- Email: [email protected]
Biochemical and physiological validation of a corticosteroid radioimmunoassay for plasma and fecal samples in oldfield mice (Peromyscus polionotus)
Physiology & Behavior 80 (2003) 405 – 411
Biochemical and physiological validation of a corticosteroid radioimmunoassay for plasma and fecal samples in oldfield mice (Peromyscus polionotus) T. Gooda,*, M.Z. Khanb, J.W. Lynchc a
Department of Ecology and Evolutionary Biology, Princeton University, 106A Guyot Hall, Washington Road, Princeton, NJ 8544-1003, USA b Department of Biology and Microbiology, University of Wisconsin-Oshkosh, Oshkosh, WI 54901, USA c Department of Anthropology, University of California, Davis, CA 95616-8522, USA Received 12 April 2003; received in revised form 24 September 2003; accepted 24 September 2003
Abstract The measurement of fecal steroids provides an increasingly important noninvasive technique for assessing reproduction, environmental stress, and aggression in populations of captive and free-living animals. In this paper, we validated the corticosterone (CORT) 125Iradioimmunoassay (ICN Pharmaceuticals) for plasma and fecal samples in a small rodent species, the oldfield mouse (Peromyscus polionotus subgriseus). The biochemical validations indicated that the assays accurately measured CORT concentrations in the plasma and corticosteroid concentrations in the feces. Physiological validation demonstrated that: (1) blood samples collected within 3 min of disturbing an animal’s cage represented ‘‘baseline’’ CORT concentrations, and (2) fecal corticosteroid concentrations collected over a 24-h period closely tracked plasma CORT concentrations approximately 4 h earlier. These results demonstrate that the plasma CORT and fecal corticosteroid assays are sensitive enough to detect biologically meaningful alterations in corticosteroid concentrations in oldfield mice. D 2003 Elsevier Inc. All rights reserved. Keywords: Baseline; Corticosteroids; Diel rhythm; Fecal samples; Lag time; Noninvasive; Oldfield mouse; Peromyscus polionotus subgriseus; Plasma samples; Radioimmunoassay; RIA; Stress response; Validation
1. Introduction Researchers have long sought an understanding of how glucocorticoids help animals respond to stressors such as predators, capture and handling, extreme weather, and social insubordination (reviewed in Refs. [27,32]). After exposure to a stressor, the hypothalamic – pituitary– adrenal axis is activated, and glucocorticoids are released from the adrenal cortex [21]. These hormones then induce a variety of behavioral and physiological effects. Changes in glucocorticoid concentrations can be measured in several ways, each associated with advantages and problems of measurement and interpretation [43]. Plasma steroid measurements reflect the endocrine status of an individual at a single point in time. Therefore, rapid and short-term changes in glucocorticoid concentrations, * Corresponding author. Tel.: +1-609-258-2543; fax: +1-609-2582712. E-mail address: [email protected] (T. Good). 0031-9384/$ – see front matter D 2003 Elsevier Inc. All rights reserved. doi:10.1016/j.physbeh.2003.09.006
such as those triggered by a novel environment or situation, are best determined using plasma samples. However, one must ascertain that the measured glucocorticoid concentrations actually reflect changes triggered by the situation of interest and not those induced by the handling and sampling procedure itself. Capturing and handling animals can be stressful and causes a rapid release of glucocorticoids into the bloodstream [4,29,31,33,44]. Moreover, the time needed to elicit the stress response differs not only among species; it may also depend on context or historical factors such as whether the animals were obtained from colonies maintained in a lab for many generations or whether the animals were obtained directly from the wild [39]. A common assumption in either case is that baseline plasma corticosterone (CORT) concentrations are obtained if samples are collected within 3 min of capture [23,28]. Fecal steroid measurements represent a cumulative secretion over a number of hours and have the advantage of integrating short-term hormonal fluctuations caused by capture and blood drawing. With the recent development
406
T. Good et al. / Physiology & Behavior 80 (2003) 405–411
of such noninvasive techniques, researchers can monitor physiological conditions in species for which more traditional endocrine methods, including blood sampling, would be precluded as too disruptive, difficult, or even impossible. Many of the species for which this is the case are either endangered [20,41], difficult to capture repeatedly [1,6,16,35], or too small to sample repeatedly (this study). Noninvasive techniques are therefore ideal for investigating the links between the expression of behaviors and circulating hormones in a diversity of species and studies. In this paper, we present biochemical and physiological validation for corticosteroid measurements in plasma and feces in a small, monogamous rodent species, the oldfield mouse (Peromyscus polionotus subgriseus). This and other Peromyscus species are widely used as laboratory animals for genetic [5,9,26,40], evolutionary [15,18], behavioral [8,17,30]; ecological [25,36], and physiological studies [13,38]. However, most Peromyscus species are too small to use in a study requiring repeated blood sampling. P. polionotus is one of the smallest Peromyscus species with an average adult weight of 15 g [14] and the amount of plasma needed for a hormonal analysis exceeds the amount allowed by the Institutional Animals Care and Use Committee (IACUC) for a daily sampling protocol (0.05% of bodyweight [12]). Furthermore, the collection of fecal samples is an easier method than blood sampling because it does not require as much special training as the collection of blood samples. In addition to conducting a biochemical validation, which demonstrates that the assay is accurately and precisely measuring the hormones of interest, we physiologically validated the blood and fecal assay. A physiological validation indicates that the assay is sensitive enough to detect biologically meaningful alterations in hormone concentrations [10]. The goals of this paper are twofold: Firstly, we examined the short-term effects of handling on plasma CORT concentrations to ascertain that samples collected within the first 3 min represented baseline CORT concentrations. Secondly, we documented an effect of time-of-day on CORT concentrations, similar to that of many species, with peak values occurring at the onset of darkness in nocturnal rodents (e.g. laboratory rats [3], and references therein), and we used this observed pattern to validate the fecal assay. We demonstrate that the corticosteroids excreted in the feces over a 24-h period predictably tracked the diel rhythm in plasma CORT concentrations.
2. Methods 2.1. Study system Our study animals were obtained from a large research colony of oldfield mice housed at the Brookfield Zoo, Brookfield, IL. This colony was founded by P. polionotus
subgriseus trapped at field sites in Ocala National Forest, Florida in 1998. All animals were housed in the same colony room at Princeton University in polycarbonate cages, 26 15 12 cm, with food and water provided ad libitum, under a 12L:12D cycle (lights on at 12:30 a.m.). All experiments were approved by Princeton’s Institutional Animals Care and Use Committee (IACUC). 2.2. Blood and fecal sample collection All blood samples were collected using the retro-orbital bleeding technique in less than 2 min (38.3 F 15.2 s) of disturbing the animal’s cage. Blood samples were centrifuged, and the resulting plasma was collected and stored at 20 jC until assayed. Fecal samples were collected by briefly holding the mice in one hand and catching the pellets in a microcentrifuge tube while avoiding urine contamination. Fecal samples were then stored at 20 jC until extraction. For fecal extraction, samples were weighed, transferred to individually labeled test tubes, homogenized with a glass rod in 1 ml of 90% methanol [41], and vortexed for 30 min on a Multi-Pulse Vortexer (Glas-Col, model 099A VB4). The samples were centrifuged at 1 g for 20 min, and the supernatant of each sample was stored at 20 jC. Prior to analysis, fecal extract (300 Al) was dried under nitrogen, reconstituted in 1 ml of 30% methanol, and passed through a solid-phase extraction cartridge (Waters, Milford, MA, WAT094226; techniques modified by T.E. Ziegler, University of Wisconsin, Madison). The steroids eluted off the column with 2 ml of 100% methanol were dried under nitrogen and reconstituted in 300 Al of 90% methanol to return the sample to its original volume in its original solute. Gravity was used to pull the solvents through the column. 2.3. Hormonal analysis 2.3.1. Biochemical assay validation for CORT in plasma and corticosteroid in fecal samples All samples were assayed in duplicate at one-half volume by using double-antibody 125I-radioimmunoassay kit (ICN Pharmaceuticals, Costa Mesa, CA; catalog no. 07-120102 for CORT) and counted on a Hewlett Packard Cobra II Auto Gamma Counter. Hormone concentrations were expressed as nanograms of hormone per gram of fecal matter. Plasma samples were diluted 1:1600 and fecal samples 1:5 in the assay buffer prior to radioimmunoassay. To ascertain biochemical validity of our assay, we followed the procedures described in Ref. [10]: (1) parallelism assay, in which pooled plasma samples and pooled fecal extracts were serially diluted (plasma: 1:200 –1:6400; fecal extract: 1:1 –1:32) and the slopes of the log – logit transformed curves of those samples were then compared to that of the standard curve. Parallelism for fecal samples was achieved by adding 10 Al of doubly charcoal-stripped fecal extract pool to the standards [35]; (2) quantitative recovery,
T. Good et al. / Physiology & Behavior 80 (2003) 405–411
in which the extraction efficiency was determined by spiking 10 individual fecal samples with 10 Al of 125ICORT before extraction. These samples were then allowed to incubate for 1 h at room temperature before fecal extraction; (3) assay precision was assessed by calculating intra- and interassay coefficients of variation (CVs) of the percentage bound of the internal controls. In our study, the controls used were low (10 Al) and high pools (25 Al) of plasma and fecal extract. Variations of duplicates within an assay were used to calculate intra-assay variation; average values from the low and high pools across assays were used to calculate interassay variation; (4) assay accuracy was determined by adding 10 Al of plasma or fecal extract to each standard curve point. 2.3.2. Physiological assay validation of plasma CORT CORT is the dominant glucocorticoid in the plasma of rats and mice [34,39]. In order to assess short-term effects of handling on plasma CORT concentrations, blood samples from four different individuals (two different males and two different females at each time period) were collected 30, 60, 90, 120, 150, 180, and 600 s after disturbing their cages and manually restraining the individuals. Each individual was only sampled once (n = 28) both to avoid the confounding influence of a potential stress response to repeated sampling and because the animals were too small to sample repeatedly. 2.3.3. Physiological validation of fecal corticosteroid assay In order to physiologically validate the fecal corticosteroid assay, blood and fecal samples were collected every 4 h for 24 h, beginning at 4:30 a.m. Samples were taken from two males and two females at each time period. Each individual was only sampled once (n = 28). 2.3.4. Statistical analysis Parallelism between standard curves and serial dilutions of plasma samples and fecal extracts were determined by testing the equality of two slopes [22]. Means are given with standard errors (S.E.) unless otherwise noted. Because the hormonal data were not normally distributed, we used nonparametric statistics for statistical testing. We fit curves to both the diel plasma CORT data and the diel fecal corticosteroid data and used the equations of both curves (see below) to calculate the lag time. Matlab curve fitter uses the Marquardt –Levenberg algorithm to find the parameters of the independent variables that give the ‘‘best fit’’ between the equation and the data. This algorithm seeks the values of the parameters that minimize the sum of squared differences between the values of the observed and predicted values of the dependent variable. Using the well-known fact CORT exhibits a diel rhythm, we chose a sinusoid curve type with a 24-h period. The following equation was used to fit the plasma CORT data: Cp ¼ a sinðp=12ðt þ bÞÞ þ c
ð1Þ
407
where a = amplitude, b = offset and c = average plasma CORT concentrations. Cp is in ng/ml, and t is in hours. The following equation was used to fit the fecal corticosteroid data: Cf ¼ d sinðp=12ðt þ eÞÞ þ f
ð2Þ
where d = amplitude, e = offset and f = average fecal corticosteroid concentrations. Cf is in ng/g, and t is in hours. The time lag between the fecal and plasma cycle is simply b e. The uncertainty of the lag is calculated as: rlag=(re2 + rb2)1=2. We used the information on lag time calculated from the curves that best fit the plasma CORT and fecal corticosteroid data to get a best estimate of lag time from the data: given that we collected fecal samples every 4 h, we used a Spearman correlation to compare fecal corticosteroid concentrations to those of the plasma CORT concentrations 4 h earlier.
3. Results 3.1. Biochemical validation for corticosteroid assay The slopes of the log – logit transformed curves generated from serially diluted plasma samples and fecal extract were not significantly different from the standard curve slopes (t = 0.74, df = 20, P=.47 for plasma samples; t = 1.56, df = 20, P=.133 for fecal extracts). Assay accuracy was 104.2 F 3.0% for plasma samples and 114.7 F 2.2% for fecal extracts. For fecal extracts, methanol extraction recoveries were 94.4% F 0.6% (n = 10). Intra- and interassay variation was 4.5% and 9.7% for the low-concentration plasma pool, 4.8% and 11.6% for the high-concentration plasma pool (n = 4), 11.7% and 20.0% for the low-concentration fecal extract pool and 6.1% and 7.4% for the highconcentration fecal extract pool (n = 4). 3.2. Physiological assay validation of plasma CORT Plasma CORT concentrations exhibited no significant change over a period of 180 s (3 min) in P. polionotus (Fig. 1, Kruskal –Wallis, v2 = 3.5, df = 5, P=.6). The average CORT concentration during this time period was 523.15 F 67.93 ng/ml (n = 24). However, CORT concentrations increased significantly between 3 and 10 min (Mann – Whitney U test, U = 1.0, P < .05). The average value of CORT was 1985.46 F 449.48 ng/ml (n = 4) after 10 min of manual restraint. 3.3. Physiological assay validation of fecal corticosteroids Over a 24-h period, plasma CORT concentrations varied significantly (Fig. 2, Kruskal – Wallis, v2 = 16.20, df = 6, P=.013) as did fecal corticosteroid concentrations (Fig. 2, Kruskal –Wallis, v2 = 15.783, df = 6, P=.015). Plasma CORT
408
T. Good et al. / Physiology & Behavior 80 (2003) 405–411
Fig. 1. Short-term effects of handling and manual restraint on plasma CORT concentrations in P. polionotus. Four different individuals were sampled at each time period (n = 28).
concentrations rose rapidly between 8:30 a.m. and 12:30 p.m. (onset of the dark phase). They peaked during the dark phase and attained concentrations of 1227.11 F 214.53 ng/ ml (n = 12) during this phase. The concentrations reached their lowest values (160.38 F 47.45 ng/ml; n = 7) at the onset of the light phase (12:30 am – 4:30 a.m.). Peak values (55.91 F 6.19 ng/g; n = 8) occurred between 4:30 p.m. and 8:30 p.m. and lowest values (24.47 F 6.10 ng/g; n = 8) between 4:30 a.m. and 8:30 a.m.
Using Matlab curve fitter, the parameters ( F 95% confidence interval) for Cp were: a = 683 ( F 413), b = 2.4 ( F 3.2), c = 895 ( F 342) with an R2 of .84. For Cf, the parameters ( F 95% confidence interval) were: d = 21.8 ( F 7.1), e = 0.50 ( F 1.65), f = 35.8 ( F 5.7) with an R2 of .95. The lag time resulting from the curves is 2.9 F 3.6 h (mean F 95% confidence interval). The diel rhythm in fecal corticosteroid concentrations exhibited a 4-h time lag to that of the plasma CORT
Plasma CORT [ng/ml]
2500
Plasma CORT Fit to plasma CORT data Fecal corticosteroids Fit to fecal corticosteroids data
80
2000 60 1500 40 1000 20
500
Fecal corticosteroids [ng/g]
100
3000
0 0 4:30 am 8:30 am 12:30 pm 4:30 pm 8:30 pm 12:30 am 4:30 am
Hours Fig. 2. Diel rhythm of plasma CORT (.) and fecal corticosteroid (D) in P. polionotus. Four different individuals were sampled at each time period, with the exception of 8:30 a.m. (n = 2) and 12:30 a.m. (n = 3) where plasma samples were lost during the centrifuging process. Means are given with standard errors (S.E.). The horizontal bar indicates the dark phase of the light/dark cycle. Between 8:30 a.m. and 12:30 p.m. plasma CORT concentrations rose to double daylight values and a comparable doubling occurred during 12:30 p.m. and 4:30 p.m. for fecal corticosteroid concentrations. Similarly, plasma CORT concentrations decreased by half between 4:30 p.m. and 12:30 a.m. and fecal corticosteroid concentrations decreased by half between 8:30 p.m. and 4:30 a.m. R2=.84 for the curve fit to the plasma CORT data; and R2=.95 for the curve fit to the fecal corticosteroid data. See text for equations. The calculated lag time is 2.9 F 3.6 h (mean F 95% confidence interval).
T. Good et al. / Physiology & Behavior 80 (2003) 405–411
concentrations and correlated with that of the plasma 4 h earlier (Spearman’s rho: .829, df = 6, P < .042).
4. Discussion 4.1. Biochemical assay validation of the plasma CORT and fecal corticosteroid assay The CORT assay precisely and accurately measured CORT concentrations in plasma and corticosteroid concentrations in fecal samples for P. polionotus. Radioimmunoassays for fecal glucocorticoids have been validated for a number of species [7,20,42]. However, to our knowledge, only one other study has validated a corticosteroid assay for fecal samples in a Peromyscus species (P. maniculatus [10]). The recoveries for their ethanol extraction of radiolabeled CORT are comparable to ours (96.0 F 3.4% vs. 94.4 F 0.6%). 4.2. Physiological assay validation of plasma CORT One of our goals was to ascertain that samples collected within the first 3 min represented baseline CORT concentrations. CORT concentrations did not vary significantly during the first 3 min after disturbing the animal’s cage. A ninefold variability in baseline CORT concentrations has been reported from laboratory studies across different Peromyscus species (P. boylii, P. californicus [39]; P. leucopus [24]; P. maniculatus [2]; P. polionotus: this study). CORT concentrations in P. polionotus are the highest reported thus far. However, cross-species comparisons are problematic because there are too many other species-specific differences in factors, such as tissue receptor numbers or plasma binding protein levels, that alter physiological function (Romero, pers. comm.). CORT concentration increased significantly between 3 and 10 min. Our sampling protocol does not permit determination of whether the CORT concentrations were at their maximum levels after 10 min. In wild-caught P. californicus, maximum CORT concentrations were reached 5 min after the stressor (electric shock) had been applied ( f 1600 ng/ml [39]). In contrast, CORT concentrations in wildcaught P. boylii continued to increase for 60 min, eventually reaching the same peak values as P. californicus [39]. Both species live in similar environments, and the micro-environmental or species-physiological differences that would produce such differences in the temporal pattern of CORT response to stress are unknown. 4.3. Physiological assay validation of the fecal corticosteroid assay Our results indicate that fecal corticosteroid concentrations accurately reflected plasma CORT concentrations with a consistent lag. Corticosteroid concentrations in-
409
creased significantly at night in both plasma and fecal samples. Therefore, these results demonstrate that the fecal corticosteroid assay is sensitive enough to detect biologically meaningful alterations in corticosteroid concentrations in oldfield mice. To our knowledge, this is the first demonstration of a diel corticosteroid rhythm measured in fecal samples for a rodent. The values obtained in our study (24.47 55.91 F 13.68 ng/g) fall in the same range as those of P. maniculatus (42.41 84.13 F 12.22 ng/g [10]) even though the method of collection differed between the two studies. We collected samples directly from each individual every 4 h for 24 h, whereas Harper and Austad [10] collected samples from the cage floor every 12 h for a total of 192 h. The same authors found that when fecal samples were collected directly from individuals every 2 h just prior to the blood collection, fecal corticosteroid concentrations were much higher, ranging from 66.99 189.58 F 27.43 ng/g (Harper, unpublished data, pers. comm., Feb. 4, 2003). This suggests that fresh fecal samples have higher corticosteroid concentrations than samples collected at a later point in time and would lead us to conclude that P. maniculatus has higher corticosteroid concentrations than P. polionotus. It is likely that bacteria in the feces degrade corticosteroid metabolites during the interval between defecation and sample collection [19]. In order to be able to generalize and compare across studies and species, details, such as method and times of collection, should be taken into consideration. The lag time calculated from the parameters derived from the best fitting curves of the diel plasma CORT and diel fecal corticosteroid data was 2.9 F 3.6 (mean F 95% confidence interval). The uncertainty of the lag is due to the high interindividual variation in diel plasma CORT and diel fecal corticosteroid concentrations, yet the range of values calculated here are in accordance with those in the literature. When calculated from the actual plasma CORT and fecal corticosteroid data, the diel rhythm in fecal corticosteroid concentration exhibited an approximately 4-h lag time from that of the plasma CORT concentrations. Fecal corticosteroids were significantly higher in deer mice (P. maniculatus) that had been confined to traps for >4 h compared to those confined to traps for 0 –4 h suggesting that the lag time could be as little as 4 h in that species [11]. The best estimate of lag time available in mice is the detection of excreted radiolabelled tracer in defecated material. Peak excretion of [3H]CORT metabolites in the feces of lab mice (Mus musculus) occurred 4 h after administration of radiolabelled CORT during the dark phase of the light cycle [37]. Peromyscus spp. are ideal mammalian models to use in investigations of questions in behavioral endocrinology and physiological ecology as well as often being important target species in a range of ecological and conservation situations. The use of a fecal assay provides a valuable and powerful noninvasive method of measuring changes in
410
T. Good et al. / Physiology & Behavior 80 (2003) 405–411
corticosteroid concentrations in small mammals such as these under both laboratory and field conditions. Acknowledgements Jeanne Altmann, Michaela Hau, and two anonymous reviewers provided helpful comments on the manuscript. We thank James Harper for providing unpublished data of P. maniculatus fecal corticosteroid concentrations over a 24-h period. Catherine Marler kindly provided access to her laboratory group and advice during training. Gordon Brackee and Pat Martin trained TG in the method of retroorbital bleeding. Brian Elliot, Wenfei Tong, and Andre´ Utzinger provided invaluable help during fecal sample collection and extraction. Chris Wyckham provided the curve-fitting expertise. Support was provided by Princeton University, the Animal Behavior Society and the Society for Integrative and Comparative Biology to TG. References [1] Asa CS, Bauman JE, Houston EW, Fisher MT, Read B, Brownfield CM, et al. Patterns of excretion of fecal estradiol and progesterone and urinary chorionic gonadotropin in Grevy’s zebra (Equus grevyi): ovulatory cycles and pregnancy. Zoo Biol 2001;20:185 – 95. [2] Bradley EL, Terman CR. A comparison of the adrenal histology, reproductive condition, and serum corticosterone concentrations of prairie deermice (Peromyscus maniculatus bairdii) in captivity. J Mammal 1981;62:353 – 61. [3] D’Agostino J, Vaeth GF, Henning SJ. Diurnal rhythm of total and free concentrations of serum corticosterone in the rat. Acta Endocrinol 1982;100:85 – 90. [4] Dawson A, Howe PD. Plasma corticosterone in wild starlings (Sturnus vulgaris) immediately following capture and in relation to body weight during the annual cycle. Gen Comp Endocrinol 1983;51: 303 – 8. [5] Foltz DW. Genetic evidence for long-term monogamy in a small rodent, Peromyscus polionotus. Am Nat 1981;117:665 – 75. [6] Garnier JN, Green DI, Pickard AR, Shaw HJ, Holt WV. Non-invasive diagnosis of pregnancy in wild black rhinoceros (Diceros bicornis minor) by faecal steroid analysis. Reprod Fertil Dev 1998;10:451 – 8. [7] Goymann W, Mo¨stl E, Van’t Hof T, East M, Hofer H. Noninvasive fecal monitoring of glucocorticoids in spotted hyenas, Crocuta crocuta. Gen Comp Endocrinol 1999;114:340 – 8. [8] Gubernick DJ, Alberts JR. Postpartum maintenance of paternal behaviour in the biparental California mouse, Peromyscus californicus. Anim Behav 1989;37:656 – 64. [9] Haig D. Genetic conflicts and the private life of Peromyscus polionotus. Nat Genet 1999;22:131. [10] Harper JM, Austad SN. Fecal glucocorticosteroids: a non-invasive method of measuring adrenal activity in wild and captive rodents. Physiol Biochem Zool 2000;73:12 – 22. [11] Harper JM, Austad SN. Effect of capture and season on fecal glucocorticoid levels in deer mice (Peromyscus maniculatus) and red-backed voles (Cleithrionomys gapperi). Gen Comp Endocrinol 2001;123: 337 – 44. [12] Hem A, Smith AJ, Solberg P. Saphenous vein puncture for blood sampling of the mouse, rat, hamster, gerbil, guinea pig, ferret and mink. Lab Anim 1998;32:364 – 8. [13] Klein SL, Nelson RJ. Adaptive immune responses are linked to the mating system of arvicoline rodents. Am Nat 1998;151:59 – 67.
[14] Layne JN. Ontogeny. In: King JA, editor. Biology of Peromyscus (Rodentia), vol. 151. Stillwater, OK: The American Society of Mammalogists; 1968. p. 148 – 253 (Special Publication No. 2). [15] Lacy RC, Ballou JD. Effectiveness of selection in reducing the genetic load in populations of Peromyscus polionotus during generations of inbreeding. Evolution 1998;52:900 – 9. [16] Lynch JW, Ziegler TE, Strier KB. Individual and seasonal variation in fecal testosterone and cortisol levels of wild tufted capuchin monkeys, Cebus apella nigritus. Horm Behav 2002;41:275 – 87. [17] Margulis SW. Relationships among parental inbreeding, parental behavior and offspring viability in oldfield mice. Anim Behav 1998;55: 427 – 38. [18] Michaux J, Reyes A, Catzeflis F. Evolutionary history of the most speciose mammals: molecular phylogeny of muroid rodents. Mol Biol Evol 2001;18:2017 – 31. [19] Mo¨stl E, Messman S, Bagu E, Robia C, Palme R. Measurement of glucocorticoid metabolite concentrations in faeces of domestic livestock. J Vet Med A 1999;46:621 – 31. [20] Monfort SL, Wasser SK, Mashburn KL, Burke M, Brewer BA, Creel SR. Steroid metabolism and validation of noninvasive endocrine monitoring in the african wild dog (Lycaon pictus). Zoo Biol 1997;16: 533 – 48. [21] Munck A, Guyre PM, Holbrook NJ. Physiological functions of glucocorticoids in stress and their relation to pharmacological effects. Endocr Rev 1984;5:25 – 44. [22] Neter J, Wasserman W, Kutner M. Applied linear statistical models: regression, analysis of variance and experimental designs. Homewood, IL: Irwin; 1990. p. 165 – 7. [23] Place NJ, Kenagy GJ. Seasonal changes in plasma testosterone and glucocorticoids in free-living male yellow-pine chipmunks and the response to capture and handling. J Comp Physiol B 2000;170:245 – 51. [24] Ransone SN, Bradley EL. Correlations between serum corticosterone concentration and reproductive conditions in the white-footed mouse (Peromyscus leucopus noveboracensis). Comp Biochem Physiol 1992;102A:453 – 7. [25] Rave EA, Holler NR. Population dynamics of beach mice (Peromyscus polionotus) in southern Alabama. J Mammal 1992;73:347 – 55. [26] Ribble DO. The monogamous mating system of Peromyscus californicus as revealed by DNA fingerprinting. Behav Ecol Sociobiol 1991; 29:161 – 6. [27] Romero LM. Seasonal changes in plasma glucocorticoid concentrations in free-living vertebrates. Gen Comp Endocrinol 2002;128:1 – 24. [28] Romero LM, Remage-Healey L. Daily and seasonal variation in response to stress in captive starlings (Sturnus vulgaris): corticosterone. Gen Comp Endocrinol 2000;119:52 – 9. [29] Romero LM, Romero RC. Corticosterone responses in wild birds: the importance of rapid initial sampling. Condor 2002;104:129 – 35. [30] Ryan KK, Altmann J. Selection for male choice based primarily on mate compatibility in the oldfield mouse, P. polionotus rhoadsi. Behav Ecol Sociobiol 2001;55:436 – 40. [31] Sabatino F, Masoro EJ, McMahan A, Kuhn RW. Assessment of the role of glucocorticoid system in aging process and in the action of food restriction. J Gerontol 1991;46B:171 – 9. [32] Saplosky RM, Romero LM, Munck AU. How do glucocorticoids influence stress-responses? Integrating permissive, suppressive, stimulatory, and adaptive actions. Endocr Rev 2000;21:55 – 89. [33] Schoech SJ, Ketterson ED, Nolan Jr V. Exogenous testosterone and the adrenocortical response in dark-eyed juncos. Auk 1999;116:64 – 72. [34] Shimizu K, Amagaya S, Ogihara Y. Analysis of corticosterone in the serum of mice and rats, using high performance liquid chromatography. J Chromatogr 1983;272:170 – 5. [35] Strier KB, Ziegler TE. Behavioral and endocrine characteristics of the reproductive cycle in wild muriqui monkeys, Brachyteles arachnoids. Am J Primatol 1997;42:299 – 310. [36] Swilling WR, Wooten MC. Subadult dispersal in a monogamous species: the Alabama beach mouse (Peromyscus polionotus ammobates). J Mammal 2002;83:252 – 9.
T. Good et al. / Physiology & Behavior 80 (2003) 405–411 [37] Touma C, Sachser N, Mo¨stl E, Palme R. Effects of sex and time of day on metabolism and excretion of corticosterone in urine and feces of mice. Gen Comp Endocrinol 2003;130:267 – 78. [38] Trainor B, Marler CA. Testosterone, parental behavior, and aggression in the monogamous California mouse (Peromyscus californicus). Horm Behav 2001;40:32 – 42. [39] Treiman DM, Levine S. Plasma corticosteroid response to stress in four species of wild mice. Endocrinology 1969;84:676 – 80. [40] Vrana PB, Guan XJ, Ingram RS, Tilghman SM. Genomic imprinting is disrupted in interspecific Peromyscus hybrids. Nat Genet 1998;20: 362 – 5. [41] Wasser SK, Monfort SL, Southers J, Wildt DE. Excretion rates and
411
metabolites of oestradiol and progesterone in baboon (Papio cynocephalus cynocephalus) faeces. J Reprod Fertil 1994;101:213 – 20. [42] Wasser SK, Velloso AdL, Rodden MD. Using fecal steroids to evaluate reproductive function in female maned wolves. J Wildl Manage 1995;59:889 – 94. [43] Wasser SK, Hunt KE, Brown JL, Cooper K, Crockett CM, Bechert U, et al. A generalized fecal glucocorticoid assay for use in a diverse array of nondomestic mammalian and avian species. Gen Comp Endocrinol 2000;120:260 – 75. [44] Whitten PL, Brockman DK, Stavisky RC. Recent advances in noninvasive techniques to monitor hormone – behavior interactions. Yearb Phys Anthropol 1998;41:1 – 23.
Biochemical and physiological validation of a corticosteroid radioimmunoassay for plasma and fecal samples in oldfield mice (Peromyscus polionotus) T. Gooda,*, M.Z. Khanb, J.W. Lynchc a
Department of Ecology and Evolutionary Biology, Princeton University, 106A Guyot Hall, Washington Road, Princeton, NJ 8544-1003, USA b Department of Biology and Microbiology, University of Wisconsin-Oshkosh, Oshkosh, WI 54901, USA c Department of Anthropology, University of California, Davis, CA 95616-8522, USA Received 12 April 2003; received in revised form 24 September 2003; accepted 24 September 2003
Abstract The measurement of fecal steroids provides an increasingly important noninvasive technique for assessing reproduction, environmental stress, and aggression in populations of captive and free-living animals. In this paper, we validated the corticosterone (CORT) 125Iradioimmunoassay (ICN Pharmaceuticals) for plasma and fecal samples in a small rodent species, the oldfield mouse (Peromyscus polionotus subgriseus). The biochemical validations indicated that the assays accurately measured CORT concentrations in the plasma and corticosteroid concentrations in the feces. Physiological validation demonstrated that: (1) blood samples collected within 3 min of disturbing an animal’s cage represented ‘‘baseline’’ CORT concentrations, and (2) fecal corticosteroid concentrations collected over a 24-h period closely tracked plasma CORT concentrations approximately 4 h earlier. These results demonstrate that the plasma CORT and fecal corticosteroid assays are sensitive enough to detect biologically meaningful alterations in corticosteroid concentrations in oldfield mice. D 2003 Elsevier Inc. All rights reserved. Keywords: Baseline; Corticosteroids; Diel rhythm; Fecal samples; Lag time; Noninvasive; Oldfield mouse; Peromyscus polionotus subgriseus; Plasma samples; Radioimmunoassay; RIA; Stress response; Validation
1. Introduction Researchers have long sought an understanding of how glucocorticoids help animals respond to stressors such as predators, capture and handling, extreme weather, and social insubordination (reviewed in Refs. [27,32]). After exposure to a stressor, the hypothalamic – pituitary– adrenal axis is activated, and glucocorticoids are released from the adrenal cortex [21]. These hormones then induce a variety of behavioral and physiological effects. Changes in glucocorticoid concentrations can be measured in several ways, each associated with advantages and problems of measurement and interpretation [43]. Plasma steroid measurements reflect the endocrine status of an individual at a single point in time. Therefore, rapid and short-term changes in glucocorticoid concentrations, * Corresponding author. Tel.: +1-609-258-2543; fax: +1-609-2582712. E-mail address: [email protected] (T. Good). 0031-9384/$ – see front matter D 2003 Elsevier Inc. All rights reserved. doi:10.1016/j.physbeh.2003.09.006
such as those triggered by a novel environment or situation, are best determined using plasma samples. However, one must ascertain that the measured glucocorticoid concentrations actually reflect changes triggered by the situation of interest and not those induced by the handling and sampling procedure itself. Capturing and handling animals can be stressful and causes a rapid release of glucocorticoids into the bloodstream [4,29,31,33,44]. Moreover, the time needed to elicit the stress response differs not only among species; it may also depend on context or historical factors such as whether the animals were obtained from colonies maintained in a lab for many generations or whether the animals were obtained directly from the wild [39]. A common assumption in either case is that baseline plasma corticosterone (CORT) concentrations are obtained if samples are collected within 3 min of capture [23,28]. Fecal steroid measurements represent a cumulative secretion over a number of hours and have the advantage of integrating short-term hormonal fluctuations caused by capture and blood drawing. With the recent development
406
T. Good et al. / Physiology & Behavior 80 (2003) 405–411
of such noninvasive techniques, researchers can monitor physiological conditions in species for which more traditional endocrine methods, including blood sampling, would be precluded as too disruptive, difficult, or even impossible. Many of the species for which this is the case are either endangered [20,41], difficult to capture repeatedly [1,6,16,35], or too small to sample repeatedly (this study). Noninvasive techniques are therefore ideal for investigating the links between the expression of behaviors and circulating hormones in a diversity of species and studies. In this paper, we present biochemical and physiological validation for corticosteroid measurements in plasma and feces in a small, monogamous rodent species, the oldfield mouse (Peromyscus polionotus subgriseus). This and other Peromyscus species are widely used as laboratory animals for genetic [5,9,26,40], evolutionary [15,18], behavioral [8,17,30]; ecological [25,36], and physiological studies [13,38]. However, most Peromyscus species are too small to use in a study requiring repeated blood sampling. P. polionotus is one of the smallest Peromyscus species with an average adult weight of 15 g [14] and the amount of plasma needed for a hormonal analysis exceeds the amount allowed by the Institutional Animals Care and Use Committee (IACUC) for a daily sampling protocol (0.05% of bodyweight [12]). Furthermore, the collection of fecal samples is an easier method than blood sampling because it does not require as much special training as the collection of blood samples. In addition to conducting a biochemical validation, which demonstrates that the assay is accurately and precisely measuring the hormones of interest, we physiologically validated the blood and fecal assay. A physiological validation indicates that the assay is sensitive enough to detect biologically meaningful alterations in hormone concentrations [10]. The goals of this paper are twofold: Firstly, we examined the short-term effects of handling on plasma CORT concentrations to ascertain that samples collected within the first 3 min represented baseline CORT concentrations. Secondly, we documented an effect of time-of-day on CORT concentrations, similar to that of many species, with peak values occurring at the onset of darkness in nocturnal rodents (e.g. laboratory rats [3], and references therein), and we used this observed pattern to validate the fecal assay. We demonstrate that the corticosteroids excreted in the feces over a 24-h period predictably tracked the diel rhythm in plasma CORT concentrations.
2. Methods 2.1. Study system Our study animals were obtained from a large research colony of oldfield mice housed at the Brookfield Zoo, Brookfield, IL. This colony was founded by P. polionotus
subgriseus trapped at field sites in Ocala National Forest, Florida in 1998. All animals were housed in the same colony room at Princeton University in polycarbonate cages, 26 15 12 cm, with food and water provided ad libitum, under a 12L:12D cycle (lights on at 12:30 a.m.). All experiments were approved by Princeton’s Institutional Animals Care and Use Committee (IACUC). 2.2. Blood and fecal sample collection All blood samples were collected using the retro-orbital bleeding technique in less than 2 min (38.3 F 15.2 s) of disturbing the animal’s cage. Blood samples were centrifuged, and the resulting plasma was collected and stored at 20 jC until assayed. Fecal samples were collected by briefly holding the mice in one hand and catching the pellets in a microcentrifuge tube while avoiding urine contamination. Fecal samples were then stored at 20 jC until extraction. For fecal extraction, samples were weighed, transferred to individually labeled test tubes, homogenized with a glass rod in 1 ml of 90% methanol [41], and vortexed for 30 min on a Multi-Pulse Vortexer (Glas-Col, model 099A VB4). The samples were centrifuged at 1 g for 20 min, and the supernatant of each sample was stored at 20 jC. Prior to analysis, fecal extract (300 Al) was dried under nitrogen, reconstituted in 1 ml of 30% methanol, and passed through a solid-phase extraction cartridge (Waters, Milford, MA, WAT094226; techniques modified by T.E. Ziegler, University of Wisconsin, Madison). The steroids eluted off the column with 2 ml of 100% methanol were dried under nitrogen and reconstituted in 300 Al of 90% methanol to return the sample to its original volume in its original solute. Gravity was used to pull the solvents through the column. 2.3. Hormonal analysis 2.3.1. Biochemical assay validation for CORT in plasma and corticosteroid in fecal samples All samples were assayed in duplicate at one-half volume by using double-antibody 125I-radioimmunoassay kit (ICN Pharmaceuticals, Costa Mesa, CA; catalog no. 07-120102 for CORT) and counted on a Hewlett Packard Cobra II Auto Gamma Counter. Hormone concentrations were expressed as nanograms of hormone per gram of fecal matter. Plasma samples were diluted 1:1600 and fecal samples 1:5 in the assay buffer prior to radioimmunoassay. To ascertain biochemical validity of our assay, we followed the procedures described in Ref. [10]: (1) parallelism assay, in which pooled plasma samples and pooled fecal extracts were serially diluted (plasma: 1:200 –1:6400; fecal extract: 1:1 –1:32) and the slopes of the log – logit transformed curves of those samples were then compared to that of the standard curve. Parallelism for fecal samples was achieved by adding 10 Al of doubly charcoal-stripped fecal extract pool to the standards [35]; (2) quantitative recovery,
T. Good et al. / Physiology & Behavior 80 (2003) 405–411
in which the extraction efficiency was determined by spiking 10 individual fecal samples with 10 Al of 125ICORT before extraction. These samples were then allowed to incubate for 1 h at room temperature before fecal extraction; (3) assay precision was assessed by calculating intra- and interassay coefficients of variation (CVs) of the percentage bound of the internal controls. In our study, the controls used were low (10 Al) and high pools (25 Al) of plasma and fecal extract. Variations of duplicates within an assay were used to calculate intra-assay variation; average values from the low and high pools across assays were used to calculate interassay variation; (4) assay accuracy was determined by adding 10 Al of plasma or fecal extract to each standard curve point. 2.3.2. Physiological assay validation of plasma CORT CORT is the dominant glucocorticoid in the plasma of rats and mice [34,39]. In order to assess short-term effects of handling on plasma CORT concentrations, blood samples from four different individuals (two different males and two different females at each time period) were collected 30, 60, 90, 120, 150, 180, and 600 s after disturbing their cages and manually restraining the individuals. Each individual was only sampled once (n = 28) both to avoid the confounding influence of a potential stress response to repeated sampling and because the animals were too small to sample repeatedly. 2.3.3. Physiological validation of fecal corticosteroid assay In order to physiologically validate the fecal corticosteroid assay, blood and fecal samples were collected every 4 h for 24 h, beginning at 4:30 a.m. Samples were taken from two males and two females at each time period. Each individual was only sampled once (n = 28). 2.3.4. Statistical analysis Parallelism between standard curves and serial dilutions of plasma samples and fecal extracts were determined by testing the equality of two slopes [22]. Means are given with standard errors (S.E.) unless otherwise noted. Because the hormonal data were not normally distributed, we used nonparametric statistics for statistical testing. We fit curves to both the diel plasma CORT data and the diel fecal corticosteroid data and used the equations of both curves (see below) to calculate the lag time. Matlab curve fitter uses the Marquardt –Levenberg algorithm to find the parameters of the independent variables that give the ‘‘best fit’’ between the equation and the data. This algorithm seeks the values of the parameters that minimize the sum of squared differences between the values of the observed and predicted values of the dependent variable. Using the well-known fact CORT exhibits a diel rhythm, we chose a sinusoid curve type with a 24-h period. The following equation was used to fit the plasma CORT data: Cp ¼ a sinðp=12ðt þ bÞÞ þ c
ð1Þ
407
where a = amplitude, b = offset and c = average plasma CORT concentrations. Cp is in ng/ml, and t is in hours. The following equation was used to fit the fecal corticosteroid data: Cf ¼ d sinðp=12ðt þ eÞÞ þ f
ð2Þ
where d = amplitude, e = offset and f = average fecal corticosteroid concentrations. Cf is in ng/g, and t is in hours. The time lag between the fecal and plasma cycle is simply b e. The uncertainty of the lag is calculated as: rlag=(re2 + rb2)1=2. We used the information on lag time calculated from the curves that best fit the plasma CORT and fecal corticosteroid data to get a best estimate of lag time from the data: given that we collected fecal samples every 4 h, we used a Spearman correlation to compare fecal corticosteroid concentrations to those of the plasma CORT concentrations 4 h earlier.
3. Results 3.1. Biochemical validation for corticosteroid assay The slopes of the log – logit transformed curves generated from serially diluted plasma samples and fecal extract were not significantly different from the standard curve slopes (t = 0.74, df = 20, P=.47 for plasma samples; t = 1.56, df = 20, P=.133 for fecal extracts). Assay accuracy was 104.2 F 3.0% for plasma samples and 114.7 F 2.2% for fecal extracts. For fecal extracts, methanol extraction recoveries were 94.4% F 0.6% (n = 10). Intra- and interassay variation was 4.5% and 9.7% for the low-concentration plasma pool, 4.8% and 11.6% for the high-concentration plasma pool (n = 4), 11.7% and 20.0% for the low-concentration fecal extract pool and 6.1% and 7.4% for the highconcentration fecal extract pool (n = 4). 3.2. Physiological assay validation of plasma CORT Plasma CORT concentrations exhibited no significant change over a period of 180 s (3 min) in P. polionotus (Fig. 1, Kruskal –Wallis, v2 = 3.5, df = 5, P=.6). The average CORT concentration during this time period was 523.15 F 67.93 ng/ml (n = 24). However, CORT concentrations increased significantly between 3 and 10 min (Mann – Whitney U test, U = 1.0, P < .05). The average value of CORT was 1985.46 F 449.48 ng/ml (n = 4) after 10 min of manual restraint. 3.3. Physiological assay validation of fecal corticosteroids Over a 24-h period, plasma CORT concentrations varied significantly (Fig. 2, Kruskal – Wallis, v2 = 16.20, df = 6, P=.013) as did fecal corticosteroid concentrations (Fig. 2, Kruskal –Wallis, v2 = 15.783, df = 6, P=.015). Plasma CORT
408
T. Good et al. / Physiology & Behavior 80 (2003) 405–411
Fig. 1. Short-term effects of handling and manual restraint on plasma CORT concentrations in P. polionotus. Four different individuals were sampled at each time period (n = 28).
concentrations rose rapidly between 8:30 a.m. and 12:30 p.m. (onset of the dark phase). They peaked during the dark phase and attained concentrations of 1227.11 F 214.53 ng/ ml (n = 12) during this phase. The concentrations reached their lowest values (160.38 F 47.45 ng/ml; n = 7) at the onset of the light phase (12:30 am – 4:30 a.m.). Peak values (55.91 F 6.19 ng/g; n = 8) occurred between 4:30 p.m. and 8:30 p.m. and lowest values (24.47 F 6.10 ng/g; n = 8) between 4:30 a.m. and 8:30 a.m.
Using Matlab curve fitter, the parameters ( F 95% confidence interval) for Cp were: a = 683 ( F 413), b = 2.4 ( F 3.2), c = 895 ( F 342) with an R2 of .84. For Cf, the parameters ( F 95% confidence interval) were: d = 21.8 ( F 7.1), e = 0.50 ( F 1.65), f = 35.8 ( F 5.7) with an R2 of .95. The lag time resulting from the curves is 2.9 F 3.6 h (mean F 95% confidence interval). The diel rhythm in fecal corticosteroid concentrations exhibited a 4-h time lag to that of the plasma CORT
Plasma CORT [ng/ml]
2500
Plasma CORT Fit to plasma CORT data Fecal corticosteroids Fit to fecal corticosteroids data
80
2000 60 1500 40 1000 20
500
Fecal corticosteroids [ng/g]
100
3000
0 0 4:30 am 8:30 am 12:30 pm 4:30 pm 8:30 pm 12:30 am 4:30 am
Hours Fig. 2. Diel rhythm of plasma CORT (.) and fecal corticosteroid (D) in P. polionotus. Four different individuals were sampled at each time period, with the exception of 8:30 a.m. (n = 2) and 12:30 a.m. (n = 3) where plasma samples were lost during the centrifuging process. Means are given with standard errors (S.E.). The horizontal bar indicates the dark phase of the light/dark cycle. Between 8:30 a.m. and 12:30 p.m. plasma CORT concentrations rose to double daylight values and a comparable doubling occurred during 12:30 p.m. and 4:30 p.m. for fecal corticosteroid concentrations. Similarly, plasma CORT concentrations decreased by half between 4:30 p.m. and 12:30 a.m. and fecal corticosteroid concentrations decreased by half between 8:30 p.m. and 4:30 a.m. R2=.84 for the curve fit to the plasma CORT data; and R2=.95 for the curve fit to the fecal corticosteroid data. See text for equations. The calculated lag time is 2.9 F 3.6 h (mean F 95% confidence interval).
T. Good et al. / Physiology & Behavior 80 (2003) 405–411
concentrations and correlated with that of the plasma 4 h earlier (Spearman’s rho: .829, df = 6, P < .042).
4. Discussion 4.1. Biochemical assay validation of the plasma CORT and fecal corticosteroid assay The CORT assay precisely and accurately measured CORT concentrations in plasma and corticosteroid concentrations in fecal samples for P. polionotus. Radioimmunoassays for fecal glucocorticoids have been validated for a number of species [7,20,42]. However, to our knowledge, only one other study has validated a corticosteroid assay for fecal samples in a Peromyscus species (P. maniculatus [10]). The recoveries for their ethanol extraction of radiolabeled CORT are comparable to ours (96.0 F 3.4% vs. 94.4 F 0.6%). 4.2. Physiological assay validation of plasma CORT One of our goals was to ascertain that samples collected within the first 3 min represented baseline CORT concentrations. CORT concentrations did not vary significantly during the first 3 min after disturbing the animal’s cage. A ninefold variability in baseline CORT concentrations has been reported from laboratory studies across different Peromyscus species (P. boylii, P. californicus [39]; P. leucopus [24]; P. maniculatus [2]; P. polionotus: this study). CORT concentrations in P. polionotus are the highest reported thus far. However, cross-species comparisons are problematic because there are too many other species-specific differences in factors, such as tissue receptor numbers or plasma binding protein levels, that alter physiological function (Romero, pers. comm.). CORT concentration increased significantly between 3 and 10 min. Our sampling protocol does not permit determination of whether the CORT concentrations were at their maximum levels after 10 min. In wild-caught P. californicus, maximum CORT concentrations were reached 5 min after the stressor (electric shock) had been applied ( f 1600 ng/ml [39]). In contrast, CORT concentrations in wildcaught P. boylii continued to increase for 60 min, eventually reaching the same peak values as P. californicus [39]. Both species live in similar environments, and the micro-environmental or species-physiological differences that would produce such differences in the temporal pattern of CORT response to stress are unknown. 4.3. Physiological assay validation of the fecal corticosteroid assay Our results indicate that fecal corticosteroid concentrations accurately reflected plasma CORT concentrations with a consistent lag. Corticosteroid concentrations in-
409
creased significantly at night in both plasma and fecal samples. Therefore, these results demonstrate that the fecal corticosteroid assay is sensitive enough to detect biologically meaningful alterations in corticosteroid concentrations in oldfield mice. To our knowledge, this is the first demonstration of a diel corticosteroid rhythm measured in fecal samples for a rodent. The values obtained in our study (24.47 55.91 F 13.68 ng/g) fall in the same range as those of P. maniculatus (42.41 84.13 F 12.22 ng/g [10]) even though the method of collection differed between the two studies. We collected samples directly from each individual every 4 h for 24 h, whereas Harper and Austad [10] collected samples from the cage floor every 12 h for a total of 192 h. The same authors found that when fecal samples were collected directly from individuals every 2 h just prior to the blood collection, fecal corticosteroid concentrations were much higher, ranging from 66.99 189.58 F 27.43 ng/g (Harper, unpublished data, pers. comm., Feb. 4, 2003). This suggests that fresh fecal samples have higher corticosteroid concentrations than samples collected at a later point in time and would lead us to conclude that P. maniculatus has higher corticosteroid concentrations than P. polionotus. It is likely that bacteria in the feces degrade corticosteroid metabolites during the interval between defecation and sample collection [19]. In order to be able to generalize and compare across studies and species, details, such as method and times of collection, should be taken into consideration. The lag time calculated from the parameters derived from the best fitting curves of the diel plasma CORT and diel fecal corticosteroid data was 2.9 F 3.6 (mean F 95% confidence interval). The uncertainty of the lag is due to the high interindividual variation in diel plasma CORT and diel fecal corticosteroid concentrations, yet the range of values calculated here are in accordance with those in the literature. When calculated from the actual plasma CORT and fecal corticosteroid data, the diel rhythm in fecal corticosteroid concentration exhibited an approximately 4-h lag time from that of the plasma CORT concentrations. Fecal corticosteroids were significantly higher in deer mice (P. maniculatus) that had been confined to traps for >4 h compared to those confined to traps for 0 –4 h suggesting that the lag time could be as little as 4 h in that species [11]. The best estimate of lag time available in mice is the detection of excreted radiolabelled tracer in defecated material. Peak excretion of [3H]CORT metabolites in the feces of lab mice (Mus musculus) occurred 4 h after administration of radiolabelled CORT during the dark phase of the light cycle [37]. Peromyscus spp. are ideal mammalian models to use in investigations of questions in behavioral endocrinology and physiological ecology as well as often being important target species in a range of ecological and conservation situations. The use of a fecal assay provides a valuable and powerful noninvasive method of measuring changes in
410
T. Good et al. / Physiology & Behavior 80 (2003) 405–411
corticosteroid concentrations in small mammals such as these under both laboratory and field conditions. Acknowledgements Jeanne Altmann, Michaela Hau, and two anonymous reviewers provided helpful comments on the manuscript. We thank James Harper for providing unpublished data of P. maniculatus fecal corticosteroid concentrations over a 24-h period. Catherine Marler kindly provided access to her laboratory group and advice during training. Gordon Brackee and Pat Martin trained TG in the method of retroorbital bleeding. Brian Elliot, Wenfei Tong, and Andre´ Utzinger provided invaluable help during fecal sample collection and extraction. Chris Wyckham provided the curve-fitting expertise. Support was provided by Princeton University, the Animal Behavior Society and the Society for Integrative and Comparative Biology to TG. References [1] Asa CS, Bauman JE, Houston EW, Fisher MT, Read B, Brownfield CM, et al. Patterns of excretion of fecal estradiol and progesterone and urinary chorionic gonadotropin in Grevy’s zebra (Equus grevyi): ovulatory cycles and pregnancy. Zoo Biol 2001;20:185 – 95. [2] Bradley EL, Terman CR. A comparison of the adrenal histology, reproductive condition, and serum corticosterone concentrations of prairie deermice (Peromyscus maniculatus bairdii) in captivity. J Mammal 1981;62:353 – 61. [3] D’Agostino J, Vaeth GF, Henning SJ. Diurnal rhythm of total and free concentrations of serum corticosterone in the rat. Acta Endocrinol 1982;100:85 – 90. [4] Dawson A, Howe PD. Plasma corticosterone in wild starlings (Sturnus vulgaris) immediately following capture and in relation to body weight during the annual cycle. Gen Comp Endocrinol 1983;51: 303 – 8. [5] Foltz DW. Genetic evidence for long-term monogamy in a small rodent, Peromyscus polionotus. Am Nat 1981;117:665 – 75. [6] Garnier JN, Green DI, Pickard AR, Shaw HJ, Holt WV. Non-invasive diagnosis of pregnancy in wild black rhinoceros (Diceros bicornis minor) by faecal steroid analysis. Reprod Fertil Dev 1998;10:451 – 8. [7] Goymann W, Mo¨stl E, Van’t Hof T, East M, Hofer H. Noninvasive fecal monitoring of glucocorticoids in spotted hyenas, Crocuta crocuta. Gen Comp Endocrinol 1999;114:340 – 8. [8] Gubernick DJ, Alberts JR. Postpartum maintenance of paternal behaviour in the biparental California mouse, Peromyscus californicus. Anim Behav 1989;37:656 – 64. [9] Haig D. Genetic conflicts and the private life of Peromyscus polionotus. Nat Genet 1999;22:131. [10] Harper JM, Austad SN. Fecal glucocorticosteroids: a non-invasive method of measuring adrenal activity in wild and captive rodents. Physiol Biochem Zool 2000;73:12 – 22. [11] Harper JM, Austad SN. Effect of capture and season on fecal glucocorticoid levels in deer mice (Peromyscus maniculatus) and red-backed voles (Cleithrionomys gapperi). Gen Comp Endocrinol 2001;123: 337 – 44. [12] Hem A, Smith AJ, Solberg P. Saphenous vein puncture for blood sampling of the mouse, rat, hamster, gerbil, guinea pig, ferret and mink. Lab Anim 1998;32:364 – 8. [13] Klein SL, Nelson RJ. Adaptive immune responses are linked to the mating system of arvicoline rodents. Am Nat 1998;151:59 – 67.
[14] Layne JN. Ontogeny. In: King JA, editor. Biology of Peromyscus (Rodentia), vol. 151. Stillwater, OK: The American Society of Mammalogists; 1968. p. 148 – 253 (Special Publication No. 2). [15] Lacy RC, Ballou JD. Effectiveness of selection in reducing the genetic load in populations of Peromyscus polionotus during generations of inbreeding. Evolution 1998;52:900 – 9. [16] Lynch JW, Ziegler TE, Strier KB. Individual and seasonal variation in fecal testosterone and cortisol levels of wild tufted capuchin monkeys, Cebus apella nigritus. Horm Behav 2002;41:275 – 87. [17] Margulis SW. Relationships among parental inbreeding, parental behavior and offspring viability in oldfield mice. Anim Behav 1998;55: 427 – 38. [18] Michaux J, Reyes A, Catzeflis F. Evolutionary history of the most speciose mammals: molecular phylogeny of muroid rodents. Mol Biol Evol 2001;18:2017 – 31. [19] Mo¨stl E, Messman S, Bagu E, Robia C, Palme R. Measurement of glucocorticoid metabolite concentrations in faeces of domestic livestock. J Vet Med A 1999;46:621 – 31. [20] Monfort SL, Wasser SK, Mashburn KL, Burke M, Brewer BA, Creel SR. Steroid metabolism and validation of noninvasive endocrine monitoring in the african wild dog (Lycaon pictus). Zoo Biol 1997;16: 533 – 48. [21] Munck A, Guyre PM, Holbrook NJ. Physiological functions of glucocorticoids in stress and their relation to pharmacological effects. Endocr Rev 1984;5:25 – 44. [22] Neter J, Wasserman W, Kutner M. Applied linear statistical models: regression, analysis of variance and experimental designs. Homewood, IL: Irwin; 1990. p. 165 – 7. [23] Place NJ, Kenagy GJ. Seasonal changes in plasma testosterone and glucocorticoids in free-living male yellow-pine chipmunks and the response to capture and handling. J Comp Physiol B 2000;170:245 – 51. [24] Ransone SN, Bradley EL. Correlations between serum corticosterone concentration and reproductive conditions in the white-footed mouse (Peromyscus leucopus noveboracensis). Comp Biochem Physiol 1992;102A:453 – 7. [25] Rave EA, Holler NR. Population dynamics of beach mice (Peromyscus polionotus) in southern Alabama. J Mammal 1992;73:347 – 55. [26] Ribble DO. The monogamous mating system of Peromyscus californicus as revealed by DNA fingerprinting. Behav Ecol Sociobiol 1991; 29:161 – 6. [27] Romero LM. Seasonal changes in plasma glucocorticoid concentrations in free-living vertebrates. Gen Comp Endocrinol 2002;128:1 – 24. [28] Romero LM, Remage-Healey L. Daily and seasonal variation in response to stress in captive starlings (Sturnus vulgaris): corticosterone. Gen Comp Endocrinol 2000;119:52 – 9. [29] Romero LM, Romero RC. Corticosterone responses in wild birds: the importance of rapid initial sampling. Condor 2002;104:129 – 35. [30] Ryan KK, Altmann J. Selection for male choice based primarily on mate compatibility in the oldfield mouse, P. polionotus rhoadsi. Behav Ecol Sociobiol 2001;55:436 – 40. [31] Sabatino F, Masoro EJ, McMahan A, Kuhn RW. Assessment of the role of glucocorticoid system in aging process and in the action of food restriction. J Gerontol 1991;46B:171 – 9. [32] Saplosky RM, Romero LM, Munck AU. How do glucocorticoids influence stress-responses? Integrating permissive, suppressive, stimulatory, and adaptive actions. Endocr Rev 2000;21:55 – 89. [33] Schoech SJ, Ketterson ED, Nolan Jr V. Exogenous testosterone and the adrenocortical response in dark-eyed juncos. Auk 1999;116:64 – 72. [34] Shimizu K, Amagaya S, Ogihara Y. Analysis of corticosterone in the serum of mice and rats, using high performance liquid chromatography. J Chromatogr 1983;272:170 – 5. [35] Strier KB, Ziegler TE. Behavioral and endocrine characteristics of the reproductive cycle in wild muriqui monkeys, Brachyteles arachnoids. Am J Primatol 1997;42:299 – 310. [36] Swilling WR, Wooten MC. Subadult dispersal in a monogamous species: the Alabama beach mouse (Peromyscus polionotus ammobates). J Mammal 2002;83:252 – 9.
T. Good et al. / Physiology & Behavior 80 (2003) 405–411 [37] Touma C, Sachser N, Mo¨stl E, Palme R. Effects of sex and time of day on metabolism and excretion of corticosterone in urine and feces of mice. Gen Comp Endocrinol 2003;130:267 – 78. [38] Trainor B, Marler CA. Testosterone, parental behavior, and aggression in the monogamous California mouse (Peromyscus californicus). Horm Behav 2001;40:32 – 42. [39] Treiman DM, Levine S. Plasma corticosteroid response to stress in four species of wild mice. Endocrinology 1969;84:676 – 80. [40] Vrana PB, Guan XJ, Ingram RS, Tilghman SM. Genomic imprinting is disrupted in interspecific Peromyscus hybrids. Nat Genet 1998;20: 362 – 5. [41] Wasser SK, Monfort SL, Southers J, Wildt DE. Excretion rates and
411
metabolites of oestradiol and progesterone in baboon (Papio cynocephalus cynocephalus) faeces. J Reprod Fertil 1994;101:213 – 20. [42] Wasser SK, Velloso AdL, Rodden MD. Using fecal steroids to evaluate reproductive function in female maned wolves. J Wildl Manage 1995;59:889 – 94. [43] Wasser SK, Hunt KE, Brown JL, Cooper K, Crockett CM, Bechert U, et al. A generalized fecal glucocorticoid assay for use in a diverse array of nondomestic mammalian and avian species. Gen Comp Endocrinol 2000;120:260 – 75. [44] Whitten PL, Brockman DK, Stavisky RC. Recent advances in noninvasive techniques to monitor hormone – behavior interactions. Yearb Phys Anthropol 1998;41:1 – 23.