Introduction

Top Abstract Introduction Materials & Methods Results Discussion References

Cyclic AMP is an ubiquitous intracellular second messenger, and an increase in its concentration causes a wide range of physiological responses (Robinson et al., 1971). By blocking cAMP degradation, inhibitors of PDE augment the cellular levels, thus mimicking and enhancing the biological effects of agents that activate adenylyl cyclase. At least seven distinct gene families of cyclic nucleotide PDE have been identified (Beavo et al., 1994; Beavo, 1995; Conti et al., 1995). Among them, we have paid particular attention to type 4 PDE. This isozyme hydrolyzes cAMP only and is the predominant isozyme present in leukocytes (Torphy and Undem, 1991). PDE4 inhibitors are highly effective in blocking the production of TNF, a proinflammatory mediator implicated in lung inflammation and injury (Semmler et al., 1993; Turner et al., 1993; 1996). It has been shown that combined treatment with the PDE4 inhibitor rolipram and either PGE or a  agonist, augments the levels of cAMP in many systems (for review, see Torphy and Undem, 1991), including U-937 cells (Cheng et al., 1994), isolated rat tail artery (Absood et al., 1992) and perfused rat lung (Barnard et al., 1994). The elevated cAMP is then transported rapidly out of the cell and tissues, which suggests that once inhibition of PDE4 occurs, extracellular levels of cAMP rise. We reasoned, therefore, that when administered systemically, a PDE4 inhibitor should increase levels of cAMP in the plasma of animals, an effect that could be correlated with its functional responses in vivo.

Therefore, we set out to examine the effect of the p.o. administration of CP-80,633, a new PDE4 inhibitor (Cohan et al., 1996; Hanifin et al., 1966), on plasma cAMP levels in mice. Because the ability of rolipram to inhibit the local production of TNF (Pettipher et al., 1996) and the formation of ear edema (Griswold et al., 1993) can be influenced by removal of adrenal glands in rodents, we also studied CP-80,633-mediated cAMP and TNF responses in adx or propranolol-treated mice.

	Materials and Methods

Top Abstract Introduction Materials & Methods Results Discussion References

Chemicals. CP-80,633, dl-rolipram, dipyridamole, piroxicam, vinpocetine and zaprinast were synthesized in the Department of Medicinal Chemistry, Pfizer Central Research (Groton, CT). dl-Propranolol and LPS (E. coli 0111:B4) were purchased from Sigma Chemical Co. (St. Louis, MO). SKF-94,120 was a gift from SKB pharmaceuticals (King of Prussia, PA).

Animals. Normal male Balb/c mice, adx mice and their sham controls (20-25 mg) were purchased from Charles River Laboratories (Raleigh, NC). The mice were housed in 23°C at a 12-hr light/dark cycle. The drinking water of the adx mice was supplemented with 0.9% NaCl.

Plasma cAMP and cGMP measurements. Mice were administered VEH p.o. (0.5% carboxymethylcellulose), CP-80,633 (1-32 mg/kg), theophylline (100 mg/kg) or other drugs (fig. 3). After 20 min (except for figs. 2 and 6), mice were exsanguinated, and blood was collected in 1.5-ml conical Eppendorff polypropylene tubes containing 50 µl of 50 mg/ml disodium EDTA. The samples were kept on ice and centrifuged at 10,000 × g for 15 min at 4°C. The plasma samples were stored at 20°C before assay. Each sample was thawed at 23°C, diluted 5-fold and assayed for cAMP in quadruplicate by cAMP RIA (New England Nuclear, Boston, MA). The sample for cGMP measurement was prepared similarly, and its level was determined by cGMP RIA (New England Nuclear). Adrenalectomized mice and sham controls were treated in the same way. In some experiments, propranolol (10 mg/kg i.p.), the CO inhibitor piroxicam (3.2 mg/kg. p.o.) or PBS was administered 15 min before treatment with CP-80,633.

View larger version (14K): [in this window] [in a new window]   Fig. 3.   Comparison of effects of PDE inhibitors on plasma cAMP levels in normal mice. The mice were dosed p.o. with compound or VEH in groups of four animals. Twenty minutes later, blood was collected and processed for cAMP measurements. Each value is the mean ± S.E. of four mice. * Significantly different from VEH control (P < .05).

View larger version (17K): [in this window] [in a new window]   Fig. 2.   Time course of plasma cAMP response in normal mice. The mice were dosed p.o. with CP-80,633 or VEH in groups of four animals. Twenty minutes after the indicated time, blood was collected and processed for cAMP measurements. Each point is the mean ± S.E. of four mice.

View larger version (17K): [in this window] [in a new window]   Fig. 6.   Effect of CP-80,633 on LPS-induced TNF production in sham controls and adx mice. CP-80,633 or VEH was administered p.o. 0.5 min before the injection of LPS (0.3 mg/mouse i.p.) or PBS. One hour later, blood was collected and processed for TNF measurements. Each sample was diluted 1 to 10-fold and assayed for TNF by ELISA. Each value is the mean ± S.E. of four mice. * Significantly different from the value of the sham control.

Serum TNF measurements. Control and adx mice were administered VEH or CP-80,633 (1-32 mg/kg p.o.) 0.5 hr before injection of LPS (0.3 mg/mouse i.p.). One hour later, blood was collected in serum separator tubes. The methods used to prepare serum samples and to assay TNF levels were identical to those reported previously (Pettipher et al., 1996). Briefly, each sample was diluted 1 to 10-fold and assayed for TNF by ELISA (Genzyme, Boston, MA).

Serum thromboxane B₂ measurements. Normal mice were bled by cardiac puncture, and the blood was transferred to a glass tube. After incubation at room temperature for 60 min, the samples were centrifuged at 3000 × g for 10 min. The serum samples were stored at 20°C before assay. Each sample was thawed at room temperature, diluted 100-fold and assayed for thromboxane B₂ by EIA (Cayman, Ann Arbor, MI).

Plasma epinephrine measurements. Epinephrine was extracted from mouse plasma by a modification of the method of Candito et al. (1993). Briefly, 100 µl of plasma was added to a 1.5-ml centrifuge tube made to contain 100 nM dihydroxybenzamine (DHBA) internal standard, and 10 mg of alumina acid type (Woelm A, activity grade I) was added, followed by 0.15 M Tris HCl, pH 8.6, containing 2 g of disodium EDTA/100 ml. After vortexed mixing, tubes were centrifuged and the supernatant removed with a pipette. The alumina pellet was washed twice with 200 µl of water, and the catecholamines were eluted from the alumina pellet using 100 µl of 0.1 N perchloric acid.

Statistics. All values presented are mean ± S.E. For statistical analysis, we performed an unpaired two-tailed Student's t test. The mean values of two groups were considered significantly different if P < .05.

	Results

Top Abstract Introduction Materials & Methods Results Discussion References

Effects of CP-80,633 and other PDE inhibitors on the plasma levels of cAMP in control and adx mice. When given p.o., CP-80,633 increased plasma cAMP in mice in a dose-dependent manner (fig. 1). The levels were increased as early as 10 min, and remained elevated for up to 250 min, after p.o. dosing (fig. 2), which demonstrates that the cAMP response is rapid and long-lasting. In contrast to its effect on the cAMP response, CP-80,633 (10 mg/kg p.o.) did not affect plasma levels of cGMP in the mice (VEH control: 48.7 ± 5.5 pmol/ml vs. CP-80,633: 35.7 ± 5.8 pmol/ml; P > .05, n = 4).

View larger version (14K): [in this window] [in a new window]   Fig. 1.   Effect of increasing doses of CP-80,633 on plasma cAMP levels in normal mice. Normal male Balb/c mice were dosed p.o. with CP-80,633 or VEH in groups of 12 animals. Twenty minutes later, blood was collected and processed for cAMP measurements. Each plasma sample was diluted 5-fold and assayed for cAMP in quadruplicate by cAMP RIA. Each point is the mean ± S.E. of 12 mice. * Significantly different from VEH control (P < .05).

View larger version (15K): [in this window] [in a new window]   Fig. 4.   Effect of increasing doses of CP-80,633 on plasma cAMP levels in sham controls and adx mice. Adrenalectomized mice and their sham control Balb/c mice were dosed p.o. with CP-80,633 or VEH in groups of four animals. Twenty minutes later, blood was collected and processed for cAMP measurements. Each value is the mean ± S.E. of eight mice. * Significantly different from the value for the sham control (P < .05); + Significantly different from the VEH control (P < .05).

Effect of CP-80,633 on the cAMP response in mice treated with propranolol and piroxicam. To investigate the involvement of beta adrenoreceptors in the response, we studied the effect of propranolol in these animals. Propranolol alone (10 mg/kg i.p.) decreased cAMP levels from 55.7 ± 6.7 to 38.9 ± 4.1 pmol/ml in control mice and from 47.3 ± 2.2 to 32.7 ± 1.5 pmol/ml in adx mice (n = 4 for each group). This suggests that basal levels of catecholamines contribute to the regulation of cAMP production.

View larger version (16K): [in this window] [in a new window]   Fig. 5.   Effect of CP-80,633 on plasma cAMP levels in sham controls and adx mice treated with propranolol. CP-80,633 (10 mg/kg) or VEH was administered p.o. 15 min before the injection of propranolol (10 mg/kg i.p.) or PBS. Twenty minutes later, blood was collected and processed for cAMP measurements. Each value is the mean ± S.E. of four mice. * Significantly different from the PBS + VEH control (P < .05); + Significantly different from the value for the PBS + CP-80,633 group (P < .05); # Significantly different from PBS + VEH control in both sham controls and adx mice (P < .05).

Effect of CP-80,633 on the systemic production of TNF in control and adx mice. It is well established that TNF is generated in the serum of mice 1 hr after challenge with a lethal dose of LPS, and this endpoint is exquisitely sensitive to inhibition by rolipram (Pettipher et al., 1996). Figure 6 shows that serum levels of TNF under these conditions were enhanced by removal of adrenal glands. As is also shown, CP-80,633 dose-dependently reduced TNF production with ED₅₀ = 1.2 mg/kg p.o. in control mice. CP-80,633 was also effective in blocking the response in adx mice (ED₅₀ = 3.9 mg/kg p.o.). This small shift in potency is similar to that seen with rolipram (Pettipher et al., 1996).

Effect of CP-80,633 on the plasma levels of epinephrine in control and adx mice. As shown in figure 7, CP-80,633 significantly (P < .05) increased plasma epinephrine when given p.o. at 10 mg/kg. The epinephrine levels were almost completely blocked by removal of adrenal glands (control: 517 ± 85 vs. adx: 2.7 ± 0.4 pmol/ml, n = 5). In adx mice, CP-80,633 (10 mg/kg p.o.) failed to elevate plasma epinephrine (1.6 ± 0.6 pmol/ml). These results suggest that CP-80,633 has a direct stimulatory effect on the release of epinephrine from adrenal glands in mice.

View larger version (14K): [in this window] [in a new window]   Fig. 7.   Effect of CP-80,633 on plasma epinephrine levels in normal mice. CP-80,633 (10 mg/kg) or VEH was administered p.o. in groups of five animals. Thirty minutes later, blood was collected and processed for epinephrine measurements. Each value is the mean ± S.E. of five mice. * Significantly different from the VEH control (P < .05).

	Discussion

Top Abstract Introduction Materials & Methods Results Discussion References

In this study, we have demonstrated that 1) when given orally, CP-80,633 can raise plasma cAMP and epinephrine concentrations in mice, 2) removal of adrenal glands causes a marked decrease in the cAMP response because of a reduction in circulating levels of epinephrine, 3) a small component of the cAMP response is independent of adrenal catecholamines and 4) the inhibitory effect of CP-80,633 on systemic TNF production is retained in adx animals, though with a slight reduction in potency.

In many cell types studied, CP-80,633 alone fails to elevate intracellular cAMP levels but potentiates the increase of cAMP levels caused by an adenylyl cyclase activator (Cohan et al., 1996). Therefore, the enhanced cAMP initiated by treatment with CP-80,633 may reflect potentiation of constitutive cAMP levels resulting from endogenous adenylyl cyclase activators. It has been shown that rolipram and denbufylline raise circulating levels of corticosterone by stimulating adrenal glands in rodents (Hadley et al., 1991; Pettipher et al., 1996). Because treatment with propranolol prevents most of the cAMP response (fig. 5), it is likely that CP-80,633 elevates cAMP primarily through a catecholamine-mediated pathway. Increased synthesis of corticosterone may have contributed to plasma cAMP response by increasing the number of beta adrenoreceptors. However, we reject this hypothesis, because elevated levels of cAMP are detected as early as 10 min after p.o. dosing (fig. 2), whereas it takes much longer (more than 4 hr) for steroids to enhance the receptor number in lung tissues (Mano et al., 1979; Cheng et al., 1980; Fraser and Venter, 1980).

Figure 3 demonstrates that among the compounds tested, only the inhibitors of PDE4, CP-80,633 and rolipram raise plasma cAMP. In this plasma cAMP test, CP-80,633 is highly effective, at least 30 times more efficacious than p.o. theophylline. The present result suggests that CP-80,633 acts as a PDE4 inhibitor in mice and that plasma cAMP can serve as a potential marker of PDE4 inhibition in vivo.

Because plasma cAMP levels after treatment with CP-80,633 are still higher in propranolol-treated, adx mice than in sham controls, endogenous factors other than adrenal catecholamines must contribute, though modestly, to the effect of CP-80,633. It has been shown that many neuropeptides are potent stimulators of plasma cAMP in mice (Absood et al., 1992). Whether these factors or others, such as adnenosince (Kaminsky et al., 1970; Nistrup Madsen, 1977; Minkoff et al., 1985; Fehr, et al., 1990) and central and peripheral nerve systems (Watchtel and Schneider, 1986; Scuvee-Moreau et al., 1987; Weinreich and Undem, 1987; O'Donnell, 1993; Underwood et al., 1996), are involved in plasma cAMP responses remains to be tested. We have ruled out products of the CO pathway as potential candidates, because piroxicam had no effect. CP-80,633 does not bind to adenosine receptors (Cohan et al., 1996), so it is unlikely that increased cAMP levels are due to an adenosine-mediated cAMP efflux as shown in pig aortic smooth muscle cells (Fehr et al., 1990). Besides stimulating the release of catecholamines from adrenal glands, theophylline can enhance the biological effects by reducing the uptake and metabolism of catecholamines in blood vessels (Kalsner, 1971, et al., 1975). However, the latter does not seem to contribute to the cAMP response induced by CP-80,633, because the adrenal gland-independent cAMP response is insensitive to propranolol (fig. 5B), a result that excludes any mechanisms mediated by beta adrenoreceptors.

Because systemically administered CP-80,633 undergoes extensive metabolism in other species with a plasma T_1/2 less than 0.5 hr in rats (data unpublished), the prolonged cAMP response in mice (5 hr after CP-80,633) was unexpected. This long-lasting response differs from the effect obtained with forskolin. Forskolin, administered i.v. to mice, elevates plasma cAMP within 10 min, but the response disappears by 30 min (Absood et al., 1992). In this regard, the in vivo plasma cAMP response is similar to the response seen with isolated cells, i.e., the intracellular cAMP rises transiently in response to an adenylyl cyclase activator, and a PDE4 inhibitor augments not only the magnitude but also the duration of the cyclase activator-mediated increase in cAMP (Beavo, 1995).

The cell types responsible for the potentiating effect of CP-80,633 are unknown. Recently, it became clear that several cells can use an alternative strategy to lower an excess of intracellular cAMP by potentiating its efflux out of the cell. This effect is particularly evident in the presence of a PDE4 inhibitor. Thus rolipram, in combination with either beta agonist or PGE, stimulates cAMP efflux from U-937 cells (a human promonocytic cell line), perfused rat lung and isolated rat tail artery (Absood et al., 1992; Cheng et al., 1994; Barnard et al., 1994). It is conceivable that macrophages, endothelial cells and leukocytes, which contain a PDE4 isozyme (Torphy and Undem, 1991), are involved in the synthesis and efflux of cAMP induced by treatment with CP-80,633. The cellular mechanism that regulates cAMP efflux in response to PDE4 inhibition is not fully understood. In a preliminary study, we found that cAMP efflux in U-937 cells occurs as early as 2 min after rolipram challenge and is inhibited by both cold temperature and chloride-channel blockade (Cheng et al., 1994). Taken together, these results suggest that elevated plasma concentrations of cAMP by CP-80,633 may result from increased cAMP export from PDE4-sensitive cell types.

Consistent with its effect in vitro (Cohan et al., 1996), p.o. administration of CP-80,633 blocks LPS-induced TNF production. In fact, several in vitro studies have demonstrated a link between PDE4 inhibition (and thus, cAMP elevation) and TNF production in isolated monocytes (Semmler et al., 1993; Prabhakar et al., 1994). In this study there appears to be a correlation between the effects of CP-80,633 on cAMP and the TNF response in vivo: 1) the dose-response curves are coincident, and 2) removal of adrenal glands decreases plasma cAMP and increases TNF production. However, inhibition of TNF by CP-80,633 is only modestly reduced by adrenalectomy, whereas the cAMP response is greatly diminished. CP-80,633 is still capable of raising plasma cAMP in adx mice, so these mice may confer a cAMP tone that is sufficient to enhance the effect of CP-80,633. This suggests that CP-80,633 may synergize with unknown adenylyl cyclase activators to inhibit TNF production in adx mice. The small decrease in potency by CP-80,633 after removal of adrenal glands is similar to that seen with rolipram (Pettipher et al., 1996). In contrast, the ability of rolipram to inhibit local TNF production is greatly reduced by adrenalectomy because of a lack of production of corticosterone (Pettipher et al., 1996).

In summary, elevated plasma levels of cAMP by CP-80,633 are largely, but not entirely, dependent on circulating epinephrine. Adrenalectomy, however, causes only a modest reduction in the ability of CP-80,633 to inhibit systemic TNF production. The clinical relevance of this observation is, at present, unknown. As yet, there are no reports indicating that a PDE4 inhibitor can cause an elevation in plasma cAMP in the human.