Effects of Thyroid Hormone Deficiency on Cyclic Adenosine 3’:5’-Monophosphate and Control of Lipolysis in Fat Cells*

SUMMARY Isolated fat cells from hypothyroid rats, in contrast with those from normal animals, lack ability to give a lipolytic response to epinephrine or glucagon. However, activation of triglyceride lipolysis was induced with (N6,02’-dibutyryl cyclic adenosine 3’ :5’-monophosphate), or a combination of epinephrine and phosphodiesterase inhibitors. Adenylyl cyclase was activated by catecholamines or glucagon in membrane ghosts prepared from fat cells obtained from hypothyroid rats. The characteristics of the enzyme, and its activation by hormones and fluoride were similar to those observed in fat cell membrane ghosts from normal animals. Thus, adenylyl cyclase appears to be functional in the hypothyroid state, although lipolysis is blocked. Whole fat cells the epinephriue of cyclic


Isolated
fat cells from hypothyroid rats, in contrast with those from normal animals, lack ability to give a lipolytic response to epinephrine or glucagon. However, activation of triglyceride lipolysis was induced with (N6,02'-dibutyryl cyclic adenosine 3' :5'-monophosphate), or a combination of epinephrine and phosphodiesterase inhibitors. Adenylyl cyclase was activated by catecholamines or glucagon in membrane ghosts prepared from fat cells obtained from hypothyroid rats. The characteristics of the enzyme, and its activation by hormones and fluoride were similar to those observed in fat cell membrane ghosts from normal animals.
Thus, adenylyl cyclase appears to be functional in the hypothyroid state, although lipolysis is blocked. Whole fat cells were incubated in the presence and absence of epinephriue or theophylline, and accumulation of cyclic AMP over 10 min was measured.
Basal, unstimulated levels were found to be similar in cells from normal and hypothyroid animals.
In the presence of theophylline alone, similar levels of cyclic AMP were again observed.
An increase in cyclic AMP accumulation in response to epinephrine stimulation, however, was not obtained in cells from hypothyroid rats except in the presence of the phosphodiesterase inhibitor. Cyclic nucleotide phosphodiesterase activities were fractionated by discontinuous sucrose gradient centrifugation. The soluble, low a%lnity form of phosphodiesterase activity did not difIer in the normal and hypothyroid states. The particulate, high affinity forms of cyclic AMP phosphodiesterase activity were elevated in fat cells from hypothyroid rats. Fat cells were obtained from three male rats, average weight 120 g, which had been fed the goitrogenic diet for 21 days. Glucagon or Bt&MP (dcAMP) was added at the concentrations indicated.
The values for glycerol product,ion are the means of either two or three incubations.
Vertical lines indicate range.
of lipolysis by BtzcAMP serves as a control and also indicates, in agreement with others (4), that hormone-sensitive lipase is present in the hypothyroid state and can be converted to the active form by an agent which bypasses the adenylyl cyclase step. Further evidence to support this is seen in the observation (Fig.  4)

Adenylyl
Cyclase Activity in Fat Cell Outer Membrane Ghosts Obtained from Normal and Hypothyroid Rats-It has been reported that adipose tissue homogenates obtained from hypothyroid rats show diminished adenylyl cyclase activity in response to catecholamines, and the conclusion was drawn that this is due to a primary effect of thyroid hormones on protein synthesis (5). It thus became important to determine the adenylyl cyclase activity, if any, which might be present in our lipolytically insensitive fat cells.
The additional question of whether or not there had been a loss in the ability of epinephrine or glucagon to effectively bind to receptors in the membrane and stimulate enzymatic production of cyclic AMP was also considered. The results shown in Fig. 6 clearly demonstrate that the metabolic block in the hypothyroid state cannot be due to the absence of adenylyl cyclase as has been presumed by these earlier studies.
The basal specific activity of the enzyme from hypothyroid rats is reduced when compared with normal animals in agreement with Krishna et al. (5), but the apparent affinity and maximal stimulatory effect of isoproterenol is similar in both cases (Fig. 6)  Values are expressed as picomoles of cyclic AMP per 106 cells Z!Z SD.
The numbers in parentheses indicate the number of incubations included for each point, using aliquots from the same pool of cells for each group of rats.
Incubations were carried out for 10 min in the same manner as described for the other fat cell ex- investigated. Discontinuous gradient fractionation studies indicate cyclic GMP phosphodiesterase activity is predominantly soluble as has been shown for other tissues (17,18). Both soluble and particulate cyclic AMP hydrolyzing activities were detected by using a 0.5 PM substrate concentration. Cyclic AMP phosphodiesterase (measured at 200 PM substrate concentration) appears in the soluble fraction (19). Previous studies indicate the fat cell soluble phosphodiesterase activity to be predominantly high K, activity and the particulate to be the low K, enzyme form displaying negative cooperativity (20).
The particulate forms of low K, activity were elevated in fat cells from hypothyroid rats when calculated either on the basis of the number of cells applied to the gradients (Fig. 7), or specific activity of each particulate fraction.
The gradient patterns shown represent experiments with four separate groups of propylthiouracil-treated rats.
The specific act.ivity of soluble, high K, cyclic AMP phosphodiesterase was unaffected, and that of the soluble cyclic GMP phosphodiesterase activity was only slightly elevated in the hypothyroid animals.
The protein distribution patterns and total enzyme activity recoveries were similar in all cases. Additional experiments will be required to characterize more fully the altered pattern of phosphodiesterase activities associated with hypothyroid status.

DISCUSSION
A physiologically significant, specific biochemical process in which thyroid hormones directly participate has been difficult to establish.
It is now apparent that many of the responses to T3 or T4 represent secondary metabolic adjustments, and are only remotely related to the immediate molecular events occurring when these hormones interact with cellular components.
Considering the wide diversity of physiological actions attributed to these hormones (21), it is probable that these responses may share common interdependent relationships rather than the hormone having different, multiple direct effects. Many of these effects appear to be associated with known cyclic AMP-dependent processes which may partially explain the frequently noted but still poorly defined relationship between thyroid hormones and catecholamines (22). Impairment of the lipolytic response to epinephrine occurs after surgical or pharmacological thyroidectomy, and conversely, administration of thyroid hormones causes an enhanced sensitivity to epinephrine (l-5, [23][24][25][26]. Rapid effects of thyroid hormones on cardiovascular function are also well known and are, in some respects, similar to those produced by the catecholamines (21,22). Stimulation of cardiac adenylyl cyclase by T3 and Tq in vitro has been observed (27), although whether this can account for the effects of these hormones on cardiac function seems problematic at present (6,(27)(28)(29)(30). Other presumed direct effects of T3 or T1 administration on cyclic AMP levels have been described (31). In general agreement with the concept of an involvement of cyclic AMP, glycogen depletion was observed by Tata et al. (31) to be one of the most immediate consequences of thyroid hormone administration.
The primary mode of action of thyroid hormones has been thought to be upon protein synthesis (32)(33)(34)(35)(36), and Krishna et al. (5) suggested that the effect of these hormones on adipose tissue might be mediated through synthesis of adenylyl cyclase.
However, such an effect could not be seen in other tissues (29,37,38), and Caldwell and Fain (25)  deficiency of the enzyme, adenylyl cyclase. It is apparent that events associated with the binding of epinephrine and glucagon and importantly, the coupling of receptor to adenylyl cyclase in the membrane preparation are unaffected by thyroid status. Not only is lipolytic activity in response to hormones blocked in the hypothyroid state, but cyclic AMP also does not accumulate.
While it is conceivable that these two events are totally independent of one another, the experimental evidence suggests that this is improbable.
Blockade can be overcome by Bt2cAMP and by a combination of epinephrine and phosphodiesterase inhibitors (Fig. 3) leading also to an increase in endogenous cyclic AMP levels (Table II).
ity for hydrolysis of the cyclic AMP formed in response to epinephrine or other hormonal stimulation. Inhibition of phosphodiesterase in vitro by addition of thyroid hormones has been reported (40)(41)(42)(43)(44), but these experiments have been largely discounted because of the excessively high levels of hormone required, and the lack of structural specificity.
Close coupling of adenylyl cyclase activity with lipase activity has been proposed by Manganiello et al. (39). Linkage of these two enzyme activities would then depend on the rate of turnover of lipolytically significant cyclic AMP. If cyclic AMP metabolism is related to the loss of hormonal response in the hypothyroid state, then the modulating effect of thyroid hormones could involve an alteration in the turnover rate of total or compartmentalized cyclic AMP.
The results of our studies suggest that activation of membranebound cyclic AMP phosphodiesterase held in close proximity to an adenylyl cyclase-lipase complex could be responsible for the insensitivity to epinephrine and glucagon, and may represent the site of the lesion in the thyroid-deficient state. Such localized changes in specific phosphodiesterases would not be readily detected in homogenates or in the supernatant fraction after centrifugation, and this may partially account for the inability of previous workers (5,25,44) to find a correlation between thyroid status and phosphodiesterase activity in adipose tissue.
Defective adenylyl cyclase activity is not likely to provide the explanation for the total lack of hormone-induced lipolysis, since we have shown that adenylyl cyclase is equally responsive in membrane ghost preparations from both normal and hypothyroid animals.
Furthermore, cyclic AMP levels in both types of intact cells can be elevated in response to hormonal stimulation if phosphodiesterase inhibitors are present. These observations lead directly to the possibility that an important factor differentiating the hypothyroid from the normal state may be an altered capac-Since normal morphogenetic development may be controlled by intracellular changes in cyclic AMP levels through altered membrane-bound phosphodiesterase activities (45), it is tempting to speculate that some of the effects of thyroid hormones on growth and development (21) might be explained in this manner. Regulation of specific phosphodiesterase activities may thus represent one critical aspect of a more general mechanism for the control of cellular metabolism by growth and developmental hormones.