No evidence of metabolic costs following adaptive immune activation or reactivation in house sparrows

Abstract

The energy cost of adaptive immune activation in endotherms is typically quantified from changes in resting metabolic rate following exposure to a novel antigen. An implicit assumption of this technique is that all variation in energy costs following antigenic challenge is due solely to adaptive immunity, while ignoring potential changes in the energy demands of ongoing bodily functions. We critically assess this assumption by measuring both basal metabolic rate (BMR) and exercise-induced maximal metabolic rate (MMR) in house sparrows before and after the primary and two subsequent vaccinations with either saline (sham) or two novel antigens (keyhole limpet haemocyanin and sheep red blood cells; KLH and SRBC, respectively). We also examined the effect of inducing male breeding levels of testosterone (T) on immune responses and their metabolic costs in both males and females. Although there was a moderate decrease in KLH antibody formation in T-treated birds, there was no effect of T on BMR, MMR or immunity to SRBC. There was no effect of vaccination on BMR but, surprisingly, all vaccinated birds maintained MMR better than sham-treated birds as the experiment progressed. Our findings caution against emphasizing energy costs or nutrient diversion as being responsible for reported fitness reductions following activation of adaptive immunity.

1. Introduction

The adaptive immune system of vertebrates is recognized as a major evolutionary advance in combating pathogenic infections [1]. The benefits of adaptive immunity are well established, but attempts to identify and measure the costs of its activation are far fewer and the results equivocal, with evidence both for [2–5] and against [6,7] reductions in fitness measures following adaptive immunity activation in endotherms. Reductions in fitness are typically ascribed to energy costs or nutrient diversion following immune activation, yet estimates of energy costs of peak antibody production following antigen exposure range from a 27% elevation in metabolic rate to a reduction of 13% [8,9]. While variations in protocols for measuring metabolic rates as well as the type, amount and number of antigenic challenges contribute to this disparity, changes in the energy demands of ongoing bodily functions may also occur. Evidence of rapid organ mass variation in rodents following novel antigen inoculation [10–12] contests the assumption that the energy demands of immune activation can be deduced from changes in basal metabolic rate (BMR) alone.

Unlike BMR, which reflects an endotherm's minimal or maintenance energy costs [13], an individual's exercise-induced maximum metabolic rate (MMR) requires the maximal performance of a suite of enzymes, organelles, cells, tissues, organs and organ systems [14], with redirection of resources from any of these organizational levels compromising attainment of maximal aerobic exertion. Thus, variation in MMR following immune activation can be assumed to reflect redistribution of energy and nutrient supply, as well as revealing its functional consequences.

Assessing the effects of antigenic challenges on MMR also permits testing assumptions underlying the provocative immunocompetence handicap hypothesis (ICHH; [15]). According to the ICHH, testosterone (T) suppresses immune responses in vertebrate males, yet females select mates based on the quality of their T-dependent sexual signals. Thus, males able to maintain superior condition and performance while experiencing a reduction in immunocompetence are likely to be those of superior genetic quality. A 2004 meta-analysis concluded that available data did not support the predictions of the ICHH [16], but the incorporation of more recent results from a broader range of studies found the reverse [17]. Nevertheless, the hypothesis does not consider potential fitness advantages that males might derive by reducing immunity-imposed demands for energy and nutrients during breeding.

Physical activities relying on aerobic fitness (e.g. repetitive songs and displays) are highly indicative of body condition and are the predominant male traits selected by females [18]. Thus, if T-induced immunosuppression improves availability of energy and nutrients for aerobically demanding activities in breeding males, then the presumed handicap imposed by elevated T would have to be reassessed. We experimentally tested this possibility by comparing MMR and BMR of male house sparrows before and after serial antigenic challenges, in T-treated and control birds. An equal number of females were used in this experiment to identify potential sex-related differences in metabolic responses to these treatments.

2. Material and methods

(a) . Experimental animals

We captured non-breeding, free-living house sparrows (Passer domesticus) during late winter in the Illawarra region of NSW, Australia. The 31 males and 31 females were distributed as mixed-sex flocks among four outdoor flight cages (4.5 × 3.6 × 2.5 m) exposed to ambient conditions. Birds had free access to commercial finch seed mix (Golden Cob, Mars Birdcare Australia), mineralized grit and water.

(b) . Experimental protocol

All male birds were bilaterally castrated, under anaesthetic, 1 or 2 days after capture and returned to the flight aviaries. After birds adjusted to captivity for two weeks, we measured BMR and exercise-induced MMR (electronic supplementary material, figure S1). Blood was then sampled to characterize hormonal and immune status, and each bird implanted with either an empty or T-filled silastic tubule (see below). Metabolic evaluations and blood sampling were undertaken two weeks later to determine plasma T levels and ‘baseline’ metabolic and immune response parameters (below). Within each sex, birds were distributed randomly into one of the four experimental groups: antigenic injection/T implant (IC(immune challenge)/T), antigenic injection/control implant (IC/C), sham injection/T implant (S/T) and sham injection/control implant (S/C).

(c) . Testosterone and control implants

Under anaesthetic (methoxyflurane), birds were implanted subcutaneously with 6 mm lengths of medical-grade silastic tubing (1.47 mm inner diameter; 1.96 mm outer diameter; Dow Corning), sealed at each end with silastic glue (Dow Corning), via a small incision on one side of their thorax. These implants were either empty (control birds; C-implants) or filled with crystalline T (Sigma Chemical Co.; T-1500; T-implants). The resultant wounds were sealed with medical-grade cyanoacrylate adhesive (Vetbond).

(d) . Immune challenge through hypervaccination

Two novel antigens, keyhole limpet haemocyanin (KLH; Sigma-Aldrich) and sheep red blood cells (SRBC; Institute of Medical and Veterinary Science, Adelaide, South Australia) were administered concurrently to ensure stimulation of adaptive immunity in the immune-challenged groups. Inoculation of KLH was followed Hasselquist et al. [19], with 100 µl of 1 mg KLH ml⁻¹ sterile water emulsified 1 : 1 with Freund's incomplete adjuvant (Sigma-Aldrich) injected into pectoral muscle. This was immediately followed by a 100 µl intra-abdominal injection of SRBC (10% by volume in phosphate-buffered saline (PBS)). Sham-treated birds were injected intra-muscularly with 100 µl sterile water emulsified 1 : 1 with Freund's incomplete immediately followed by an intra-abdominal injection of 100 µl PBS.

Injections were administered three times, two to three weeks apart, to achieve hypervaccination (electronic supplementary material, figure S1). Metabolic measurements and blood sampling occurred 12 days after the first injection and 6 days after the second and third injections in accordance with documented periods of peak KLH antibody titres after multiple injections in starlings ([19]; electronic supplementary material, figure S1).

(e) . Quantifying adaptive immunity

Following venepuncture, approximately 200 µl of blood was collected from a brachial vein in heparinized microhaematocrit tubes. Blood samples were immediately centrifuged at 6000 r.p.m. for 5 min and plasma stored at −20°C for hormonal (see electronic supplementary material) and immune analyses. Plasma concentrations of anti-KLH antibodies were determined using enzyme-linked immunosorbent assay (ELISA; [19,20]; see electronic supplementary material for details). Constitutive innate immunity and specific immunity to SRBC were determined by quantifying agglutination and lysis activity of plasma before and after exposure to SRBC, respectively, following Matson et al. [21], modified to use 1% suspension of SRBC instead of rabbit RBC.

(f) . Basal metabolic rate

Measurement of BMR was conducted when birds were post-absorptive, during the birds' rest-phase, and at temperatures within the thermoneutral zone for this species (29–31°C; [22]). At least 3 h after they had last fed, birds were placed individually in 2 l metal chambers fitted with a perch and supplied with room air at 500 ml min⁻¹ (Tylan mass flow controllers FC-280S). Oxygen content of inlet and outlet air for each chamber was measured using a Sable Systems Oxzilla II oxygen analyser in combination with an electronic stream selector (Sable Systems Respirometer Multiplexer V 2.0; further details in [23]). Basal rates of oxygen consumption (BMR) were defined as the mean of the two lowest 5 min periods of oxygen uptake recorded during the last half of the 12 h measurement period, which ensured all birds were post-absorptive [24,25].

(g) . Maximum metabolic rate

Oxygen consumption rates (V̇O₂) were measured during intense exercise within an enclosed 5 l drum with clear sides and carpet lining the inner rim [26]. These measurements took place in the morning following completion of the overnight BMR determinations, with all birds held in small cages with free access to food and water for 4 h prior to exercise measurements. A mass flow controller (Tylan Corp.) supplied air to the chamber at 5 l min⁻¹ and the oxygen content of inlet and outlet ports was measured with an oxygen analyser (Sable Systems FC-1). A single bird was placed in the drum and, once settled, the cover was removed and the drum rotated. The Ping-pong balls within the drum encouraged birds to maintain a series of rapid take-offs and short-term flights interspersed with vigorous hopping. The V̇O₂ data were adjusted with ‘instantaneous’ conversion procedures to account for gas mixing characteristics of the wheel and accurately resolve short-term oxygen content variation [23,26]. The highest continuous 60 s instantaneous oxygen consumption rate was designated as MMR.

(h) . Statistical analysis

We considered the performance (body mass, metabolic measures and immune responses) of animals two weeks after receiving their implant, but prior to the series of immune challenges (electronic supplementary material, figure S1) to represent individual baseline measures. We tested for differences in baseline performance measures between treatment groups across a set of hypothesis-driven candidate general linear models including fixed effects of T treatment (T/C), sex (M/F), immune challenge (IC/S—although no immune challenges had yet taken place) and their biologically relevant interactions, in R ([27]; electronic supplementary material, tables S1, S4, S7, S12 and S16).

To verify that immune-challenged individuals elicited adaptive immune responses, immune parameters after each injection were quantified as individual's change, positive or negative, in a given immune parameter (lysis, agglutination, KLH) compared to their ‘baseline’ for that immune parameter (measured two weeks post-implant; electronic supplementary material, figure S1). Across the study, the effects of injection stage (post-injection 1, 2 and 3) and experimental treatments on change in each immune parameter were assessed across a set of candidate linear mixed models including random effects for individual and fixed effects of immune treatment, T treatment, sex and their biologically relevant interactions, using ML estimation in the lme4 package in R ([27]; electronic supplementary material, tables S20, S23 and S25). Linear mixed models (with ML estimation in the lme4 package in R) were also used to verify T treatments elevated plasma T concentration, and that there were no effects of sex or immune treatments (electronic supplementary material, table S18). To test the effects of experimental treatments on change, in each metabolic parameter (BMR, MMR and body mass), as difference relative to baseline, a set of candidate linear mixed models were assessed in the lme4 package in R (with ML estimation), including random effects for individual and fixed effects for experimental treatments and stages (electronic supplementary material, tables S28, S31 and S33). Results are reported from all supported models, based on AICc comparisons across models. All data and code used in these analyses are publicly available through Dryad [28] and Zenodo [29], respectively.

3. Results

(a) . Evaluation of experimental treatments

Prior to the series of immune challenges, body mass and all immune measures were statistically indistinguishable between sexes, immune treatments and T treatments (electronic supplementary material, tables S1–S11). However, birds with T-implants had slightly lower BMR across all supported models (ca 1.5%; t = −2.36; p = 0.022; electronic supplementary material, tables S13–S15) and males had higher MMR (10.4%; t = 2.63; p = 0.011; electronic supplementary material, tables S16 and S17) than females. Plasma T content differed between implant groups (F_1,135 = 90.019; p = <2.2 × 10⁻¹⁶; electronic supplementary material, table S18 and S19; figure S2), averaging 5.15 ± 0.36 and 0.56 ± 0.07 ng ml⁻¹ in T-implanted and C-implanted birds, respectively, but was unaffected by sex, experimental stage or immune challenge (electronic supplementary material, table S18 and S19). All T-treated birds, including females, developed blackened bills, suggesting the elevated T in plasma was biologically effective [30,31]. Bills of all control birds retained a pale, non-reproductive colour throughout the experiment.

(b) . Immune responses to hypervaccination

All birds injected with SRBC and KLH displayed significantly greater immune responses to these antigens than sham-injected individuals, with the extent of these differences becoming greater with each injection (figure 1), indicated by significant effects of stage (p = 0.036) and immune treatment (p < 0001) for lysis scores (electronic supplementary material, tables S21 and S22), and significant interactions between stage and immune treatment (p < 0.009) for agglutination (electronic supplementary material, table S24) and KLH (electronic supplementary material, tables S26 and S27). Neither body mass nor sex affected these results (electronic supplementary material, tables S21, S22, S24, S26 and S27) and T-treatment only had an effect on KLH response in one of the two supported models (electronic supplementary material, tables S26 and S27), whereby T-implants resulted in antigen-injected birds having 50% lower responses to KLH, on average, compared to C-implants (electronic supplementary material, table S26).

Figure 1.

Change in immune responses of antigen-injected (immune-challenged) (n = 32) and sham-injected (n = 30) birds following serial injections, in relation to pre-injection ‘baseline’ scores, where immune responses were measured as the highest twofold dilution of plasma producing lysis (a) or agglutination (b) to SRBC, and the optical density (at 450 nm) of plasma in an ELISA to KLH as a percentage of positive control (c). Immune responses of birds given a sham injection did not deviate from pre-injection values, whereas immune responses of vaccinated birds increased significantly after each injection.

(c) . Metabolic responses to hypervaccination

Throughout the experiment, BMR remained stable (figure 2b), and the supported model indicated individual change in BMR relative to baseline was unaffected by sex, plasma T content, immune challenge or measurement stage (electronic supplementary material, table S31 and S32). By contrast, individual changes in body mass were significantly higher in immune- compared to sham-challenged birds (p = 0.007, electronic supplementary material, tables S29 and S30; figure 2a). Similarly, although MMR decreased over the experimental period (as expected in captivity; figure 2c; ‘stage’ p < 0.001, electronic supplementary material, tables S34–S36), two of the three supported models included a significant interaction between injection type and stage (p < 0.05, electronic supplementary material, tables S34 and S36), indicating immune-challenged birds experienced significantly less decrease in their MMR over time compared to sham-injected birds (figure 2c).

Figure 2.

Change in body mass (a) and metabolic rates ((b) BMR; (c) MMR) of antigen-injected (immune-challenged) (n = 32) and sham-injected (n = 30) birds following serial injections, in relation to pre-injection ‘baseline’ scores. Body mass and BMR were unaffected by injection type and did not vary over time. By contrast, MMR declined in all treatment groups over the experimental period, but to a lesser extent in immune-challenged birds compared to sham-injected birds.

4. Discussion

Most studies examining the energy costs of antibody production are based on a single antigenic challenge. While these are valid measures, they miss the far greater antibody production that follows subsequent antigen exposures [32]. Our use of a hypervaccination protocol, paired with repeated metabolic evaluations, permitted us to evaluate multiple stages of adaptive immunity. As expected (e.g. [19]), immune measures increased substantially following sequential injections of novel antigens, as exemplified by the six- and sevenfold greater responses to the second and third injections of KLH compared with the first. (figure 1c).

All hormone-implanted birds had T levels corresponding to free-living male house sparrows during breeding (26), and approximately fivefold higher than in breeding free-living females [33]. Thus, although the T-treatment reduced responsiveness to KLH in all birds, this result in females stems from hyper-physiological T levels, that have been found to be immunosuppressive in females of other species [34,35]. By contrast, T-treatment had no effect on adaptive immune responses to serial SRBC vaccination. This calls into question the generality of T-mediated immunosuppression underlying the ICHH [17], particularly since our study incorporated all criteria associated with the highest effect sizes across previous studies of the ICHH, namely humoral immunity, experimental immune challenge, castrated males and animals sourced from free-living populations [17].

Our experimental design met all requirements for determining BMR [36] and clearly found that the immune challenges had no effect on the birds' maintenance energy requirements. These findings corroborate other studies reporting nil to negative metabolic costs of adaptive immune activation in endotherms [9,11,37–39]. If, however, immune activation promotes reallocation of resources among bodily functions, this effect is likely to be detected by evaluating changes in MMR. For instance, during avian moult, the protein demands of feather replacement are accompanied by increased rates of muscle protein turnover [40,41] and associated decreases in MMR, in direct proportion to the mass of feathers being replaced [23]. In our study, both immune-challenged and sham-treated birds showed a gradual decline in MMR over the duration of the experiment, likely a consequence of aerobic detraining due to reduced flight activities imposed by captivity [42,43]. Surprisingly, however, immune-challenged birds maintained significantly higher MMR than their unvaccinated counterparts as the study progressed, irrespective of their sex or T-treatment. Thus, the demands of antibody production did not constrain peak aerobic performance following immunization and, more remarkably, appeared to be protective of such activities in both sexes. Whether this was a consequence of differences between immune-challenged and sham-treatment groups in the extent of flight activity and/or in physiological processes forestalling aerobic decline requires further experimentation.

In conclusion, activation and reactivation of adaptive immunity did not provoke changes in BMR nor decrease maximal aerobic capacity. Plasma T levels had moderate, but inconsistent, effects on measured immune responses, but no effect on MMR. The absence of additional maintenance energy costs or reductions in maximum aerobic capability during peak antibody circulation following hypervaccination reinforces the notion that antibody production following adaptive immunity activation is a low-cost mechanism in the immunological armoury of vertebrates [44].

Ethics

All aspects of animal care and their use in these experimental procedures were approved by the University of Wollongong Animal Ethics Committee (AE04/01).

Data accessibility

Data are available from the Dryad Digital Repository: https://doi.org/10.5061/dryad.4xgxd25br [28]. Statistical code used in this study is linked to the Dryad record, but also available directly from Zenodo: https://doi.org/10.5281/zenodo.5904580 [29].

The data are provided in the electronic supplementary material [45].

Authors' contributions

W.A.B.: conceptualization, data curation, funding acquisition, investigation, methodology, project administration, resources, supervision, validation, visualization and writing—original draft; T.W.O.: data curation, investigation, methodology, project administration, validation and writing—review and editing; L.B.A.: conceptualization, data curation, funding acquisition, investigation, methodology, project administration, resources, software, supervision, validation and writing—review and editing; K.C.K.: conceptualization, methodology, supervision, visualization and writing—review and editing; B.J.H.: data curation, formal analysis, investigation, project administration, software, validation, visualization, writing—original draft and writing—review and editing.

All authors gave final approval for publication and agreed to be held accountable for the work performed therein.

Conflict of interest declaration

We declare we have no competing interests.

Funding

This research was funded by the Australian Research Council grant DP0453021 awarded to W.A.B. and L.B.A.