Effect of Rickettsia felis Strain Variation on Infection, Transmission, and Fitness in the Cat Flea (Siphonaptera: Pulicidae).

Rickettsia felis is a human pathogen transmitted by the cat flea, Ctenocephalides felis (Bouché) (str. LSU), as well as an obligate symbiont of the parthenogenic booklouse Liposcelis bostrychophila (Badonnel) (str. LSU-Lb). The influence of genetic variability in these two strains of R. felis on host specialization and fitness and possible resulting differences on infection and transmission kinetics in C. felis is unknown. Utilizing an artificial host system, cat fleas were exposed to a R. felis str. LSU-Lb-infected bloodmeal and monitored for infection at 7-d intervals for 28 d. Quantitative real-time PCR was used to determine rickettsial load and infection density in newly exposed cat fleas, and transmission frequency between cat fleas. The effect of persistent R. felis infection on cat flea F1 progeny was also assessed. At 7 d postexposure 76.7% of the cat fleas successfully acquired R. felis str. LSU-Lb. In R. felis str. LSU-Lb-exposed cat fleas, the mean infection load (6.15 × 106), infection density (0.76), and infection prevalence (91/114) were significantly greater than R. felis str. LSU infection load (3.09 × 106), infection density (0.68), and infection prevalence (76/113). A persistent R. felis str. LSU-Lb infection was detected for 28 d in adult cat fleas but neither female:male ratio distortion nor vertical transmission was observed in F1 progeny. While infection kinetics differed, with higher intensity associated with R. felis str. LSU-Lb, no distinct phenotype was observed in the F1 progeny.

Insect-borne Rickettsia encompasses several agents of human disease across the globe, including Rickettsia felis, which is recognized as an emerging human pathogen in distinct regions of the world (e.g., sub-Saharan Africa and Southeast Asia; Richards et al. 2010, Socolovschi et al. 2010, Parola 2011, Maina et al. 2012). Originally identified in the cat flea, Ctenocephalides felis (Bouché) (Adams et al. 1990), the presence of R. felis and related genotypes has been reported from a wide range of arthropod hosts, including at least 40 additional species of fleas, mosquitoes, ticks, and mites (Brown and Macaluso 2016). Although there has been molecular detection in numerous hematophagous arthropods, identification of competent biological transmission vectors has been limited (Dieme et al. 2015, Brown and Macaluso 2016). Consequentially, the occurrence of R. felis in most hematophagous arthropods is credited to ingestion of an R. felis-infected bloodmeal via cofeeding transmission between infected and naïve vectors (Brown et al. 2015).

In addition to molecular detection in multiple hematophagous hosts, a novel strain of R. felis (str. LSU-Lb) has been identified within the non–blood-feeding booklouse, Liposcelis bostrychophila (Badonnel) (Insecta: Psocoptera) (Yusuf and Turner 2004, Behar et al. 2010, Thepparit et al. 2011). Although considered a facultative parasite of cat fleas, as an obligate mutualist in the booklouse host R. felis is required for the early maturation of the oocyte, maintained 100% transovarially, and is associated with a parthenogenic phenotype (Perotti et al. 2006). Previous studies demonstrated that the loss of R. felis from L. bostrychophila, either as larvae by increased temperature or as adults via antibiotics, results in diminished longevity and fecundity, as well as the production of nonviable eggs (Yusuf and Turner 2004, Perotti et al. 2006). Currently, no measurable effect on the fitness of fleas infected with R. felis has been recognized, and transovarial transmission is inconsistent (ranging from 0 to 100% in laboratory colonies; Higgins et al. 1994, Hirunkanokpun et al. 2011, Reif et al. 2011). Therefore, heterogeneity within each R. felis genotype may influence the transmission routes utilized for sustained infection among distinct arthropod populations.

Several genotypes of R. felis have been found in colonized and wild-caught cat fleas and propagated using various cell lines (Horta et al. 2006, Pornwiroon et al. 2006), including the reference strain Marseille-URRWXCal2 derived from fleas of California origin (La Scola et al. 2002). A recent study showed genomic diversity across R. felis strains, and identified several factors that differentiate the R. felis isolated from cat fleas (str. URRWXCal2 and str. LSU) from that of booklice (str. LSU-Lb; Gillespie et al. 2015). In addition to the pRF plasmid, common to all R. felis genomes, R. felis str. LSU-Lb also contains the pLbaR plasmid. Furthermore, phylogenomics analysis demonstrated that the booklouse strain of R. felis diverged from the flea-associated strains. Despite the recent detailed report of the phylogenetic relationship between strains of R. felis, little is known about the biology or transmission phenotype of such strains.

In an effort to conclude if genetic variability in R. felis underlies host specialization and fitness, and results in strain-specific infection and transmission kinetics we examined: 1) the capacity of cat fleas to acquire R. felis str. LSU-Lb via an infectious bloodmeal; 2) the prevalence and infection load dynamics of R. felis str. LSU-Lb in cat fleas; 3) the horizontal transmission of R. felis str. LSU-Lb between cat fleas; 4) the vertical transmission of R. felis str. LSU-Lb in the cat flea; and 5) the effect of R. felis str. LSU-Lb on the development and reproductive fitness of the cat flea.

Materials and Methods

Source of Fleas and Rickettsia

Freshly emerged, unfed cat fleas were obtained from Elward II (El-Labs, Soquel, CA). Cat fleas from this colony have previously been described as being negative for R. felis (Pornwiroon et al. 2007). Fleas were given a bloodmeal using an artificial dog (Wade and Georgi 1988), and eggs were raised to adults on sand with artificial diet as previously described (Lawrence and Foil 2000). Rickettsia felis str. LSU (Pornwiroon et al. 2006) and R. felis str. LSU-Lb (Thepparit et al. 2011), were isolated from LSU colonies of cat fleas and booklice (L. bostrychophila), respectively, and were grown in ISE6 cells as described previously (Pornwiroon et al. 2006).

Rickettsia felis-Infected Bloodmeal and Flea Infection

Bloodmeals for infection were created by resuspending ISE6 cells infected with R. felis str. LSU or R. felis str. LSU-Lb. The R. felis exposure dose was prepared using the BacLight viability stain kit (Molecular Probes, Carlsbad, CA) to assess viability and enumerate rickettsiae (Sunyakumthorn et al. 2008). For the infection kinetics experiment, concentrations of R. felis (passage 6) were adjusted to 5 × 109 rickettsiae in 600 µl of heat-inactivated, defibrinated bovine blood (HemoStat Laboratories, Dixon, CA). For the horizontal transmission experiment, the concentration of R. felis str. LSU-Lb (passage 6) was adjusted to 2.5 × 1010 rickettsiae in 600 µl of heat-inactivated, defibrinated bovine blood (Hirunkanokpun et al. 2011).

Sample Preparation and Rickettsial Quantification by PCR

All flea samples were washed with 10% bleach for 5 min, 70% ethanol for 5 min and three times sterile distilled water (5 min). Fleas were then placed in 1.5-ml tubes and crushed with sterile plastic pestles in a liquid nitrogen bath. Extraction of genomic DNA (gDNA) was accomplished using the DNeasy Blood and Tissue Kit (Qiagen, Germantown, MD) according to the manufacturer’s instructions for tissue samples and eluted into 25 µl of UltraPure DNAse/RNAse-free distilled water (dH2O; Invitrogen, Grand Island, NY). A negative environmental control was utilized for each DNA extraction process.

Fleas were assessed for the presence of R. felis str. LSU and R. felis str. LSU-Lb by qPCR amplification of a 157-bp portion of the 17-kDa antigen gene (Reif et al. 2008). The absence of rickettsial infection before exposure to R. felis, was shown by testing a subset of fleas by qPCR. The qPCR was performed with a LightCycler 480 Real-Time PCR system (Roche, Indianapolis, IN) and R. felis infection load was quantified as the copy number of Rf17kDa per individual flea lysate. Quantitative PCR analyses used serial 10-fold dilutions of the plasmid pCR4-TOPO-Rf17kda + Cf18SrDNA to generate a standard curve and quantify R. felis gene copies for individual flea lysates (Reif et al. 2008). Rickettsia felis-infection density was calculated as the ratio of log transformed Rf17kDa and log transformed Cf18SrDNA copy numbers (Rf17kD/Cf18SrDNA) per each flea (Reif et al. 2008).

Infection of R. felis str. LSU-Lb Through Infectious Blood

In two trials, both containing two experimental groups and one uninfected (control) group, fleas were challenged for 24 h to a R. felis str. LSU, R. felis str. LSU-Lb, or an uninfected (control) bloodmeal. After exposure to the blood, cat fleas were sustained on defibrinated bovine blood that was replaced every 2–3 d for the entirety of the trial. Fleas were examined every 7 d postexposure (dpe) for acquisition (ability to acquire R. felis from an infected bloodmeal), persistence (detection of R. felis in fleas over 28 d period), prevalence (no. positive/no. tested), infection load (copy number of Rf17kDa per individual flea lysate), and infection density (ratio of Rf17kDa/Cf18S per flea). At each time point (7, 14, 21, and 28 dpe), gDNA was extracted from 10 female and 5 male viable fleas and R. felis infection was tested by qPCR. For each trial the prevalence of R. felis infection, R. felis infection load per flea lysate, and R. felis infection density were compared between treatments at each time point.

Horizontal Transmission of R. felis str. LSU-Lb

To distinguish R. felis-infected donor and uninfected recipient fleas in the bioassay, rhodamine B (RB) (Sigma-Aldrich, St. Louis, MO) was used as a biomarker (Hirunkanokpun et al. 2011). The donor fleas were fed a R. felis str. LSU-Lb infected bloodmeal as described earlier. The RB labeled bloodmeal was created with 500 µl of a 0.1% solution of RB in heat-inactivated bovine serum added to 100 µl of heat-inactivated defibrinated bovine blood and fed to recipient fleas for 24 h, as previously described (Hirunkanokpun et al. 2011). Rickettsia felis str. LSU-Lb-infected donor fleas (n = 70) were placed in the same cage with RB-labeled uninfected recipient fleas (n = 70). Fleas were then fed on normal defibrinated bovine blood and assessed for R. felis str. LSU-Lb infection at days 1, 7, 14, and 21 post cofeeding.

Vertical Transmission of R. felis str. LSU-Lb to Flea Progeny and F1 Sex Ratio

In both trials, eggs from fleas exposed to R. felis str. LSU and R. felis str. LSU Lb-infected bloodmeals were collected every 5 or 10 d post parental bloodmeal contact, and reared to adults in an incubator maintained at 27°C and 90% relative humidity. Adult fleas were sexed and resulting female-to-male sex ratio determined. Total gDNA was then extracted from individual (LSU n = 80, LSU-Lb n = 79, Control n = 15) or pools of 10 (LSU n = 1,050, LSU-Lb n = 840, Control n = 30) F1 cat fleas and the presence of R. felis was determined by qPCR as described previously.

Statistical Analysis

The ratio of Rf17kD/Cf18SrDNA was assessed after the logarithmic transformation of the quantity of the genes of interest (Rf17kD and Cf18SrDNA). Data were analyzed with one-way analysis of variance (ANOVA) to examine potential differences between R. felis str. LSU and str. LSU-Lb infection load (copy number of Rf17kDa per individual flea lysate), and infection density (ratio of Rf17kDa/Cf18S per flea) over the course of the experiment (IBM SPSS Statistics for Windows, Version 23.0, Armonk, NY). When overall significance was found, Tukey’s honestly significant difference (HSD) post hoc test was used to examine pairwise differences of means (Figs. 2 and 3). An independent samples t-test was used for general comparisons of grouped means (Fig. 1). Differences in the proportion (prevalence) of infected fleas in R. felis str. LSU vs. R. felis str. LSU-Lb (Table 1) and donor vs. recipient fleas (Table 2) were analyzed using Fisher’s Exact test (Brown et al. 2015). Differences in sex ratios in the F1 generations (Table 3) were analyzed using Fisher’s exact test (Morick et al. 2013). For all comparisons, a P value of < 0.05 was considered significantly different.

Fig. 2

Rickettsial load and infection density in fleas exposed to a R. felis str. LSU- or R. felis str. LSU-Lb-infected bloodmeal 7–28 d postexposure (dpe). (A) Rickettsial load in adult fleas infected with R. felis str. LSU or R. felis str. LSU-Lb collected at different time points. (B) Rickettsial infection density in adult fleas infected with R. felis str. LSU or R. felis str. LSU-Lb collected at different time points. Bars represent means (±SEM) of Rf17kDa copy number and Rf17kDa/Cf18S ratio. Means with different letters are significantly different (P < 0.05).

Fig. 3

Mean Rickettsia felis str. LSU-Lb infection in donor and recipient fleas at different time points. (A) Rickettsial load in R. felis str. LSU Lb-infected donor and recipient fleas collected at days 1, 7, 14, and 21 post cofeeding. (B) Rickettsial infection density in R. felis str. LSU Lb-infected donor and recipient fleas collected at days 1, 7, 14, and 21 post cofeeding. Donor fleas typically had greater rickettsial loads and densities, compared with recipient fleas. Bars represent means (±SEM) of Rf17kDa copy number and Rf17kDa/Cf18S ratio. Means with different letters are significantly different (P < 0.05).

Fig. 1

Overall mean Rickettsia felis str. LSU-Lb infection in fleas. (A) Overall Rickettsial load in R. felis str. LSU Lb-infected fleas compared to R. felis str. LSU. (B) Overall Rickettsial infection density in R. felis str. LSU Lb-infected fleas. Overall R. felis str. LSU Lb-infected fleas had significantly greater rickettsial loads (females and mixed sex adults) and densities (mixed sex adults), compared with R. felis str. LSU-infected fleas. Bars represent means (±SEM) of Rf17kDa copy number and Rf17kDa/Cf18S ratio. Significant differences (P < 0.05) between strains are marked with an asterisk.

Table 1

Prevalence (no. positive/no. tested) of R. felis str. LSU and R. felis str. LSU-Lb in cat fleas exposed to an infected bloodmeal

Fleas	Day 7 (%)		Day 14 (%)		Day 21 (%)		Day 28 (%)		All Days (%)
	R. felis	R. felis	R. felis	R. felis	R. felis	R. felis	R. felis	R. felis	R. felis	R. felis
	str. LSU	str. LSU-Lb	str. LSU	str. LSU-Lb	str. LSU	str. LSU-Lb	str. LSU	str. LSU-Lb	str. LSU	str. LSU-Lb
Female	13/20	16/20	15/20	17/20	7/20	19/20	12/13	13/14	47/73	65/74
	(65.0)	(80.0)	(75.0)	(85.0)	(35.0)	(95.0)	(92.3)	(92.9)	(64.4)	(87.8)
Male	8/10	7/10	7/10	5/10	7/10	7/10	7/10	7/10	29/40	26/40
	(80.0)	(70.0)	(70.0)	(50.0)	(70.0)	(70.0)	(70.0)	(70.0)	(72.5)	(65.0)
Total	21/30	23/30	22/30	22/30	14/30*	26/30*	19/23	20/24	76/113*	91/114*
	(70.0)	(76.7)	(73.3)	(73.3)	(46.7)	(86.7)	(82.6)	(83.3)	(67.3)	(79.8)

A significant difference was detected in the total prevalence between strains.

Table 2

Donor and recipient fleas (no. positive/no. tested) infected with R. felis str. LSU-Lb in horizontal transmission bioassay

Fleas	Day 1 (%)	Day 7 (%)	Day 14 (%)	Day 21 (%)	All days (%)
Donor	20/21 (95.2)	3/5 (60.0)	3/5 (60.0)	5/5 (100.0)	31/36* (86.1)
Recipient	5/21 (23.8)	3/15 (20.0)	6/14 (42.9)	5/14 (35.7)	19/64* (29.7)

A significant difference was observed in the prevalence between recipient and donor fleas.

Table 3

Sex ratio (male: female) of F1 adult fleas (no. males/no. females)

Group	Sex ratio m:f (n male/n female)
	Days eggs collected post infection
	1–5	6–10	11–15	16–20	21–30	31–40	Total
R. felis str. LSU	1:1.05 (20/21)	1:1.07 (196/209)	1:1.06 (165/175)	1:1.06 (50/63)	1:0.92 (75/69)	1:0.83 (6/5)	1:1.06 (512/542)
R. felis str. LSU-Lb	1:2.78 (9/25)	1:0.89 (130/115)	1:0.75 (166/124)	1:1.56 (48/75)	1:0.96 (73/70)	1:1.17 (6/7)	1:0.96 (416/432)
Uninfected	1:0.78 (9/7)	1:1.13 (192/216)	1:1.02 (127/129)	1:1.01 (102/103)	1:0.82 (62/51)	1:0.73 (11/8)	1:1.02 (502/514)

Results

Acquisition of R. felis str. LSU-Lb and Infection in Cat Fleas Exposed to an Infectious Bloodmeal

The C. felis colony tested negative for R. felis as had been shown in previous accounts (Pornwiroon et al. 2007, Reif et al. 2011, Thepparit et al. 2013). In two trials, both groups fed on an infected bloodmeal had positive fleas for 4 wk postexposure, with a 67.3% and 79.8% mean prevalence of infection in R. felis str. LSU and R. felis str. LSU-Lb, respectively. The total Day 21 infection prevalence for R. felis str. LSU-Lb-infected fleas (26/30) was significantly higher than that of R. felis str. LSU-infected fleas (14/30). In addition, the overall total infection prevalence for R. felis str. LSU-Lb-infected fleas (91/114) was significantly higher than that of R. felis str. LSU-infected fleas (76/113) (Table 1). The prevalence of infection ranged from 46.7% to 82.6% and 73.3% to 86.7% within R. felis str. LSU- and R. felis str. LSU-Lb-exposed fleas, respectively.

The overall mean (±SEM) infection load per flea lysate was significantly higher in R. felis str. LSU-Lb-exposed fleas (6.15 × 106 ± 7.40 × 105), compared to fleas exposed to R. felis str. LSU (3.09 × 106 ± 4.80 × 105; Fig. 1A). The range of infection load for R. felis str. LSU was 2.65 × 102 to 1.73 × 107 rickettsiae per flea lysate. Similarly, the infection load for the R. felis str. LSU-Lb-exposed fleas ranged from 2.65 × 102 to 2.92 × 107 rickettsiae per flea lysate. When comparing the density of R. felis infection in fleas by calculating the mean (±SEM) Rf17kD/Cf18S ratio for each collection point, the overall mean ratio was significantly higher in R. felis str. LSU-Lb (0.76 ± 0.02) than in R. felis str. LSU (0.68 ± 0.02) exposed cat fleas (Fig. 1B). The ranges of the infection density were 0.29 to 0.93 and 0.29 to 0.96 in R. felis str. LSU and R. felis str. LSU-Lb-exposed fleas, respectively.

Assessing infection dynamics at weekly intervals over the course of infection, significant differences in the mean (±SEM) quantities of R. felis str. LSU and R. felis str. LSU-Lb at 21 and 28 dpe were observed, with a significant increase in the rickettsial load in R. felis str. LSU-Lb-exposed fleas after 14 dpe (Fig. 2A). These differences were not associated with sex of the host, except for males exposed to R. felis str. LSU-Lb at 28 dpe (data not shown). In contrast, the mean infection density of R. felis str. LSU and R. felis str. LSU-Lb did not significantly differ after 7 dpe (Fig. 2B). The increase in rickettsial density at the 7dpe time point is associated with a significant difference observed in female fleas exposed to R. felis str. LSU-Lb (data not shown).

Horizontal transmission of R. felis str. LSU-Lb to uninfected cat fleas occurred after 1 to 21 d of cofeeding with infected donor cat fleas with a total of 29.7% (19/64) of RB-labeled recipient cat fleas becoming positive for R. felis str. LSU-Lb. The total infection prevalence for R. felis str. LSU-Lb infected donor fleas (31/36) was significantly higher than that of recipient fleas (19/64) (Table 2). The mean (±SEM) rickettsial load for R. felis str. LSU-Lb-infected donor fleas (3.76 × 106 ± 1.69 × 106) was significantly higher than that of infected recipient fleas (1.17 × 103 ± 2.94 × 102). Likewise, the mean (±SEM) infection density (Rf17kDa/Cf18S ratio) for R. felis str. LSU-Lb infected donor fleas (.54 ± .04%) was significantly higher than that of infected recipient fleas (.34 ± .01%). The mean (±SEM) load of R. felis str. LSU-Lb infected donor fleas on day 14 (2.69 × 107 ± 1.03 × 107) was significantly higher than infected recipient cat fleas at all time points and donor fleas at days 1, 7, and 21 (Fig. 3A). The mean (±SEM) infection density of R. felis str. LSU-Lb infected donor fleas on day 7 (.82 ± .03) and day 14 (.88 ± .01) were significantly higher than R. felis str. LSU-Lb infected recipient fleas at all time points and other donor fleas on days 1 and 21 (Fig. 3B).

Eggs from cat fleas (n = 200) exposed to R. felis str. LSU and R. felis str. LSU Lb-infected bloodmeals were collected every 5 or 10 d post parental bloodmeal exposure, and reared to adults. The R. felis str. LSU-infected fleas produced 542 females and 512 males (n = 1,054). The R. felis str. LSU-Lb-infected fleas produced 416 females and 432 males (n = 848). The uninfected fleas produced 514 females and 502 males (n = 1,016). The difference between the F1 sex ratios in the R. felis str. LSU-infected fleas and the R. felis str. LSU-Lb-infected fleas was not significant. Additionally, neither of the F1 sex ratios in the R. felis str. LSU-infected fleas and the R. felis str. LSU-Lb-infected fleas were significantly different from the uninfected flea sex ratio (Table 3). Thus, R. felis str. LSU-Lb did not influence F1 sex determination as it does in L. bostrychophila (Yusuf and Turner 2004).

Total gDNA was then extracted from individual (LSU n = 80, LSU-Lb n = 79, Control n = 15) or pools of 10 (LSU n = 1,050, LSU-Lb n = 840, Control n = 30) F1 cat fleas and the presence of R. felis was assessed by qPCR. Despite R. felis str. LSU and R. felis str. LSU-Lb infections in adults, infection from either genotype was not detected in any F1 progeny.

Discussion

Several genotypes of R. felis have been isolated from colonized and wild-caught arthropods, with unique genotypes for specific host populations (Gillespie et al. 2015, Brown and Macaluso 2016). Aside from the number of plasmids, there is minimal genomic divergence between R. felis str. LSU (isolated from a cat flea colony) and R. felis str. LSU-Lb (isolated from a booklice colony; Gillespie et al. 2015). Because of the niche overlap of C. felis and L. bostrychophila, it is hypothesized that the host-specific strain of R. felis from the nonhematophagous booklouse arose via the lice feeding on infectious flea detritus (Gillespie et al. 2015). The capability of R. felis str. LSU-Lb to infect vertebrates is unknown, but rickettsial DNA has been detected in dust samples (containing L. bostrychophila and other dust-associated arthropods) collected from the beds of R. felis-infected patients in Senegal (Mediannikov et al. 2014). While not confirmed, it is postulated that these individuals may have acquired R. felis by inhalation or dermal inoculation of booklouse feces (Parola et al. 2015). Therefore, although R. felis str. LSU-Lb is considered an obligate mutualist of L. bostrychophila, certain qualities suggest that the booklouse strain has the potential to be a human pathogen. As such, it is critical to determine whether R. felis str. LSU-Lb can infect cat fleas, which would provide an additional route for human exposure. To determine whether vector competence varies as a function of the genetic characteristics of the infecting strain, we aimed to determine if cat fleas could acquire, maintain, and transmit R. felis str. LSU-Lb.

In principle, the normal feeding behavior of cat fleas may lead to the acquisition of any number of microorganisms present in the bloodstream of a vertebrate host (Mehlhorn 2012); however, the rapid turnover of midgut contents in actively feeding cat fleas generally disrupts progression to an established microbial infection (Bland and Hinnebusch 2016). For example, cat fleas are competent vectors for the plague bacterium (Yersinia pestis), but the frequency with which cat fleas feed results in bacterial clearance and disruption of biofilm accumulation (Bland and Hinnebusch 2016). As cat fleas are not associated with a diverse range of other rickettsial species, and transmission biology for R. felis genotypes in cat fleas has not been assessed, the present study demonstrated acquisition and maintenance of R. felis str. LSU-Lb by cat fleas after feeding on an infectious bloodmeal. Molecular analyses revealed that more than 75% of the cat fleas exposed to R. felis str. LSU-Lb were positive for rickettsial DNA after 1 wk, and more than 80% of the cat fleas were infected at 4 wk post exposure. Interestingly, the overall mean infection load and density, as well as infection prevalence, of R. felis str. LSU-Lb within cat fleas was significantly higher when compared to the flea-associated strain of R. felis. This observation perhaps highlights the underlying genetic factors that differentiate infecting strains of R. felis as facultative parasites of fleas from obligate mutualists of booklice. Indeed, short-term infection of cat fleas by either strain of R. felis demonstrates a distinct transcriptional profile compared to culture conditions, suggesting that vector adaptation occurs (Verhoeve et al. 2016). Although less studied in rickettsial species, the efficiency of vector infection is known to differ among viral strains within a genotype (e.g., dengue virus), which often results in variable transmission phenotypes (Armstrong and Rico-Hesse 2003). The selective factors that induce host-specific genetic variability of R. felis, including R. felis interaction with host microbiota, require further study.

Acquisition of a pathogen by an arthropod is not sufficient alone to confirm its transmission potential to other hosts. Given that rickettsial transmission by arthropods can be vertical or horizontal, bioassays were generated to examine the transmission of R. felis str. LSU-Lb from adult cat fleas to their progeny (vertical), as well as transmission between infected and uninfected cofeeding cat fleas (horizontal). Similar to recent experiments with R. felis str. LSU (Hirunkanokpun et al. 2011, Reif et al. 2011), vertical transmission of R. felis str. LSU-Lb was not observed in cat fleas, nor was there an impact of R. felis str. LSU-Lb infection on the development or reproductive fitness of cat fleas. Additional studies are required to characterize the dissemination of horizontally acquired R. felis str. LSU-Lb in cat fleas and other factors that might influence vertical transmission to progeny. Remarkably, using comparable experimental parameters as with R. felis str. LSU (Hirunkanokpun et al. 2011, Brown et al. 2015), horizontal transmission of R. felis str. LSU-Lb was observed between cofeeding cat fleas in an artificial host system. Rickettsial DNA was detected in both donor and recipient fleas 1-day post cofeeding on a shared bloodmeal, and transmission continued over the course of 21 d. Regardless of whether infection load was quantified as total rickettsiae per flea or as a ratio of R. felis to C. felis genes, rickettsial load was significantly decreased between donor and recipient fleas. In addition, infection prevalence was also significantly higher in donor fleas than in recipient fleas, similar to previous work with R. felis str. LSU (Hirunkanokpun et al. 2011, Brown et al. 2015). The transmission of R. felis str. LSU-Lb between cat fleas via cofeeding on vertebrate blood has broad implications toward the potential exposure risk for humans. The ability of fleas to transmit different genotypes may contribute to the high proportion of R. felis infections detected from patients and vertebrate hosts (Maina et al. 2016).

Although these R. felis isolates possess unique genotypes, there were no profound consequences for their subsequent infection of and transmission by the cat flea vector. Intriguingly, R. felis str. LSU-Lb actually infected cat fleas more efficiently despite adaptation in the booklouse host. Rickettsial strain variation in infection density is consistent with the previous observation of higher density of R. felis str. LSU-Lb, compared to strain LSU, in booklouse and flea hosts, respectively (Thepparit et al. 2011). However, the higher load did not result in a distinct flea biological phenotype, as seen with R. felis str. LSU-Lb infection of booklice. The results of the current study suggest that although genetic variability determines R. felis host specialization, cat fleas maintain a universal ability to acquire and transmit horizontally multiple genotypes of R. felis. Additional flea transmission studies examining novel R. felis isolates are needed to elucidate the role of other R. felis genotypes in the epidemiology of flea-borne spotted fever.