(Alligator Mississippiensis): Hepatic and Renal Se Accumulation and Its Effects on Growth and Body Condition

Arch Environ Contam Toxicol (2017) 72:439–448 DOI 10.1007/s00244-017-0370-4

Dietary Selenomethionine Administration in the American Alligator (Alligator mississippiensis): Hepatic and Renal Se Accumulation and Its Effects on Growth and Body Condition

1,2,3,4 2 1,3 John W. Finger Jr. • Matthew T. Hamilton • Travis C. Glenn • Tracey D. Tuberville2,3

Received: 1 December 2016 / Accepted: 17 January 2017 / Published online: 1 February 2017 Ó Springer Science+Business Media New York 2017

Abstract Selenium (Se) is an essential trace nutrient, but p \ 0.0001; 2000 ppm, p = 0.0316). Body condition and in excess, it can induce toxicity. Incomplete combustion of growth remained unchanged in control alligators coal produces coal combustion wastes, which are enriched (p [ 0.05). Our results demonstrate alligators are capable in Se and often disposed of in aquatic basins. While a of accumulating high levels of Se through trophic transfer. multitude of studies have investigated Se accumulation in The positive effects of accumulation on growth may vertebrates, few studies have examined its effects on demonstrate Se essentiality, whereas the negative effects longer-lived top trophic carnivores, such as the American on condition may demonstrate toxicity. Accumulation also alligator (Alligator mississippiensis). In this study, alliga- was associated with mortality, further demonstrating toxi- tors were fed one of three Dietary Treatments: mice city. Future studies should further investigate the physio- injected with water (controls) or water supplemented with logical effects of Se accumulation in long-lived, top- 1000 or 2000 ppm selenomethionine (SeMet). Dietary trophic carnivores. Treatment significantly affected Se levels in both the liver (p \ 0.0001; raw mean ± SE: 1000 ppm group, 35.20 ± 6.32 ppm; 2000 ppm group, 49.97 ± 4.00 ppm) and kid- Selenium (Se) is an essential trace element, with deficiency ney (p \ 0.0001; raw mean ± SE: 1000 ppm group, leading to disease and excess causing toxicity. While Se 101.60 ± 8.64 ppm; 2000 ppm, 96.38 ± 5.81 ppm), deficiency is a concern in some areas of the world (Haug which were significantly higher in alligators fed SeMet et al. 2007), anthropogenic activities, including fossil fuel than in controls. Post-treatment head length, used to control production, smelting, landfills, and agricultural activities, for size variation, was negatively related to both kidney are increasing the risk of Se exposure (Lemly 2004; Lemly (p = 0.0142) and liver (p = 0.0010) Se concentrations. and Skorupa 2012). In particular, incomplete combustion Dietary treatment with SeMet significantly reduced body of coal for energy production creates coal combustion condition (1000 ppm, p \ 0.0029; 2000 ppm, p = 0.0075), wastes (CCWs), which are rich in Se (Lemly 2004; Rowe but it significantly increased growth (1000 ppm, et al. 2002). Approximately 20% of all CCWs are disposed in aquatic settling basins (Lemly and Skorupa 2012), in which particulates settle out from the CCW slurry in a & John W. Finger Jr. primary basin before surface waters are drained into sub- [email protected] sequent basins or into wetland systems. Remarkably, a 1 Department of Environmental Health Science, University of number of organisms are known to inhabit or forage in ash Georgia, Athens, GA 30602, USA basins or the surrounding terrestrial habitat (Hopkins et al. 2 Savannah River Ecology Laboratory, University of Georgia, 1999a, b; Roe et al. 2004; Rowe et al. 2002), increasing PO Drawer E, Aiken, SC 29802, USA their risk of exposure to CCWs, including Se. 3 Interdisciplinary Toxicology Program, University of Georgia, As an essential nutrient, Se is an important component Athens, GA 30602, USA of a suite of proteins known as selenoproteins. Incorpora- 4 Present Address: Department of Biological Sciences, Auburn tion of Se into selenoproteins occurs both specifically, University, Auburn, AL 36849, USA whereby the UGA codon specifies selenocysteine, and 123 440 Arch Environ Contam Toxicol (2017) 72:439–448 nonspecifically, through incorporation of selenomethionine of Se in the liver and kidney were quantified and associated (SeMet; Kohrle et al. 2005; Lu and Holmgren 2009). effects on growth and condition were determined. Selenoproteins serve many functions within the cell, including prevention of oxidative stress and in thyroid metabolism (Janz et al. 2010; Kohrle et al. 2005). How- Methods ever, in excess, Se leads to toxicity. Although the exact mechanism of Se toxicity is not known, it is thought that Se Study Subjects may exert its toxic effects through promotion of oxidative stress or amino acid substitution with sulfur, leading to Twenty-four, sexually immature alligators originally protein misfolding and loss of functionality (Janz et al. obtained from Rockefeller Wildlife Refuge (RWR) in 2010; Lemly 2002). Regardless of the toxicological mech- Grand Chenier, LA, USA were randomly allocated to one anism, accumulation has been shown to affect health and of three pens (corresponding to eventual SeMet treatment reproductive activities in a number of species, including groups: Pen 1, 1000 ppm SeMet; Pen 2, 2000 ppm SeMet; reptiles and birds (Heinz et al. 1987, 1989; Janz et al. 2010; Pen 3, control) within an enclosed aquatic animal facility Rowe et al. 2002; Roe et al. 2004). However, most of these on the Savannah River Site (SRS) in west-central SC as studies have focused on short-lived or lower-trophic level described previously (Finger et al. 2015b; Hamilton et al. organisms (Hopkins et al. 1999a, b, 2002, 2004). 2016a). Briefly, unheated water was continuously filtered Se accumulation and its consequent effects in top through each pen and ambient air temperature was main- trophic organisms are less well known. The American tained at 22.7 °C throughout the study. Alligators were alligator (Alligator mississippiensis) is a top trophic, long- maintained on natural circadian cycles via light filtration lived, aquatic organism found in the southeastern United through translucent fiberglass roof panels (Finger et al. States. Because of this trophic placement, as well as their 2015b). Prior to commencement of the experiment, alli- longevity, alligators may serve as unique indicators of gators were fed to satiation thrice weekly with Mazuri environmental quality and how anthropogenic activities Crocodilian Diet pellets (PMI Nutrition International LLC, affect the environment (Finger and Gogal 2013; Milnes and Brentwood, MO, USA). Thereafter, they were fed pellets Guillette 2008). While most studies have investigated the two times weekly, in addition to the injected mice provided effects of endocrine disrupting chemicals on alligators once a week as described below. Alligators were allowed to (references in Milnes and Guillette 2008), relatively few acclimatize to pens for [5 months prior to SeMet studies have investigated the effects of trace elements, like administration. those present in CCWs, on alligators (Finger et al. 2016; Tuberville et al. 2016). However, because alligators are Dietary Treatment known to frequent ash basins and surrounding areas (Roe et al. 2004; personal observation), they may be more This study was modeled after Hopkins et al. (2004)in readily exposed and serve as relevant models for investi- which house snakes (Lamprophis fuliginosus) were fed gating the effects of CCWs, such as Se, on higher trophic mice-injected with SeMet for 10 months. Due to space organisms. Indeed, alligators have been shown to accu- limitations, alligators were housed communally (n = 8in mulate Se from the environment or their diet (Campbell each of the three pens). Therefore, to make sure every et al. 2010; Roe et al. 2004; Tuberville et al. 2016). In alligator received an equivalent dose (within a treatment) hatchling alligators collected downstream from ash basins, and prevent Se diffusion into pen water, each alligator was increased Se accumulation was associated with decreased force-fed previously thawed small ‘‘fuzzy’’ mice (2–3 g; fecundity and increased embryonic mortality (Roe et al. Layne Labs, Arroyo Grande, CA, USA) spiked with 2004). A previous study documented that dietary exposure SeMet. Due to the large difference in size between vehicle to CCW-contaminated prey for 25 months led to signifi- (i.e., spiked mouse) and alligator, and to minimize the cant accumulation of Se in juvenile alligators (Tuberville number of times an alligator was force-fed a week (i.e., to et al. 2016). However to our knowledge, no controlled minimize stress), the amount of SeMet injected into each laboratory study has investigated accumulation, tissue mouse was greatly increased relative to that used previ- burdens, and impact of Se on a top-trophic carnivore, ously (Hopkins et al. 2004). Briefly, 1 g of seleno-D,L- without the possible interference of other trace elements methionine (Sigma S3875, St. Louis, MO, USA) was dis- present in CCWs. solved in 100 mL of DI water for a stock concentration of The impetus of this study was to better understand Se 10,000 ppm (based on dry mouse weight). accumulation through trophic transfer in the American To determine the volume of the 10,000 ppm stock alligator. Alligators were fed prey spiked with SeMet (1000 solution of SeMet to inject to achieve the desired target and 2000 ppm) for 7 weeks. Subsequently, tissue burdens concentration (in ppm dry weight of mouse), a sample of 123 Arch Environ Contam Toxicol (2017) 72:439–448 441 mice were freeze dried prior to experimentation. A stan- One individual treated with 2000 ppm SeMet died dur- dard curve based on mouse dry weight versus mouse wet ing the course of this study and was only fed for a total of weight was generated using CurveExpert 1.4. This standard 6 weeks. Four other individuals (1 in 1000 ppm group; 3 in curve was subsequently used to estimate dry weights of 2000 ppm group) died before euthanasia, but all received mice for SeMet injection. Briefly, thawed dead mice were the entire, 7-week dosage regimen. Of the individuals that weighed and subsequently injected with either 1000 or died prior to euthanasia, three individuals from the 2000 ppm SeMet through the buccal cavity using a 1-CC 2000 ppm treatment group were not weighed and two were syringe on treatment days. Control mice were injected with not measured for HL. As such, these individuals were 0.5 mL of DI water. To verify that our target doses were excluded from respective analyses. However, because all achieved, we submitted an injected mouse from each tissues were dissected immediately after death, none of dietary treatment for trace element analysis. Actual Se these five individuals were excluded from Se accumulation concentrations (mean ± SE) for mice injected with 1000 analysis. and 2000 ppm SeMet were 1297.55 (±33.09) ppm and 2104.80 (±35.05) ppm, respectively, whereas Se concen- Blood Sampling for Se Analysis trations of control mice (DI water) were below minimum detection limits (MDLs; 2.15 ppm). Alligators fed mice Blood samples were obtained from the occipital sinus injected with either 1000 or 2000 ppm SeMet (lg SeMet/g 1 month before treatment to determine baseline Se levels mouse dry weight) received an average dose per week (Finger et al. 2015a). At least 1 mL of whole blood was (based on their own body weight) of 1.22 (range stored at -60 °C for subsequent trace element analysis. 0.90–1.51 ppm) or 1.65 (range 1.47–1.90 ppm) ppm Se, respectively (lg SeMet/g alligator), which is similar to Se Concentration Analysis Hopkins et al. (2004). All alligators were subjected to 7 weekly doses (Octo- To determine blood Se concentrations, 1 mL of blood was ber–December 2014). On feeding days, alligators were aliquoted into a 15-mL trace metal-free tube. Blood sam- caught and each individual alligator was administered one ples were then digested in 1 mL of 70% trace metal grade mouse injected with a solution corresponding to their nitric acid (HNO3) and 1 mL of 30% hydrogen peroxide dietary treatment. Swallowing was facilitated by manually (H2O2)at70°C for 1 h. Following digestion, samples were rubbing along the pharynx to prevent regurgitation. Alli- brought to a total volume of 4 mL and then allowed to gators were then released into their respective pens. cool. After cooling, all samples were brought to a total volume of 15 mL with 18-MX DI water and then analyzed on an inductively coupled plasma mass spectrometer (ICP- Morphometrics, Euthanasia, and Mortalities MS; Perkin-Elmer, Norwalk, CT) to determine blood Se. The MDL for Se in blood samples was 0.23 ppm. Blood Se Alligators were captured and sampled periodically values are reported on wet mass basis. throughout this study to determine the effects of SeMet on Liver and kidney tissue samples were freeze-dried, size, growth, and body condition. After alligators had finely ground, and subsequently digested in 10 mL of 70% resided in pens for more than 3 months to allow for HNO3 using microwave digestion (MarsExpress, CEM habituation (April 2014; 6 months before SeMet treat- Corp., Matthews, NC; Tuberville et al. 2016). Following ment), they were captured and measured for head length digestion, all tissue samples were brought to a final volume (HL; nearest cm). One week before commencing dietary of 15 mL with 18-MX DI water. Subsequently, samples treatment administration and 1 week after the last admin- were analyzed using three separate runs on an ICP-MS to istration, all alligators were measured (HL) and weighed determine tissue Se. Mean Se MDL across the three sep- (kg). A body condition index (BCI) using Fulton’s condi- arate runs was 0.222 ppm (dry mass). Liver and kidney Se tion factor (K = W/L3 9 10n, where W = mass in kg, concentrations are reported in ppm on a dry-mass basis. L = HL in cm, and n = 3) was generated for each indi- Certified reference material (TORT-3; National vidual based on their HL and mass (Hamilton et al. 2016b; Research Council) was included in each analysis for Zweig 2003). quality control. Mean percent recovery for Se in reference One week after the seventh and final feeding with 1000 material was 143.66%. and 2000 SeMet (4–5 December 2014), alligators were euthanized via a blow to the head and subsequent cervical Statistical Analysis dislocation and pithing (Nevarez et al. 2014). Liver and kidney tissues were then removed and stored at -60 °C for All statistical analyses were performed in JMP Pro 12.01 subsequent trace element analysis. (SAS, Cary, NC, USA). Linear regression was used when 123 442 Arch Environ Contam Toxicol (2017) 72:439–448 comparing Se concentrations in the liver and kidney, examine interpen size variation. There were significant dif- growth rates, and for certain morphometrics. Linear mixed ferences in HL among the treatment groups (p = 0.0106) models (LMM) with a random effect of AnimalID (ex- 6 months before starting SeMet treatment: control alligators plained below) were used to analyze repeated morphome- were significantly longer in HL (p = 0.0030; simple effect tric samplings. Binary logistic regression was used to size: 0.788 ± 0.235 cm) than 1000 ppm alligators. How- analyze how Dietary Treatment affected alligator ever, no other differences (p [ 0.05) were observed (raw Mortality. mean HL ± SE: 1000 ppm group, 10.99 ± 0.20 cm; Dietary treatment (control, 1000 and 2000 ppm) of 2000 ppm group, 11.45 ± 0.17 cm; control group, 11.78 ± SeMet was used as a factor in analysis to account for any 0.12 cm). treatment differences. Time was included as a factor to investigate morphometric changes in alligators throughout Pre-treatment Blood Selenium Concentrations the course of the study (October vs. December). Dietary treatment also was used as a factor to determine inter-pen Expectedly, no pretreatment differences (p = 0.1639) in variation in blood Se concentrations prior to starting SeMet blood Se were observed among treatment groups (raw blood treatment. AnimalID was included as a random effect in Se mean ± SE: 1000 ppm group, 0.52 ± 0.04 ppm; analyses to link multiple samplings on an individual (Fin- 2000 ppm group, 0.47 ± 0.04 ppm; control group, 0.26 ± ger et al. 2013, 2015a). Sex was not included in modelling 0.14 ppm) 1 month before commencing SeMet treatment. due to the disproportionate number of females (n = 20) to males (n = 4). Liver and Kidney Selenium Levels When analyzing Se accumulation, outcome variables included kidney Se and liver Se. Post-treatment HL (i.e., Liver Se levels were significantly affected by dietary treat- HL at euthanasia) was used as a covariate to account for ment (p \ 0.0001) and the covariate post-treatment HL differences in final size among Dietary Treatments when (p = 0.0010). Levels of Se in the liver of alligators fed examining only liver and kidney Se accumulation. HL 2000 ppm were significantly higher than those fed 6 months before commencing SeMet treatment served as 1000 ppm (p = 0.0007; simple effect size: 24.09 ± an outcome variable to examine inter-pen size variation. 5.88 ppm) or no SeMet (p \ 0.0001; simple effect size: BCI and HL (as an indicator of growth) were included as 57.28 ± 5.91 ppm; Fig. 1). Se levels in the liver also were outcome variables in LMM analysis. HL growth rate was significantly higher in alligators fed 1000 ppm than those fed calculated by the following equation: no SeMet (p \ 0.0001; simple effect size: 33.19 ± ðÞFinal HL À Initial HL 4.78 ppm; Fig. 1). The covariate post-treatment HL had a : Initial HL negative effect on liver Se: for every 1-cm increase in post- treatment HL there was a decrease of 16.04 (±4.10) ppm Se When differences were detected among groups (i.e., dietary in the liver (RC). treatment or time), post hoc multiple comparisons were Both dietary treatment (p \ 0.0001) and the covariate made using t tests. Unstandardized effect sizes, including post-treatment HL (p = 0.0142) significantly affected regression coefficients (hereafter denoted as ‘‘RC’’) ± s- kidney Se. Kidney Se levels were significantly higher in tandard errors (1 SE) and simple effect sizes (i.e., differences alligators fed both 1000 (p \ 0.0001; simple effect size: in group means) ± 1 SEs, are presented and distinguished 95.31 ± 7.62 ppm) and 2000 ppm (p \ 0.0001; simple below to indicate the size of a particular effect. When pre- effect size: 102.63 ± 9.41 ppm) SeMet than those fed no sented, raw or predicted means ± 1 SEs are explicitly SeMet (Fig. 1). However, there was no difference in kid- indicated below to distinguish between regression coeffi- ney Se between alligators fed 1000 and 2000 (p = 0.4454) cients and effects. Since confidence is asymmetrical in ppm SeMet (Fig. 1). Similar to liver Se, the covariate post- logistic regression, odds ratios (b) are indicated with corre- treatment HL also had a negative effect on kidney Se: for sponding 95% confidence limits (CLs; b ± upper CL— every 1-cm increase in post-treatment HL, there was a lower CL). The significance level was set at a = 0.05. decrease of 17.73 (±6.53) ppm Se in the kidney (RC).

Effect of SeMet Treatment on Condition Results A linear mixed model with the random effect of GatorID Morphometrics After Habituation (i.e., a unique identiﬁer) to link repeated observations (i.e., pre- vs. post-treatment) on an individual was used to After alligators were allowed to habituate to pens, HL was understand the effect SeMet Dietary Treatment on alligator measured 6 months before starting SeMet treatment to better body condition. BCI was negatively affected by time 123 Arch Environ Contam Toxicol (2017) 72:439–448 443

Effect of SeMet Treatment on Growth

A linear mixed model (GatorID as random effect; see above) was used to understand the effect SeMet Dietary Treatment on growth. Used as an indicator of growth, this analysis incorporated both pre-treatment and post-treatment HL measurements. Repeated measures analysis of growth revealed a significant effect of both time (p = 0.0245) and dietary treatment (p = 0.0472), as well as a significant time 9 dietary treatment interaction (p = 0.0003). In contrast to BCI, treatment with 1000 (p \ 0.0001; simple effect size: 0.325 ± 0.06 cm) or 2000 ppm (p = 0.0075; simple Fig. 1 Effect of dietary treatment with selenomethionine on selenium effect size: 0.221 ± 0.07 cm) SeMet for 7 weeks signifi- accumulation in alligators. Bars are raw means (±1 SE) of respective cantly increased alligator HL growth (Fig. 3). However, treatments. Selenium (Se) levels are reported in ppm (dry weight) feeding alligators control mice for 7 weeks had no effect on HL growth (p = 0.0967; Fig. 3). Variations in HL among dietary treatment groups were (p = 0.0354); alligators had a higher condition before observed both before and after treatment. Before treatment, treatment than after treatment (simple effect size: HL in alligators fed 2000 ppm SeMet was already longer 0.07 ± 0.03), but BCI was not affected by SeMet dietary than in those fed 1000 ppm SeMet (p = 0.0081; simple treatment (p = 0.6708). To better understand which dietary effect size: 0.725 ± 0.25 cm). After 7 weeks of treatment, treatment groups contributed to this negative effect of Time HL in alligators fed 2000 ppm SeMet was still significantly on BCI, we examined the interaction of dietary treatment longer than in those fed 1000 ppm SeMet (p = 0.0216; and time. The dietary treatment 9 time interaction had a simple effect size: 0.621 ± 0.25). Control alligators and significant effect on BCI (p = 0.0118). Post-hoc analysis 2000 ppm SeMet alligators (p = 0.2419) were similar in (individual t tests) revealed treatment differences between size before treatment, but after 7 weeks of treatment, HL in each time period that were masked when dietary treatment alligators fed 2000 ppm SeMet was significantly longer was examined alone. Treatment with 1000 (p = 0.0029; than those fed control mice (p = 0.0193; simple effect simple effect size: 0.14 ± 0.04) or 2000 ppm (p = 0.0316; size: 0.634 ± 0.25 cm). No difference in HL was observed simple effect size: 0.12 ± 0.05) SeMet for 7 weeks sig- between alligators administered 1000 ppm SeMet or con- nificantly reduced BCI (Fig. 2). In contrast, condition of trol mice before the start (p = 0.1024) or after 7 weeks of control alligators did not change over the 7-week treatment treatment (p = 0.9605). period (p = 0.3184; Fig. 2). There were no differences in To account for size variations observed between treat- BCI among the three dietary treatment groups before the ment groups both before and after SeMet treatment, HL start of the 7-week treatment (p [ 0.05). growth rates were computed and the effect of dietary treatment on growth rate was examined. During the 7-week treatment period, the HL growth rate of alligators fed 2000 (p = 0.0036) or 1000 ppm SeMet (p \ 0.0001) was

Fig. 2 Effect of dietary treatment with selenomethionine on body Fig. 3 Effect of dietary treatment with selenomethionine on growth condition index in alligators. Circles are raw mean (±1 SE) body in alligators. Circles are raw mean (±1 SE) head lengths (HLs) of condition indices (BCIs) of alligators from each dietary treatments alligators from each dietary treatments before and after the 7-week before and after the 7-week feeding treatment feeding treatment 123 444 Arch Environ Contam Toxicol (2017) 72:439–448 significantly higher than that of control alligators. How- treatment HL had a negative effect on Se concentrations in ever, no difference in growth rate was observed between both the liver and kidney, whereby larger alligators alligators fed 2000 or 1000 ppm SeMet (p = 0.2288). exhibited lower Se levels in both tissues. Similarly, we also observed a negative relationship between Se levels and size Effect of Dietary Treatment on Mortality in our previous study where alligators were fed CCW- contaminated prey for 25 months (Tuberville et al. 2016). Binary logistic regression was used to investigate the effect One possible reason for the negative effect of post-treat- of dietary treatment on mortality. Control alligators were ment HL on Se levels observed in this study is that all significantly more likely to survive than alligators adminis- individuals within a respective treatment received an tered 1000 (p \ 0.0001; b ± upper CL—lower CL, equivalent dose (lg SeMet/g mouse dry weight), regardless 3370.21 ± 415.55—34,206.12) or 2000 ppm SeMet of size. In other words, smaller alligators would have (p \ 0.0001; b ± upper CL—lower CL, 26,347.70 ± received a relatively higher dose (in relation to body size; 559.04—124,074.27). While alligators administered lg SeMet/g alligator) than larger alligators and, as such, 1000 ppm SeMet were more likely to survive than alligators exhibited higher organ Se concentrations. The lack of a administered 2000 ppm SeMet, this was not significant dose-dependent increase in kidney Se may be related to (p = 0.1440; b ± upper CL—lower CL, 6.99 ± 0.57— this, as alligators fed 1000 ppm SeMet were shorter in 86.05). post-treatment HL (raw mean ± SE, 12.81 ± 0.26 cm) and weighed less at the end of treatment (raw mean weight ± SE, 3.27 ± 0.32) than those fed 2000 ppm Discussion SeMet (raw mean HL ± SE, 13.5 ± 0.11 cm; raw mean weight ± SE 3.99 ± 0.16). Another possible contributor to Selenium Accumulation this negative effect of post-treatment HL may be related to somatic growth dilution (i.e., faster growing alligators This is the first study to investigate ingestion of dietary Se accumulated less Se relative to their body weight than in the form of SeMet and subsequent Se accumulation in slower growing alligators; see Tuberville et al. 2016). the American alligator, a top-trophic, long-lived species in The Se levels observed in this study are substantially the southeastern United States. Our results demonstrate that higher than those reported previously in alligators the feeding of alligators 1000 or 2000 ppm SeMet for a (Campbell et al. 2010; Roe et al. 2004; Tuberville et al. period of 7 weeks caused significant Se accumulation in 2016; Fig. 1). Se concentrations from hatchling alligators both the liver and the kidney through trophic transfer inhabiting a CCW-polluted swamp were [7 ppm (dry (Fig. 1). Similar to our previous study where alligators mass; Roe et al. 2004), whereas the highest liver and kid- were fed CCWs for 2 years (Tuberville et al. 2016), Se ney Se concentrations from alligators fed CCW-contami- accumulation was more pronounced in the kidney relative nated prey for 25 months were 17.47 and 28.61 ppm (dry to the liver (Fig. 1). Higher Se partitioning in the kidney, mass), respectively (Tuberville et al. 2016; Finger et al. relative to liver, also has been observed in brown house 2016). Because of the high doses of SeMet employed in snakes (Lamprophis fuliginosus) fed SeMet for 10 months this study, questions may arise about the biological rele- and slider turtles (Trachemys scripta) fed SeMet for vance of this dosage regimen. For example, liver Se in 8 weeks (Dyc et al. 2016; Hopkins et al. 2004). water snakes (Nerodia fasciata) inhabiting CCW-contam- There was a dose-dependent increase in liver Se, with inated wetlands on the SRS were [140 ppm Se (Hopkins alligators fed 2000 ppm SeMet accumulating higher Se et al. 1999b). Therefore, alligators would need to consume levels in the liver than those fed 1000 ppm SeMet. How- almost 10 of these snakes to equate to mouse dosages ever, this was not observed in the kidney. No difference in employed in this study. Levels of Se in other possible prey kidney Se accumulation was observed between alligators items in CCW-contaminated sites have been described fed 1000 or 2000 ppm SeMet. Similarly, we previously previously in a number of studies (summarized in Rowe observed no difference in kidney Se between alligators fed et al. 2002). For example, mosquitofish (Gambusia hol- CCWs 2 or 3 times a week for 25 months (Tuberville et al. brooki) and crayfish (Procambarus and Cambarus sp.) 2016). In contrast, previous studies in turtles and snakes captured from ash basins on the SRS were more than 15 have observed dose-dependent increases in kidney Se fol- and 52 ppm Se, respectively (Tuberville et al. 2016). Even lowing SeMet administration (Hopkins et al. 2004; Dyc though our overall study design was modelled after Hop- et al. 2016). kins et al. (2004), whereby snakes were fed 10 or 20 ppm Post-treatment HL was used as a covariate to account for SeMet (lg of SeMet/g mouse), necessary adjustments were size differences between individual alligators when ana- employed to investigate a Se accumulation in a longer- lyzing Se tissue burdens in the liver and kidney. Post- lived, much larger species. Briefly, alligators were 123 Arch Environ Contam Toxicol (2017) 72:439–448 445 administered mice injected with a much higher dose (1000 an important role in combating oxidative stress and in and 2000 ppm SeMet; lg of SeMet/g mouse) than used thyroid hormone metabolism (Kohrle et al. 2005; Lu and previously to mitigate size differences (i.e., mass) between Holmgren 2008). In particular, the iodothyronine deiode- snakes and alligators (Hopkins et al. 2004), enable efficient nases are an important Se-containing family of enzymes and reliable feeding of a target dose (i.e., force feeding), that mediate thyroid metabolism. Type 1 (D1) and Type 2 and accelerate accumulation in a long-lived vertebrate and (D2) iodothyronine deiodenases convert inactive, prohor- consequently allow for the determination of the level of Se mone thyroxine (T4) into active triiodothyronine (T3; required to induce detrimental effects. This higher dosage Beckett and Arthur 2005; Drutel et al. 2013; Kohrle et al. regimen also minimized stressors associated with capture 2005). Because T3 is an important regulator of growth and and force-feeding because feeding took place only once a metabolism, SeMet treatment may have increased T3 week. This study provides a firm foundation for future levels through increased expression of D1 and D2 (Gross investigation of Se accumulation and its consequent effects et al. 1995; Kohrle et al. 2005; Lin et al. 2014), conse- in alligators. However, the high dosages employed herein quently leading to an increase in alligator size (Thompson also warrant further investigation of how more relevant et al. 1995). Although T3 was not measured in this study, in dosages, administered for a longer period of time, may chickens (Gallus gallus), T3 levels increased and growth affect Se accumulation. was promoted following supplementation with Se (Jianhua et al. 2000). It also is possible that SeMet treatment Effects of Selenium on Growth, Condition, increased appetite (Bunk and Combs 1980), leading to an and Mortality increase in size of alligators fed SeMet. In alligators supplemented with sodium selenite, no differences in growth Throughout the course of this study, alligators were mea- rate were observed relative to controls (Elsey and Lance sured for HL and mass to better understand the effects of 1983). However, this may have been related to the lower Se accumulation on growth and condition. Our results dosage (1 ppm) and the species of Se provided to alliga- suggest that Se accumulation affected both alligator growth tors, as SeMet is more readily accumulated than inorganic and condition. However, the effects are disparate. Although forms of Se, which are more easily excreted (Fan et al. alligators fed 1000 or 2000 ppm SeMet increased in HL 2002; Schrauzer 2000; Thomson 1998). (Fig. 3; i.e., their growth rate was higher than control In contrast, the negative effects of SeMet treatment on alligators), the condition of alligators fed 1000 or condition seem likely to reflect Se toxicity. Although Se 2000 ppm SeMet declined over the 7-week period (Fig. 2), enhances skeletal growth (see above), in excess, Se can which means that they either lost weight over the feeding inhibit growth and reduce condition (Ackerman and period or did not gain weight at the same rate as they Eagles-Smith 2009; De Riu et al. 2014; Heinz et al. 1987; gained length. This suggests that although SeMet had no Hopkins et al. 2005). In fact, high levels of Se are asso- detrimental impacts on skeletal growth and in fact ciated with decreased levels of growth hormone, insulin enhanced skeletal growth, this increased growth was not like growth factor-1, and insulin-like growth factor-binding associated with a concomitant increase in body condition. proteins (Grønbæk et al. 1995; Kohrle et al. 2005), leading Because this study was conducted in the cooler fall to growth retardation. In male fence lizards (Sceloporus months in the southeastern United States, a possible occidentalis) fed SeMet-contaminated prey, Se accumula- explanation for decreased condition is that ambient factors tion was associated with reduced condition (Hopkins et al. induced behavioral changes in appetite. However, it is 2005). There also was a negative relationship between liver highly unlikely that any ambient factors influenced condi- Se and body condition in wild-caught Forster’s terns tion because changing photoperiods appear to have no (Sterna forsteri; Ackerman and Eagles-Smith 2009). effect on alligator appetite (references within Chabreck and Exposure to high levels of Se may affect the hypothalamic– Joanen 1979), all alligators were housed in a climate- pituitary–adrenal axis by increasing levels of glucocorti- controlled facility (Finger et al. 2015b), and no differences coids (Hopkins et al. 1999a), which in turn may suppress in condition were observed in control alligators after the condition. 7-week period. Therefore, this suggests that SeMet During the latter stages of this study, five alligators died. administration over the 7-week period, and the subsequent While toxicity thresholds are not established for alligators, accumulation of Se was the primary cause of reduced these mortalities indicate that Se accumulated to toxic condition. levels in alligators fed SeMet. In fact, alligators fed 1000 or The increase in growth observed after SeMet treatment 2000 ppm SeMet were significantly more likely to die than may demonstrate the hormetic relationship of Se on control alligators. In other taxa, much lower liver and growth, whereby lower levels of Se are beneficial and kidney Se levels have been associated with mortality and necessary for growth (Harding 2008). Selenoproteins play reproductive impairment (Heinz et al. 1987, 1989; Janz 123 446 Arch Environ Contam Toxicol (2017) 72:439–448 et al. 2010). For example, almost 100% mortality was impact interpretations of growth and condition in studies observed after feeding mallards (Anas platrhynchos) using morphometrics as toxicological endpoints. 100 ppm sodium selenite for 40 days even though liver levels were almost 1/6th the concentration of that observed in alligators in our study (Heinz et al. 1987). This may Conclusions and Future Directions indicate that alligators are less susceptible to Se toxicity. Hypotheses of reptilian tolerance to Se (and other trace This is the first study to administer directly a Se-containing elements) have been suggested by other studies as well compound and measure subsequent accumulation in a (Hopkins et al. 1999b, 2004; Tuberville et al. 2016); crocodilian. The results suggest that chronic SeMet however, this has yet to be fully elucidated. Because administration led to significant Se accumulation. Se reproductive endpoints may be a more sensitive indicator accumulation was associated with a concomitant increase of Se toxicity (Hopkins et al. 2004), if possible, these in size but not condition. Mortalities were associated with should be incorporated into future studies. This would not Se accumulation, especially in alligators fed the highest have been feasible in this study, because all alligators were SeM dosage, suggesting that the levels accumulated by sexually immature. alligators were toxic. Because this is one of the first studies to investigate Se Alligator Size Variation before Treatment in a crocodilian directly, this study should provide a firm foundation for future research. While this study only Alligators were measured 6 months (HL only) and 1 week examined Se accumulation in the kidney and liver, clearly (HL and BCI) before starting SeMet treatment to determine demonstrating their importance in long-term monitoring possible size variations in treatment groups. Variations in (Grillitsch and Schiesari 2010; Tuberville et al. 2016), size (i.e., pre-treatment HL) were observed both 6 months future investigations should aim for the validation and (control alligators larger than 1000 ppm) and 1 week eventual use of nondestructive tissues in monitoring (2000 ppm alligators larger than 1000 ppm) before com- (Hopkins et al. 2001). The disparate effects of Se accu- mencing treatment. Because all alligators were housed for mulation on HL and condition are particularly interesting 2 months within one communal pen (n = 24) for quarantine and necessitate further studies that should be focused at purposes and to simulate husbandry conditions employed at understanding better the relationship between Se and RWR before allocation among the three pens for dietary endocrine hormones, Se hormesis in aquatic organisms, treatment, it is possible that a dominance hierarchy emerged and the impacts of Se on alligator health, immune function, that allowed one (or a few other) individual to monopolize and reproduction. food, resulting in increased growth relative to other conspecifics (Finger et al. 2015a, b; Tuberville et al. 2016). In Acknowledgements We would like extend a special thanks to Ruth fact, in agricultural production settings, crocodiles often are Elsey and her staff at Rockefeller Wildlife Refuge for providing alligators used in this study. Sharon L. Finger assisted in transport of graded (i.e., larger individuals are removed) to facilitate the alligators from LA to SC. William Hopkins (Virginia Tech) provided growth of smaller conspecifics to counteract this (Finger insight on dosage regimens. John Seaman and his lab at SREL ana- et al. 2015a). However, because alligators were randomly lyzed Se concentrations in blood and tissues. Bess Harris, Nicole allocated (and thus, true grading could not be done) to one of White, Nick Bossenbroak (NB), and Caitlin Kupar assisted in feeding. Megan Winzeler, Daniel Quinn, Sam Dean, and David Haskins three pens, we may have unintentionally biased a particular helped with alligator blood sampling and dissections. NB also assisted pen via randomly placing larger individuals in it. Likely, this with processing samples for digestions and trace element analysis. accounts for the initial size discrepancy observed 6 months Meghan D. Kelley helped with statistical analysis. JWF was funded before treatment. by the Department of Environmental Health Science and the Inter- disciplinary Toxicology Program. MTH was funded by a grant from There were no apparent differences in condition among the Area Closures Project to TDT. Support for this work was provided the three groups before commencing SeMet treatment, in part by Award No. DE-FC09-07SR22506 from Department of suggesting that both nutrition and space requirements were Energy to the University of Georgia Research Foundation. All satisfactory, regardless of pen. Alligators were fed in experimental protocols were approved by the Institutional Animal Care and Use Committee at the University of Georgia (Approval No. excess (i.e., until satiation) three times per week before A2014 01-030-Y1-A3). treatment (and 2 times/week throughout treatment; Finger et al. 2015b). At every feeding, attempts were made to spread food equally to all individuals (i.e., to mitigate dominance hierarchies), because more dominant individu- References als (i.e., larger alligators) may exclude subordinates from Ackerman JT, Eagles-Smith C (2009) Selenium bioaccumulation and eating (Morpurgo et al. 1993). Future studies should body condition in shorebirds and terns breeding San Francisco investigate intra-pen social interactions and how this may Bay, California, USA. Environ Toxicol Chem 28:2134–2141 123 Arch Environ Contam Toxicol (2017) 72:439–448 447

Beckett GJ, Arthur JR (2005) Selenium and endocrine systems. Hamilton MT, Kupar CA, Kelley MD, Finger JW Jr, Tuberville TD J Endocrinol 184:455–465 (2016b) Blood and plasma biochemistry reference intervals for Bunk MJ, Combs GF (1980) Effect of selenium on appetite in the wild juvenile American alligators (Alligator mississippiensis). selenium-deficient chick. J Nutr 110:743–749 J Wildl Dis 52:631–635 Campbell JW, Waters MN, Tarter A, Jackson J (2010) Heavy metal Harding LE (2008) Non-linear uptake and hormesis effects of and selenium concentrations in liver tissue from wild American selenium in red-winged blackbirds (Agelaius phoeniceus). Sci alligator (Alligator mississippiensis) livers near Charleston, Total Environ 389:350–366 South Carolina. J Wildl Dis 46:1234–1241 Haug A, Graham RD, Christophersen OA, Lyons GH (2007) How to Chabreck RH, Joanen T (1979) Growth rates of American alligators use the world’s scarce selenium resources efficiently to increase in Louisiana. Herpetologica 35:51–57 the selenium concentration in food. Microb Ecol Health Dis De Riu N, Lee JW, Huang SS, Moniello G, Hung SS (2014) Effect of 19:209–228 dietary selenomethionine on growth performance, tissue burden, Heinz GH, Hoffman DJ, Krynitsky AJ, Weller DMG (1987) and histopathology in green and white sturgeon. Aquat Toxicol Reproduction in mallards fed selenium. Environ Toxicol Chem 148:65–73 6:423–433 Drutel A, Archambeaud F, Caron P (2013) Selenium and the thyroid Heinz GH, Hoffman DJ, Gold LG (1989) Impaired reproduction of gland: more good news for clinicians. Clin Endocrinol (Oxf) mallards fed an organic form of selenium. J Wildl Manag 78:155–164 53:418–428 Dyc C, Far J, Gandar F, Poulipoulis A, Greco A, Eppe G, Das K Hopkins WA, Mendonca MT, Congdon JD (1999a) Responsiveness (2016) Toxicokinetics of selenium in the slider turtle, Trachemys of the hypothalamo-pituitary-interrenal axis in an amphibian scripta. Ecotoxicology 25:727–744 (Bufo terrestris) exposed to coal combustion wastes. Comp Elsey RM, Lance VA (1983) Effect of diet on blood selenium and Biochem Physiol 122:191–196 glutathione peroxidase activity in the alligator. Comp Biochem Hopkins WA, Rowe CL, Congdon JD (1999b) Elevated trace element Physiol 76B:831–837 concentrations and standard metabolic rate in banded water Fan TWM, Teh SJ, Hinton DE, Higashi RM (2002) Selenium snakes (Nerodia fasciata) exposed to coal combustion wastes. biotransformations in proteinaceous forms by foodweb organ- Environ Toxicol Chem 18:1258–1263 isms of selenium-laden drainage waters in California. Aquat Hopkins WA, Roe JH, Snodgrass J, Jackson BP, Kling DE, Rowe CL, Toxicol 57:65–84 Congdon JD (2001) Nondestructive indices of trace element Finger JW Jr, Gogal RM Jr (2013) Endocrine-disrupting chemical exposure in squamate reptiles. Environ Pollut 115:1–7 exposure and the American alligator: a review of the potential Hopkins WA, Roe JH, Snodgrass J, Staub BP, Jackson BP, Congdon role of environmental estrogens on the immune system of a top JD (2002) Effects of chronic dietary exposure to trace elements trophic carnivore. Arch Environ Contam Toxicol 65:704–714 on banded water snakes (Nerodia fasciata). Environ Toxicol Finger JW Jr et al (2013) Using phytohaemagglutinin to determine Chem 21:906–913 immune responsiveness in saltwater crocodiles (Crocodylus Hopkins WA, Staub BP, Baionno JA, Jackson BP, Roe JH, Ford NB porosus). Aust J Zool 61:301–311 (2004) Trophic and maternal transfer of selenium in brown house Finger JW Jr, Thomson PC, Adams AL, Benedict S, Moran C, Isberg snakes (Lamprophis fuliginosus). Ecotox Environ Safe SR (2015a) Reference levels for corticosterone and immune 58:285–293 function in farmed saltwater crocodiles (Crocodylus porosus) Hopkins WA, Snodgrass J, Baionno JA, Roe JH, Staub BP, Jackson hatchlings using current Code of Practice guidelines. Gen Comp BP (2005) Functional relationships among selenium concentra- Endocrinol 212:63–72 tions in the diet, target tissues, and nondestructive tissue samples Finger JW Jr, Williams RJ, Hamilton MT, Elsey RM, Oppenheimer of two species of snakes. Environ Toxicol Chem 24:334–351 VA, Holladay SD, Gogal RM Jr (2015b) Influence of collection Janz D, DeForest DK, Brooks ML, Chapman PM, Gilron G, Hoff D, time on hematologic and immune markers in the American Hopkins WA, McIntyre DO, Mebane CA, Palace VP, Skorupa alligator (Alligator mississippiensis). J Immunoassay Immuno- JP, Wayland M (2010) Selenium toxicity to aquatic organisms. chem 36:496–509 In: Chapman PM et al (eds) Ecological assessment of selenium Finger JW Jr, Hamilton MT, Metts BS, Glenn TC, Tuberville TD in the aquatic environment. Society of Environmental Toxicol- (2016) Chronic ingestion of coal fly-ash contaminated prey and ogy and Chemistry, Pensacola, pp 139–230 its effects on health and immune parameters in juvenile Jianhua H, Ohtsuka A, Hayashi K (2000) Selenium influences growth American alligators (Alligator mississippiensis). Arch Environ via thyroid hormone status in broiler chickens. Br J Nutr Contam Toxicol 71:347–358 84:727–732 Grillitsch B, Schiesari L (2010) The ecotoxicology of metals in Kohrle J, Jakob F, Contempre B, Dumont JE (2005) Selenium, the reptiles. In: Sparling DW, Linder G, Bishop CA, Krest S (eds) thyroid, and the endocrine system. Endocr Rev 26:944–984 Ecotoxicology of amphibians and reptiles, 2nd edn. CRC Press, Lemly AD (2002) Symptoms and implications of selenium toxicity in Boca Raton, pp 337–448 fish: the Belews Lake case example. Aquat Toxicol 57:39–49 Grønbæk H, Frystyk J, Ørskov H, Flyvbjerg A (1995) Effect of Lemly AD (2004) Aquatic selenium pollution is a global environ- sodium selenite on growth, insulin-like growth factor-binding mental safety issue. Ecotox Environ Safe 59:44–56 proteins and insulin-like growth factor-I in rats. J Endocrinol Lemly AD, Skorupa JP (2012) Wildlife and the coal waste policy 145:105–112 debate: proposed rules for coal waste disposal ignore lessons Gross M, Oertel M, Kohrle J (1995) Differential selenium-dependent from 45 years of wildlife poisoning. Environ Sci Technol expression of type I 50-deiodinase and glutathione peroxidase in 46:8595–8600. doi:10.1021/es301467q the porcine epithelial kidney cell line LLC-PK1. Biochem J Lin SL, Wang CW, Tan SR, Liang Y, Yao HD, Zhang ZW, Xu SW 306:851–856 (2014) Selenium deficiency inhibits the conversion of thyroidal Hamilton MT, Finger JW, Winzeler ME, Tuberville TD (2016a) thyroxine (T4) to triiodothyronine (T3) in chicken thyroids. Biol Evaluating the effect of sample type on American alligator Trace Elem Res 161:263–271 (Alligator mississippiensis) analyte values in a point-of-care Lu J, Holmgren A (2009) Selenoproteins. J Biol Chem 284:723–727 blood analyser. Conserv Physiol 4:1–7

123 448 Arch Environ Contam Toxicol (2017) 72:439–448

Milnes MR, Guillette JLJ (2008) Alligator tales: new lessons about Schrauzer GN (2000) Selenomethionine: a review of its nutritional environmental contaminants from a sentinel species. Bioscience significance, metabolism and toxicity. J Nutr 130:1653–1656 58:1027–1036 Thompson KM, Haibach H, Sunde RA (1995) Growth and plasma Morpurgo B, Gvaryahu G, Robinzon B (1993) Aggressive behaviour triiodothyronine concentrations are modified by selenium defi- in immature captive Nile crocodiles (Crocodylus niloticus)in ciency and repletion in second-generation selenium-deficient relation to feeding. Physiol Behav 53:1157–1161 rats. J Nutr 125:864–873 Nevarez JG, Strain GM, da Cunha AF, Beaufrere H (2014) Evaluation Thomson CD (1998) Selenium speciation in human body fluids. of four methods for inducing death during slaughter of American Analyst 123:827–831 alligators (Alligator mississippiensis). Am J Vet Res 75:536–543 Tuberville TD, Scott DE, Metts BS, Finger JW Jr, Hamilton MT Roe JH, Hopkins WA, Baionno JA, Staub BP, Rowe CL, Jackson BP (2016) Hepatic and renal trace element concentrations in (2004) Maternal transfer of selenium in Alligator mississippien- American alligators (Alligator mississippiensis) following sis nesting downstream from a coal-burning power plant. chronic dietary exposure to coal fly ash contaminated prey. Environ Toxicol Chem 23:1969–1972 Environ Pollut 214:680–689 Rowe CL, Hopkins WA, Congdon JD (2002) Ecotoxicological Zweig CL (2003) Body condition index analysis for the American implications of aquatic disposal of coal combustion residues in alligator (Alligator mississippiensis). University of Florida, the United States: a review. Environ Monit Assess 80:207–276 Gainesville

123