Inhibition of cyclooxygenase-2 prevents adverse effects induced by phosphodiesterase type 4 inhibitors in rats.

BACKGROUND AND PURPOSE: Phosphodiesterase type 4 (PDE4) inhibitors such as roflumilast are currently being developed as anti-inflammatory treatments for chronic airway disorders. However, high doses of PDE4 inhibitors have also been linked to several side effects in different animal species, including pro-inflammatory effects in the rat. Here, we analysed PDE4-related toxicological findings in a rat model and how these side effects might be therapeutically prevented. EXPERIMENTAL APPROACH: Wistar rats were treated orally once daily with 10 mg·kg⁻¹ roflumilast for 4 days. Macroscopic changes were monitored throughout the study and further parameters were analysed at the end of the experiment on day 5. In addition, the effects of concomitant treatment with cyclooxygenase (COX) inhibitors were assessed. KEY RESULTS: Supratherapeutical treatment with roflumilast induced marked body and spleen weight loss, diarrhea, increased secretory activity of the harderian glands, leukocytosis, increased serum cytokine-induced neutrophil chemoattractant-1 (CINC-1) levels, and histopathological changes in thymus, spleen, mesentery and mesenteric lymph nodes. All these toxicological findings could be prevented by the non-steroidal anti-inflammatory drug (NSAID) and non-selective COX inhibitor, diclofenac, given orally. Similar protective effects could be achieved by the COX-2 selective inhibitor lumiracoxib, whereas the COX-1 selective inhibitor SC-560 was generally not effective. CONCLUSIONS AND IMPLICATIONS: Treatment with an NSAID inhibiting COX-2 prevents the major effects found after subchronic overdosing with the PDE4-specific inhibitor roflumilast. If this effect translates into humans, such combined treatment may increase the therapeutic window of PDE4 inhibitors, currently under clinical development.

The inositol trisphosphate pathway mediates platelet-activating-factor-induced pulmonary oedema.

Platelet-activating factor (PAF) is a pro-inflammatory lipid mediator that increases vascular permeability by simultaneous activation of two pathways, one dependent on the cyclooxygenase metabolite prostaglandin E2 and the other on the sphingomyelinase metabolite ceramide. The hypothesis that part of the PAF-induced oedema is mediated via the inositol 1,4,5-trisphosphate (IP3) pathway or Rho kinase pathway was investigated. Oedema formation was induced in isolated perfused rat lungs by injection of 5 nmol PAF into the pulmonary artery. Lungs were pre-treated with specific inhibitors: edelfosine (L108) to block phosphatidyl-inositol-specific phospholipase C, xestospongin to block the IP3 receptor, 5-iodonaphthalene-1-sulphonyl-homopiperazine (ML-7) to block myosin light chain kinase, and (+)-R-trans-4-(aminoethyl)-N-(4-pyridyl)cyclohexanecarboxamide (Y27632) to block Rho-associated protein kinase. Pre-treatment with L108 or xestospongin reduced PAF-induced oedema formation by 58 and 56%, respectively. The effect of L108 was additive to that of the cyclooxygenase inhibitor acetyl salicylic acid (88% oedema reduction). PAF-induced oedema formation was also reduced if extracellular calcium concentrations were lowered. Furthermore, treatment with ML-7 reduced oedema formation by 54%, whereas Y27632 was without effect. It is concluded that platelet-activating-factor-triggered oedema is mediated by activation of the inositol 1,4,5-trisphosphate pathway, influx of extracellular calcium and subsequent activation of a myosin light chain kinase-dependent and Rho-associated-protein-kinase-independent mechanism.

Pressor responses to platelet-activating factor and thromboxane are mediated by Rho-kinase.

Platelet-activating factor (PAF) contracts smooth muscle of airways and vessels primarily via release of thromboxane. Contraction of smooth muscle is thought to be mediated either by calcium and inositol trisphosphate (IP(3))-dependent activation of the myosin light chain kinase or, alternatively, via the recently discovered Rho-kinase pathway. Here we investigated the contribution of these two pathways to PAF and thromboxane receptor-mediated broncho- and vasoconstriction in two different rat models: the isolated perfused lung (IPL) and precision-cut lung slices. Inhibition of the IP(3) receptor (1-10 microM xestospongin C) or inhibition of phosphatidylinositol-specific PLC (30 microM L-108) did not affect bronchoconstriction but attenuated the sustained vasoconstriction by PAF. Inhibition of myosin light chain kinase (35 microM ML-7) or of calmodulin kinase kinase (26 microM STO609), which regulates the phosphorylation of the myosin light chain, had only a small effect on PAF- or thromboxane-induced pressor responses. Similarly, calmidazolium (10 microM), which inhibits calmodulin-dependent proteins, only weakly reduced the airway responses. In contrast, Y-27632 (10 microM), a Rho-kinase inhibitor, attenuated the thromboxane release triggered by PAF and provided partial or complete inhibition against PAF- and thromboxane-induced pressor responses, respectively. Together, our data indicate that PAF- and thus thromboxane receptor-mediated smooth muscle contraction depends largely on the Rho-kinase pathway.

Ventilator-induced lung injury leads to loss of alveolar and systemic compartmentalization of tumor necrosis factor-alpha.

OBJECTIVES: To determine the effect on compartmentalization of the tumor necrosis factor (TNF)-alpha response in the lung and systemically after ventilation with high peak inspiratory pressure with and without positive end-expiratory pressure (PEEP). DESIGN AND SETTING: Prospective, randomized, animal study in an experimental laboratory of a university. SUBJECTS AND INTERVENTIONS: 85 male Sprague-Dawley rats. Lipopolysaccharide was given intratracheally or intraperitoneally to stimulate TNF-alpha production; control animals received a similar amount of saline. Animals were subsequently ventilated for 20 min in a pressure control mode with peak inspiratory pressure/PEEP ratio of either 45/0 or 45/10 (frequency 30 bpm, I/E ratio 1:2, FIO2 = 1). MEASUREMENTS AND RESULTS: Blood gas tension and arterial pressures were recorded at 1, 10, and 20 min after start of mechanical ventilation. After killing of the animals pressure-volume curves were recorded, and bronchoalveolar lavage (BAL) was performed for assessment of protein content and the small/large surfactant aggregate ratio. TNF-alpha was determined in serum and BAL. TNF-alpha levels were significantly increased after lipopolysaccharide stimulation; furthermore ventilation without PEEP resulted in a significant shift of TNF-alpha to the nonstimulated compartment as opposed to ventilation with a PEEP level of 10 cmH2O. CONCLUSIONS: Ventilation strategies which are known to induce ventilation-induced lung injury (VILI) disturb the compartmentalization of the early cytokines response in the lung and systemically. Furthermore, the loss of compartmentalization is a two-way disturbance, with cytokines shifting from the vascular side to the alveolar side and vice versa. A ventilation strategy (PEEP level of 10 cmH2O) which prevents VILI significantly diminished this shift in cytokines.

Quinolines attenuate PAF-induced pulmonary pressor responses and edema formation.

In the present study we have investigated the mechanisms of pulmonary edema caused by platelet-activating factor (PAF) in isolated rat lungs as well as in mice in vivo. In blood-free perfused and ventilated rat lungs, PAF increased lung weight by 0.59 +/- 0.18 g. The cyclooxygenase inhibitor aspirin (500 microM) blocked this response by one-third, and the quinolines quinine (330 microM), quinidine (100 microM), and chloroquine (100 microM) by two-thirds. Lipoxygenase inhibition (10 microM AA861) alone or in combination with thromboxane receptor antagonism (10 microM SQ29548) had no effect on PAF-induced weight gain. In combination with aspirin, quinine or quinidine completely prevented PAF-induced weight gain and the concomitant increase of the capillary filtration coefficient (K(f,c)). Pretreatment with quinine in vivo prevented not only PAF-, but also endotoxin-induced edema formation as assessed by Evans Blue extravasation. In addition, in vivo quinine prevented the endotoxin-induced release of tumor neurosis factor (TNF). Furthermore, in perfused lungs quinine reduced the PAF-induced increases in airway and vascular resistance, as well as thromboxane release. These findings demonstrate the following anti-inflammatory properties of quinolines: reduction of thromboxane and TNF formation; reduction of PAF-induced vasoconstriction and bronchoconstriction; and attenuation of PAF- and lipopolysaccharide (LPS)-induced edema formation. We conclude that the PAF- induced edema consists of two separate mechanisms, one dependent on an unknown cyclooxygenase metabolite, the other one sensitive to quinolines.