Physiological validation of enzyme immunoassay of fecal glucocorticoid metabolite levels and diurnal variation measured in captive Black-tufted Marmoset Callithrix penicillata (Mammalia: Primates: Callitrichidae)

Measuring stress responses is an important aspect for the conservation of endangered wild species. Non-invasive measuring of glucocorticoid metabolite levels has become an important tool to measure stress intensity. The aims of the present study were as follows: to validate the enzyme immunoassay to measure the concentration of fecal metabolites of glucocorticoids (FGM) after stressful stimuli and to determine whether FGM concentrations fluctuate diurnally in Black-tufted Marmosets Callithrix penicillata in captivity. Eight captive healthy adult Black-tufted Marmosets (four males and four females) were included in the study. The animals were subjected to three treatments: (1) hormone challenge with adrenocorticotropic hormone (ACTH), (2) saline administration and (3) control treatment to monitor diurnal changes of FGM. Fecal samples were collected on days -1, 0, +1 and +2, with intramuscular administration of ACTH and saline performed on day 0. To control diurnal variations, all feces from all animals were collected over six consecutive days and identified using the time of defecation and animal identification number. There were four designated two-hour periods per day (8–10 h, 10–12 h, 12–14 h and 14–16 h), and the samples were grouped for each two-hour period to obtain a representative pool. The samples were frozen, and the metabolite concentrations were measured by enzyme immunoassay following extraction. The results show that immunoassay measurements of FGM concentrations in C. penicillata can be validated physiologically. Diurnal variation of the FGM concentration was observed, with significantly increased FGM levels in the early afternoon in both sexes. The mean FGM concentration was higher in captive females than in males. Physical restraint followed by saline administration led to adrenocortical stimulation similar to that observed following ACTH hormone challenge, a finding that has not previously been reported in C. penicillata. Our results show immunoassay measurements of FGM concentrations provide a valuable tool for the non-invasive study of the endocrine correlates of behavior and wellbeing of this species.


INTRODUCTION
Many studies in the past, have focussed on understanding the implications of stress on vertebrate species better. As described by Heistermann et al. (2006), chronic stress may have deleterious effects on physiological, psychosocial and behavioral parameters. Chronic stress may also compromise immune functions by inhibiting the release of cytokines and lymphocytes and consequently increasing disease susceptibility (Sapolsky 2002;Wingfield 2005). Other studies have reported suppressed reproductive functions (Wingfield & Sapolsky 2003) and effects on cognitive abilities (Ohl & Fuchs 1999) and behavior (Wielebnowski et al. 2002) resulting from chronic stress. The combination of animal behavior analysis and the measurement of glucocorticoid levels is a promising approach (Pizzutto et al. 2010) to study the impacts of stress (Romero 2004).
Glucocorticoids are steroid hormones produced by the adrenal glands, and the major glucocorticoids are cortisol and corticosterone. Following the immediate adrenergic response that releases catecholamines in situations of acute stress, glucocorticoids increase the amount of energy available to the body for necessary physiological functions, such as locomotion (Sapolsky 2002;Reeder & Kramer 2005).
In addition to mediating the stress response, glucocorticoids are essential to regulate energy availability based on activity patterns. For example, diurnal animals with highly circadian activity patterns have increased glucocorticoid levels at the end of the dark cycle, which provides energy for locomotion and foraging activities. In nocturnal animals, the rise in glucocorticoid levels occurs at the end of the light cycle (Reeder & Kramer 2005;. It is, therefore, critical to take into account the circadian rhythms of a species when standardizing the time of sample collection for glucocorticoid measurements (Heintz et al. 2011). In this vein, the application and validation of non-invasive hormone measurement techniques reflect the need to complement behavioral data with endocrine-physiological information (Heintz et al. 2011).
Glucocorticoids and their metabolites can be reliably measured in the blood, urine and feces of many species using immunoenzymatic assays (Romano et al. 2010;Sheriff et al. 2011). For the study of stress physiology in wild animals, non-invasive methods based on the measurement of glucocorticoid metabolites in urine or feces have become the preferred approach because many samples can be collected over long periods of time in a safe and non-stressful manner. The physical restraints required for the serial collection of blood samples are undesirable because they induce stress and are sometimes difficult to perform due to the physical and behavioral characteristics of a given species (Heistermann et al. 2006;Rimbach et al. 2013). When analyzing changes in the concentrations of glucocorticoid metabolites, particularly in the feces, the pulsatile changes that occur in the blood levels of these metabolites over a period of hours or days have to be taken into account by frequently measuring metabolite concentrations (Whitten et al. 1998;Rimbach et al. 2013). These non-invasive methods have been widely used in reproductive (Strier & Ziegler 1994Brown et al. 1995;Schwarzenberger et al. 1996;Ziegler et al. 1997), behavioral and welfare studies (Pizzutto et al. 2008a(Pizzutto et al. ,b, 2010. Despite the tremendous progress in the application of non-invasive monitoring techniques, several problems related to urinary or fecal metabolite measurements remain unresolved (Higham et al. 2010). For example, the considerable differences in steroid metabolism in different species make it quite difficult to determine the prevalent metabolite (fecal or urinary) and the appropriate antibody for measurement assays ). Therefore, validation in different species is critical for the different methods to measure glucocorticoid metabolites , particularly for techniques involving the application of exogenous adrenocorticotropic hormone (ACTH).
There is a lack of information in literature regarding the physiology of the Black-tufted Marmoset Callithrix penicillata (Image 1) a neotropical primate species with a great potential as a model for reproductive and Image 1. Black-tufted Marmoset Callithrix penicillata behavioral studies. The aims of the present study were as follows: (1) to validate the assay used to measure fecal glucocorticoid metabolites (FGM) in Blacktufted Marmosets and (2) to determine whether the concentrations of these metabolites fluctuate diurnally.

Animals
Eight adult Black-tufted Marmosets (C. penicillata) (four males and four females) were included in the study. All had intact reproductive systems and had been kept in the Wild Animal Rehabilitation Center (Centro de Reabilitação de Animais Silvestres, CRAS) of the Department of Parks and Green Areas (Departamento de Parques e Áreas Verdes, DEPAVE III) of the municipality of São Paulo, São Paulo, Brazil. The animals were separated into pairs and kept in 120cm x 60cm x 60cm cages. While all of the animals had auditory, visual and olfactory contact with each other, physical contact was restricted to their respective partners. Following the standard protocols of the supporting institution food was provided twice a day, in the early morning and late afternoon. The diet consisted primarily of seasonal fruits, primate feed (MegaZoo®, Minas Gerais, MG, Brazil), meal worms, crickets and eggs. Water was available ad libitum. All of the animals had been in captivity for at least two years by the time the study was performed.

Physiological validation
All of the animals were subjected consecutively to the three following treatments: Treatment 1-ACTH: hormone challenge with ACTH administration; Treatment 2-saline: saline administration; and Treatment 3-control: no intervention. Treatment 3 was used to monitor diurnal changes.
Treatment 1-hormone challenge with ACTH administration: Over four consecutive days, all feces from all animals were collected and labelled with the time of defecation and animal identification number. With day 0 corresponding to the day of ACTH administration, the collection of fecal samples began on day-1 and continued on days 0, +1 and +2. As it was unpredictable to determine when each individual would defecate, there were four designated two-hour periods per day (8-10 h, 10-12 h, 12-14 h and 14-16 h), and the samples were grouped for each two-hour period from each animal to obtain a representative pool. On day 0 at 8-10 h, the animals were physically restrained for intramuscular (IM) administration of 0.3 ml Synacthen Depot® of a 0.1 % solution (0.1g/100ml), which corresponds to 31 IU/kg ACTH (Heistermann et al. 2006).
Treatment 2-saline administration: Over four consecutive days, all feces from all animals were collected and labelled with the time of defecation and animal identification number. Collection began on day-1 and continued on days 0, +1 and +2. There were four designated two-hour periods per day (8-10 h, 10-12 h, 12-14 h and 14-16 h), and the samples from each animal were grouped for each two-hour period to obtain a representative pool. On day 0 at 8-10 h, the animals were physically restrained for intramuscular administration of 0.3ml of a saline solution 0.9%.
Treatment 3-control to monitor diurnal changes: Over six consecutive days, all feces from all animals were collected and labelled with the time of defecation and animal identification number. There were four designated two-hour periods per day (8-10 h, 10-12 h, 12-14 h and 14-16 h), and the samples were grouped for each two-hour period to obtain a representative pool. During these six days, the animals were not restrained, and no substances were administered to them. To avoid interference between consecutive treatments and to provide a resting period for the animals, the treatments were separated by one-week intervals.
A 2ml dose of Arcolor® food coloring for chocolate was mixed with fruit juice orally administered twice a day to all females to differentiate their feces from that of the males in the study. Samples were collected from 8-16 h, at the end of the designated two hour periods; after this the cages were placed in a closed space to allow the animals to sleep. Feces were not found during night-time, and the first morning samples were produced shortly after the animals awoke.
All samples were stored in a freezer at -20 0 C until they were processed at the Hormone Dosage Laboratory (

Fecal extraction and sample processing
The metabolites were extracted according to the technique described by Ziegler et al. (1996) with some modifications. For each sample, 0.2g of the lyophilized (freeze-dried) feces were weighed and transferred to an appropriately labeled test tube. A 5ml volume of an 80% ethanol (MERCK®) solution prepared with type I water (Milli-Q®) was added. The tubes were placed in a multivortex device (VWR Scientific Products, VX-2500) for 15 minutes for homogenization, followed by centrifugation (CENTRIBIO, TDL80-2B) for 15 minutes at 3000 rpm. A 1.5ml volume of supernatant (fecal extract) was transferred to a 2.0ml plastic tube and stored in a freezer at -30 0 C.
Prior to the analysis, the samples were diluted in gelatin buffer (pH 7.0) (13.8g NaPO 4 , 9.0g NaCl, 1.0g sodium azide and 1.0 g gelatin in 1000ml distilled water).

Hormone analyses
The samples were analyzed using the parallelism test. Twenty samples with high concentrations were diluted to obtain concentration values similar to those used to develop the calibration curve for the assay. For this purpose fecal samples were serially diluted in the same proportions as the calibration curves for the assay. Depleted fecal extract steroids were used as the matrix. The curve obtained was compared to the calibration curve using a linear regression to assess parallelism of the two curves.
FGM concentrations were measured using an enzyme immunoassay method (EIM) following procedures described for other mammals (Heistermann et al. 2006), using cortisol R4866 monoclonal antibodies (University of California, Davis, USA) as described by Munro & Stabenfeldt (1984).

Statistical analysis
The data were analyzed to assess whether the residues exhibited normal distributions, and a homogeneity of variance test was performed using the Guided Data Analysis application with SAS/LAB® Guided Data Analysis application.
The mean glucocorticoid concentration was determined using all samples of each individual in each group. All samples with values two standard deviations (SD) above the mean were removed. The mean was recalculated, and this process was repeated until all values were within two SD from the mean. The resulting mean value was designated the baseline value, and all values greater than two SD from the mean were considered peak values. This process allows for a statistical assessment of baseline values for biologically relevant hormone concentrations (Heintz et al. 2011).
For each period of diurnal variation studied, the mean FGM concentration and standard deviation were calculated. The means were compared using ANOVA and Tukey's test, both with P<0.05.
The present study was approved by the Bioethics Commission of the School of Veterinary Medicine and Animal Science, São Paulo University (FMVZ, USP), Brazil under procedure number 1470/2008.

RESULTS
The values obtained for the FGM parallelism test were Y = 5.485+0.93 * X and R 2 = 0.984 showing that the antibody used in the assays was highly capable to detect the searched metabolite in the fecal matrix.
For the hormone tests, the low and high inter-assay variation values were 5.5% and 10.1%, respectively, and the low and high intra-assay variation coefficients were 2.3% and 5.6%, respectively. The sensitivity of the assay was 2.1 ng/well, measured in the calibration curve.

Treatment 1 (hormone challenge with ACTH administration)
The mean basal FGM concentrations and standard deviations for the females were 636.41 (±314.58) and for the males 255.20 (±113.50) ng/g dry feces. Thus, prior to ACTH administration, females had mean FGM levels 2.49 times higher than males.
After ACTH administration, the FGM levels reached the peak values after approximately 9.25 hours in females and 8.75 hours in males (Fig. 1). The peak FGM values were 2.47 and 2.20 times greater than the basal values in females and males, and the FGM concentrations returned to basal levels after 32 hours in both sexes.

Treatment 2 (saline administration)
The mean basal FGM concentrations in females were 506.18 (±470.12) and in males 278.95 (±89.42) ng/g dried feces, indicating that the FGM values were 1.81 times greater in females than in males prior to saline administration.
Following saline administration, the FGM levels exceeded the basal values after six hours in both sexes (Fig. 2), and the peak values were observed after approximately 9.25 hours in females and 13.50 hours in males. The peak FGM values were 1.70 times higher than the basal values in females and 1.92 times higher in males. The FGM concentrations returned to basal levels after 34.5 hours in females and after 28 hours in males.

Treatment 3 (control for monitoring diurnal changes)
In terms of diurnal fluctuation, the FGM Concentration of fecal metabolites of glucocorticoid (x 10 3 ng/g of dry feces) Concentration of fecal metabolites of glucocorticoid (x 10 3 ng/g of dry feces) Concentration of fecal metabolites of glucocorticoid (x 10 3 ng/g of dry feces) Concentration of fecal metabolites of glucocorticoid (x 10 3 ng/g of dry feces) Concentration of fecal metabolites of glucocorticoid (x 10 3 ng/g of dry feces) Concentration of fecal metabolites of glucocorticoid (x 10 3 ng/g of dry feces) Concentration of fecal metabolites of glucocorticoid (x 10 3 ng/g of dry feces) Concentration of fecal metabolites of glucocorticoid (x 10 3 ng/g of dry feces)  Concentration of fecal metabolites of glucocorticoid (x 10 3 ng/g of dry feces) Concentration of fecal metabolites of glucocorticoid (x 10 3 ng/g of dry feces) Concentration of fecal metabolites of glucocorticoid (x 10 3 ng/g of dry feces) Concentration of fecal metabolites of glucocorticoid (x 10 3 ng/g of dry feces) Concentration of fecal metabolites of glucocorticoid (x 10 3 ng/g of dry feces) Concentration of fecal metabolites of glucocorticoid (x 10 3 ng/g of dry feces) Concentration of fecal metabolites of glucocorticoid (x 10 3 ng/g of dry feces) Another important finding of the present study were the consistently higher FGM concentrations in females than in males. Possible explanations for this observation could be differences in the metabolic rate between males and females as discussed by Palme et al. (2005), Touma et al. (2003) and Belz et al. (2003). We also hypothesize that these results may reflect the relatively small sample size of the study or, considering the high dependence of the physiological impact of stressors on the perception of environmental stimuli, a greater difficulty to adapt to captivity among the females studied (Natelson et al. 1987).
In conclusion, the enzyme immunoassay used is a reliable method to measure FGM in captive C. penicillata.