Xanthines for asthma - present status

Purinergic signaling in the airways

Evidence for a significant role and impact of purinergic signaling in normal and diseased airways is now beyond dispute. The present review intends to provide the current state of knowledge of the involvement of purinergic pathways in the upper and lower airways and lungs, thereby differentiating the involvement of different tissues, such as the epithelial lining, immune cells, airway smooth muscle, vasculature, peripheral and central innervation, and neuroendocrine system. In addition to the vast number of well illustrated functions for purinergic signaling in the healthy respiratory tract, increasing data pointing to enhanced levels of ATP and/or adenosine in airway secretions of patients with airway damage and respiratory diseases corroborates the emerging view that purines act as clinically important mediators resulting in either proinflammatory or protective responses. Purinergic signaling has been implicated in lung injury and in the pathogenesis of a wide range of respiratory disorders and diseases, including asthma, chronic obstructive pulmonary disease, inflammation, cystic fibrosis, lung cancer, and pulmonary hypertension. These ostensibly enigmatic actions are based on widely different mechanisms, which are influenced by the cellular microenvironment, but especially the subtypes of purine receptors involved and the activity of distinct members of the ectonucleotidase family, the latter being potential protein targets for therapeutic implementation.

Methylxanthines in surgery: a bright future?

Methylxanthines have been used in clinical practice for over 100 years, and although understanding of their mechanisms of action is growing their effects are not fully understood. Nevertheless the knowledge to date has brought about a general upsurge of interest in methylxanthines and the development of novel derivatives. Methylxanthines are poised to escape the confines of their traditional role as these agents are applied in novel ways to surgical illnesses such as septic shock, the adult respiratory distress syndrome, cancer cachexia and functional neutrophil disorders. Methylxanthines, alone or in combination with other compounds, may well become part of the surgeon's future stock-in-trade.

Structure-pharmacokinetic relationships among the N1,N3-alkylxanthines in rats

The pharmacokinetics of four N3-alkylxanthine and four N1-methyl-N3-alkylxanthine derivatives has been investigated in rats after intravenous administration of the individual alkylxanthines. The concentration of N1,N3-alkylxanthine in plasma and urine was determined by HPLC. A one-compartment model adequately described the plasma concentration time data. The steady-state volume of distribution (Vss) was calculated using model-independent methods. The relation between Vss and unbound drug fraction in plasma (fu) was significantly correlated (Vss = 0.844fu + 0.119; r = 0.999, P less than 0.01), indicating that the differences in fu among these xanthine derivatives is mainly responsible for differences in Vss. The decrease in Vss and increase in plasma protein binding with lipophilicity reflected a relatively constant tissue affinity. The total body clearance increased with lipophilicity with the exception of the first three lower congeners which were almost completely excreted unchanged in urine, mainly via active tubular secretion. Renal elimination was markedly reduced by the presence of a methyl group at the N1-position. Renal clearance decreased with increasing lipophilicity, due to increased tubular reabsorption whereas non-renal (hepatic) clearance increased with increasing lipophilicity.

The actions of some beta-receptor agonists and xanthines on isolated muscle strips from the human oesophago-gastric junction

Isolated preparations from the circular muscle layer of the human oesophago-gastric junction were mounted in organ baths and isometric tension recorded. During an equilibration period, active resting tension developed suggesting that the preparations were representing the lower oesophageal sphincter. Active tension was abolished by exposing the preparations to Ca(++)-free medium. The two xanthines theophylline and enprofylline almost equipotently relaxed the preparations in a concentration-dependent manner (10(-7)-10(-3) M). Within therapeutic concentrations, theophylline inhibited active resting tension by 30-60%, while enprofylline lowered tension by less than 20%. Inhibitory actions of adenosine were demonstrated, and this suggests that adenosine antagonism is not the mechanism of action for xanthines in the oesophagus. Non-selective beta-receptor stimulation with isoprenaline inhibited active tension by 70% (10(-7) M), while beta 2-receptor stimulation with terbutaline inhibited tension by 47% (10(-5) M). Dobutamine, believed to preferentially stimulate beta 1-receptors, inhibited active tension in a concentration-dependent manner (10(-7)-10(-4) M). Metoprolol (10(-6) M), a selective beta 1-receptor antagonist, shifted the concentration-response curve for isoprenaline to the right, but left the maximal response unchanged. It is concluded that xanthines and beta-receptor agonists have inhibitory actions on circular muscle from the human oesophagogastric junction. The experimental data suggest the presence of beta 1- as well as beta 2-receptors, both mediating inhibition of active resting tension.

Activities of caffeine, theophylline, and enprofylline analogs as tracheal relaxants

A variety of xanthines cause tracheal relaxation, an activity predictive of antiasthmatic potential. Structural analogs of caffeine, theophylline, and enprofylline were examined for their abilities to relax carbamylcholine-stimulated guinea pig trachea in vitro. All caffeine analogs tested were more potent than caffeine (EC50 = 551 +/- 81 microM) except the 8-p-sulfophenyl analog. 1,3,7-Tripropylxanthine and 1,3,7-tripropargylxanthine were among the more potent analogs with EC50 values of 12 +/- 1.3 and 65 +/- 11 microM respectively. Increasing the polarity at the 1- or 3-position by substituting a propargyl group for an n-propyl group decreased relaxant activity, an effect not observed at the 7-position. The 8-cyclohexyl-, 8-cyclopentyl- and 8-phenyl-derivatives of caffeine were relatively potent (EC 50 = approximately 75 microM). The theophylline analog 1,3-di-n-propylxanthine was approximately two times more active than theophylline (EC50 = 162 +/- 17 microM). 3-Isobutyl-1-methylxanthine (EC50 = 7.1 +/- 1.8 microM) and 1-isoamyl-3-isobutylxanthine (EC50 = 37 +/- 4.2 microM) were among the most potent tracheal relaxants. The 8-substituted theophylline analogs were weak or inactive relaxants except for 8-cyclopentyl- and 8-cyclohexyltheophylline, which were more potent or as potent as theophylline. In contrast to enprofylline (EC50 = 56 +/- 9 microM), 8-substituted enprofylline analogs were weak or inactive as relaxants with the exception of the 8-cyclohexyl analog. The potency of xanthines as tracheal relaxants was unrelated to potency as adenosine receptor antagonists and may reflect activity as phosphodiesterase inhibitors.

Effects of anti-asthma drugs on PAF-induced death in mice

A number of anti-asthma drugs was assayed for the ability to protect mice from platelet-activating factor (PAF)-induced death, which has been suggested to be dependent on the bronchoconstrictive features of this autacoid. Salbutamol, dexamethasone, theophylline, ketotifen and zindotrine, administered parenterally, produced a dose-dependent protection, while forskolin, enprofylline, disodium cromoglycate, nedocromil, azelastine and antagonists of acetylcholine, histamine and serotonin were devoid of protective effects. Theophylline, contrary to salbutamol, lost its protective effects in adrenalectomized mice, suggesting that these effects are dependent on the release of adrenomedullary cathecolamines. Salbutamol, theophylline and dexamethasone, at doses capable of inhibiting PAF-induced death, did not affect PAF-induced haemoconcentration in mice, thus excluding widespread changes in vascular permeability as a major cause of death. These results are generally in agreement with the hypothesis that airway obstruction is an important determinant of PAF-induced death in mice.

Enprofylline pharmacokinetics in children with asthma

The pharmacokinetics of intravenous enprofylline has been studied in 8 children with asthma. The mean plasma half-life of enprofylline (1.0 h) was considerably shorter than that previously reported in adults. The half-life determined from log urine excretion rate data was identical to the plasma half-life, so urine excretion could be used as a non-invasive method to study the elimination rate. As in adults, urinary recovery of unchanged drug averaged 89%, and the volume of distribution, Vz, averaged 0.58 l/kg. Clearance was higher in children than in adults when calculated per kg body weight, but not when calculated per m2 body surface area. The dosage of enprofylline in children would be more accurate if calculated in proportion to surface area rather than to body weight. Data agree with published information on creatinine clearance, which, adjusted for body surface area, stays constant from the age of 3 years until early adult life.

Additive bronchodilator effects of terbutaline and enprofylline in asthma

Enprofylline is a novel xanthine derivative with negligible adenosine antagonizing ability. It is eliminated almost exclusively by renal clearance with a half-life of about 2 h. Three i.v. infusions of enprofylline (1 mg/kg body weight over 10 min) were given at hourly intervals to 16 patients with stable, reversible airway obstruction. The patients were pretreated at random with i.v. terbutaline (4 micrograms/kg body weight) or placebo according to a double blind cross-over design. Lung function and drug concentrations in plasma were followed. Enprofylline produced significant and concentration-dependent bronchodilation between plasma levels of 1.24 and 3.22 mg/l. The improvement in ventilatory function was significantly enhanced by terbutaline pretreatment. At the highest plasma levels of enprofylline nausea and headache were found as subjective side effects. The results suggest that enprofylline and terbutaline might best be used in a low dose combination in the treatment of bronchial asthma.

Microcalorimetric studies on the effect of adenosine receptor agonists and xanthine derivatives on overall metabolism in human platelets

The direct overall metabolic effects of drugs acting on adenosine receptors in human platelets were evaluated using a sensitive microcalorimetric method. Adenosine induced a concentration-dependent increase in the heat production rate at concentrations above 100 microM. The adenosine uptake inhibitor, dipyridamole (3 microM) did not modify the effect of adenosine. Two putative adenosine receptor agonists were tested: NECA (5-N-ethyl carboxamide adenosine) and PIA (L-N6-phenylisopropyl-adenosine). NECA induced, at significantly lower concentrations than adenosine, an enhanced heat production rate. Concentrations above 1 mM had no effect. PIA, on the other hand, invariably induced a reduction in the heat production rate already at a concentration of 100 microM. The two xanthine derivatives enprofylline (25 microM) and theophylline (100 microM) were tested at concentrations found during antiasthmatic therapy. Neither had any thermogenic effect by themselves nor showed any significant modification of the heat production rate induced by adenosine (300 microM). These results indicate that adenosine and NECA increase human platelet metabolism, whereas PIA has an opposite effect. The proposed adenosine receptor antagonists enprofylline and theophylline were without effects. This microcalorimetric study gives new insights into the complex nature of adenosine mechanisms in a human test system and indicates that the thermogenic effect of adenosine is unrelated to adenosine receptors.

Effects of enprofylline and theophylline may show the role of adenosine

It is well established that at low and clinically relevant concentrations theophylline (and caffeine) exerts antagonism at cell surface receptor sites for adenosine. However, it is not known which actions of theophylline are due to adenosine antagonism, because theophylline apparently activates other cellular mechanisms at the same low concentrations. Investigations into the actions of xanthines and their structure activity relationships have identified xanthine compounds like enprofylline (3-propylxanthine) that only has some actions in common with theophylline and that has a negligible ability to antagonize adenosine. Enprofylline is a more potent smooth muscle relaxant and antiasthmatic drug than theophylline but does not produce, e.g., theophylline-like diuretic effects, CNS-stimulant behavioural effects (restlessness - seizures), gastric secretory effects and release of free fatty acids. It is proposed that pharmacodynamic dissimilarities between enprofylline and theophylline may indicate physiological roles of adenosine.

The xanthine derivative 1-(5'-oxohexyl)-3-methyl-7-propyl xanthine (HWA 285) enhances the actions of adenosine

The effect of two closely related xanthine derivatives, pentoxifylline and HWA 285, on cyclic AMP accumulation in rat hippocampal slices and on adenosine uptake in erythrocytes was examined. Pentoxifylline was a weak competitive antagonist of adenosine effects on cyclic AMP accumulation. HWA 285, by contrast, had a small stimulatory effect per se and also potentiated the effect of adenosine (10-30 microM). Neither pentoxifylline nor HWA 285 significantly affected the cyclic AMP accumulation induced by the stable adenosine analogue NECA or by alpha- or beta-adrenoceptor activation. HWA 285 was a much more potent inhibitor of adenosine uptake into human erythrocytes than pentoxifylline and other examined xanthines including thiocaffeine, 8-p-sulphophenyltheophylline, theophylline, caffeine and enprofylline. It is suggested that HWA 285 may potentiate, rather than antagonize, the effects of endogenous as well as exogenous adenosine, partly as a consequence of adenosine uptake inhibition.

Effects of enprofylline on A1 and A2 adenosine receptors

The effects of enprofylline were studied on A1 adenosine receptors of rat fat cells and on A2 adenosine receptors of human platelets and of guinea-pig lung. Enprofylline antagonized the 5'-N-ethylcarboxamidoadenosine (NECA)-induced stimulation of platelet adenylate cyclase activity with a KB of 130 microM. In human platelets, enprofylline did not antagonize but potentiated the NECA-induced inhibition of aggregation. This potentiation was abolished in the presence of the phosphodiesterase inhibitor papaverine. An adenosine antagonistic effect of enprofylline could not be evaluated on A2 receptors of guinea-pig lung because the xanthine enhanced basal and NECA-stimulated cyclic AMP accumulation. Enprofylline antagonized the N6-R-(-)-phenylisopropyladenosine (R-PIA)-induced inhibition of rat fat cell adenylate cyclase with a KB of 32 microM. The Ki value for inhibition of [3H]PIA binding to fat cell membranes was 45 microM. Enprofylline inhibited cyclic AMP phosphodiesterase activity of human platelets, guinea-pig lung and rat fat cells with Ki values of 15, 130 and 110 microM, respectively. The results show that enprofylline was nearly equipotent as antagonist at A1 and A2 adenosine receptors. Mechanisms other than adenosine antagonism or phosphodiesterase inhibition may be involved in the pharmacological effects of enprofylline.

Effects of enprofylline and theophylline on exercise-induced asthma

Eight asthmatic out-patients with a history of exercise-induced asthma (EIA) were randomly treated with intravenously administered enprofylline 1.5 mg/kg b.wt., theophylline 5 mg/kg b.wt., and placebo immediately prior to a 6-min exercise provocation in this double-blind crossover comparison. A reduction in peak flow of more than 20% was seen in all patients after placebo pre-treatment. Mean plasma concentrations at the start of the exercise test were 3.3 mg/l and 13.2 mg/l after 20 min infusion of enprofylline and theophylline, respectively. The corresponding figures 25 min later were 2.3 and 11.7, respectively. Maximal fall in peak expiratory flow (PEF) after exercise in percent of pre-exercise PEF was 49% +/- 6% (mean +/- SEM), 39% +/- 6% and 24% +/- 5% after infusion of placebo, enprofylline, and theophylline, respectively. Theophylline produced a statistically significant better protection against EIA compared to enprofylline and placebo. Enprofylline produced a slight protection from EIA not statistically significantly different from placebo.

A caffeine analogue (1,3,7-trimethyl-6-thioxo-2-oxopurine) with a negative inotropic and chronotropic effect

Cardiac effects of thio-xanthine derivatives, S-caffeine and S-theophylline, were studied on isolated guinea-pig atria and on partially purified cardiac cAMP phosphodiesterase enzymes. Theophylline and caffeine were taken as reference compounds. On electrically driven left atria S-caffeine (0.01-1 mmol/l) decreased contractile tension in a concentration dependent manner. On spontaneously beating atria, the same concentrations of S-caffeine showed negative inotropic as well as negative chronotropic effects. On electrically driven left atria, S-theophylline (0.01-1 mmol/l) increased heart contractile tension but, at higher concentrations, a reversal of the stimulating effect was observed. Both S-caffeine and S-theophylline inhibited bovine heart cAMP phosphodiesterase activity to a comparable extent. Their inhibitory potencies were about three and nine times higher than those of theophylline or caffeine but consistently lower than that of IBMX. The results show that the replacement of O with S in the methylxanthine molecule drastically modifies the effects induced by the drugs on cardiac function without changing those on cAMP phosphodiesterase.

Correlation between inhibition of a cyclic GMP phosphodiesterase and relaxation of canine tracheal smooth muscle

Inhibition of partially purified cyclic nucleotide phosphodiesterase activity as well as pharmacologically induced relaxation of respiratory airways smooth muscle was examined to determine whether any correlation between these two effects could be found. The phosphodiesterase in extracts of canine tracheal smooth muscle was chromatographed on a DEAE Bio-Gel A column and eluted with a sodium chloride gradient. The peak I activity hydrolyzed cGMP at a higher rate than cAMP although the apparent Km values for these two cyclic nucleotides were relatively close. Comparison of the Ki values for alkylxanthine inhibitors of peak I activity correlated remarkably well with the EC50 values of the same compounds as relaxants of canine tracheal smooth muscle strips. It is concluded that inhibition of the peak I enzyme may cause accumulation of an intracellular pool of cyclic nucleotide and thus produce or contribute to the muscle relaxant effects that were observed.

Effects of bronchoconstrictors and bronchodilators on a novel human small airway preparation

Human lung bronchiolar segments (about 2 mm long and with a diameter of 0.6-1.5 mm) were dissected and circular muscle tension recorded. Airways were identified by histology and in some preparations by relaxant responses to noradrenaline (0.1-10 microM). Adenosine (1-100 microM) produced only very weak contractions, whereas carbachol (EC50 = 0.40 microM), histamine (EC50 = 0.63 microM), prostaglandin D2 (EC50 = 0.50 microM), substance P (EC50 = 4.6 microM) and ATP (1-100 microM) produced much greater ones. The contractions generally developed rapidly and were stable. The mean maximum increase in tension achieved with the most efficient constrictor, carbachol, was 0.5 g. ATP was the least efficient producing only about 40% of carbachol's maximum. Terbutaline, theophylline and enprofylline relaxed carbachol (2.0 microM = EC70)-contracted preparations. Terbutaline (3-3000 nM) relaxed 4 out of 11 bronchioles. Theophylline (10-4000 microM) and enprofylline (1-400 microM) consistently relaxed the bronchiolar preparations including those exhibiting little responsiveness to the beta 2-adrenoceptor agonist. Since enprofylline (which does not block adenosine receptors) was a five times more potent relaxant than theophylline and since adenosine produced only weak contractions, antagonism of adenosine receptors is probably not involved in relaxation of the small airways. It is suggested that the present data, which apparently differ from those obtained with lung parenchymal strips, are of relevance for human small airways responsiveness.

Adenosine mechanisms in the regulation of breathing in the rat

The central respiratory effects of various adenosine (A) analogues were studied in halothane-anesthetized rats. Intracerebroventricular (i.c.v.) and intraperitoneal (i.p.) injections of the A analogues (2-Cla, L-PIA, CHA and NECA) reduced minute ventilation (VE) due to decreases in respiratory frequency (f) as well as tidal volume (VT). Dose-dependent effects were seen after i.c.v. L-PIA in both normal and vagotomized rats. Analysis of the A-induced changes using the occluded breath technique revealed an increase in expiratory time (TE) as well as a decrease in inspiratory drive. NECA, a relatively specific A2 agonist seemed to be somewhat more potent in eliciting respiratory depression than a relatively specific A1 agonist like L-PIA. Pretreatment with the methylxanthine theophylline completely antagonized the respiratory depression induced by L-PIA. It is concluded that central A receptors are involved in the central regulation of breathing and that A interacts with the respiratory control system mainly by decreasing inspiratory neural drive and prolonging expiratory time.

Increase in plasma free fatty acids and natriuresis by xanthines may reflect adenosine antagonism

The hypothesis has been examined that adenosine is involved in the diuretic and free fatty acid (FFA) - releasing action of xanthines. The effects of theophylline (T), a potent adenosine antagonist, were compared with those of enprofylline (3-propyl xanthine, E), which exerts negligible antagonism of adenosine. Eight healthy male volunteers were given E 1.5 mg/kg, T 5.0 mg/kg or placebo 0.9% saline (P) intravenously in a double-blind, randomized, cross-over investigation. Blood samples were analyzed for E, T, catecholamines (CA: adrenaline, noradrenaline and dopamine), FFA, renin, glucose, glucagon and insulin, and urine was collected at 2-h intervals. T (plasma concentration 53 +/- 8 mumol/l) but not E (11 +/- 2 mumol/l) caused an increase in FFA from 0.42 to 0.86 mmol/l after 90 min. Without affecting the urinary excretion of potassium, T doubled natriuresis and the urine volume as compared to E and P. Neither T nor E had any effect on plasma CA, or on any other of the metabolic parameters studied. E, but not T, produced a small but statistically significant decrease in diastolic blood pressure (5 mmHg) and an increase in heart rate (3 beats/min). It is suggested that the difference between E and T in terms of stimulation of FFA-release and natriuresis may be related to their different ability to antagonize adenosine.

Effects on aggregation of human platelets of two xanthines and their interactions with adenosine

Adenosine receptor antagonism has been suggested to be the cellular basis for many extrapulmonary actions of xanthine derivatives, such as theophylline. Enprofylline (3-propylxanthine) is a poor adenosine antagonist but is five times as potent as theophylline as a bronchodilator in man. Adenosine is a potent inhibitor of platelet aggregation, but also xanthines are considered to exert this action. In the present study, effects of theophylline and enprofylline on ADP-induced aggregation of human platelets were studied in vitro. Theophylline alone in concentrations exceeding 280 microM inhibited platelet aggregation concentration-dependently. Enprofylline alone mimicked this effect but was about five times more potent than theophylline. At the lowest concentrations used, corresponding to upper therapeutic levels, neither of the two xanthines affected platelet aggregation by ADP. The interactions between these low concentrations of the xanthines and adenosine were then evaluated. In the presence of theophylline 110 microM the inhibitory effect of adenosine 4 microM was attenuated, whereas the presence of enprofylline 21 microM enforced the inhibitory effect of adenosine. Thus, at low concentrations where neither theophylline nor enprofylline inhibits platelet aggregation theophylline antagonizes the antiaggregatory effect of adenosine, whereas enprofylline acts in synergy with this nucleoside.

Tolerance and some circulatory effects of intravenous and oral enprofylline in healthy volunteers

Enprofylline (3-propylxanthine), a novel bronchodilating xanthine derivative that seems to lack adenosine antagonistic potency was given intravenously to eight recumbent healthy male volunteers in the doses 0.5, 1 and 1.5 mg/kg body-weight and to six of them also orally in the doses 2, 4 and 6 mg/kg. Mean enprofylline plasma levels ranged between 1.6 and 4.4 mg/1 (8.2-22.2 mumol/1) after intravenous, and between 1.9 and 5.5 mg/1 (9.8-27.9 mumol/1) after oral administration. Enprofylline was rapidly and completely absorbed and had an elimination half-life of approximately 2 hrs. About 90% of the dose given by either route was recovered as unchanged drug in the urine. A slight but significant increase in heart rate was seen at peak plasma levels after each of the highest intravenous and oral doses. At these dose levels the heart rate response to orthostatic tests was significantly increased by enprofylline. Adverse reactions were mild and short-lasting and occurred most frequently after the two highest intravenous and oral doses. Headache and nausea were noted in 5 of the 24 intravenous experiments and in 9 of the 17 times that enprofylline was given orally. In conclusion, the circulatory effects of enprofylline were small and the adverse reactions mild. Further clinical studies with enprofylline seem warranted.

New derivatives of methyl-xanthines: effect of thiocaffeine thiotheophylline and 8-phenyltheophylline on lipolysis and on phosphodiesterase activities

The effects of some old and new methylxanthines and 6-thioxanthines, i.e. theophylline (TH), caffeine (CAFF), thiotheophyl line (S-TH), thiocaffeine (S-CAFF), 8-phenyltheophylline (8-PT), 3-isobutyl-1-methyl-xanthine (IBMX) were compared at the level of adipose tissue on spontaneous and norepinephrine-induced lipolysis as well as on fat cell phosphodiesterases. These agents stimulated lipolysis. 8-PT was the most potent; thiocaffeine and thiotheophylline the least potent; IBMX and theophylline had intermediate potencies. The order of potency of the same drugs in potentiating norepinephrine-stimulated lipolysis was: IBMX greater than 8-PT greater than S-CAFF greater than greater than S-TH greater than CAFF greater than TH. The rank order of potency to inhibit cAMP phosphodiesterases was: IBMX greater than S-TH and S-CAFF greater than TH much greater than 8-PT (uneffective). Thus (a) thiocaffeine and thiotheophylline were more potent than the parent compound theophylline in inhibiting cAMP-PDE, although their per se stimulating effect on lipolysis was much lower. This indicates that in the thioxanthines the stimulus per se on lipolysis can be dissociated from their effectiveness as inhibitors of PDE. (b) In contrast, 8-PT (a blocker of adenosine receptors) has a potent lipolytic action per se and a potentiating effect on norepinephrine-induced lipolysis, even if deprived of effect on PDE. This indicates that the potentiating power of methylxanthines on norepinephrine-stimulated lipolysis is not strictly dependent on their anti-PDE activities. These results suggest that in adipose tissue (more similar than adipocytes to the in vivo conditions) the anti-adenosine potency of methylxanthines is the prominent factor in stimulating basal lipolysis. Both anti-adenosine and anti-PDE activities are involved in modulating hormone-induced lipolysis.