| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
 |
ABSTRACT |
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
|
INTRODUCTION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Our lack of knowledge of local plague transmission cycles is particularly acute in much of Africa, where the majority of human cases have occurred in recent decades.5,6 In many parts of East Africa, rat fleas (Xenopsylla cheopis, X. brasiliensis) are believed to play a crucial role in plague epizootics because they commonly infest susceptible rodent hosts,1,7,8 readily feed on humans when encountered,9–11 and laboratory studies have demonstrated that they are efficient vectors of Y. pestis.1,2,4,12–15 However, in some regions, other fleas are also commonly encountered in huts. For example, recent studies have indicated that Pulex irritans is commonly found in human habitations in Tanzania.16 In northwestern Uganda, the cat flea (Ctenocephalies felis) has been reported as the most common species in human habitations. This last flea species is typically perceived as a nuisance biter, rather than as an insect of public health importance, despite limited studies indicating that it can transmit plague bacteria under some circumstances.1,7,17
Although there have been surprisingly few empirical evaluations of its vector competency,11,18,19 C. felis is typically referred to as incapable of transmitting Y. pestis or as an inefficient vector.1,3,20,21 However, this flea species is a catholic feeder and will readily bite humans and rats.10,11 It therefore could serve as a bridging vector between zoonotic and anthroponotic cycles. High flea densities and frequent feeding could compensate for a low transmission efficiency rate and result in high vectorial capacity for C. felis22–26; thus, cat fleas may pose a greater risk to public health than previously believed.
In the West Nile region of Uganda, Arua and Nebbi Districts have consistently reported human plague cases. From 1999 through 2007, for example, clinics from these districts reported to the Ugandan Ministry of Health a mean of approximately 223 (range: 76–467) suspect human cases per year (CDC, unpublished data). In the field component of this study, we demonstrate that C. felis is the most common flea in human habitations in this plague-endemic region, and that it occasionally infests potential rodent reservoirs of Y. pestis. Our laboratory study shows that C. felis is a competent vector of plague bacteria, but that efficiency is low compared with another flea species collected in the same area, Xenopsylla cheopis. We introduce a preliminary model of vectorial capacity for C. felis, which estimates that the number of cat fleas per host required to maintain Y. pestis in human habitations is within the range of field-derived values.
|
MATERIALS AND METHODS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
). These trap locations are later referred to as "sylvatic." These areas represent a mixture of agricultural plots, fallow fields, and natural vegetation.
|
Field evaluation of flea infestations in human habitations. In the villages sampled in Arua and Nebbi districts, a typical residence consists of a mud and brick structure with a grass thatched roof and earthen floors. Within each of the 10 residences in villages selected for small mammal trapping, two flea traps placed on the floor were set for two nights per trapping session. The traps were a modification of the Kilonzo flea trap31 for collection of photosensitive fleas. One trap consisted of a flashlight suspended over a metal pan (25.4 cm diameter) containing 2% saline with Tween 80. The other trap targeted non-photosensitive fleas and consisted of a metal pan containing 2% saline with Tween 80. On both types of traps, Vaseline was applied to the rim of the pan to prevent fleas from exiting. Village residents were instructed to turn the flashlights on at night and to leave on throughout the night. Fresh batteries were used for each night of collection. After two days, fleas were collected from the pans and stored in microcentrifuge tubes containing 2% saline with Tween 80 and later identified to species following published taxonomic keys.17,28,29 In Nebbi district, five villages were sampled on two occasions (August 2006 and February–March 2007) and in Arua district four villages were sampled on two occasions (December 2006 and March 2007).
Laboratory evaluation of flea-borne transmission of Y. pestis. The ability of C. felis to transmit Y. pestis 1–4 d p.i. (early-phase transmission) was tested following methods published previously.14,22,32 Oropsylla montana, a species evaluated previously,22 was used as a positive control for flea-borne transmission efficiency. This species was selected because it has demonstrated the highest transmission efficiency 1–4 d p.i. of all species examined in our laboratory. Methods for infecting C. felis and O. montana with Y. pestis (CO96-3188), confirming transmission from fleas to naive Swiss Webster mice, quantification of bacterial loads in fleas and evaluation of vector competency 1–4 d post infection (p.i.) were described in detail previously.22 The strain was selected so vector-efficiency could be directly compared among flea species that were evaluated previously14,22,32–34 and infected with the same bacterial strain. Briefly, on day 0 four batches of colony-reared female C. felis obtained from a commercial vendor (Heska, Fort Collins, CO) and one batch of O. montana from a colony maintained by the Division of Vector-Borne Infectious Diseases, Centers for Disease Control and Prevention (N = 50–60 fleas per feeder) were allowed to feed for 1 h on artificial feeders containing defibrinated Sprague-Dawley strain rat blood (Bioreclamation, Jericho, NY) heated to 37°C and containing a fully virulent North American strain of Y. pestis, biovar orientalis, designated CO96-3188 (1.09 x 109 –2.61 x 109 cfu/mL). Aliquots of bacteria used in the present study came from the same source population as those used previously.14,22,32–34 Similarly, the O. montana controls were derived from the same colony as those used previously.22,32 Fed fleas containing red blood in the proventriculus or midgut were differentiated from unfed fleas using light microscopy.22 Any flea that did not take a potentially infectious bloodmeal was discarded; the remaining fleas were held at 23°C and 85% relative humidity for 1–4 d p.i. On each of the days 1–4 p.i., pools of approximately 10 potentially infectious C. felis were placed in feeding capsules for 1 h on anesthetized 6-wk-old Swiss-Webster strain mice. On day 1 p.i. only, the procedure was replicated with pools of 10 O. montana. After 1 h, fleas were removed from the feeding capsule using a mechanical aspirator. Flea feeding success was determined using light microscopy22; fed fleas were stored in individual microcentrifuge tubes at –80°C until infection status and bacterial loads were determined by serial dilutions of fleas triturated in heart infusion broth supplemented with 10% glycerol and plated in duplicate on blood agar plates containing 6% sheep blood.22 Following exposure to fleas, recipient mice were housed individually in filter-top cages, monitored daily and euthanized when signs of Y. pestis exposure were evident (e.g., slow response to stimuli, shivering, ruffled fur). Transmission from flea to mouse was demonstrated if the exposed mouse showed clinical symptoms of plague followed by presumptive identification of Y. pestis in the liver or spleen by direct fluorescence antibody test targeting the F1 antigen and was confirmed by culture isolation and subsequent bacteriophage lysis. Alternatively, mice surviving to 21 d p.i. were euthanized and blood was collected for testing for serological evidence of exposure to Y. pestis using passive hemagluttination and inhibition tests.30 Animal procedures were approved by the Division of Vector-Borne Infectious Diseases (Centers for Disease Control and Prevention) Institutional Animal Care and Use Committee.
Estimate of vectorial capacity. The density of fleas in relation to humans that are required for person-to-person transmission of Y. pestis was determined following recently published studies of vectorial capacity.14,22,25 The model assumes that 1) host density is sufficiently high for every infected flea to find a susceptible host upon which to feed, and 2) the first bite following infection results in transmission. The resulting model predicts the number of fleas required per host to maintain an infection in a population, as:
 |
where R0 represents vectorial capacity or force of infection (e.g., the number of secondary infections arising from a focal infection). When person-to-person transmission of Y. pestis by cat fleas could be maintained, R0 = 1. The flea density per person, m, represents the infestation in human habitations. The daily biting rate of infected fleas, a was set at 0.92 following previously published studies.22 The probability of a flea acquiring an infection after feeding on a septicemic host and transmitting the infection during a subsequent feeding on a susceptible host, b, was derived from the range in vector efficiency yielded from our laboratory study. The probability of the flea surviving the extrinsic incubation period (defined as the duration of time from which a flea is infected until it can transmit) is symbolized as pn. This was estimated to be 1.0 because the extrinsic incubation period is believed to be very short, allowing all or nearly all fleas to survive to become infectious. Finally, 1/r defines the life expectancy of the host after it reaches threshold septicemia and following previous studies was set as equaling 2.22,25
Statistical analysis. For C. felis tested during each time point (1–4 d p.i.), transmission efficiency per individual flea was estimated using maximum likelihood based on the number of infected fleas that fed on an individual mouse and whether transmission was observed for that recipient mouse using the Microsoft® Excel® Add-In PooledInfRate, Version 3.0.14 Mean maximum bacterial loads for infected fleas fed per animal at each time point were compared using ANOVA. Kruskal-Wallis and Wilcoxon rank sums tests were used to compare median numbers of colony-forming units per flea among and between treatments, respectively. Wilcoxon signed rank tests were used to compare the percentage of fleas classified as C. felis between trapping sessions. All statistical comparisons were run using JMP statistical software (SAS Institute, Cary, NC) and results were considered significant if P < 0.05.
|
RESULTS |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
).
|
).
|
). Significantly more C. felis were collected from Nebbi than Arua district (Wilcoxon rank sum test with 
2 approximation; 2 = 9.4, d.f. = 1, P = 0.0021). In Arua, the remaining 54 fleas were classified as Ctenophthalmus spp. (N = 2), Dinopsyllus spp. (N = 4), Pulex irritans (N = 2), Tunga penetrans (N = 5), and X. cheopis (N = 41). In Nebbi, additional fleas collected in pan traps included Ctenocephalides canis (N = 6), Echidnophaga gallinacea (N = 6), T. penetrans (N = 12), and X. brasiliensis (N = 17). In Arua district, the proportion of fleas classified as C. felis was significantly lower in session 1 compared with session 2 (one-tailed Wilcoxon signed rank test; P = 0.03). No differences were detected between trapping sessions in Nebbi district.
|
). Transmission of Y. pestis to mice was confirmed by recovery of plague bacteria from the liver and spleen of a mouse that showed clinical symptoms of infection. On that day (day 2 p.i.), transmission efficiency was estimated to be 2.15% (95% CI: 0.13–10.64). Because transmission efficiency was similar from 1–4 d p.i. (Table 4), we estimated the per flea transmission efficiency for the entire early-phase observation period using maximum likelihood. This yielded an estimate for transmission efficiency of 0.57% (95% CI: 0.03–2.78%). Demonstrating the ability of this strain of Y. pestis to be transmitted from flea to mouse, each of the five O. montana flea pools infected with plague bacteria from the same aliquot of CO96-3188 transmitted Y. pestis to susceptible mice (Table 4).
|
). Mean maximum bacterial loads for O. montana harvested 24 h p.i. were similar to those in C. felis for 48 to 96 h p.i., but were significantly higher than for C. felis at 24 h p.i. (F = 9.08, d.f. = 4, 14, P = 0.0004).
|
2 = 68.5, d.f. = 4, P < 0.0001; Table 4
). Bacterial loads for C. felis were lower at 24 h p.i. compared with all other time points (2 = 18.12, d.f. = 1, P < 0.0001), but similar among each of the other time points. Compared with O. montana, bacterial loads were lower in C. felis for each of the different time points (2 > 7.88 d.f. = 1, P < 0.005).
Estimate of vectorial capacity.
We sought to determine if flea loads observed in human habitations would be sufficient to sustain person to person transmission if a single individual in the hut became infectious. Using the range in transmission rates estimated in our laboratory study (Table 4), we determined that to maintain person to person transmission of Y. pestis, an average of 25 cat fleas per person would be required (95% CI: 5–418). When vectorial capacity is based on the pooled early-phase transmission estimate (days 1–4 p.i. combined where transmission efficiency was 0.57% (95% CI: 0.03–2.78%)), the average number of fleas per host required for a single focal infection to give rise to a secondary infection was estimated to be approximately 95 fleas per person (95% CI: 20–1,811).
|
DISCUSSION |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
Previous studies have speculated that P. irritans, a human-biting flea that is commonly found in human habitations in Tanzania, Kenya and the Democratic Republic of Congo, may serve as a vector in anthroponotic cycles of Y. pestis in Africa and elsewhere.16,35–37 It is possible that C. felis, which replaces P. irritans as the most common flea in human habitations in Uganda,17,35 plays a similar role in Y. pestis transmission. A high proportion of human habitations in the Arua and Nebbi districts were infested with host-seeking fleas and the overwhelming majority of those fleas were C. felis. Although pan trapping gave us an estimate of the relative abundance of fleas in human habitations among villages, we are uncertain what proportion of the population is captured by this method and therefore the abundance values reported in this study represent a minimum estimate of flea abundance. Interestingly, our estimates of the number of cat fleas per person required for maintenance of Y. pestis by person-to-person transmission (approximately 25 fleas per person) was within the range observed in our study. Among sampled huts that yielded fleas, the number of C. felis ranged from 1 to 62 cat fleas for a single sampling occasion. To more precisely estimate the density of fleas required for maintenance of the pathogen in human habitations, field-derived values of daily biting rates, host preference, and daily survivorship are needed. Our data revealed a great deal of variability among huts and among villages in the abundances of fleas captured. Perhaps understanding the underlying mechanisms explaining flea abundance could be used to make recommendations on how to reduce flea numbers and disrupt transmission in the home environment.
Cat fleas were rarely collected from rodents susceptible to plague infection, such as R. rattus and A. niloticus, and transmission efficiency for Y. pestis was very low for C. felis. Thus, it is unlikely that this flea alone could sustain enzootic transmission among rodents, and it is an improbable bridging vector to humans. However, it is quite likely that X. cheopis could serve as a bridging vector to humans because of its high efficiency of transmission and willingness to bite humans when its typical rodent hosts perish from plague infection.1,2,4,9,12–15,29,38 Furthermore, because of the high abundance of rats and their heavy infestations with X. cheopis, rat fleas are probably contributing more strongly than cat fleas to transmission in peridomestic settings. Previous studies have estimated that to sustain enzootic transmission of Y. pestis by X. cheopis, an average of 4.7–7.8 fleas per host are needed.14,25 In Arua district, R. rattus harbored an average of 8.3 X. cheopis in sylvatic areas, a quantity theoretically high enough to sustain transmission in an X. cheopis-R. rattus cycle. In peridomestic areas, X. cheopis infestation levels were slightly above one25 and slightly below another14 theoretical estimate of the threshold infestation needed for enzootic transmission. Because several small mammal species that are commonly infested with X. cheopis were collected in sylvatic and peridomestic areas (Tables 1 and 2), it is possible that X. cheopis could serve as a bridging species between sylvatic and peridomestic cycles in the Arua district. In the Nebbi district, X. cheopis was often replaced by another competent vector of Y. pestis, X. brasiliensis,1,2 and probably plays a similar role in that district. A future study will evaluate ecological differences between these districts that could account for the observed distribution.
Future studies are needed to elucidate the roles of X. cheopis, X. brasiliensis, and C. felis in the transmission of plague bacteria in Arua and Nebbi districts. First, our field collection was conducted during an inter-epizootic period. It is possible that flea loads and the composition of flea species on hosts differ between epizootic and inter-epizootic periods. Thus, a similar study conducted during a plague epizootic is needed. Second, the abundance of host-seeking fleas in human habitations may increase during epizootics and X. cheopis and X. brasiliensis could be more common in these areas when their rodent hosts succumb to plague infection. Interestingly, the proportion of fleas captured in pan traps in Arua district that were classified as C. felis was significantly lower during the first trapping session, which represented the peak of the plague season than the second session when few human cases are typically reported. The majority of non-cat flea captures included X. cheopis. These data show that flea species composition is not temporally static and comparisons between epizootic and inter-epizootic periods are necessary. Although C. felis is the most common flea in human habitations, it is unclear how frequently they feed on humans in this setting. Molecular identification of blood meals of fleas, similar to those produced for ticks and mosquitoes,39–44 would be useful for assessing how commonly cat fleas bite humans compared with other potential hosts such as rodents in the homes or livestock in peridomestic settings. Third, we made the assumption that most human exposures occur in peridomestic areas, which is reportedly the situation in many of the world’s plague foci, but this has not been confirmed by epidemiological studies for our study sites in Uganda. Finally, our vector efficiency study focused on C. felis felis, whereas C. felis strongylus is the most common subspecies in East Africa.17 It is possible that vector competency could differ between these subspecies. On the other hand, a molecular study questioned the validity of a subspecies characterization,45 raising the question of how dissimilar they are biologically. Regardless of taxonomic classification, use of fleas and bacteria from different geographical regions could result in differing vector efficiency outcomes. Therefore, future studies using fleas and Y. pestis strains originating from Uganda are necessary.
Infrequent transmission of Y. pestis by C. felis compared with X. cheopis 1–4 d p.i. (Table 4) sheds light on possible mechanisms of early-phase transmission. In 1914, Bacot and Martin12 described a flea-borne transmission mechanism that has been the dominant paradigm for nearly a century. Under this scenario, the plague bacilli multiply and form a blockage in the proventriculus of the infected flea. This blockage prevents newly ingested blood from reaching the midgut, causing the flea to starve. As a result, the flea increases its feeding attempts and ingested blood combined with infectious material cleaved from the blockage are occasionally regurgitated into the bite site resulting in transmission. Although Bacot and Martin12 proposed that a partially blocked flea may be even more efficient as a vector, the blocked-flea scenario has dominated the plague literature. For example, vector efficiency is often equated with a flea’s ability to block.1,3,4,16,21,25,46 However, recent studies using aliquots from the same stock of Y. pestis used in the present study have demonstrated high vector efficiency by unblocked fleas of all species tested to date using the experimental system described in this study (i.e., X. cheopis, O. montana, and O. tuberculata cynomuris, O. hirsuta.14,32–34,47). Among these studies, bacterial loads were similarly as high as those observed in C. felis in the present study. As suggested previously,14,32,33 bacterial loads in fleas are not predictive of a flea’s ability to transmit plague bacteria. Unlike other species evaluated 1–4 d p.i., C. felis appeared to have a higher digestive rate, as indicated by copious excretion of partially digested blood while feeding, and other researchers have noted that this species commonly defecates large quantities of partially digested blood on the host while feeding.9 It is possible that more rapid digestion or increased peristalsis could force bacteria further posterior in the digestive tract, making transmission by regurgitation less probable than when it is in a more anterior position.32 Interestingly, bacterial loads were significantly lower in cat fleas 24 h p.i. compared with all other time points. In addition, at 24 h p.i., bacterial loads were significantly lower in C. felis than O. montana fed the same source of infectious blood. These observations suggest that bacteria may have been eliminated in the feces within 24 h of feeding, followed by replication of the bacteria that remained in the digestive tract. Although useful in providing data contributing to a continuum of vector efficiency that could be used in comparative physiological studies, the mechanism of early-phase transmission remains the focus of future work.
Our study suggests that plague control programs should continue to focus on reducing the abundance of rat fleas (X. cheopis and X. brasiliensis). However, because of their abundance in human habitations, catholic feeding habits, and ability to transmit Y. pestis, cat fleas could serve as secondary vectors and should not be ignored. In the United States, plague prevention and control focuses primarily on reducing rodent abundance through elimination of harborage (e.g., piles of wood, brush, or debris), food sources (e.g., pet food, garbage), and access to human dwellings.48–50 Due to differences in housing structure and cultural practices, this is less feasible in rural Uganda, where rodent-targeted vector control or indoor residual spraying may be more effective.50 Initiatives aimed at reducing the incidence of malaria, such as use of insecticide-treated bednets and indoor residual spraying, may also be effective in reducing human contact with fleas in human habitations, providing flea vectors do not develop resistance to the given insecticides. In some areas where malaria control programs were implemented in the absence of modifying housing conditions the number of human plague cases have been reduced.51 Integrated vector control programs aimed at reducing transmission of multiple vector-borne pathogens may be cost effective, but the efficacy of such a strategy for reducing the incidence of plague requires further investigation and will require the kind of knowledge provided by our study.
Received January 22, 2008. Accepted for publication March 20, 2008.
Acknowledgments: The authors thank L. Eisen for comments on the manuscript and P. Collins, C. Williams, A. Ogen, N. Owor for logistical support. The authors are grateful to Asaph Ogen-Odoi, plague project manager, Uganda Virus Research Institute, who worked tirelessly on this project and passed away on December 14, 2006, while conducting plague field work in Arua.
* Address correspondence to Rebecca J. Eisen, Division of Vector-Borne Infectious Diseases, NCID/CDC 3150 Rampart Rd., Fort Collins, CO 80522. E-mail: dyn2{at}cdc.gov 
Authors’ addresses: Rebecca J. Eisen, Jeff N. Borchert, Jennifer L. Holmes, Kristen Van Wyk, Russell E. Enscore, Aryn P. Wilder, Sara M. Vetter, Scott W. Bearden, John A. Montenieri, and Kenneth L. Gage, Bacterial Diseases Branch, Division of Vector-Borne Infectious Diseases NCID/CDC 3150 Rampart Rd., Fort Collins, CO 80522, Tel: 970-266-3523, Fax: 970-225-4257, E-mail: dyn2{at}cdc.gov. Gerald Amatre, Nackson Babi, and Linda A. Atiku, Uganda Virus Research Institute, P.O. Box 49, Entebbe Uganda, Tel: 256-041320385, E-mail: plague{at}ug.cdc.gov.
|
REFERENCES |
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  G. Amatre, N. Babi, R. E. Enscore, A. Ogen-Odoi, L. A. Atiku, A. Akol, K. L. Gage, and R. J. Eisen Flea Diversity and Infestation Prevalence on Rodents in a Plague-Endemic Region of Uganda Am J Trop Med Hyg, October 1, 2009; 81(4): 718 - 724. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. E. Woods, J. A. Montenieri, R. J. Eisen, N. S. Zeidner, J. N. Borchert, A. Laudisoit, N. Babi, L. A. Atiku, R. E. Enscore, and K. L. Gage Identification of Flea Blood Meals Using Multiplexed Real-Time Polymerase Chain Reaction Targeting Mitochondrial Gene Fragments Am J Trop Med Hyg, June 1, 2009; 80(6): 998 - 1003. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
A. M. Winters, J. E. Staples, A. Ogen-Odoi, P. S. Mead, K. Griffith, N. Owor, N. Babi, R. E. Enscore, L. Eisen, K. L. Gage, et al. Spatial Risk Models for Human Plague in the West Nile Region of Uganda Am J Trop Med Hyg, June 1, 2009; 80(6): 1014 - 1022. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 F. Sebbane, C. Jarrett, D. Gardner, D. Long, and B. J. Hinnebusch The Yersinia pestis caf1M1A1 Fimbrial Capsule Operon Promotes Transmission by Flea Bite in a Mouse Model of Bubonic Plague Infect. Immun., March 1, 2009; 77(3): 1222 - 1229. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
