Survival for Immunity: The Price of Immune System Activation for Bumblebee Workers

See allHide authors and affiliations

Science  10 Nov 2000:
Vol. 290, Issue 5494, pp. 1166-1168
DOI: 10.1126/science.290.5494.1166


Parasites do not always harm their hosts because the immune system keeps an infection at bay. Ironically, the cost of using immune defenses could itself reduce host fitness. This indirect cost of parasitism is often not visible because of compensatory resource intake. Here, workers of the bumblebee, Bombus terrestris, were challenged with lipopolysaccharides and micro–latex beads to induce their immune system under starvation (i.e., not allowing compensatory intake). Compared with controls, survival of induced workers was significantly reduced (by 50 to 70%).

Parasitic infections are pervasive, but hosts often show no obvious effects. Alas, this does not mean that parasites impose no fitness costs on their host, because the immune system is often able to keep the infection within bounds. Recent discussions in the field of evolutionary ecology have concentrated on the idea that the evolution of the immune system is traded off against other fitness components (1). In addition, the activation and use of the immune system are thought to be costly and therefore cannot be sustained simultaneously with other demanding activities (2). With such costs, the main effect of infection is not the direct damage by the parasite itself but the cost imposed when the host immune system is activated. Why, if these costs exist, are they not more often evident? One reason is that hosts may compensate for increased demand by increased resource intake. Costs are thus masked and no outward signs of a parasitic infection are observed, although the host pays a cost to prevent the establishment and spread of the parasite. To date, such fitness costs have only been shown indirectly, for example, by forcing the individual to increase its parental effort and measuring the corresponding decrease in the immune response (2).

Here, the survival cost for the activation of the immune system was analyzed when the host was denied compensation for increased demand. In particular, the host's condition was experimentally “frozen” by adopting a starvation protocol at the point when an “infection”—a standardized immunogenic challenge—occurred (3). When an individual is starved, any future allocation to defense reduces the resources available for other needs and thus eventually for maintenance and survival. The starvation paradigm also mimics some important ecological conditions, such as the natural occurrence of adverse weather and limited food availability, that are typical for most animal populations.

In this study, workers of the bumblebee, Bombus terrestrisL., were used as hosts. Bumblebees are primitively eusocial insects inhabiting temperate habitats where weather conditions often vary over short time periods. Foraging activity is often interrupted by spells of rain and cold weather, leading to the starvation and demise of the colony if workers fail to collect sufficient amounts of pollen and nectar (4). Starved workers cannot survive for long (20 to 30 hours). In field populations, most workers are infected by some parasite but nevertheless show normal behaviors and activities (5). Bumblebee workers usually do not reproduce themselves. Hence, worker (inclusive) fitness is determined by their survival, and therefore any cost of immunity that reduces survival also reduces fitness (6). As in other insects, immunity in bumblebees is innate and based on both cellular (7) and humoral mechanisms (8). An immune response starts with the recognition of immunogens released by or present on the surface of parasites entering the host hemocoel. Various pathways of the immune system then become activated (9), leading to the destruction of the parasite and its removal by cellular reactions such as phagocytosis or encapsulation.

To measure the survival cost of the immune reaction, we experimentally activated the worker's immune system with two kinds of established immune elicitors. (i) Lipopolysaccharides (LPS; Sigma L-2755), i.e., surface molecules extracted from Escherichia coli. This nonpathogenic and nonliving elicitor is specifically recognized by pattern recognition proteins of the invertebrate immune system (9). LPS induces several pathways of the immune response (8, 10) that persist over many hours (11). LPS is cleared from insect hemolymph by lipophorin, a transport protein that shuffles LPS to the fat body (11,12). Hence, the clearance of LPS should not involve processes that are responsible for clearing bacteria from the hemolymph such as phagocytosis. (ii) Sterile micro–latex beads (Polysciences; diameter 4.5 μm). These beads are a similar size to bacteria and are cleared from the hemolymph by a combination of processes (8, 10), including phagocytosis. In both cases, the immune system is activated, but the artificial “parasite” is unable to generate any pathogenic effect.

A first experiment tested whether decreased survival might result from a toxic side effect of the immune elicitors, assuming that such effects would also decrease the survival of nonstarved animals. In addition, survival should then correlate with dose (13). Survival of nonstarved workers for 72 hours after injection did not depend on dose of LPS [mean survival of workers in relation to dose:r = −0.312; F(1,6) = 0.65,P = 0.45, N = 235 workers] or on numbers of micro–latex beads [r = 0.410;F(1,5) = 1.02, P = 0.36,N = 202 workers] and, in particular, was not different from noninduced control animals (13). In addition, survival rate never dropped below 90%. Hence, neither LPS nor micro–latex beads appear to exert any toxic side effect, even when concentrations are 10-fold (LPS) or 100-fold (beads) as high as in the subsequent experiments.

Activation of the immune system by LPS and latex beads was measured by the antibacterial activity of hemolymph with a zone-inhibition assay (14). Antibacterial activity increased with dose of LPS [mean activity in relation to dose: r = 0.895;F(1,6) = 24.25, P = 0.003,N = 216 workers]. Induction, but no relation with dose [r = 0.144; F(1,5) = 0.11,P = 0.76, N = 188 workers], was found for latex beads. LPS and micro–latex beads may thus induce different pathways of the immune system.

The survival of challenged and control bees under starvation was tested in the same way. Two LPS treatments (low-LPS and high-LPS) and one bead treatment (beads) were used, in addition to a combined challenge of micro–latex beads and LPS (15). All inferences were made with Cox regression and with respect to the survival observed under the control (16). Three hundred workers were used in the analysis. As before, the activation of the immune system was checked by the zone-inhibition assay. The induction lasted as long as the experiment. Not surprisingly, starvation decreased survival time to a few hours (mean: 20.8 ± 0.67 hours, SE, n = 51, for control bees). However, survival time was reduced, by factors of 1.5 to 1.7 (odds ratios), for workers that were challenged by LPS or beads (Fig. 1). There was no significant difference in survival between high-LPS and low-LPS treatments. The addition of beads to LPS had an additive survival cost compared with LPS and beads alone (Fig. 1) (no significant two-way interaction terms in the Cox regression were found). Given the experimental paradigm and the biology of the species, the differences among colonies probably reflect genotypic differences among colonies.

Figure 1

Survival of workers under different treatments. The top, nearly horizontal lines refer to nonstarved animals (no difference found among treatments). Low-LPS + beads and high-LPS + beads refer to the combined injection of LPS and beads. Cox regression analysis for the starved animals (sloping lines) shows that injection of beads reduces survival by a factor of 1.56 (odds ratio; Wald statistic = 13.58, df = 1,P < 0.001). Similarly, injecting a low dose of LPS (odds ratio = 1.73; Wald = 13.18, df = 1, P < 0.001) or a high dose of LPS (odds ratio = 1.75; Wald = 13.86, df = 1, P < 0.001) reduces survival, but both the two doses produce a similar effect (comparing low-LPS with high-LPS: P = 0.85). Survival was also affected by colony of origin (P < 0.001) and its interaction with injection of beads (P < 0.001) and LPS (P = 0.002) and the corresponding three-way interaction (i.e., colony, beads, LPS: P < 0.001). All other terms were nonsignificant; in particular, there was no two-way interaction of beads and LPS [for further statistical details, see supplementary data available at Science Online (18)].

Here, a hidden survival cost is demonstrated that is continuously paid to keep infections in check. It will go unnoticed when enough resources are available to compensate. Hence, even when no externally visible infections are observed in natural populations (17), such hidden costs can still have major effects on host survival strategies and ramify into many aspects of population biology and host-parasite coevolution. Future studies must therefore clarify to what extent resource allocation to defense is varied and whether such variation is in agreement with predictions of life history theory. Our results also suggest that the relative efficiencies of innate immune response in insects and acquired immunity in vertebrates (2) may be similar when costs are taken into account.

  • * To whom correspondence should be addressed. E-mail: moret{at}


View Abstract

Navigate This Article