Association between body size and reservoir competence of mammals bearing Borrelia burgdorferi at an endemic site in the northeastern United States

Background The reservoirs for the Lyme disease agent, Borrelia burgdorferi, are dominated by several different small to medium sized mammals in eastern North America. Findings To experimentally assess the competence of different mammalian species to transmit this pathogen to ticks, we carried out quantitative species-specific PCR of individual nymphal Ixodes scapularis ticks, which had been collected as replete larvae from animals captured at a field site in eastern Connecticut and then allowed to molt in the laboratory. The mammals, in order of increasing body mass, were the white-footed mouse, pine vole, eastern chipmunk, gray squirrel, Virginia opossum, striped skunk, and common raccoon. The prevalence of infection in the nymphs and the counts of spirochetes in infected ticks allometrically scaled with body mass with exponents of −0.28 and −0.29, respectively. By species, the captured animals from the site differed significantly in the mean counts of spirochetes in the ticks recovered from them, but these associations could not be distinguished from an effect of body size per se. Conclusions These empirical findings as well as inferences from modeling suggest that small mammals on the basis of their sizes are more competent as reservoirs of B. burgdorferi in this environment than medium-to large-sized mammals.

Borrelia burgdorferi, the predominant agent of Lyme disease in North America, is a generalist microparasite and exploits several different species of mammals and birds as reservoirs [1]. Transmission between vertebrate hosts in the eastern and central United States and adjoining areas of Canada is accomplished by the tick Ixodes scapularis. As a step towards further defining relative contributions of different hosts in B. burgdorferi's natural cycle, we compared the competence of the whitefooted mouse Peromyscus leucopus and selected other mammalian species in transmitting the pathogen to ticks. We used quantitative PCR to identify and measure the burden of B. burgdorferi in the nymphs derived from larvae that had naturally infested and fed on different hosts captured at a field site, Lake Gaillard, in Connecticut [2]. Reservoir competence was defined as the proportion of molted nymphs bearing B. burgdorferi after having fed as larvae upon a B. burgdorferi-infected animal. Quantitation of the spirochetes in each tick provided an additional assessment of competence. Nymphal ticks may be infected, as revealed by qualitative PCR or immunofluorescence, but they may be inadequate as vectors if spirochete counts are low [3].
We noted that species and individual animals represented in this survey varied not only in the prevalence of infections in their ticks but also in the counts of spirochetes in the ticks. For instance, for the majority of infected P. leucopus every tick recovered from an animal had B. burgdorferi and the spirochete burdens in these mouse-derived ticks were 3-to 9-fold higher than the counts in the ticks from opossums, raccoons, squirrels, and the skunk (Table 1). Voles and chipmunks resembled the mice in having high average counts per infected tick but, like the four larger mammals, tended to have lower prevalences of infections among the nymphs than was observed with the mice. The observations suggested an association with size of the mammal. We examined this possible relationship using the midpoint of the ranges of body masses in grams given by the Smithsonian National Museum of Natural History (http://www.mnh.si.edu/mna/ main.cfm): vole (25.5), chipmunk (115), squirrel (544), opossum (2800), skunk (3250), and raccoon (6100). The value of 20 g for midpoint mass given for P. leucopus was same value as the mean of 298 adult P. leucopus captured at the field site in other collections (unpublished findings). These values closely corresponded with those provided by the AnAge database (http://genomics.senescence.info/ species).
Nymphal infection prevalence (i P ) decreased with body size (w): i P = 1.8w −0.29 ; R 2 = 0.43; F 1,29 = 21.5; p <0.0001. Spirochete counts (n) in individual infected ticks similarly decreased with w: n = 15,284w −0.28 ; R 2 = 0.37; F 1,29 = 17.2; p <0.0001. The product of infection prevalence and mean spirochete load normalized the counts across every tick collected from an individual animal of each of the 6 species (Table 1). The log-transformed normalized counts differed by species (ANOVA; R 2 = 0.66; F 6,24 = 7.9; p <0.0001). But the species association for this estimate of reservoir competence was not distinguishable from an equally strong association with body mass alone (Fig. 1).
LoGiudice et al. performed a similar study of mammals and their attached larval ticks, which were collected in New York [10]. B. burgdorferi and B. miyamotoi were detected by genus-specific antibodies and immunofluorescence and not quantitatively, but the study assessed the prevalence of infection in all the ticks obtained from each animal. Analysis of the summarized data shows a similar negative association between body size over several orders of magnitude and a measure the authors called reservoir competence, which effectively was nymphal infection prevalence (Fig. 2). The species represented were, with exception of voles, the same as those examined here, with the addition of the common shrew (Sorex cinereus), shorttailed shrew (Blarina brevicauda), and white-tailed deer (Odocoileus virginianus). In an earlier study, which used the same criterion of reservoir competence, P. leucopus mice had comparatively greater competence than the larger chipmunk in an experimental study of this species and chipmunks [11]. We do not discount the roles of demography or life history, nor various species-specific resistances and susceptibilities to infection [4], in accounting for the observations reported here and by others. But appreciative of the heuristic tool of Occam's Razor, i.e. among competing hypotheses, the one with the fewest assumptions should be selected, we propose consideration of an allometric explanation as well. One way in which body size may cash out in reservoir competence is in differing time periods for a vertebrate to achieve a state of infectiousness for ticks, namely, to reach a tissue or blood density above which feeding ticks would acquire the pathogen [1]. Noting that minimal infectious doses of B. burgdorferi appear to be 200 bacteria or fewer for dogs, laboratory rats, and mice [12][13][14], we posit that inocula required for acquiring this pathogen do not appreciably scale with body size of the host. With a discrete deterministic model we estimated the time (t) in days it took for a pathogen to reach a particular density (c) per gram of body mass when infections began with one organism. For this exercise, body size, w, was in grams and the pathogen's doubling rate, v, was in hours. Figure 3 shows the association between body size and time to infectiousness in this simple model for growth rates that ranged from a 6 h doubling time typical of in vitro cultivation of B. burgdorferi [15] to 1 of 12 h, which may more accurately represent proliferation in a naïve vertebrate host [16]. The target density was 10 5 spirochetes per gram of tissue, which was the median density of spirochetes in ear tissue in naturally-infected white-footed mice [5]. The body sizes were those for species of this study. Between the smallest and largest animals there was a difference of 1.5 to 4 days before the putative threshold for infectiousness in this model was reached, depending on the in vivo growth rate.
In a dynamical model of epidemics of De Leo and Dobson, transmission scaled allometrically with body size, increasing with size when transmission was densitydependent and decreasing with size for frequencydependent transmission, which more suitably applies to vector-borne infections [17]. See text for the description of the model. The body masses correspond to the masses of the species presented in Fig. 1 infection was also explored by Morand and Poulin [18] and by Bolzoni et al. [19]. Taken together with these other perspectives on the phenomenon, our findings prompt further study of the effect of body size on reservoir competence. A priority is on examining this relationship with other vector-borne pathogens and their natural hosts.