas or to a susceptible mammalian host.