Veterinary pharmaceuticals contained in dead livestock may be ingested by avian scavengers and negatively affect their health and consequently their population dynamics and conservation. We evaluated the potential role of antibiotics as immunodepressors using multiple parameters measuring the condition of the cellular and humoral immune system in griffon (Gyps fulvus), cinereous (Aegypius monachus) and Egyptian vultures (Neophron percnopterus). We confirmed the presence of circulating antimicrobial residues, especially quinolones, in nestlings of the three vulture species breeding in central Spain. Individuals ingesting antibiotics showed clearly depressed cellular and humoral immune systems compared with nestlings from the control areas, which did not ingest antibiotics. Within central Spain, we found that individuals with circulating antibiotics showed depressed cellular (especially CD4+and CD8+T-lymphocyte subsets) and humoral (especially acellular APV complement and IL8-like) immune systems compared with nestlings without circulating antibiotics. This suggests that ingestion of antibiotics together with food may depress the immune system of developing nestlings, temporarily reducing their resistance to opportunistic pathogens, which require experimental confirmation. Medicated livestock carrion should be considered inadequate food for vultures due to their detrimental consequences on health derived from the ingestion and potential effects of the veterinary drugs contained in them and for this reason rejected as a management tool in conservation programmes.
Modern intensive livestock production uses huge quantities of antibiotics and generates large amounts of residues containing these drugs (Sarmah et al. 2006). Antibiotics, especially quinolones, take a long time to degrade in livestock meat, even taking more than two weeks to reach levels safe for human consumption (EMEA 2004). Carcasses of diseased livestock often medicated until death on the farms are usually dumped outdoors for consumption by scavenger wildlife in Spain (Blanco et al. 2006; Lemus et al. 2008). Therefore, the consumption of recently dumped livestock carcasses may imply the ingestion of active antibiotics (Lemus et al. 2008), of which the proximate effects on the health of avian scavengers remain generally unknown. Despite this, carrion of medicated livestock is currently used as food in management programmes for conservation of threatened scavengers in Spain, without any toxicological or veterinary control (Blanco et al. 2006, 2007a,b; Lemus et al. 2008).
Antibiotics are the most commonly used agents in intensive stabled livestock facilities (Friendship 2000; Tanner 2000). Among them, the more commonly used 5-fluoroquinolones and specifically enrofloxacin and its active metabolite ciprofloxacin are very stable molecules that have been found to contaminate the environment (Sarmah et al. 2006) and meat destined for human consumption after the withdrawal period (Okerman et al. 2001; Juan-García et al. 2007). Previously, a clear association between quinolone ingestion, pathogen acquisition and mortality has been found in vultures (Lemus et al. 2008). However, there are no data regarding how antibiotics may negatively affect vulture health, although it has been suggested that the effects are partially due to the immunodepressive impact of antibiotics under conditions of chronic but discontinuous ingestion (Lemus et al. 2008), similar to the effects derived from misuse of these drugs in humans (Levy 2002; Chide & Orisakwe 2007). In fact, the immunodepressive effects of antibiotics have long been established (Norrby & Lietman 1993; Brown 1996; Walker 2000; Lathers 2002), especially in continued treatments, causing fungal overgrowth in place of protective normal bacterial flora due to their selective elimination, as well as direct effects on normal production and activity of lymphocytes (Levy 2002; Chide & Orisakwe 2007).
Antibiotics are administered only when an individual is not able to eliminate an infection. If antibiotics are ingested under inappropriate conditions in terms of concentration, timing and duration of treatment, the active molecules can cause primary suppression of lymphocyte proliferation and bone marrow (Jiménez-Valera et al. 1995; Chide & Orisakwe 2007). Other detrimental effects include the blocking of the specific target molecules for macrophage activity, impeding their proper function (Azuma et al. 2001). The influence of fluoroquinolones on immunoglobulin production, even in the developing egg (Tokarzewski 2002), seems to indicate this same mechanism. The increase in free radicals, especially of hydrogenous peroxide and deleterious hydroxyl radicals are other action mechanisms that could explain immunodepression caused by quinolones (Azuma et al. 2001; Kohanski et al. 2007). The recently proposed phage antibiotic synergy hypothesis suggests a massive deleterious growth of bacteriophages due to quinolone application (Comeau et al. 2007). Therefore, the potentially detrimental impact of antibiotics on the immune system, generally associated with their misuse, may be complex as several very different action mechanisms, immunological subsystems and metabolic routes may be involved (Levy 2002; Chide & Orisakwe 2007). All of these negative effects of antibiotics may be exacerbated in developing animals because of their partial dependence on passive maternal immunity, active development of immune organs and, hence, limited functional activity of their immune system (Tizard 2000; Fellah et al. 2008).
In this paper, we evaluated the potential role of antibiotics as immunodepressors using multiple parameters measuring the condition of the cellular and humoral immune system in griffon (Gyps fulvus), cinereous (Aegypius monachus) and Egyptian vultures (Neophron percnopterus). We use flow cytometry techniques to evaluate cell immunity by direct account of the three most representative T-cell subsets (CD4+, CD5+ and CD8+) and different techniques for the determination of humoral immunity molecules, including complement, interleukins (ILs) and γ interferon (IFNγ). We hypothesized that the previously reported association between the presence of circulating antibiotics and detrimental pathogens found in the three vulture species (Lemus et al. 2008) may be partially due to the potential immunodepressive impact of antibiotics, especially under conditions of chronic but irregular ingestion at variable concentrations of different antibiotics, primarily quinolones, during nestling development.
This hypothesis leads to the prediction that the immunological parameters of vultures feeding upon carrion from intensively medicated livestock, which then show antibiotic residues in the blood, should show lower values (i.e. immunodepression) than those of vultures showing no circulating antibiotics. We tested this prediction by comparing immunological parameters of vultures from central Spain, in which circulating antibiotics were previously found (Lemus et al. 2008), with those from control areas where antibiotics should not be ingested because of their lack in the wild prey and unstabled livestock that constitutes their food supply (Donázar 1993). We suggested previously that the lack of antibiotics in a portion of nestlings feeding upon stabled medicated livestock in central Spain could be due to the frequency of ingestion of medicated carrion, the concentration of antibiotics contained in carrion, as well as to individual differences in metabolization and excretion of ingested antibiotics (Lemus et al. 2008). Therefore, we hypothesized that antibiotics may exert immediate depressive effects on the immune system during the time when they may be detected in blood. This leads to the prediction that individuals with detectable circulating antibiotics should show more severe immunodepression than those without circulating antibiotics despite the presumably chronic ingestion of antibiotics by all individuals in central Spain (Lemus et al. 2008). We tested this prediction by comparing immunological parameters of vultures in central Spain with and without circulating antibiotics at the time of sampling.
2. Material and methods
During the breeding seasons of 2006 and 2007, nestlings of the three vulture species were sampled for blood in their nests in central Spain (Segovia, Ávila and Madrid provinces). In this area, a large population of avian scavengers depends on livestock farming operations providing livestock carrion, especially swine (Blanco et al. 2006, 2007a,b; Lemus et al. 2008). Control individuals were sampled from two additional areas (Extremadura and Andalucia) in western and southern Spain where scavengers feed upon wild prey or unstabled livestock (Donázar 1993; Benítez et al. 2003; Costillo et al. 2007), and should not contain antibiotics in high prevalence and concentrations.
Vulture nests were accessed by climbing and nestlings were sampled when they were 60–80 days old, depending on the species. A sample of blood (5 ml) was taken from the brachial vein; 1 ml was preserved in EDTA for flow cytometry and the remaining 4 ml was preserved in lithium heparin, centrifuged and plasma frozen until analysis.
(b) Antibiotic determination
The presence and concentrations of residues of quinolones (enrofloxacin and ciprofloxacin), amoxicillin and oxytetracycline in plasma were determined using high-performance liquid chromatography and microbiological techniques and standard protocols, as described previously (Lemus et al. 2008). The limits of quantification, percentage recoveries and inter- and intra-assay reproducibility were adequate (see Lemus et al. 2008).
(c) Cellular immunity
For the peripheral lymphocyte isolation, cell sorting and flow cytometry, lymphocytes of peripheral blood were isolated as described previously (Finkelstein et al. 2003; Lavoie & Grasman 2005). The mononuclear cells (lymphocytes and monocytes) of the supernatant were washed twice in phosphate-buffered saline (PBS), resuspended in 1 ml of PBS and subsequently used for flow cytometry. For cell sorting, T-cell characterization and analysis of the dynamics of γδ T-cell subsets, we used methods described previously (Tella et al. 2008).
(d) Humoral immunity
Complement activity was determined with a haemolytic technique as described previously (Demey et al. 1993; Parmentier et al. 2002) using an adapted light-scattering method. In brief, sera were diluted serially in appropriate buffers in flat-bottomed 96-well microtitre plates and incubated with sensitized (haemolysin, Biomerieux, ref. no. 72202) sheep erythrocytes to measure complement classical pathway (CPV) or rabbit erythrocytes to measure alternative complement pathway (APV). Plates were shaken in a Titertek (Flow Laboratories) every 30 min during the period of incubation. The results (the amount of light scattering by erythrocytes upon lysis) were read at 655 nm in a microtitre reader (BioRad model 3550). Readings were transformed by log–log equations (Demey et al. 1993; Parmentier et al. 2002) and the haemolytic titre was expressed as the titre that lysed 50 per cent of the erythrocytes (CH50 U ml−1).
IL determination was made by using enzyme immunoabsorbent assay for the IL1β-, IL2- and IL8-like (Kogut 2002) measurements, and enzyme-linked immunoabsorbent assay (ELISA) for the IL6 measurement. All the reagents and kits were purchased from Biosource Europe (Nivelles, Belgium). Sensitivity was 0.35 pg ml−1 for the IL1β, 0.025 IU ml−1 for the IL2, 1.5 pg ml−1 for the IL6 and 1.1 pg ml−1 for the IL8-like.
IFNγ was determined by ELISA. Briefly, 96-well flat microtitre plates were coated with 100 μl per well of a solution consisting of 2 μg ml−1 coating antibody (antichicken IFNγ mAb, Biosource) in PBS (pH 7.2) for 18 hours at 4°C, and then blocked for 2 hours at room temperature by adding 300 μl per well PBS-1 per cent bovine serum albumin (BSA). After three washes with PBS containing 0.2 per cent Tween-20 (PBS-T), diluted standards (natural chicken IFN-γ, Biosource), controls and samples were added at 100 μl per well, followed immediately by the addition of 50 μl per well of biotinylated antibody (mouse anti-chicken IFN-γ mAb, 0.1 μg ml−1, Biosource) in PBS–T1 per cent BSA. Plates were incubated for 2 hours at room temperature with continuous shaking (700g), washed three times and incubated with HRP-labelled streptavidin (1 : 2500, Biosource) for 30 min. After washes, 100 μl per well of the chromogen tetramethylbenzidine (Sigma) was added and incubated for 30 min. Substrate reactions were stopped with 2 N H2SO4, and the optical density (OD) values were measured at 450 nm within 30 min by plotting each standard OD (ordinates) versus the standard concentration (abcisses). Concentration of unknown samples was determined from the standard curve.
(a) Prevalence and concentrations of circulating antibiotics
Antimicrobials were detected in 63.6 per cent (n=22), 80.9 per cent (n=21) and 39.2 per cent (n=28) of the nestling griffon, cinereous and Egyptian vultures sampled in central Spain, respectively. No nestling showed circulating antibiotic residues in the control regions (n=6, 10 and 6, respectively). Enrofloxacin and ciprofloxacin were the most frequent antibiotics found in vultures (50 and 38.5%, 81 and 52.4% and 29.7 and 22.2% in griffon, cinereous and Egyptian vultures, respectively) at high concentrations (mean±s.d., 0.1331±0.0511 μg ml−1, n=13 and 0.0860±0.0589 μg ml−1, n=9 for griffon vultures; 0.0852±0.0516 μg ml−1, n=18 and 0.0764±0.0495 μg ml−1, n=11 for cinereous vultures; and 0.0288±0.0330 μg ml−1, n=8 and 0.0.0133±0.0171 μg ml−1, n=6 for Egyptian vultures). Prevalence of amoxicillin was 11.5 per cent (0.0600±0.0200 μg ml−1, n=3), 28.6 per cent (0.0480±0.0236 μg ml−1, n=6) and 3.7 per cent (0.1150±0.0354 μg ml−1, n=6) in griffon, cinereous and Egyptian vultures, respectively. Prevalence and concentration of oxytetracycline was 3.8 per cent (0.03 μg ml−1, n=1) in griffon vultures and 14.3 per cent (0.0417±0.0.0176 μg ml−1, n=3) in cinereous vultures, but was not detected in Egyptian vultures. Most (72.5%) individuals of the three species (n=41) showed combinations of two or three different antibiotics in blood.
(b) Immunological parameters
We found significant differences in all immunological parameters between nestling vultures sampled in central Spain, distinguishing between nestlings with circulating residues of antibiotics and those without residues, and nestlings from the control area, where vultures showed no circulating antibiotics (MANOVA, all p<0.0001 in the three vulture species). Mean values for each parameter in each group are shown in table 1. Post hoc differences (Tukey test) between the three groups showed significant differences in all immunological parameters between nestlings from central Spain and the control areas in the three species, independent of the effect derived from the presence of circulating antibiotics in nestlings from central Spain (all p<0.0001). Significance (p-values) of the differences (Tukey post hoc test) in the immunological parameters between nestlings with and without circulating antibiotics is shown in table 1 for each vulture species.
CD4+ and CD8+ showed lower values in nestling griffon and Egyptian vultures with antibiotics than in those without antibiotics in central Spain (table 1). These differences were not significant in the cinereous vulture probably due to the small sample of individuals without antibiotics. APV and IL8-like showed higher values in individuals without antibiotics than with circulating antibiotics in the three species (table 1). Remaining parameters vary from those that show significant variation due to the presence of antibiotics (complement CPV and IL2) to those affected by their presence in some species but not in others (IL1β, IL2 and IL6).
Because all the immunological parameters were intercorrelated (Pearson correlation, all p<0.001), we attempted to objectively reduce the original database to smaller, mutually uncorrelated composite variables or factors, by conducting principal components factor analysis using the correlation matrix. We obtained a single axis with an eigenvalue greater than 1 (PC1=7.61) that accounted for 76.12 per cent of the variance. This component revealed high loading for all immunological variables, with factor loadings varying between 0.573 (CD8+) and 0.889 (complement APV) and hence, it defines a gradient of immunological state. Factor scores of the principal component differed between groups (nestlings with and without antibiotics in central Spain and nestlings in the control areas; ANOVA, F2,93=1475.03, p<0.0001) and vulture species (F2,93=131.74, p<0.0001; interaction between groups and species: F4,93=7.63, p<0.0001). Overall, nestlings of the three vulture species showing circulating antibiotics were more immunodepressed than those without circulating antibiotics in central Spain (figure 1), and both of these groups of nestlings were immunodepressed compared with nestlings of the three vulture species from the control areas (figure 1).
The results of this study confirm the presence of circulating antibiotic residues in nestlings of the three vulture species breeding in central Spain, especially quinolones ingested with carrion of intensively medicated livestock (Lemus et al. 2008). On the contrary, no nestling of the three vulture species feeding primarily upon unstabled livestock and wild prey in southern and western Spain showed antibiotics in its blood. Individuals chronically ingesting antibiotics showed clearly depressed cellular and humoral immune systems compared with nestlings from the control areas, which did not ingest antibiotics. Within central Spain, we found that individuals with circulating antibiotics were more immunodepressed than nestlings without circulating antibiotics. This suggests that continued ingestion of different antibiotics at variable concentrations over long periods (the nestling stage) may play a role in depressing the immune system of developing nestlings. Given the correlative nature of this study, the results need to be confirmed experimentally in order to evaluate the magnitude of the potential immunodepressing effects of antibiotics, as well as the particular mechanisms and components of the immune system activated to respond against them. Previous experiments about the effects of veterinary drugs on the immune system of domestic animals have involved a controlled standard ‘treatment’ or exposure rather than an uncontrolled exposure to these drugs, as occurs with vultures in the wild. Therefore, experiments conducted to evaluate the effects of antibiotics on vultures should simulate the completely caothic ingestion of antibiotics from livestock carrion, including variable concentrations, simultaneous ingestion of different antibiotics, variable timing of ingestion, etc., in young birds with a developing immune system.
All immunological parameters from nestlings of the three vulture species sampled in central Spain reached values lower than in individuals not ingesting antibiotics in control areas from southern and western Spain. This striking result arises even when including in the comparison nestlings from central Spain, which do not present circulating antibiotic residues at sampling due to the likely chronic but discontinuous ingestion of antibiotics by nestlings of the three vulture species in central Spain. The probability of finding circulating antibiotics at sampling may depend on the amount of time since the last ingestion of food by nestlings, which may be fed by parental vultures every several days (Donázar 1993). This probability may also depend on parental differences in their dependence on stabled medicated livestock versus unstabled livestock and wild prey, the concentration of antibiotics contained in carrion ingested in the last feeding bout, as well as individual differences in metabolization and excretion of ingested antibiotics (Lemus et al. 2008). These factors may explain the lack of circulating antibiotics in a portion of individuals from central Spain at sampling, despite all of them depending to some extent on medicated livestock carrion as demonstrated by the presence of remains of such food in all sampled nests. The ingestion of antibiotics probably occurs during the entire development period of nestlings, as suggested both by the presence of circulating residues of antibiotics in nestlings sampled at different ages (Lemus et al. 2008) and by the remains of stabled livestock found in nests accessed at different stages of nestling development (G. Blanco & J. A. Lemus 2001–2007, unpublished data). Overall, the chronic discontinuous antibiotic ingestion may promote a generalized immunodepression, which was not apparent in control area individuals that had not ingested antibiotics.
Uncontrolled antibiotic ingestion may exert immediate depressing effects on the immune system at the time that they are detected in blood (Brown 1996; Lathers 2002). This may explain why individuals of the three species with circulating antibiotics showed a more severe immunodepression than those without circulating antibiotics, despite the presumably chronic ingestion of antibiotics by all individuals in central Spain (Lemus et al. 2008). In contrast with other toxicants that cause an irreversible effect on the immune system, some effects of antibiotics may be quickly reversed (Brown 1996; Walker 2000; Lu et al. 2008). In addition, the developing immune system of nestling vultures may be able to recover more quickly than a fully developed immune system (Tizard 2000; Fellah et al. 2008). This suggests that immunodepression may occur shortly after antibiotic ingestion, but also that the immune system may recover quickly after antibiotic ingestion ceases. This may also explain the general immunodepression of all individuals (with or without antibiotics at sampling) from central Spain compared with those from southern and western Spain.
Many potential action mechanisms may be simultaneously involved in immunodepression due to antibiotic ingestion (Levy 2002; Chide & Orisakwe 2007), as indicated by the fact that both cellular and humoral systems were affected to variable degrees by the presence of circulating antibiotics. The complex inter-relationships between cellular and humoral immunities may explain the patterns of variation of the different immunological parameters due to chronic but discontinuous or recent ingestion of antibiotics. Regarding cellular immunity, subsets promoting active defence (CD4+) and possessing cytotoxicity (CD8+) were highly reduced in nestlings showing antibiotics, in contrast to the subset with helper and coordinative activity (CD5+), which was not so clearly affected by antibiotics probably due to their low number and higher longevity (Jeurissen et al. 2000; Campello et al. 2006). Regarding humoral immunity, APV complement and IL8-like were especially lowered associated with the presence of antibiotics in the three species. APV complement is mainly related to the specific immunity, as well as to pathogen identification by heterophils and macrophages (Parmentier et al. 2002). Lower values may be ineffective at avoiding the immediate heterophil defence against pathogens, thus allowing their acquisition and proliferation. Although we did not study the activity of macrophages, the frequent infection of the nestlings by opportunistic pathogens such as Candida albicans and Mycobacterium avium (Lemus et al. 2008) may be indicative of their lack of or reduced activity.
IL8-like is related to heterophil chemotaxis (Kogut 2002). This cytokine is produced by macrophages and endothelial cells close to or at the site of inflammation or infection (Kaiser et al. 1999; Kaiser & Staheli 2008). If the macrophage activity is reduced, very low IL8 secretion is expected, as occurred in strongly selected, highly immunodepressed broiler hens (Cheng et al. 2001; Bridle et al. 2006). The remaining parameters (IL1, IL6 and IFNγ) also showed low values associated with antibiotic ingestion in some species but not in others. This suggests that some parameters involved in the immunodepression associated with antibiotics may be species specific and requires further research.
Humoral immunocompetence is mostly cell dependent, especially in developing individuals with an intense activity of lymphocytes, which act as mediators or effectors of most of the activation and activities of humoral molecules when the immune organs and their function are still developing (Tizard 2000; Kaiser & Staheli 2008). In this sense, macrophages, mast cells, plasma cells, natural killer cells, lymphocytes, monocytes, heterophils and even non-leucocyte cells, such as endothelial cells, possess the capacity to activate several cytokines, interferon or complement activity (Okamura et al. 2004; Kaiser & Staheli 2008). The observed low levels of T cells and cytokines associated with the presence of antibiotics seem to adjust to the previously mentioned feedback pattern. Thus, low levels of T lymphocytes can secrete low levels of cytokines and chemotactic molecules which may be low enough to contribute to their feedback participation promoting the presence of T lymphocytes (Jiménez-Valera et al. 1995; Riesbeck & Forsgren 1995; Campello et al. 2006).
The immunosuppressive effects of antibiotics temporarily reducing host resistance to opportunistic pathogens may explain the previously reported marked association between antibiotic residues and disease in the three vulture species (Lemus et al. 2008). A strongly depressed immune system may allow the access of other more detrimental pathogens that may cause dramatic epizootic episodes associated with population declines, especially in the social foragers and colonially breeding griffon and cinereous vultures. These results are of great concern, and could indicate the beginning of detrimental effects on vulture populations, especially on the reduction in breeding success by an increased mortality of nestlings (Carrete et al. 2006; Lemus et al. 2008) but also in a reduction in the number of breeding pairs (J. A. Lemus & G. Blanco 2001–2007, unpublished data). In Spain, the main stronghold of vultures in Europe, legal measures to mitigate the spread of bovine spongiform encephalopathy have caused a lack or scarcity of unstabled livestock carcasses available for avian scavengers, and a parallel increase in the use of dumps of livestock carcasses supplied by farms, especially of intensively medicated pigs and poultry (Tella 2001; Blanco et al. 2006, 2007a,b). However, pathogen or drug residue controls are not conducted on food provided by carcasses of medicated stabled livestock disposed of in refuse dumps for vultures and other scavengers. Livestock carrion proven to contain veterinary drugs should be considered as inadequate food for vultures due to their potential detrimental consequences on health and for this reason rejected as a management tool in conservation programmes. Such a precautionary approach could also provide important experimental information on this issue.
We thank F. Polo, J. C. Rincón, B. Arroyo, F. Martínez, A. Tejedor, O. Frías and N. Baniandrés for their help with fieldwork and two reviewers for their constructive comments and corrections on the manuscript. Funds were provided by the project CGL2007-61395/BOS of Spanish Ministerio de Educación y Ciencia.
- Received January 14, 2009.
- Accepted February 27, 2009.
- © 2009 The Royal Society