| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
COMPLEX FUNCTION OF THE CENTRAL NERVOUS SYSTEM, SLEEP AND LOCOMOTION
1Integrative Biology, University of Texas, Austin, Texas 78712; 2Department of Biology, Arizona State University, Tempe, Arizona 85287; 3Department of Zoology, University of Washington, Seattle, Washington 98915; 4Neurobiology, Physiology, and Behavior, University of California, Davis, California 95616; and 5Division of Biomedical and Clinical Laboratory Sciences, University of Edinburgh Medical School, Edinburgh EH8 9XD, Scotland, United Kingdom
Submitted 9 December 2002 ; accepted in final form 18 May 2003
 |
ABSTRACT |
|---|
 TOP ABSTRACT MATERIALS AND METHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
white-crowned sparrow; corticosterone; corticosteroid-binding globulin; corticosteroid receptor; reproduction
This series of experiments examined plasticity of the stress response among three populations of white-crowned sparrow (Zonotrichia leucophrys). The three populations have different distributions that are likely to affect the evolutionary pressures shaping the stress response. In particular, these populations breed at different elevations and latitudes and thus have breeding seasons that differ markedly in length (see Fig. 1). High-latitude birds (from the subspecies Z. l. gambelii) were sampled north of the Brooks Range in the Alaskan Arctic, where the short breeding season allows only one clutch per season (10). High-elevation birds (from the subspecies Z. l. oriantha) were sampled in subalpine meadows of the Sierra Nevada, where two broods per season are sometimes possible (10, 17). Low-elevation temperate zone birds (from the subspecies Z. l. pugetensis) were sampled in western Washington State, where they often raise three broods per season (10, 28).
|
We hypothesized that the length of breeding season has a strong evolutionary influence on how birds respond to stress, particularly in short-lived taxa such as white-crowned sparrows where reproductive opportunities in subsequent years are relatively unlikely. Specifically, in birds with time to raise only one or rarely two broods in a season, the fitness costs of abandoning a nest may be substantially larger than in birds that can raise up to three broods per season. There is evidence that both environmental and physiological stressors can induce nest abandonment (11, 26, 31) and that these effects may be mediated through increased Cort (Refs. 21 and 31 and O. Love, unpublished data). Thus individuals with short breeding seasons should be less responsive to stressors and therefore less likely to abandon their young.
Surprisingly, the adrenocortical response to stress (secretion of glucocorticoids from the adrenal gland) does not differ in the two subspecies measured during nesting (oriantha and pugetensis; Refs. 4 and 30). However, Cort secretion is only one component of a complex pathway from stressor to behavioral and physiological responses. Although it appears that oriantha and pugetensis do not show the predicted differences in Cort secretion, there could be regulation further downstream in the system, with changes occurring in levels of corticosteroid-binding globulin (affecting free hormone levels available to cells; Ref. 16) as well as tissue-specific changes in the three types of corticosteroid receptors (2 intracellular and 1 membrane bound).
Within this study, we investigated several components of the stress response to test the hypothesis that populations with shorter breeding seasons will be less sensitive to stressors. Toward this end, we caught birds on their territories during the nesting phase of the breeding season, measuring baseline and handling-induced Cort secretion, plasma corticosteroid-binding globulin capacity, and intracellular corticosteroid receptor capacity in liver and brain tissue. Samples were collected from the oriantha and gambelii during their first clutch and from pugetensis during their second to third clutch. Our predictions, based on simple extrapolation from the hypotheses, are 1) although baseline and stress-induced Cort levels may be similar among populations during nesting, corticosteroid-binding globulin levels will be higher in gambelii (the population with the shortest breeding season), decreasing the amount of stress-induced free Cort; 2) intracellular glucocorticoid-like receptors (the low-affinity receptor that responds to stress-induced levels of Cort) will be present in lower levels in the brain of the gambelii population during nesting, potentially resulting in a reduced behavioral sensitivity to stress; and, finally, 3) levels of both intracellular receptor types in the liver will be similar in each population, because metabolic need for glucose in response to stress should not vary geographically.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
Males at each location were caught between 7:00 and 11:00 AM with the use
of mist nets and conspecific playback on their territories during the nesting
phase of the breeding season. A baseline blood sample was obtained within 3
min of capture (alar vein puncture with a 26-gauge needle; 
80 µlof
blood collected into heparinized microcapillary tubes). Males were held in a
cloth bag until another blood sample was taken at 30 min. Blood samples were
kept on ice until plasma was separated by centrifugation and stored at
-20°C. For liver and brain tissue collection, males were brought into
captivity and immediately given a mitotane injection (0.71 mg/g body wt) to
reduce endogenous glucocorticoid production
(5). Captive photoperiods were
matched to current day length. Approximately 36 h after injection, 3- and
30-min blood samples were taken to assess the efficacy of the mitotane
treatment; animals were then anesthetized with Nembutal (0.09 mg/kg) and
perfused transcardially with heparinized saline. Brain and liver tissue was
removed, snap frozen on dry ice, and stored at -75°C. All procedures
complied with university and federal regulations.
Z. l. pugetensis. Eight male pugetensis were captured at Charles Lathrop Pack Forest Station (Puget Sound region, south of Seattle, WA; 47° north, 275 m elevation) on July 7 and 8, 2000. After capture, birds were held in environmental chambers at the Univ. of Washington, given mitotane injections, and perfused 36 h later.
Z. l. oriantha. All males were captured at Sonora Pass in the Sierra Nevada, CA (38° north, 2,940 m elevation). Blood samples were collected from eight free-living males captured June 2 and 3, 2000. Eight males were brought into captivity May 25 and 26, 2001 (at the field station near Lee Vining, CA), given mitotane injections, and processed for tissue collection 36 h later.
Z. l. gambelii. All males were captured at the Toolik Field Station, north of the Brooks Range in Arkansas (68° north, 720 m elevation). Eight males were brought to the field station June 8 and 16, 2000, for mitotane injections and tissue collection. Blood samples were collected from eight free-living males June 17-24, 2001.
Corticosterone Assay
Plasma Cort levels were determined following the methods of Wingfield et al. (34). Briefly, samples were allowed to equilibrate overnight with 2,000 counts/min (cpm) of corticosterone for determination of individual recoveries. Each sample was extracted with 4.0 ml of dichloromethane, dried under nitrogen, and resuspended in phosphate-buffered saline with 1% gelatin. Samples were assayed in duplicate, and assay values were corrected for plasma volume and individual recoveries after extraction (recoveries after extraction, 71-100%; standard curve range, 2,000-7 pg; accuracy, 83%; detectability, 7.8 pg/tube). Intra-assay coefficients of variation were 4 and 15% (oriantha and pugetensis samples were run in one assay, and gambelii samples were run in a second assay); interassay coefficient of variation was 15%.
Corticosteroid Receptor Assays
Corticoid receptor assays were performed as per Breuner and Orchinik (8). Temperature, rinse volume, and tissue concentration were optimized for each receptor assay to maximize specific binding, and the time to reach equilibrium was empirically determined for each receptor. All assays contained 50 µl [3H]Cort, 50 µl buffer or unlabeled Cort, and 50 µl tissue or plasma preparation. Nonspecific binding was determined by use of 1 µM unlabeled Cort. All samples were run in triplicate. Bound and free radioligands were separated with the use of rapid vacuum filtration over glass fiber filters (Brandel Harvester). After filtration, radioactivity bound to filters was measured by standard liquid scintillation spectroscopy.
Corticosteroid-Binding Globulin
Plasma collected to determine baseline Cort levels was also used to measure
corticosteroid-binding globulin (CBG) affinity and capacity. Plasma was
stripped of endogenous steroid in a 20-min room-temperature incubation with 2
vol dextran-coated charcoal solution (0.1% dextran, 1% Norit A charcoal in 50
mM Tris). Plasma was maintained at <4°C at all times outside of this
stripping process. Plasma samples were assayed at a final dilution of 1:900;
assays were performed at 4°C in 50 mM Tris buffer and terminated after 2
h. Glass fiber filters were soaked in 25 mM Tris with 0.3% polyethylenimine
for 1 h before filtering. Filters were rapidly rinsed with 9 ml ice-cold 25 mM
Tris (3 rinses of 3 ml each). Point sample analysis was run on individual
plasma samples, whereas saturation analyses were run on pooled samples. For
saturation analysis, pooled plasma from each subspecies was incubated with
0.25-12 nM [3H]Cort in the presence or absence of unlabeled Cort.
CBG capacity in individual birds was estimated by use of either 14.4 nM (2001
pugetensis and oriantha) or 12.6 nM (2002 gambelii)
[3H]Cort. On the basis of affinity estimates derived from
equilibrium saturation analysis, this ligand concentration should occupy
74, 80, and 78% of total binding sites in pugetensis, oriantha,
and gambelii, respectively. To account for interassay variation, the
pugetensis samples were run in both assays (2001 and 2002) and
corrected to 100% capacity within each assay, and gambelii capacity
was corrected for the variation in pugetensis samples between
assays.
Cytosolic Receptors
Saline-perfused brains were hemisectioned with a midsagittal cut, and
one-half of the brain was homogenized in a TEGMD buffer (10 mM Tris, 1 mM
EDTA, 10% glycerol, 20 mM molybdic acid, and 5 mM dithiothreitol) and
centrifuged at 104,000 g for 1 h at 4°C to produce cytosol with a
protein concentration of 4-8 mg/ml (determined using Bradford reagent and a
standard curve of BSA). Assays were performed at room temperature in TEGM
buffer and terminated after 4 h. Glass fiber filters were soaked in TEM buffer
with 0.3% polyethylenimine for 1 h before filtering. Tissue on filters was
rapidly rinsed with 9 ml ice-cold TEM buffer (3 rinses of 3 ml each).
Competition and saturation experiments were performed with the use of tissue
pooled from every individual in the population. For single-point assays, each
one-half brain was processed separately. For saturation binding analysis,
pooled cytosol was incubated with [3H]Cort ranging from 0.05 to 12
nM. The high- and low-affinity cytosolic receptors were distinguished by
incubation of cytosol with radioligand and 8 nM RU486, a mammalian
glucocorticoid receptor antagonist that occupies lower-affinity Cort
receptors, or 1 µM unlabeled Cort to define nonspecific binding. Therefore,
at each [3H]Cort concentration, we had measures of total specific
binding to all cytosolic corticosteroid receptors and specific binding to
high-affinity receptors (in presence of 8 nM unlabeled RU486). We calculated
binding to lower-affinity receptors as the difference between the total
specific binding and specific binding to high-affinity sites. Individual
estimations of the concentrations of high- and low-affinity receptors were
made by using 10 nM [3H]Cort in the presence of either buffer
(labeling both receptors), 8 nM RU486 (blocking the low-affinity receptor), or
1 µM Cort (blocking both receptors to determine nonspecific binding). On
the basis of affinity estimates derived from our equilibrium saturation
analysis, mass action predicts that 10 nM [3H]Cort should occupy
>95% of high-affinity receptors and 63% of lower-affinity receptors.
To avoid interassay variation, receptor number was determined for all
individuals in the same assay.
Statistics
Population-dependent differences in total Cort levels were identified by
use of a repeated-measures ANOVA (with time as the repeated factor; StatView
5, SAS Institute, Cary, NC). Free Cort levels were log transformed to correct
for heteroscedasticity, and population differences were identified with the
use of factorial ANOVA on baseline and stress-induced samples separately,
followed post hoc by Fisher's protected least-significant difference test
(PLSD). Binding parameter estimates from the saturation analysis were obtained
by fitting untransformed data to appropriate equations using iterative,
least-squares curve-fitting techniques (GraphPad Prism, San Diego, CA). For
analysis, CBG and high- and low-affinity receptor capacity data were brought
to 100% and compared by use of factorial ANOVAs followed by Fisher's PLSD. A
familywise 
= 0.05 significance level was used for all tests.
Free Cort titers were estimated from total Cort concentrations and CBG binding parameters by use of the equation of Barsano and Baumann (3) where Hfree is free hormone, Htotal is total hormone, Bmax is total binding capacity of CBG, and Ka = 1/dissociation constant (Kd) (all values in nM). Baseline Cort, stress-induced Cort, and CBG capacity were measured in each individual (from blood samples taken at capture), allowing free Cort estimations for each individual, and mean ± SE was calculated for each population.
 |
|
RESULTS |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
Thirty minutes of capture and handling significantly elevated total Cort in all three populations (ANOVA: F2,21 = 398.9, P < 0.0001). However, there were no significant differences among populations in either baseline or stress-induced total Cort levels (Fig. 2A; ANOVA: F2,21 = 0.639, P = 0.54).
|
CBG
In white-crowned sparrow plasma, we found a single high-affinity (3
nM) binding site for corticosterone (data not shown), consistent with Lynn et
al. (13). CBG affinity for
corticosterone and total binding capacity differed among the populations
(Fig. 3). CBG from the
gambelii population had higher capacity than CBG from either the
pugetensis or the oriantha populations (ANOVA:
F2,21 = 6.94, P < 0.005), whereas CBG from the
pugetensis population had lower affinity for corticosterone than CBG
from the oriantha or gambelii populations (ANOVA:
F2,21 = 4.99, P < 0.02).
|
Free Corticosterone
Stress-induced free Cort estimates were lowest in the gambelii population, twofold higher in the oriantha population, and fivefold higher in the pugetensis population (Fig. 2B; ANOVA: F2,21 = 17.32, P < 0.0001). Baseline free Cort estimates were lower in the gambelii population than in the pugetensis or oriantha populations (ANOVA: F2,21 = 4.24, P < 0.03).
Intracellular Receptors
Thirty-six hours of mitotane treatment effectively blocked the
adrenocortical response to stress in both pugetensis (0- to 3-min
sample, 5.09 ± 1.16 ng/ml; 30-min sample, 6.46 ± 1.80 ng/ml) and
oriantha (0- to 3-min sample, 12.58 ± 4.19 ng/ml; 30-min
sample, 13.51 ± 3.40 ng/ml). In gambelii, mitotane treatment
was less effective (0- to 3-min sample, 4.50 ± 1.26 ng/ml; 30-min
sample, 17.43 ± 3.28 ng/ml). However, estimates of stress-induced free
Cort concentrations in mitotane-treated gambelii are low (0.76
nM). Mass action predicts that this level of free Cort would occupy only 10%
of receptors. As a consequence, our estimate of the total number of
low-affinity receptors measured is probably only 10% low.
In both brain and liver cytosol, there are two specific binding sites for corticosterone: one high-affinity mineralocorticoid receptor (MR)-like site and one lower-affinity glucocorticoid receptor (GR)-like site (Fig. 4). Receptor capacities differed significantly among populations (see Table 1). In liver cytosol, there were no differences in high-affinity receptor capacity, but low-affinity receptor capacity was highest in oriantha, 30% lower in pugetensis, and 52% lower in gambelii. Cytosolic low-affinity receptors from brain tissue showed similar population differences in capacity. In brain cytosol, receptor affinity ranged from 0.08 ± 0.05 to 0.32 ± 0.22 nM (high-affinity receptor) and from 5.5 ± 1.4 to 14.2 ± 23.8 nM (low-affinity receptor). In liver cytosol, receptor affinity ranged from 0.21 ± 0.04 to 0.35 ± 0.10 nM (high affinity) and from 12.0 ± 1.4 to 23.2 ± 13.6 nM (low affinity). Within tissue type, however, there were no population-dependent differences in receptor affinities.
|
|
|
DISCUSSION |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
In avian cytosol, there are two intracellular corticosteroid receptors. High-affinity receptor capacities in liver tissue were similar among the three populations but differed in the brain tissue. Within this study, however, we are more interested in concentration of low-affinity glucocorticoid receptors in avian tissue, as this receptor is most likely to play a prominent role in stress-related functions (the low affinity of this receptor indicates it will not be significantly activated until Cort reaches stress-induced levels; Ref. 6). Low-affinity GR-like receptor capacities differed among the three populations in both brain and liver tissue: oriantha had the most low-affinity receptors in both liver and brain, whereas gambelii had the fewest.
It is hypothesized that in birds breeding in extreme habitats (e.g., Z. l. gambelii in the Arctic), the central nervous system is "desensitized" to high corticosterone levels (2, 15, 32) so that harsh environmental conditions do not interrupt breeding. One mechanism to decrease sensitivity to circulating glucocorticoids would be a reduction in neural glucocorticoid receptors; our data indicate that GR-like receptors are lower in gambelii than in their temperate conspecifics. With decreased numbers of brain corticosteroid receptors, one would predict reduced behavioral responses in Arctic vs. temperate species, given the same levels of glucocorticoids.
This "Cort insensitivity hypothesis" has been tested on aggressive responses to territorial intrusion; in several temperate species, Cort implants decrease the aggressive response to conspecific playback (29, 33). In Arctic-breeding American tree sparrows (Spizella arborea) and Smith's longspurs (Calcarius pictus), Cort implants have no effect on aggressive behavior as predicted by the hypothesis (2, 14). However, Arctic Z. l. gambelii do appear to be sensitive to Cort implants (15). To investigate whether there is differential sensitivity to Cort among white-crowned sparrow populations, we need to create similar free Cort levels in each population by Cort implant studies; data presented here indicate that it would be difficult to compare neural sensitivity to increased Cort among populations without accounting for differences in binding globulin parameters. It would also be useful to determine how low-affinity "glucocorticoid" receptor number changes in specific brain regions (this study measured whole brain receptor capacity) and relate this to behavioral effects of exogenous Cort administration.
The data in this study allow us to develop a simple model of corticosteroid activity based on integration of Cort, CBG, and receptor levels in each population. Using mass action, we have estimated the fractional occupancy of the low-affinity receptor at stress-induced free Cort levels in both liver and brain tissue (Fig. 5). This model predicts that the functional output of the stress response varies with the number of broods white-crowned sparrows can typically raise in one season.
|
What are the origins of differences in stress responses among populations? First, evolutionary pressures are likely to have shaped the stress response to optimize breeding success in particular environments. Shorter breeding seasons should lead to greater fitness costs when nests are abandoned. In this case, selection would favor individuals who are less sensitive to stressors while there are young in the nest and therefore less likely to abandon when environmental conditions deteriorate. The least responsive animals may not survive the breeding effort, leading to stabilizing selection at an intermediate level of sensitivity; however, populations breeding under extreme time constraints should stabilize at a lower sensitivity than those with time to raise several broods per season.
Alternatively, observed differences in stress sensitivity among populations may be related to clutch number at sampling. The stress series were collected from gambelii during their first and only clutch and from oriantha during their first and (most likely) only clutch. However, stress series from pugetensis were collected in early July, when pairs were probably on their second or third clutch of the season. It is possible that the birds become more sensitive to environmental perturbations once they have successfully completed a clutch in one season, resulting in a more responsive HPA axis as they raise their second or third clutch. However, glucocorticoid levels measured in pugetensis in this study are similar to those obtained from pugetensis in other substages of nesting (Wingfield, unpublished observations). It is possible that CBG capacity changes with clutch order; this has not been tested in any species we are aware of and would present an interesting mechanism for short-term (among-clutch) plasticity in stress responsivity. Regardless of whether the observed differences reflect inherent differences among groups sampled (phylogenetic or geographic) or a seasonally plastic response that varies with brood number, our results provide novel evidence of within-species variation in several components of stress physiology and suggest numerous productive lines of future investigation.
Perspectives
The most striking result in this study is the difference in CBG among populations. Why increase CBG capacity instead of reducing Cort secretion rate? In general, CBG is thought to regulate bioavailability and metabolic clearance of Cort (16, 19), so CBG-bound Cort are unavailable to enter and activate peripheral tissues. However, higher CBG capacity means a larger pool of Cort is available in the blood if it is needed (9); CBG can be broken down in response to numerous stressors, ranging from social subordination to food deprivation (1, 23, 25) to increased free Cort without a concurrent increase in total Cort. Additionally, binding sites for CBG have been identified in multiple tissues (i.e., Refs. 12, 22, and 24), implicating CBG in site-specific delivery of Cort. Taken together, CBG appears to increase the flexibility and specificity with which animals can respond to various environmental and physiological stressors (7). In the present study, total Cort levels do not vary among populations, but CBG levels do, leading to significant differences in free Cort where none were expected. In house sparrows (Passer domesticus), Cort levels vary seasonally but so does CBG capacity, leading to surprisingly static free Cort levels (8). In future studies, consideration must be taken of the role CBG plays in regulating the stress response and how both total and free Cort levels differ among individuals, sexes, populations, and taxa.
|
DISCLOSURES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
|
ACKNOWLEDGMENTS |
|---|
|
FOOTNOTES |
|---|
The costs of publication of this article were defrayed in part by the payment of page charges. The article must therefore be hereby marked "advertisement" in accordance with 18 U.S.C. Section 1734 solely to indicate this fact.
|
REFERENCES |
|---|
|
TOPABSTRACTMATERIALS AND METHODSRESULTSDISCUSSIONDISCLOSURESREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  Z. G Hodgson, S. L Meddle, M. L Roberts, K. L Buchanan, M. R Evans, R. Metzdorf, M. Gahr, and S. D Healy Spatial ability is impaired and hippocampal mineralocorticoid receptor mRNA expression reduced in zebra finches (Taeniopygia guttata) selected for acute high corticosterone response to stress Proc R Soc B, January 22, 2007; 274(1607): 239 - 245. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
B. J Heidinger, I. C.T Nisbet, and E. D Ketterson Older parents are less responsive to a stressor in a long-lived seabird: a mechanism for increased reproductive performance with age? Proc R Soc B, September 7, 2006; 273(1598): 2227 - 2231. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
| Visit Other APS Journals Online |
)or absence (
) of 8 nM RU486. Without RU486, data
are best fit by a 2-site model with Kd values of 0.28
± 0.04 and 11.96 ± 1.43 nM. With RU486, the data are best fit by
a 1-site model with Kd = 0.21 ± 0.04 nM.
Inset: Scatchard-Rosenthal replot of the data. Analysis of
pugetensis and oriantha tissue gave similar results.