Function of cyclo-oxygenase-1 and cyclo-oxygenase-2 in the ductus arteriosus from foetal lamb: differential development and change by oxygen and endotoxin

Abstract

1. Prenatal patency of the ductus arteriosus is maintained mainly by prostaglandin(PG) E(2). Here we have examined the relative importance of cyclo-oxygenase-1 (COX1) and cyclo-oxygenase-2 (COX2) for PGE(2) formation in the foetal lamb ductus (0.65 gestation onwards). 2. Using fluorescence microscopy and immunogold staining, COX1 appeared more abundant than COX2 in endothelial and smooth muscle cells, and this difference was greater before-term. Inside muscle cells, COX1 and COX2 immunoreactivity was located primarily in the perinuclear region. Endotoxin, given to the lamb in utero (approximately 0.1 microg kg(-1)), caused COX2 upregulation, while an opposite effect with disappearance of the enzyme followed endotoxin treatment in vitro (100 ng ml(-1)). COX1 immunoreactivity remained virtually unchanged with either treatment; however, this isoform as well as any induced COX2 migrated towards the outer cytoplasm. 3. The COX2 inhibitor L-745,337 (1--10 microM) contracted the isolated ductus at term, the response being almost as high as that to indomethacin (dual COX1/COX2 inhibitor) over the same dose-range. Conversely, L-745,337 was relatively less effective in the premature. 4. Pretreatment of the premature in vivo with endotoxin enhanced the contraction of the ductus to L-745,337, while in vitro endotoxin had a variable effect. 5. The premature ductus exhibited a stronger contraction to L-745,337 following exposure to oxygen. On the other hand, the oxygen contraction, which is modest before-term, was enhanced by L-745,337. 6. We conclude that COX1 and COX2 develop unevenly in the ductus. While both enzymes contribute to PGE(2) formation at term, COX1 is the major isoform in the premature. COX2, however, may acquire greater importance before-term following physiological and pathophysiological stimuli.

Figures

Figure 1

Immunofluorescence micrographs of the ductus arteriosus from control vs endotoxin-treated (treatment in vivo) foetal lambs at 0.7 gestation and at term. (A) Control preparation at 0.7 gestation, section incubated with COX1 antiserum; note the labelling of endothelial (arrowheads in this and the following panels) and smooth muscle cells. (B) Control preparation at term gestation, section incubated with COX1 antiserum. (C) Sequential section to (A) incubated with COX2 antiserum; note the faint staining compared to COX1. (D) Sequential section to (B) incubated with COX2 antiserum; COX2 immunoreactivity increases with gestation age, although it still appears weak compared to COX1 immunoreactivity. (E) Endotoxin-treated preparation at 0.7 gestation, section incubated with COX1 antiserum; note that immunolabelling is essentially the same as in the control preparation. (F) Endotoxin-treated preparation at term gestation, section incubated with COX1 antiserum; immunolabelling virtually unchanged compared to controls. (G) Sequential section to (E) incubated with COX2 antiserum; note the increased immunoreactivity relative to control of the same age. (H) Sequential section to (F) incubated with COX2 antiserum; immunoreactivity is stronger compared to both the control of the same age and the endotoxin-treated premature. Note that under control conditions immunoreactivity for both COX1 and COX2 showed no obvious difference at 0.65 vs 0.7 gestation (results not presented). Bar represents 250 μm.

Figure 2

Transmission electron micrograph of ultrathin cryosections from the muscle layer of the foetal ductus arteriosus at term (untreated animal vs animal treated with endotoxin in vivo). In all panels, sections have been reacted with antiserum against COX1. (A) Immunogold labelling of the perinuclear region in the untreated tissue; note that label is found close to the nucleus (N) in Golgi apparatus and Golgi-derived vesicles (arrowheads). (B) Immunogold labelling in the untreated tissue; note the sparse label in the peripheral cytoplasm along the plasma membrane (arrowheads). (C) Immunogold labelling in the endotoxin-treated tissue; label is found in vesicular and membraneous structures distant from the nucleus (N). (D) Immunogold labelling in the endotoxin-treated tissue; note the abundance of label in the peripheral cytoplasm in proximity of the plasma membrane (indicated with arrows). Bar represents 0.2 μm.

Figure 3

Transmission electron micrograph of ultrathin cryosections from the muscle layer of the foetal ductus arteriosus at term (untreated animal vs animal treated with endotoxin in vivo). In all panels, sections have been reacted with antiserum against COX2. (A) Immunogold labelling of the perinuclear region in the untreated tissue; note the abundance of label (arrowheads) in the Golgi apparatus, Golgi-derived vesicles and sarcoplasmic reticulum surrounding the nucleus (N). (B) Immunogold labelling in the untreated tissue; note the distribution of sparse label (arrowhead) in the peripheral cytoplasm of two adjacent muscle cells. (C) Immunogold labelling in the endotoxin-treated tissue; the cytoplasm adjacent to the nucleus (N) is void of any label. (D) Immunogold labelling in the endotoxin-treated tissue; note label (arrowheads) in the peripheral cytoplasm close to the plasma membrane (indicated with arrows). (E) Immunogold labelling in the endotoxin-treated tissue; note the high label density in the sarcoplasmic reticulum (arrowheads). Bar represents 0.2 μm.

Figure 4

Intact strip of ductus arteriosus from near-term foetal lamb. Contractile responses to L-745,337 (A), indomethacin (B), and excess potassium (C). L-745,337 effect was measured before and during treatment with endotoxin (LPS) in vitro (for details, see Methods). For each group, number of experiments are given above the columns, and a significant difference in panel (A) between control and treatment groups is indicated with *P<0.05 or **P<0.01 (ANOVA).

Figure 5

Intact strip of ductus arteriosus from foetal lamb at 103 – 107 days gestation. Contractile responses to L-745,337 (A), indomethacin (B), and excess potassium (C). L-745,337 effect was measured in control tissues and tissues treated with endotoxin (LPS) in vitro and in vivo (for details, see Methods). For each group, number of experiments are given above the columns, and a significant difference in panel (A) between control and treatment groups is indicated with *P<0.05 or **P<0.01 (ANOVA). Note that the inhibitory effect of in vitro endotoxin became significant when the analysis was limited to the subgroup of experiments in which responses were measured in the same tissue before and during treatment (1 μM, P<0.01; 2.8 μM, P<0.05; n=8).

Figure 6

Intact strip of ductus arteriosus from foetal lamb at 103 – 107 days gestation. (A) Contractile response to L-745,337 before and after exposure to oxygen (for details, see Methods). (B) Contractile response to oxygen before, during, and after treatment with L-745,337 (2.8 μM) (for details, see Methods). Note that in panel (B) L-745,337 alone caused a contraction of 0.78±0.2 g and that the response to oxygen is measured from this new baseline. For each group, number of experiments are given above the columns, and a significant difference relative to control is indicated with *P<0.01 (ANOVA).

Figure 7

Intact strip of ductus arteriosus from foetal lamb at 94 – 97 days gestation. (A) Contractile response to L-745,337 before and during treatment with endotoxin (LPS) in vitro (for details, see Methods). (B) Contractile response to L-745,337 before and after exposure to oxygen (for details, see Methods). For each group, number of experiments are given above the columns, and a significant difference in panel (B) relative to the control before oxygen exposure is indicated with *P<0.05 or **P<0.01 (ANOVA). Note that in this age group, unlike the 103 – 107 days group (see Figure 5), the inhibitory effect of endotoxin did not reach significance even when comparing responses to L-745,337 in the same tissue (n=7).