Circadian Patterns of Stress-Induced ACTH Secretion Are Modified by Corticosterone Responses WILLIAM C. ENGELAND, JEANETTE SHINSAKO, CHARLES M. WINGET, JOAN VERNIKOS-DANELLIS, AND MARY F. DALLMAN Department of Physiology and the Metabolic Unit, University of California, San Francisco, and NASA-Ames Research Center, Moffett Field, California ABSTRACT. To test whether there is a circadian rhythm in the ACTH response to stress, young female rats were exposed to a variety of ACTH-releasing stimuli at 0600 and 1800 h and changes in circulating ACTH and corticosterone were measured. The results of these experiments suggested that after the high intensity stimuli of laparotomy with intestinal traction or 250 fig histamine ip/100 g BW, the morning ACTH response was greater than the
7.5 and 15 min after the start of surgery and after injection with either 2% EtOH-saline, or 50 fig corticosterone at operation, or with 30 fig corticosterone at 5 min. Compared with ACTH levels in rats treated with vehicle, ACTH levels were decreased 7.5 min after 50 fig corticosterone at operation (P < 0.01), but not after 30 fig corticosterone at 5 min. ACTH levels were slightly reduced 10 min after 30 fig corticosterone at 5 min compared with
evening response. However, the ACTH response
those of rats injected
to ip saline was greater in the evening in one experiment and greater in the morning in a second experiment. Plasma corticosterone responses were faster and greater in the morning in the first experiment and in the evening in the second experiment. The ACTH response to 125 fig histamine ip/100 g BW was greater in the evening and the change in corticosterone was greater in the morning. Thus, after low intensity stimuli, the ACTH responses appeared to depend on both the lag time prior to the corticosterone response, and its magnitude. To test this possibility, rats were adrenalectomized and the ACTH response was measured
(P < 0.05). These results are consistent with the interpretation that corticosterone secretion modifies stress-induced ACTH secretion via the fast-feedback effect. Comparison of the ACTH responses to acute adrenalectomy plus injection with EtOH-saline at 0600 and 1800 h demonstrated that, in the absence of a corticosterone response to the stress, the ACTH response is greater in the morning than in the evening (P < 0.01). Finally, this group of experiments suggests strongly that the responsiveness of rat adrenal glands to ACTH increases markedly between 0600 and 1800 h. (Endocrinology 100: 138, 1977)
with vehicle at operation
M
ARKED circadian fluctuations in logical range of plasma corticosterone plasma corticosteroid levels have whereas trough levels approach or define the been thoroughly documented in man, bottom of the range, and may be zero. It has monkeys and rodents (1-3). There is general been shown that elevations of corticosterone agreement that under normal conditions within the physiological range can act to corticosteroids rise to a peak at the start inhibit stress-induced ACTH secretion (4); of the daily activity cycle (1). Rats, which and corticosterone has been shown to act on are active nocturnally, exhibit peak plasma the brain to inhibit stress-induced CRF corticosterone levels at or just after dark (3). formation and secretion (5). From these obThere is a 5 to 10-fold difference between servations one might predict that the ACTH peak and trough levels of plasma corti- response to stress applied in the morning costerone in rats. Peak concentrations are when circulating corticosterone levels are approximately at the midpoint of the physio- low would be greater than the ACTH response to stress applied in the evening when plasma corticosterone levels are elevated. Received July 24, 1975. Supported by NIH grants AM 06704, GM 00927, This point has been tested repeatedly; NS 09528, 2 NASA-University Consortium agreeresponsiveness to stress has been found to ments NCAR-665-401 and NCA2-OR665-502, as well as be greater in the morning than in the eveby UCSF research funds. Mary F. Dallman is the ning (6), of equal magnitude morning and recipient of NIH Research Career Development Award evening (7), and of greater magnitude in the AM 00072. 138
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 May 2016. at 04:39 For personal use only. No other uses without permission. . All rights reserved.
ORCADIAN STRESS RESPONSES
evening than in the morning (8). Generally in these studies, changes in plasma corticosterone concentrations were measured and these changes were presumed to reflect ACTH secretion. Use of corticosterone as an index of ACTH secretion assumes that the response of the adrenal cortex to ACTH does not change between morning and evening. If changes in peripheral plasma corticosterone, rather than adrenal corticosterone content or secretion rate, are used as an endpoint, the further assumptions are made that the metabolic clearance rate and distribution volume of corticosterone are constant throughout the day. Additionally, it is difficult to compare quantitatively plasma corticosterone responses that start from different initial values. If initial levels are low, infusion of corticosterone at a given rate results in a greater elevation of plasma corticosterone than if the initial levels are high (4), probably because in the low range much of the infused steroid is bound with high affinity to circulating transcortin. Because of this, neither the incremental responses nor the maximum value attained after stress can be used directly to compare the magnitude of the stress response in the morning and evening. Although many of the same considerations apply to quantifying the stress response by measuring circulating ACTH levels morning and evening, ACTH is at least one step closer to the site of the neural initiation of stress responses, and it is likely that there are less dramatic nonlinearities in the ACTH distribution-binding-metabolism system than those in the corticosterone system (9). In this study, we have measured stress-induced changes in both circulating ACTH and corticosterone at frequent intervals after each of 3 stresses. In addition, the responses of adrenalectomized rats to the same stresses were determined. We hoped to define the role, if any, of circadian changes in circulating corticosterone levels on stressinduced ACTH secretion. Portions of the data have been reported in an abstract (10).
139
Materials and Methods Female Sprague-Dawley rats 100-130 g BW (Simonsen, Gilroy, Ca.) were received 7 to 10 days before use and were housed in animal rooms with controlled temperature (22-24 C) and light (fluorescent lighting, lights on 0600, lights off 1800 h). Two studies were performed at the NASA-Ames research facilities at Moffett Field, and 6 studies were performed in the animal facility at the University of California, San Francisco. At Ames, rats were housed in plastic pans, 6-8/cage, and were moved into the laboratory 24 h before the experiments were conducted. In San Francisco, rats were housed 2-3/cage in hanging wire mesh cages and the experiments were performed in the animal room. In all experiments, animals were weighed, numbered and apportioned to the correct cages 1-2 days before the experiment and were not subjected to further handling or unusual environmental noise thereafter. In 3 experiments, adrenalectomy was performed under ether anesthesia by the dorsal approach 3 days before the study. Rats assigned to the evening groups were adrenalectomized in the evening. Treatments Either 0.9% saline or histamine injected intraperitoneally was used as a stimulus to ACTH secretion. For the injections, rats were picked up firmly by the skin of the neck and back, and were held vertically while injected with substances through a 27 gauge Vi in needle into the midline peritoneal region about 2 cm below the xyphoid process. The dose of saline was 0.2 ml/ 100 g BW. Histamine was given in two forms. In one experiment histamine phosphate, 250 ixg free base/100 g BW, was given; in the second experiment, histamine -2HC1, 125 fig free base/ 100 g BW, was given. Dexamethasone phosphate (Decadron®, Merck, Sharp and Dohme) when used was diluted and rats were injected with 2, 5 or 10 ;ug/100 g BW 2.5 h before stress was applied. The response to laparotomy with intestinal traction under ether anesthesia was also studied. A 2 cm midline abdominal incision was made and the length of the small intestine was gently exteriorized and handled. The intestine was returned to the peritoneal cavity, the skin incision was clipped, and rats were returned to their home cages until sacrificed.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 May 2016. at 04:39 For personal use only. No other uses without permission. . All rights reserved.
Endo • 1977 Vol 100 • No 1
ENGELAND ET AL.
140 TABLE
1. Plasma ACTH and corticosterone responses to ip saline given at 5 different times of the
Time after
day
Plasma corticosterones (fig/100 ml)
Plasma ACTH (pg/ml)
(min)
0600 h
1200 h
1800 h
2400 h
0300 h
0600 h
1200 h
1800 h
2400 h
0300 h
0
63* ±7 212 ±45 193 ±58 171
95 ±20
109 ± 19
88 ± 10
110 ± 15
4.3 ± 1.0
20.5 ± 4.8
49.6 ±8.1
12.6 ± 2.8
171 ± 36
219 ± 22
212 ± 33
283 ±44
170 ±26
216 ± 27
355 ±53
263 ±45 266 ±50
60.8 ±5.4 67.2 ±8.8
± 19
83.7 ± 5.0
72.3 ± 7.8
63.0 ± 5.6
184 ± 44
162 ± 12
173 ±26 163 ± 39
176 ±39 204 ±77 187 ±34
18.8 ±5.5 53.4 ± 6.1 51.5 ±4.6 49.2 ±3.3
37.0 ±4.3
175 ± 20
23.6 ± 3.9 35.7 ±3.6 42.9 ±3.1 48.1 ±5.0 54.3 ±4.3
61.1 ±4.8 51.0 ±2.2 81.7 ± 8.4
74.7 ±6.3 48.3 ±9.3
56.2 ± 6.6 65.0 ± 4.4
19.8 ±5.7 17.5 ±4.4
36.5 ±7.9
84.8 ± 7.7 57.3 ± 10.0
37.3 ±5.0
2.5 5.0 7.5
10.0 15.0
171 ± 18
238 ± 34
179 ± 17
239
±51 357
±97
30.0 60.0
57.9 ± 6.1
24.5 ±8.4
40.8 ±9.2
49.8 ±4.0
60.0 ± 2.7
62.7 ± 6.6 62.7 ±4.5 19.2 ±4.3
Mean of 6 rats/group ± SEM.
The acute response to bilateral adrenalectomy was studied in two groups of rats at 0600 h and at 1800 h. In these experiments rats were anesthetized with sodium pentobarbital, 4.5 jiig/100 g BW (Nembutal®, Abbott), bilaterally adrenalectomized and were killed at the times indicated. In one experiment corticosterone dissolved in a small volume of EtOH and diluted with 0.9% saline to an EtOH concentration of 2% was given in two regimens. One group of animals was given 50 fig corticosterone immediately after adrenalectomy; another group was given 30 fig corticosterone SC 5 min after the start of adrenalectomy. Both groups were killed at 7.5 and 15 min and the results were compared to those from adrenalectomized rats treated at operation with the vehicle. All rats were killed by decapitation and trunk blood was collected in iced, heparinized plastic centrifuge tubes. The blood was rapidly centrifuged at 4 C and the plasma was separated and frozen for subsequent determinations of ACTH and corticosterone concentrations. ACTH was measured by radioimmunoassay (11,12). Corticosterone was measured by a competitive protein binding technique (13), using human plasma as the source of transcortin. In each experiment, samples from each group were included in each ACTH and corticosterone assay.
In some experiments up to 180 rats were stressed at each time. No experiment required more than 90 min with the animals. The time required to complete the experiment was calculated, and the experiment was begun so that the nominal hour of the experiment occurred halfway through the period. Non-stress, 0 time control rats were either killed in the middle of the working period, or half were killed at the beginning and half at the end of the period. The latter design was adopted to assure that the baseline condition did not drift during the experimental period. Significance of the difference between ACTH and corticosterone responses was determined by analysis of variance and the multiple range test of Duncan (14) or Student's unpaired t test.
Results Plasma ACTH and corticosterone responses to ip saline at 5 times of day An analysis of variance of the data shown in Table 1 indicated that both plasma ACTH and corticosterone levels were influenced by the time of day (ACTH: F = 36.32, df= 4/141, P < 0.001; corticosterone: F = 51.20, df = 4/200, P < 0.001) and by the
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 May 2016. at 04:39 For personal use only. No other uses without permission. . All rights reserved.
CIRCADIAN STRESS RESPONSES time following injection stress (ACTH: F = 8.49, df = 5/141, P < 0.001; corticosterone: F = 25.10, df = 7/200, P < 0.001). A rhythm in both ACTH and corticosterone levels was observed in the 0 time control samples. Minima coincided at 0600 h and peak levels occurred at 1800 h, as anticipated. Resting ACTH levels doubled between 0600 and 1800 h (F < 0.05) whereas there was a 15-fold increase in corticosterone levels (P < 0.01). An unexpected finding was the high resting ACTH level at 0300 h with low steroid values at this time. ACTH levels increased by 2.5 min after injection at all times of day and more or less plateaued thereafter. Surprisingly, highly significant increases in corticosterone concentrations occurred by 2.5 min at 0300 and 0600 h (Pig Bot operation
°
60 Plasma 40 corticosterone (>ig/IOOml)
20
Of Saline or 50 jig B
| 7.5 30>igB
15 Time (min)
FIG. 3. Plasma ACTH and corticosterone levels after adrenalectomy plus sc vehicle (solid circles and line); or 50 fig corticosterone immediately after adrenalectomy (open triangles and dashed line); or 30 fig corticosterone 5 min after adrenalectomy (open circles and dot-dashed line). Each point represents the mean response of 8 rats and the bars represent ± 1 SEM.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 May 2016. at 04:39 For personal use only. No other uses without permission. . All rights reserved.
CIRCADIAN STRESS RESPONSES ACTH levels at 7.5 min after vehicle injected at operation and 30 fig corticosterone injected at 5 min (P > 0.7); at this time plasma corticosterone levels were less than 30 /xg/100 ml in both groups. By 15 min there was a significant decrease in ACTH levels in the group of rats treated with 30 fig corticosterone at 5 min (P < 0.05), at a time when plasma corticosterone levels were greater than 40 fig/100 ml. In contrast, 7.5 min after 50 fig corticosterone injected at operation, plasma ACTH was significantly decreased compared to control at 7.5 min (P < 0.01) when plasma corticosterone levels were greater than 50 fig/100 ml. Fifteen minutes after operation and 50 fig corticosterone the results fell into two groups (not shown). In one group plasma corticosterone levels rose above the 7.5 min value to 61.9 ± 4.3 fig/100 ml and ACTH levels were 166 ± 26 pg/ml (n = 4). In the other group, plasma corticosterone levels fell below the 7.5 min value to 37.5 ± 0.6 fig/100 ml and ACTH levels were 275 ± 36 pg/ml. In summary, these results are entirely consistent with what is known about the fast-feedback action of corticosterone on ACTH secretion. Discussion After low intensity stimuli, the morning ACTH and corticosterone responses to stress can be greater, less than or the same as the responses to the same stimuli applied to rats in the evening. The data from individual experiments (Tables 1-3) suggested that there might be an interaction between the potency of the stimulus, the corticosterone response to stress, and the duration and the magnitude of the ACTH response. Figure 4 shows the results of the saline andhistamine experiments in intact rats, as well as the response to saline in adrenalectomized rats, plotted as increases in plasma ACTH and corticosterone concentrations above initial levels. Rows a and b represent two sets of responses to the supposedly identical stimulus of saline injection. In experiment a (data from Table 1) the 0600 h ACTH response peaked at 2.5 min and was lower
0 6 0 0 hour
145 1800 hour
J
5
10 15 0 5 10 Time after stress (minutes)
0
15
FIG. 4. ACTH responses (shaded areas) and corticosterone responses (lines) at 0600 and 1800 h to each of five stimuli, a = ip saline; b = ip saline; c = ip histamine, 125 /xg/100 g BW; d = ip histamine, 250 /Ag/100 g BW, in rats pretreated with 2 /u,g/100 g BW dexamethasone; e = ip saline in adrenalectomized rats.
thereafter in the face of a rapid (2.5 min) and large (45 /u.g/100 ml) response in plasma corticosterone. In experiment b (data from Table 2) at 0600 h when there was a slower rise in plasma corticosterone (5 min) and a smaller response (16 fig)100 ml), plasma ACTH did not peak until 7.5 min. The evening responses in experiments a and b are opposite to those observed in the morning. In experiment b there was a rapid (2.5 min) and large (38 figllOO ml) increase in plasma corticosterone, and plasma ACTH peaked at 5 min, declining thereafter; in experiment a the increase in plasma corticosterone was slow (5 min) and small (24 fig/100 ml) and plasma ACTH did not peak until 7.5 min. Similar trends can be seen in the results plotted in experiments c (125 fig histamine) and d (250 fig histamine after 2 fig dexamethasone). In the absence of an increase in plasma corticosterone in response to low
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 May 2016. at 04:39 For personal use only. No other uses without permission. . All rights reserved.
146
ENGELAND ET AL.
intensity stress, the 0600 h ACTH response is probably greater than that at 1800 h as shown in experiment e (adrenalectomized rats, saline ip). This conclusion is supported by the results of experiments on the acute ACTH response to adrenalectomy in the morning and the evening (Fig. 2). After adrenalectomy, a low-intensity stimulus (16) that removes the source of corticosterone, the ACTH response in the morning is clearly greater than that in the evening. We interpreted these results to mean that after low intensity stimuli the magnitude and pattern of the ACTH response will be determined to a large extent by the rapidity and magnitude of the corticosterone response. The test of this interpretation (Fig. 3) showed that a large and rapid corticosterone signal (supplied by 50 /xg corticosterone at operation) inhibited plasma ACTH levels 7.5 min after adrenalectomy compared to vehicle treated controls, and that a small and sluggish corticosterone signal (supplied by 30 fig corticosterone 5 min after operation) did not inhibit ACTH levels at 7.5 min, and only slightly reduced ACTH levels at 15 min compared with vehicle injected controls. We conclude from the total of these experiments utilizing lowintensity stimuli to ACTH secretion that the basic pattern of ACTH secretion is greater in the morning than in the evening, but that when the corticosterone response is rapid and large, ACTH secretion is curtailed by the rate-sensitive feedback action of corticosterone. The corticosterone response may be rapid and large or sluggish and small either in the morning or in the evening; we have not identified a variable that will predict its magnitude. An interesting consequence of these results is that if the adrenals respond to ACTH rapidly with a "healthy" increase in corticosterone, they are exposed to a relatively small amount of ACTH; if, on the other hand, there is a sluggish and small corticosterone response to the initial ACTH secretion, then the adrenals are exposed to prolonged elevations in ACTH. Thus, it seems that there is
Endo • 1977 Vol 100 • No 1
an automatic adrenal conditioning mechanism inherent in the total stress response. After the more intense, steroid-resistant stimuli of 250 fig histamine/100 g BW given to saline pre-treated rats (Table 3) or laparotomy with intestinal traction under ether anesthesia in intact or adrenalectomized rats (Fig. 1), the ACTH response observed at 0600 h was significantly greater than that at 1800 h. Plasma corticosterone responses in the evening were of similar or lesser magnitude, if they occurred, and the increase in corticosterone was faster in the morning. As with the ACTH response to acute adrenalectomy, the responses to high intensity stimuli appear to be greater in the morning than in the evening. Surprisingly, in adrenalectomized rats subjected to laparotomy with traction in the evening ACTH levels were not increased above 0 time values at either 7.5 or 15 min, although there was a clear response to this stimulus when it was applied at 0600 h (Fig. 1). The purpose of measuring the ACTH response to stress applied at 0600 h and 1800 h to adrenalectomized rats was to determine whether the circadian rhythm in resting corticosterone levels affected the morning and evening ACTH responses to stress. We were unable in these experiments to answer the question clearly. There were no significant differences in ACTH levels after either ip histamine or ip saline applied in the morning and the evening, although the incremental response to saline but not to histamine appeared to be greater in the morning (Table 2, Fig. 4e). Additionally, there was an ACTH response to laparotomy with intestinal traction in the morning, but no response was detected in the evening in rats that had been adrenalectomized for 3 days. Results of stimulating CRF secretion from the hypothalamus in vitro using serotonin suggest that in adrenalectomized rats there is significantly greater CRF secreted in the morning than in the evening in response to a standard serotonin stimulus (Mahmoud, Dallman, and Jones, Unpublished results). These results suggest that
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 May 2016. at 04:39 For personal use only. No other uses without permission. . All rights reserved.
147
ORCADIAN STRESS RESPONSES underlying changes in corticosterone levels in intact rats (low in the morning and high in the evening) are not responsible for the greater ACTH response to stress in the morning. The intention of testing the interaction between dexamethasone and histamine injection was to determine the relative sensitivity of rats to the feedback effects of dexamethasone at two times of day. The results were uninterpretable because of the large difference in maximal responses, and because of the potent rate-sensitive feedback effect observed in 1800 h, 2 /xg dexamethasone pretreatment group (Table 3, Fig. 4d). The resting levels of ACTH and corticosterone showed considerable variation across experiments. For example, in the initial saline stress experiment, the evening resting levels of corticosterone were extremely high (Table 1); this was probably a consequence of animal illness since other animals in the same animal colony were dying from respiratory ailments. Because the experiments reported were completed over a period of many months, variation in resting levels may have resulted from changes in animal room conditions (e.g., feeding times, noise level, etc.). When the results of all experiments were averaged, the resting levels of ACTH were twice as high in the evening as in the morning, whereas the corresponding corticosterone levels were 9 times as high in the evening. That is, rather than the log dose ACTH-linear corticosterone response usually seen in bioassay for ACTH, between morning and evening the relationship appears to be more nearly linear dose-log response. The disproportionately greater increase in corticosterone than in ACTH suggests that an altered adrenal sensitivity to ACTH amplifies the moderate diurnal changes in circulating levels of this hormone to yield the marked fluctuation in corticosterone characteristic of this system. In support of this interpretation is the finding
that when initial levels were made equal with 2 fig dexamethasone, the evening corticosterone response to histamine was twice that observed in the morning, although the ACTH response was greater in the morning (Table 3, Fig. 4d). More detailed evidence for a circadian variation in adrenal responsiveness to ACTH is described in another report (17). Acknowledgment We are grateful for the expert technical assistance of Mr. Brian Matsumura.
References 1. Orth, D. N., D. P. Island, and G. W. Liddle, J Clin Endocrinol Metab 27: 549, 1967. 2. Migeon, C. J., A. B. French, L. T. Samuels, and J. Z. Bowers, Am J Physiol 182: 462, 1955. 3. Guillemin, D., W. E. Dean, and R. A. Liebelt, Proc Soc Exp Biol Med 101: 394, 1959. 4. Dallman, M. F., and F. E. Yates, Ann NY Acad Sci 156: 696, 1969. 5. Sato, T., M. Sato, J. Shinsako, and M. F. Dallman, Endocrinology 97: 265, 1975. 6. Dunn, J., L. Scheving, and P. Miller, Am J Physiol 223: 402, 1972. 7. Zimmerman, E., and V. Critchlow, Proc Soc Exp Biol Med 125: 658, 1967. 8. Gibbs, F. P., Am J Physiol 219: 288, 1970. 9. Urquhart, J., In Greep, R. O., and E. B. Astwood (eds), Handbook of Physiology, section 7, vol. IV, part 2, American Physiological Society, Washington, D.C., 1974, p. 133. 10. Engeland, W. C., M. F. Dallman, J. Shinsako, C. M. Winget, and J. Vernikos-Danellis, Soc Neurosci 201: 1974 (Abstract). 11. Rees, L. H., D. M. Cook, J. W. Kendall, C. F. Allen, R. M. Kramer, J. G. Ratcliff, and R. A. Knight, Endocrinology 89: 254, 1971. 12. Dallman, M. F., D. DeManicor, and J. Shinsako, Endocrinology 95: 65, 1974. 13. Murphy, B. E. P., / Clin Endocrinol Metab 27: 973, 1967. 14. Duncan, D. B., Biometrics 11: 1, 1955. 15. Dallman, M. F., and F. E. Yates, Mem Soc Endocrinol 17: 39, 1968. 16. Dallman, M. F., M. T. Jones, J. VernikosDanellis, and W. F. Ganong, Endocrinology 91: 961, 1972. 17. Dallman, M. F., W. C. Engeland, and J. Shinsako, Prog 58th Ann Mtg Endocrine Soc, San Francisco, 1976, p 58.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 May 2016. at 04:39 For personal use only. No other uses without permission. . All rights reserved.
7.5 and 15 min after the start of surgery and after injection with either 2% EtOH-saline, or 50 fig corticosterone at operation, or with 30 fig corticosterone at 5 min. Compared with ACTH levels in rats treated with vehicle, ACTH levels were decreased 7.5 min after 50 fig corticosterone at operation (P < 0.01), but not after 30 fig corticosterone at 5 min. ACTH levels were slightly reduced 10 min after 30 fig corticosterone at 5 min compared with
evening response. However, the ACTH response
those of rats injected
to ip saline was greater in the evening in one experiment and greater in the morning in a second experiment. Plasma corticosterone responses were faster and greater in the morning in the first experiment and in the evening in the second experiment. The ACTH response to 125 fig histamine ip/100 g BW was greater in the evening and the change in corticosterone was greater in the morning. Thus, after low intensity stimuli, the ACTH responses appeared to depend on both the lag time prior to the corticosterone response, and its magnitude. To test this possibility, rats were adrenalectomized and the ACTH response was measured
(P < 0.05). These results are consistent with the interpretation that corticosterone secretion modifies stress-induced ACTH secretion via the fast-feedback effect. Comparison of the ACTH responses to acute adrenalectomy plus injection with EtOH-saline at 0600 and 1800 h demonstrated that, in the absence of a corticosterone response to the stress, the ACTH response is greater in the morning than in the evening (P < 0.01). Finally, this group of experiments suggests strongly that the responsiveness of rat adrenal glands to ACTH increases markedly between 0600 and 1800 h. (Endocrinology 100: 138, 1977)
with vehicle at operation
M
ARKED circadian fluctuations in logical range of plasma corticosterone plasma corticosteroid levels have whereas trough levels approach or define the been thoroughly documented in man, bottom of the range, and may be zero. It has monkeys and rodents (1-3). There is general been shown that elevations of corticosterone agreement that under normal conditions within the physiological range can act to corticosteroids rise to a peak at the start inhibit stress-induced ACTH secretion (4); of the daily activity cycle (1). Rats, which and corticosterone has been shown to act on are active nocturnally, exhibit peak plasma the brain to inhibit stress-induced CRF corticosterone levels at or just after dark (3). formation and secretion (5). From these obThere is a 5 to 10-fold difference between servations one might predict that the ACTH peak and trough levels of plasma corti- response to stress applied in the morning costerone in rats. Peak concentrations are when circulating corticosterone levels are approximately at the midpoint of the physio- low would be greater than the ACTH response to stress applied in the evening when plasma corticosterone levels are elevated. Received July 24, 1975. Supported by NIH grants AM 06704, GM 00927, This point has been tested repeatedly; NS 09528, 2 NASA-University Consortium agreeresponsiveness to stress has been found to ments NCAR-665-401 and NCA2-OR665-502, as well as be greater in the morning than in the eveby UCSF research funds. Mary F. Dallman is the ning (6), of equal magnitude morning and recipient of NIH Research Career Development Award evening (7), and of greater magnitude in the AM 00072. 138
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 May 2016. at 04:39 For personal use only. No other uses without permission. . All rights reserved.
ORCADIAN STRESS RESPONSES
evening than in the morning (8). Generally in these studies, changes in plasma corticosterone concentrations were measured and these changes were presumed to reflect ACTH secretion. Use of corticosterone as an index of ACTH secretion assumes that the response of the adrenal cortex to ACTH does not change between morning and evening. If changes in peripheral plasma corticosterone, rather than adrenal corticosterone content or secretion rate, are used as an endpoint, the further assumptions are made that the metabolic clearance rate and distribution volume of corticosterone are constant throughout the day. Additionally, it is difficult to compare quantitatively plasma corticosterone responses that start from different initial values. If initial levels are low, infusion of corticosterone at a given rate results in a greater elevation of plasma corticosterone than if the initial levels are high (4), probably because in the low range much of the infused steroid is bound with high affinity to circulating transcortin. Because of this, neither the incremental responses nor the maximum value attained after stress can be used directly to compare the magnitude of the stress response in the morning and evening. Although many of the same considerations apply to quantifying the stress response by measuring circulating ACTH levels morning and evening, ACTH is at least one step closer to the site of the neural initiation of stress responses, and it is likely that there are less dramatic nonlinearities in the ACTH distribution-binding-metabolism system than those in the corticosterone system (9). In this study, we have measured stress-induced changes in both circulating ACTH and corticosterone at frequent intervals after each of 3 stresses. In addition, the responses of adrenalectomized rats to the same stresses were determined. We hoped to define the role, if any, of circadian changes in circulating corticosterone levels on stressinduced ACTH secretion. Portions of the data have been reported in an abstract (10).
139
Materials and Methods Female Sprague-Dawley rats 100-130 g BW (Simonsen, Gilroy, Ca.) were received 7 to 10 days before use and were housed in animal rooms with controlled temperature (22-24 C) and light (fluorescent lighting, lights on 0600, lights off 1800 h). Two studies were performed at the NASA-Ames research facilities at Moffett Field, and 6 studies were performed in the animal facility at the University of California, San Francisco. At Ames, rats were housed in plastic pans, 6-8/cage, and were moved into the laboratory 24 h before the experiments were conducted. In San Francisco, rats were housed 2-3/cage in hanging wire mesh cages and the experiments were performed in the animal room. In all experiments, animals were weighed, numbered and apportioned to the correct cages 1-2 days before the experiment and were not subjected to further handling or unusual environmental noise thereafter. In 3 experiments, adrenalectomy was performed under ether anesthesia by the dorsal approach 3 days before the study. Rats assigned to the evening groups were adrenalectomized in the evening. Treatments Either 0.9% saline or histamine injected intraperitoneally was used as a stimulus to ACTH secretion. For the injections, rats were picked up firmly by the skin of the neck and back, and were held vertically while injected with substances through a 27 gauge Vi in needle into the midline peritoneal region about 2 cm below the xyphoid process. The dose of saline was 0.2 ml/ 100 g BW. Histamine was given in two forms. In one experiment histamine phosphate, 250 ixg free base/100 g BW, was given; in the second experiment, histamine -2HC1, 125 fig free base/ 100 g BW, was given. Dexamethasone phosphate (Decadron®, Merck, Sharp and Dohme) when used was diluted and rats were injected with 2, 5 or 10 ;ug/100 g BW 2.5 h before stress was applied. The response to laparotomy with intestinal traction under ether anesthesia was also studied. A 2 cm midline abdominal incision was made and the length of the small intestine was gently exteriorized and handled. The intestine was returned to the peritoneal cavity, the skin incision was clipped, and rats were returned to their home cages until sacrificed.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 May 2016. at 04:39 For personal use only. No other uses without permission. . All rights reserved.
Endo • 1977 Vol 100 • No 1
ENGELAND ET AL.
140 TABLE
1. Plasma ACTH and corticosterone responses to ip saline given at 5 different times of the
Time after
day
Plasma corticosterones (fig/100 ml)
Plasma ACTH (pg/ml)
(min)
0600 h
1200 h
1800 h
2400 h
0300 h
0600 h
1200 h
1800 h
2400 h
0300 h
0
63* ±7 212 ±45 193 ±58 171
95 ±20
109 ± 19
88 ± 10
110 ± 15
4.3 ± 1.0
20.5 ± 4.8
49.6 ±8.1
12.6 ± 2.8
171 ± 36
219 ± 22
212 ± 33
283 ±44
170 ±26
216 ± 27
355 ±53
263 ±45 266 ±50
60.8 ±5.4 67.2 ±8.8
± 19
83.7 ± 5.0
72.3 ± 7.8
63.0 ± 5.6
184 ± 44
162 ± 12
173 ±26 163 ± 39
176 ±39 204 ±77 187 ±34
18.8 ±5.5 53.4 ± 6.1 51.5 ±4.6 49.2 ±3.3
37.0 ±4.3
175 ± 20
23.6 ± 3.9 35.7 ±3.6 42.9 ±3.1 48.1 ±5.0 54.3 ±4.3
61.1 ±4.8 51.0 ±2.2 81.7 ± 8.4
74.7 ±6.3 48.3 ±9.3
56.2 ± 6.6 65.0 ± 4.4
19.8 ±5.7 17.5 ±4.4
36.5 ±7.9
84.8 ± 7.7 57.3 ± 10.0
37.3 ±5.0
2.5 5.0 7.5
10.0 15.0
171 ± 18
238 ± 34
179 ± 17
239
±51 357
±97
30.0 60.0
57.9 ± 6.1
24.5 ±8.4
40.8 ±9.2
49.8 ±4.0
60.0 ± 2.7
62.7 ± 6.6 62.7 ±4.5 19.2 ±4.3
Mean of 6 rats/group ± SEM.
The acute response to bilateral adrenalectomy was studied in two groups of rats at 0600 h and at 1800 h. In these experiments rats were anesthetized with sodium pentobarbital, 4.5 jiig/100 g BW (Nembutal®, Abbott), bilaterally adrenalectomized and were killed at the times indicated. In one experiment corticosterone dissolved in a small volume of EtOH and diluted with 0.9% saline to an EtOH concentration of 2% was given in two regimens. One group of animals was given 50 fig corticosterone immediately after adrenalectomy; another group was given 30 fig corticosterone SC 5 min after the start of adrenalectomy. Both groups were killed at 7.5 and 15 min and the results were compared to those from adrenalectomized rats treated at operation with the vehicle. All rats were killed by decapitation and trunk blood was collected in iced, heparinized plastic centrifuge tubes. The blood was rapidly centrifuged at 4 C and the plasma was separated and frozen for subsequent determinations of ACTH and corticosterone concentrations. ACTH was measured by radioimmunoassay (11,12). Corticosterone was measured by a competitive protein binding technique (13), using human plasma as the source of transcortin. In each experiment, samples from each group were included in each ACTH and corticosterone assay.
In some experiments up to 180 rats were stressed at each time. No experiment required more than 90 min with the animals. The time required to complete the experiment was calculated, and the experiment was begun so that the nominal hour of the experiment occurred halfway through the period. Non-stress, 0 time control rats were either killed in the middle of the working period, or half were killed at the beginning and half at the end of the period. The latter design was adopted to assure that the baseline condition did not drift during the experimental period. Significance of the difference between ACTH and corticosterone responses was determined by analysis of variance and the multiple range test of Duncan (14) or Student's unpaired t test.
Results Plasma ACTH and corticosterone responses to ip saline at 5 times of day An analysis of variance of the data shown in Table 1 indicated that both plasma ACTH and corticosterone levels were influenced by the time of day (ACTH: F = 36.32, df= 4/141, P < 0.001; corticosterone: F = 51.20, df = 4/200, P < 0.001) and by the
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 May 2016. at 04:39 For personal use only. No other uses without permission. . All rights reserved.
CIRCADIAN STRESS RESPONSES time following injection stress (ACTH: F = 8.49, df = 5/141, P < 0.001; corticosterone: F = 25.10, df = 7/200, P < 0.001). A rhythm in both ACTH and corticosterone levels was observed in the 0 time control samples. Minima coincided at 0600 h and peak levels occurred at 1800 h, as anticipated. Resting ACTH levels doubled between 0600 and 1800 h (F < 0.05) whereas there was a 15-fold increase in corticosterone levels (P < 0.01). An unexpected finding was the high resting ACTH level at 0300 h with low steroid values at this time. ACTH levels increased by 2.5 min after injection at all times of day and more or less plateaued thereafter. Surprisingly, highly significant increases in corticosterone concentrations occurred by 2.5 min at 0300 and 0600 h (Pig Bot operation
°
60 Plasma 40 corticosterone (>ig/IOOml)
20
Of Saline or 50 jig B
| 7.5 30>igB
15 Time (min)
FIG. 3. Plasma ACTH and corticosterone levels after adrenalectomy plus sc vehicle (solid circles and line); or 50 fig corticosterone immediately after adrenalectomy (open triangles and dashed line); or 30 fig corticosterone 5 min after adrenalectomy (open circles and dot-dashed line). Each point represents the mean response of 8 rats and the bars represent ± 1 SEM.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 May 2016. at 04:39 For personal use only. No other uses without permission. . All rights reserved.
CIRCADIAN STRESS RESPONSES ACTH levels at 7.5 min after vehicle injected at operation and 30 fig corticosterone injected at 5 min (P > 0.7); at this time plasma corticosterone levels were less than 30 /xg/100 ml in both groups. By 15 min there was a significant decrease in ACTH levels in the group of rats treated with 30 fig corticosterone at 5 min (P < 0.05), at a time when plasma corticosterone levels were greater than 40 fig/100 ml. In contrast, 7.5 min after 50 fig corticosterone injected at operation, plasma ACTH was significantly decreased compared to control at 7.5 min (P < 0.01) when plasma corticosterone levels were greater than 50 fig/100 ml. Fifteen minutes after operation and 50 fig corticosterone the results fell into two groups (not shown). In one group plasma corticosterone levels rose above the 7.5 min value to 61.9 ± 4.3 fig/100 ml and ACTH levels were 166 ± 26 pg/ml (n = 4). In the other group, plasma corticosterone levels fell below the 7.5 min value to 37.5 ± 0.6 fig/100 ml and ACTH levels were 275 ± 36 pg/ml. In summary, these results are entirely consistent with what is known about the fast-feedback action of corticosterone on ACTH secretion. Discussion After low intensity stimuli, the morning ACTH and corticosterone responses to stress can be greater, less than or the same as the responses to the same stimuli applied to rats in the evening. The data from individual experiments (Tables 1-3) suggested that there might be an interaction between the potency of the stimulus, the corticosterone response to stress, and the duration and the magnitude of the ACTH response. Figure 4 shows the results of the saline andhistamine experiments in intact rats, as well as the response to saline in adrenalectomized rats, plotted as increases in plasma ACTH and corticosterone concentrations above initial levels. Rows a and b represent two sets of responses to the supposedly identical stimulus of saline injection. In experiment a (data from Table 1) the 0600 h ACTH response peaked at 2.5 min and was lower
0 6 0 0 hour
145 1800 hour
J
5
10 15 0 5 10 Time after stress (minutes)
0
15
FIG. 4. ACTH responses (shaded areas) and corticosterone responses (lines) at 0600 and 1800 h to each of five stimuli, a = ip saline; b = ip saline; c = ip histamine, 125 /xg/100 g BW; d = ip histamine, 250 /Ag/100 g BW, in rats pretreated with 2 /u,g/100 g BW dexamethasone; e = ip saline in adrenalectomized rats.
thereafter in the face of a rapid (2.5 min) and large (45 /u.g/100 ml) response in plasma corticosterone. In experiment b (data from Table 2) at 0600 h when there was a slower rise in plasma corticosterone (5 min) and a smaller response (16 fig)100 ml), plasma ACTH did not peak until 7.5 min. The evening responses in experiments a and b are opposite to those observed in the morning. In experiment b there was a rapid (2.5 min) and large (38 figllOO ml) increase in plasma corticosterone, and plasma ACTH peaked at 5 min, declining thereafter; in experiment a the increase in plasma corticosterone was slow (5 min) and small (24 fig/100 ml) and plasma ACTH did not peak until 7.5 min. Similar trends can be seen in the results plotted in experiments c (125 fig histamine) and d (250 fig histamine after 2 fig dexamethasone). In the absence of an increase in plasma corticosterone in response to low
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 May 2016. at 04:39 For personal use only. No other uses without permission. . All rights reserved.
146
ENGELAND ET AL.
intensity stress, the 0600 h ACTH response is probably greater than that at 1800 h as shown in experiment e (adrenalectomized rats, saline ip). This conclusion is supported by the results of experiments on the acute ACTH response to adrenalectomy in the morning and the evening (Fig. 2). After adrenalectomy, a low-intensity stimulus (16) that removes the source of corticosterone, the ACTH response in the morning is clearly greater than that in the evening. We interpreted these results to mean that after low intensity stimuli the magnitude and pattern of the ACTH response will be determined to a large extent by the rapidity and magnitude of the corticosterone response. The test of this interpretation (Fig. 3) showed that a large and rapid corticosterone signal (supplied by 50 /xg corticosterone at operation) inhibited plasma ACTH levels 7.5 min after adrenalectomy compared to vehicle treated controls, and that a small and sluggish corticosterone signal (supplied by 30 fig corticosterone 5 min after operation) did not inhibit ACTH levels at 7.5 min, and only slightly reduced ACTH levels at 15 min compared with vehicle injected controls. We conclude from the total of these experiments utilizing lowintensity stimuli to ACTH secretion that the basic pattern of ACTH secretion is greater in the morning than in the evening, but that when the corticosterone response is rapid and large, ACTH secretion is curtailed by the rate-sensitive feedback action of corticosterone. The corticosterone response may be rapid and large or sluggish and small either in the morning or in the evening; we have not identified a variable that will predict its magnitude. An interesting consequence of these results is that if the adrenals respond to ACTH rapidly with a "healthy" increase in corticosterone, they are exposed to a relatively small amount of ACTH; if, on the other hand, there is a sluggish and small corticosterone response to the initial ACTH secretion, then the adrenals are exposed to prolonged elevations in ACTH. Thus, it seems that there is
Endo • 1977 Vol 100 • No 1
an automatic adrenal conditioning mechanism inherent in the total stress response. After the more intense, steroid-resistant stimuli of 250 fig histamine/100 g BW given to saline pre-treated rats (Table 3) or laparotomy with intestinal traction under ether anesthesia in intact or adrenalectomized rats (Fig. 1), the ACTH response observed at 0600 h was significantly greater than that at 1800 h. Plasma corticosterone responses in the evening were of similar or lesser magnitude, if they occurred, and the increase in corticosterone was faster in the morning. As with the ACTH response to acute adrenalectomy, the responses to high intensity stimuli appear to be greater in the morning than in the evening. Surprisingly, in adrenalectomized rats subjected to laparotomy with traction in the evening ACTH levels were not increased above 0 time values at either 7.5 or 15 min, although there was a clear response to this stimulus when it was applied at 0600 h (Fig. 1). The purpose of measuring the ACTH response to stress applied at 0600 h and 1800 h to adrenalectomized rats was to determine whether the circadian rhythm in resting corticosterone levels affected the morning and evening ACTH responses to stress. We were unable in these experiments to answer the question clearly. There were no significant differences in ACTH levels after either ip histamine or ip saline applied in the morning and the evening, although the incremental response to saline but not to histamine appeared to be greater in the morning (Table 2, Fig. 4e). Additionally, there was an ACTH response to laparotomy with intestinal traction in the morning, but no response was detected in the evening in rats that had been adrenalectomized for 3 days. Results of stimulating CRF secretion from the hypothalamus in vitro using serotonin suggest that in adrenalectomized rats there is significantly greater CRF secreted in the morning than in the evening in response to a standard serotonin stimulus (Mahmoud, Dallman, and Jones, Unpublished results). These results suggest that
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 May 2016. at 04:39 For personal use only. No other uses without permission. . All rights reserved.
147
ORCADIAN STRESS RESPONSES underlying changes in corticosterone levels in intact rats (low in the morning and high in the evening) are not responsible for the greater ACTH response to stress in the morning. The intention of testing the interaction between dexamethasone and histamine injection was to determine the relative sensitivity of rats to the feedback effects of dexamethasone at two times of day. The results were uninterpretable because of the large difference in maximal responses, and because of the potent rate-sensitive feedback effect observed in 1800 h, 2 /xg dexamethasone pretreatment group (Table 3, Fig. 4d). The resting levels of ACTH and corticosterone showed considerable variation across experiments. For example, in the initial saline stress experiment, the evening resting levels of corticosterone were extremely high (Table 1); this was probably a consequence of animal illness since other animals in the same animal colony were dying from respiratory ailments. Because the experiments reported were completed over a period of many months, variation in resting levels may have resulted from changes in animal room conditions (e.g., feeding times, noise level, etc.). When the results of all experiments were averaged, the resting levels of ACTH were twice as high in the evening as in the morning, whereas the corresponding corticosterone levels were 9 times as high in the evening. That is, rather than the log dose ACTH-linear corticosterone response usually seen in bioassay for ACTH, between morning and evening the relationship appears to be more nearly linear dose-log response. The disproportionately greater increase in corticosterone than in ACTH suggests that an altered adrenal sensitivity to ACTH amplifies the moderate diurnal changes in circulating levels of this hormone to yield the marked fluctuation in corticosterone characteristic of this system. In support of this interpretation is the finding
that when initial levels were made equal with 2 fig dexamethasone, the evening corticosterone response to histamine was twice that observed in the morning, although the ACTH response was greater in the morning (Table 3, Fig. 4d). More detailed evidence for a circadian variation in adrenal responsiveness to ACTH is described in another report (17). Acknowledgment We are grateful for the expert technical assistance of Mr. Brian Matsumura.
References 1. Orth, D. N., D. P. Island, and G. W. Liddle, J Clin Endocrinol Metab 27: 549, 1967. 2. Migeon, C. J., A. B. French, L. T. Samuels, and J. Z. Bowers, Am J Physiol 182: 462, 1955. 3. Guillemin, D., W. E. Dean, and R. A. Liebelt, Proc Soc Exp Biol Med 101: 394, 1959. 4. Dallman, M. F., and F. E. Yates, Ann NY Acad Sci 156: 696, 1969. 5. Sato, T., M. Sato, J. Shinsako, and M. F. Dallman, Endocrinology 97: 265, 1975. 6. Dunn, J., L. Scheving, and P. Miller, Am J Physiol 223: 402, 1972. 7. Zimmerman, E., and V. Critchlow, Proc Soc Exp Biol Med 125: 658, 1967. 8. Gibbs, F. P., Am J Physiol 219: 288, 1970. 9. Urquhart, J., In Greep, R. O., and E. B. Astwood (eds), Handbook of Physiology, section 7, vol. IV, part 2, American Physiological Society, Washington, D.C., 1974, p. 133. 10. Engeland, W. C., M. F. Dallman, J. Shinsako, C. M. Winget, and J. Vernikos-Danellis, Soc Neurosci 201: 1974 (Abstract). 11. Rees, L. H., D. M. Cook, J. W. Kendall, C. F. Allen, R. M. Kramer, J. G. Ratcliff, and R. A. Knight, Endocrinology 89: 254, 1971. 12. Dallman, M. F., D. DeManicor, and J. Shinsako, Endocrinology 95: 65, 1974. 13. Murphy, B. E. P., / Clin Endocrinol Metab 27: 973, 1967. 14. Duncan, D. B., Biometrics 11: 1, 1955. 15. Dallman, M. F., and F. E. Yates, Mem Soc Endocrinol 17: 39, 1968. 16. Dallman, M. F., M. T. Jones, J. VernikosDanellis, and W. F. Ganong, Endocrinology 91: 961, 1972. 17. Dallman, M. F., W. C. Engeland, and J. Shinsako, Prog 58th Ann Mtg Endocrine Soc, San Francisco, 1976, p 58.
The Endocrine Society. Downloaded from press.endocrine.org by [${individualUser.displayName}] on 09 May 2016. at 04:39 For personal use only. No other uses without permission. . All rights reserved.
