Frequent and seasonally variable sublethal anthrax infections are accompanied by short-lived immunity in an endemic system

Abstract

Few studies have examined host-pathogen interactions in wildlife from an immunological perspective, particularly in the context of seasonal and longitudinal dynamics. In addition, though most ecological immunology studies employ serological antibody assays, endpoint titre determination is usually based on subjective criteria and needs to be made more objective. Despite the fact that anthrax is an ancient and emerging zoonotic infectious disease found world-wide, its natural ecology is not well understood. In particular, little is known about the adaptive immune responses of wild herbivore hosts against Bacillus anthracis. Working in the natural anthrax system of Etosha National Park, Namibia, we collected 154 serum samples from plains zebra (Equus quagga), 21 from springbok (Antidorcas marsupialis) and 45 from African elephants (Loxodonta africana) over 2-3 years, resampling individuals when possible for seasonal and longitudinal comparisons. We used enzyme-linked immunosorbent assays to measure anti-anthrax antibody titres and developed three increasingly conservative models to determine endpoint titres with more rigourous, objective mensuration. Between 52 and 87% of zebra, 0-15% of springbok and 3-52% of elephants had measurable anti-anthrax antibody titres, depending on the model used. While the ability of elephants and springbok to mount anti-anthrax adaptive immune responses is still equivocal, our results indicate that zebra in ENP often survive sublethal anthrax infections, encounter most B. anthracis in the wet season and can partially booster their immunity to B. anthracis. Thus, rather than being solely a lethal disease, anthrax often occurs as a sublethal infection in some susceptible hosts. Though we found that adaptive immunity to anthrax wanes rapidly, subsequent and frequent sublethal B. anthracis infections cause maturation of anti-anthrax immunity. By triggering host immune responses, these common sublethal infections may act as immunomodulators and affect population dynamics through indirect immunological and co-infection effects. In addition, with our three endpoint titre models, we introduce more mensuration rigour into serological antibody assays, even under the often-restrictive conditions that come with adapting laboratory immunology methods to wild systems. With these methods, we identified significantly more zebras responding immunologically to anthrax than have previous studies using less comprehensive titre analyses.

Keywords: ELISA; antibody persistence; bacteria; ecological immunology; endoparasites; host‐parasite interactions; microbiology; microparasites; seroconversion dynamics; serology.

Figure 1

Etosha National Park in northern Namibia. Most animal sampling occurred in the plains within a 20-60km radius from the Etosha Ecological Institute. Perennial watering points are marked with black circles.

Figure 2

Total culture-confirmed anthrax cases for elephant, springbok, and plains zebra by month for 2008-2010 (years of this study). The graph insert indicates mean monthly rainfall ± standard deviation in the Okaukuejo region for this same time period.

Figure 3

Antibody endpoint titers to B. anthracis PA toxin in a) plains zebra, b) springbok, and c) elephant, determined by each of the three titer cutoff rules. Titers are shown as log₂ of the reciprocal of dilution cut-off for clarity.

Figure 4

Optical density titer cutoff rules by species. Colored lines represent the difference between the optical densities (OD) of a sample Etosha animal (averaged between duplicates) and the OD of the negative control sample (averaged between duplicates) run on the same plate. The thick black lines show the threshold cutoffs for these differences that determined the titer assigned to an animal. Each sample’s OD curve appears only once in this figure (i.e. there are 154 colored lines in the left column representing each unique zebra serum sample). Any animals whose OD curve was never greater than the negative control on its plate are considered seronegative by all three Rules and appear only in the top row. Animals are considered positive by Rules 1, 2, and 3 when their OD curve was greater than the negative control (dotted line), greater than the negative control plus the 1.96 times the standard deviation between duplicate samples (dashed line), or greater than the negative control plus 1.96 times the standard deviation amongst the negative control population (run on separate plates; solid line), respectively. Thus, animals that are positive by Rule 3, the most conservative criterion, appear only in the bottom (fourth) row. The third row shows the animals that are positive by both Rules 1-2 but not Rule 3 and, similarly, the second row shows animals that are positive by Rule 1 but not by Rules 2-3.

Figure 5

Change over time of zebra antibody titers to B. anthracis PA toxin. Titers are shown as the log₂ of the reciprocal of the dilution endpoint as determined by Rule 2. Only individuals sampled 4 or 5 times are shown, with each plot showing 4-5 different animals for 13 zebra total across the 3 plots. Sampling seasons 1, 2, 3, and 4 are roughly 6 months apart from each other, whereas S5 is approximately 12 months after S4. Overlapping portions of some plots have been jittered slightly for clarity.

Figure 6

Percent (± 95% CI) of zebra samples with positive antibody titers or no titers against B. anthracis PA toxin, compared between wet and dry seasons.