Evidence for irreversible, actinomycin D-sensitive, and temperature-sensitive steps following binding of cortisol to glucocorticoid receptors and preceding effects on glucose metabolism in rat thymus cells.

Abstract Following cortisol binding to the specific glucocorticoid receptors of thymus cells at 37° (a process which takes about 7 min), there is a 5- to 10-min lag period before a cortisol effect can be observed on glucose 6-phosphate accumulation after a glucose pulse. Evidence is presented that during this period the signal initiated by cortisol binding traverses an irreversible, an actinomycin D-sensitive, and a temperature-sensitive step. The irreversible step is shown by the fact that removal of cortisol from the glucocorticoid receptors (by displacement with cortexolone, a metabolically inactive glucocorticoid competitor, or by washing) well before any metabolic effect has appeared does not prevent the subsequent appearance of the cortisol effect. The actinomycin D-sensitive step is shown by the fact that, whereas addition of actinomycin D to thymus cells together with cortisol prevents the cortisol effect from developing subsequently, addition of actinomycin D 5 min after cortisol does not prevent the cortisol effect. To produce these actions, actinomycin D must be used at 100 µg per ml. This same concentration is necessary to inhibit RNA synthesis rapidly in thymus cells by about 80%. The temperature-sensitive step is shown by the fact that the duration of the lag period preceding the appearance of a cortisol effect increases markedly at temperatures below 37° to more than 120 min at 20°. The duration of the lag period at 37° can be shortened if cells are first incubated with cortisol at 20°. The temperature dependence of the irreversible step is such that it cannot be identical with, but must precede, the temperature-dependent step. The actinomycin D-sensitive step may or may not be identical with one of the other two steps.

The earliest metabolic effect of cortisol on rat thymus lymphocytes in o&o that has so far been reported is a block in the conversion of glucose to glucose 6-phosphate (1,2). At 37" the block appears 15 to 20 min after cells are exposed to hormone (2). A number of observations (3), including the demonstration that cortisol inhibits uptake of 3-O-methyl g,lucose, which is not phosphorylated by thymus cells, suggest that the hormone acts on glucose transport.
Much recent evidence (4)(5)(6)(7)(8) supports the hypothesis (cf. References 2 and 9) that the later inhibitory effects that cortisol exerts on macromolecular metabolism in lymphoid cells, and perhaps the ultimate lymphocytolytic effects of cortisol, are direct consequences of this early inhibition of glucose metabolism.
Preceding the effect on glucose metabolism, the only event that has been clearly shown to take place is the association of cortisol with specific glucocorticoid receptors (1,10) to form hormonereceptor complexes that at 37" and 21' are found largely in nuclei and at 3" in the cytoplasm (ll).' At 37" this association is complete by about 7 min. Between the primary hormonereceptor interaction and the subsequent effect on glucose metabolism, there is thus a latent period of at least 5 mm during which the actions of the hormone are unaccounted for. The purpose of the present work is to show that during this latent period the stimulus initiated by the hormone when it interacts with the receptors traverses two or possibly three distinct steps before it influences glucose metabolism. These steps are: (a) an irreversible step; (b) an actinomycin D-sensitive step; and (c) a temperature-sensitive step.
Steps a and b may be identical, and may be the first step that follows formation of hormone-receptor complexes.
The molecular nature of these steps remains unknown, but the sensitivity to actinomycin D suggests that Step b may involve RNA synthesis. Some of these results have been reported briefly elsewhere ( Table  I, cell concentrations, V (milliliters of packed cells per ml of suspension), measured by microhematocrit (2), were between 0.12 and 0.30.
For measurement of glucose g-phosphate, incubations were stopped by adding 1 ml of 12% cold perchloric acid per ml of incubation mixture. The perchloric acid suspensions were treated as described previously (2). Values of glucose 6-phosphate are referred to l-ml packed cell volume (2).
Steroid concentrations (in water or buffer) were determined spectrophotometrically (12). Except when noted, cortisol was present from the start of the incubation at a concentration in the cell suspension of 10m6 M. Actinomycin D was added in aqueous solution (2.5 mg per ml) to give a concentration in the cell suspension of 1, 10, or 100 pg per ml.
Steroids were obtained from Calbiochem, and enzymes and cofactors from Boehringer. tions and analyze for glucose g-phosphate.
Previous studies (2) have shown that under these conditions a glucose pulse at 37" raises glucose 6-phosphate to levels that remain fairly constant between 5 and 40 min.
The early time course of such pulses, at 37" and 20", is shown in Fig. 1. At 37" about 5 min is required to reach a constant glucose g-phosphate level. It should be noted that the cortisol effect is evident from the 1st min.
At 20", by contrast, the level of glucose 6-phosphate continues to rise for at least 30 min, and is distinctly higher than at 37". In other experiments we have found that higher concentrations of glucose do not lead to higher levels of glucose 6-phosphate and do not alter the cortisol effect. The magnitude of the cortisol effect given in subsequent figures, referred to as "per cent inhibition," represents the difference between glucose 6-phosphate levels of control and cortisol-treated cells 5 or 6 min after addition of glucose, expressed as per cent of the control level.
Control and cortisol-treated samples are paired (12), per cent inhibition is calculated for each pair, and means and standard errors are calculated from replicate pairs.

AND DISCUSSION
Temperature-sensitive Xtep-The cytolytic effects that cortisol produces after several hours of incubation with thymus cells at 37" (13) do not take place at 22". This observation, and our own finding of a temperature-induced artifact in glucose uptake (2), has led us to study the time and temperature dependence of the early cortisol effect on glucose metabolism. Fig. 2  were first incubated with or without cortisol (10-e M), and without substrate, at 20" for the periods indicated. At zero time the suspensions were warmed to 37" and the incubations continued.
At various times after warming to 37" the presence of a cortisol effect was tested for by adding glucose and measuring glucose 6-phosphate 5 min later (see "Methods" for details). The data for lo-and 120-min preliminary incubation with cortisol are from a single experiment in which all cells were first incubated at 20" for 120 min, but cortisol was added to the IO-min flasks 10 min before warming to 37". For the other data cortisol was present for the full preliminary incubation period.
Values represent means f 1 S.E. for three pairs of flasks. several different time periods and at various temperatures. The solid symbols on the solid curve show that the magnitude of the cortisol effect after a 25-min incubation at various temperatures diminishes sharply as the temperature is reduced below 37", disappearing altogether at 25". As can be seen from the upper curve of Fig. 1, this diminution as the temperature is lowered is not, due to a lack of accumulation of glucose g-phosphate at low temperatures; nor can the diminution be accounted for by slow binding of cortisol to the specific glucocorticoid receptors of thymus cells, since the rate of specific binding is almost the same at 22' as at 37" (lo), and at both of these temperatures most of the specific binding is accounted for by a nuclear cortisol-receptor complex (11) .1 Finally, the diminution is not due to the inability of cells at low temperatures to exhibit a cortisol effect (owing to a change in rate-limiting step, for example) since after a 210-min incubation a distinct cortisol effect appears even at 20" (Fig. 2) and since (as can be seen from the 60-min incubation, broken curve in Fig. 2) at all temperatures the magnitude of the effect increases with time. Fig. 3 illustrates the progress of the action of cortisol at 20". All glucose pulses in these experiments were at 37". The data show that the time that it takes for a co&sol effect to appear at 37" is progressively shortened by progressively longer periods of preliminary incubation with cortisol at 20". The time course at 37" after lo-min preliminary incubation at 20" (top rcnu) is very A, all of the incubations were at 37" and lasted 31 min. CortexoIone (lop5 M), when used, was added 5 min after the incubations started.
Glucose was added to all flasks 20 min after the cortexolone addition and 6 min later the incubations were stopped for glucose 6-phosphate determination.
Twenty minutes after cortexolone addition all flasks were warmed to 37"; 20 min later, glucose was added, and 6 min later the incubat,ions were stopped for glucose g-phosphate determination. See "Methods" for details. Values represent means f 1 S.E. for three pairs of flasks. similar to the time course determined previously (2) without any prior incubation; in both instances the earliest significant cortisol effect appears with a glucose pulse at 15 min.
Our interpretation of these results is that on the path between the initiating events of cortisol action-binding to glucocorticoid receptors-and the effect on glucose metabolism that appears 5 to 10 min after binding is complete there is a temperaturesensitive step. At, temperatures below 37", this step (or steps) can still proceed, but at a slower rate. Judging from the results in Fig. 3, at, 20" the rate is roughly 10 times smaller than at 37".

Irreversible
Step-It is well known that in viva some of the effects of an injected dose of cortisol may not appear until well after the hormone has practically disappeared from the blood. In vitro, with incubated thymus cells, the continued presence of hormone does not appear to be necessary either, since the slow effects of cortisol on cytolysis (13) and on incorporation of uridine into RNA (8) are not removed by washing cells free of cortisol after l-or Zhour incubation.
In preliminary experiments in which cortisol added initially at 10m6 M was removed by repeated washing (reducing the cortisol concentration to less than 5 x 10mg M) we found that even a few minutes exposure to cortisol at either 37" or 20" was s&icient to produce a subsequent, cortisol effect on a glucose pulse. Because of the unwieldiness of the washing procedure, for more precise studies we made use of cortexolone,2 a steroid that has been shown to compete with cortisol for binding to the specific glucocorticoid receptors of thymus cells, and thereby to reduce or abolish the effects of cortisol on glucose uptake (10). CORTISOL ADDED (Omin) 4 I In Fig. 4A the Jirst three bars (0 min) illustrate, respectively, the inhibitory effect of cortisol at 1W6 M, of cortexolone at 10-b M, and of both these steroids when added simultaneously. These data, in which inhibition is determined with respect to glucose 6-phosphate accumulation following a 6-min glucose pulse, agree with our earlier results (10) showing that cortexolone drastically reduces the effect of cortisol on inhibition of glucose uptake. The magnitude of the reduction is about what would be expected from the dose-response curve for cortisol (12) and the relative binding affinities of cortisol and cortexolone for glucocorticoid receptors (10). 0 5 The next pair of bars in Fig. 4A (5 min) show that if instead of being added together with cortisol cortexolone is added 5 min later, it hardly reduces the effect of cortisol at all. By 5 min the binding of cortisol to glucocorticoid receptors is approaching completion (10). Cortexolone displaces cortisol from the glucocorticoid receptors within approximately another 5 min,3 well before a cortisol effect on glucose metabolism can be detected (2). The subsequent appearance of a cortisol effect, which we know from other measurements persists for at least 40 min, therefore shows that specific cortisol binding is rapidly followed, or is accompanied, by an irreversible step--i.e. a step that results in a change that is not reversed by removing cortisol. when used, was either present from the start of tie incudation (dj or added after 5 min (5).

TIME OF ACTINOMYCIN-D ADDITION
In both cases, glucose was added 20 min after actinomycin D and the incubation was stopped 6 min later for glucose 6-phosphate determination.
The groups without actinomycin D were incubated along with each actinomycin D group.
All incubations were at 37". Values are the mean & 1 S.E. for three or four flasks. Fig. 4B shows that this irreversible step probably proceeds more slowly at 20". The experimental design here is similar to that in Fig. 4A, except that the initial period of exposure to cortisol alone was at 20"; after addition of cortexolone the incubations were continued at 20" for 20 min, sufficient time for cortexolone to displace cortisol almost completely. Then, in order to allow the temperature-sensitive step to proceed, all flasks were rewarmed to 37" for 20 min, and finally the presence of a cortisol effect, was tested for with a glucose pulse. Since the only difference between the various experiments in which cortisol and cortexolone were both added was the amount of time that the cells were exposed at 20" to cortisol alone, it may reasonably be concluded that the increase in the cortisol effect with time of exposure to cortisol represents the rate at which the irreversible step proceeds at 20". Part of the decrease in rate at 20" (as compared to 37") may be ascribed to a slight decrease in rate of formation of the nuclear cortisol-receptor complex (10, II),' so that if anything the rate of the irreversible step is greater at 20" than would appear from Fig. 4B. on a glucose pulse 20 min later; if it is added 5 min after cortisol (5) a normal cortisol effect develops.
These experiments have been repeated many times, with completely consistent results. In other experiments we have found that the same dose of actinomycin D has no effect on the rate or magnitude of specific binding of cortisol to whole thymus cells or to the nuclear receptors, so we conclude from these results that specific cortisol binding is rapidly followed by an actinomycin D-sensitive step. Values in Fig. 5 are in absolute amounts of glucose B-phosphate rather than per cent inhibition, to show that actinomycin D by itself increases glucose 6-phosphate levels.
From other evidence, it is likely that this action of actinomycin D is due to a decreased rate of utilization of glucose 6-phosphate. It can be seen to be separate from the action that leads to elimination of the cortisol effect,, because when actinomycin D is added 5 min after cortisol a conspicuous cortisol effect develops despite the elevation of glucose 6-phosphate levels.
The main conclusion from these experiments is that, although the rate of the irreversible step at 20" is slower than at 37", it is not nearly slow enough to account for the temperature-sensitive step discussed above (at 20" the cortisol effect takes more than 120 min to appear).
Consequently, the irreversible and the temperature-sensitive steps must be separate steps. Furthermore, since the irreversible step can be initiated rapidly by adding cortisol at 20", it is clear that the irreversible step must precede the temperature-sensitive step. Actinornycin D-sensitive Step-The data in Fig. 5, obtained with incubations at 37", show that if actinomycin D is added to cells at the same time as cortisol (0) no cortisol effect is observed 2 The trivial name used is: cortexolone, 17,21 -dihydroxypregnlene-3,20-dione.
Using an approach similar to that which allows us to distinguish between the temperature-sensitive and the irreversible step, we have attempted, unsuccessfully, to see whether the actinomycin D-sensitive step can be distinguished from the irreversible or the temperature-sensitive step. A series of experiments in which cortisol and actinomycin D were added at various times relative to one another at various temperatures has shown that actinomycin D at 20" fails to block the cortisol effect,. These results can be explained by supposing that actinomycin D has no metabolic effect on thymus cells at 20", an explanation which is supported by separate experiments in which we have found that: (a) in incubations at 20" actinomycin D has no effect on RNA synthesis at least during the first 30 min, and (b) after preliminary incubation with actinomycin for 10 min at, 37" the antibiotic subsequently inhibits the incorporation of uridine into RNA at 20".  (5,6,8), as well as on amino acid incorporation into protein -- (3,8,14). With thymus cells it is clear that these effects have are ineffective. There is therefore a distinct possibility that little to do with the primary actions of the hormones, but rather we may be dealing with nonspecific effects of actinomycin D, are probably secondary to preceding effects on glucose metaborather than solely with an effect on RNA synthesis. However,lism (46,15).
as shown by the data in Table I (data not presented) we have found that this inhibition increases it occurs at about the time of the appearance of the cortisol to 90 to 95% by 10 min, whereas with 10 pg of the antibiotic the effect on glucose metabolism6). It is in any case not unreasoninhibition is only 40 to 50% by 15 min. Moreover, as also able to suppose that these steps involve macromolecular metabshown in Table I Issue of February 10, 1971 K. M. Mosher,D. A. Young,and A. Munck 659 many complex events take place in the 5 to 10 min between the initial actions of the hormone and the ultimate effect on glucose transport.