Introduction

Top Abstract Introduction Materials and Methods Results Discussion References

Preeclampsia is one of the most significant health problems of human pregnancy. It is a leading cause of fetal growth retardation, premature birth, and low birth weight babies. Preeclampsia is characterized by maternal hypertension, reduced uteroplacental blood flow, increased platelet aggregation, endothelial cell dysfunction, proteinuria, and edema. Biochemically preeclampsia is associated with two significant imbalances: 1) an imbalance of increased thromboxane and decreased prostacyclin (Walsh, 1985; Friedman, 1988), and 2) an imbalance of increased lipid peroxides and decreased antioxidants (Walsh, 1994).

These imbalances may be significant because many of the physiologic and biochemical actions of lipid peroxides pertain to the abnormalities that are seen with preeclampsia, as recently reviewed in Walsh (1994). Some of the effects of lipid peroxides include the following. 1) Stimulation of prostaglandin endoperoxide synthase to increase production of thromboxane, but inhibition of prostacyclin synthase to decrease production of prostacyclin. This imbalance is believed to contribute to the increased platelet aggregation, maternal hypertension, and reduced uteroplacental blood flow that occur in preeclampsia because thromboxane is a potent vasoconstrictor and a stimulator of platelet aggregation, whereas prostacyclin is a potent vasodilator and an inhibitor of platelet aggregation. 2) Increased cell membrane permeability to proteins and increased incorporation of fatty acids into endothelial cell membranes. Alteration of endothelial cell membranes in the renal and systemic vasculatures with increased permeability to proteins could explain proteinuria and edema. 3) Increased thrombin generation and decreased antithrombin III levels to trigger thrombus formation. This, along with increased thromboxane synthesis, could explain coagulation abnormalities, such as disseminated intravascular coagulation and platelet consumption.

AA-2414 [(±)-7-(3,5,6-trimethyl 1-1,1,4-benzoquinon-2-yl)-7-phenylheptanoic acid] is a novel, long-acting, potent stereospecific thromboxane A₂/prostaglandin endoperoxide type 2 receptor antagonist (Hussein et al., 1994). AA-2414 inhibits the contractile response of aortic and saphenous vein preparations to the thromboxane analog, U-46619, and it is a potent inhibitor of platelet aggregation. Its effects are thought to be primarily by blocking thromboxane receptors rather than by affecting the synthesis of thromboxane, because AA-2414 has been shown to have only weak inhibitory effects on the activity of the cyclooxygenase enzyme. In addition to its thromboxane receptor blocking effects, AA-2414 also has antioxidant properties. AA-2414 inhibits the generation of reactive oxygen species by alveolar macrophages and polymorphonuclear leukocytes (Matsumoto and Ashida, 1996). These properties of AA-2414 might make it a potentially useful drug to consider for the treatment of women with preeclampsia.

The isolated perfused human placental cotyledon presents a useful model to study the potential effects of this drug in pregnancy. This model offers the advantage of studying the effects in a tissue unique to pregnancy, and one in which a physiological event, such as vasoconstriction, can be correlated to a biochemical event, such as secretion of vasoactive compounds. The following study was done to evaluate the ability of AA-2414 to block peroxide-induced vasoconstriction in the isolated perfused human placental cotyledon. Maternal and fetal effluent samples were also collected and analyzed for lipid peroxides, thromboxane, and prostacyclin to determine whether AA-2414 affects their secretion rates by the placenta.

	Materials and Methods

Top Abstract Introduction Materials and Methods Results Discussion References

Placentas were obtained immediately after term delivery from normally pregnant women delivering at the Medical College of Virginia main hospital. Institutional approval to conduct this study was granted by the Committee on the Conduct of Human Research at the Medical College of Virginia.

Isolated Perfused Placental Cotyledon Methodology. This methodology was used as previously described (Walsh et al., 1993). Briefly, a chorionic plate artery leading to a single placental cotyledon and a chorionic plate vein draining the cotyledon were catheterized and perfusion was begun immediately. Krebs-Ringer-bicarbonate (KRB) buffer gassed with 95% O₂, 5% CO₂ and warmed to 37°C was used for perfusion. The composition of the KRB buffer was 125 mM NaCl, 4.5 mM KCl, 0.2 mM Na₂HPO₄, 0.7 mM NaH₂PO₄, 2.5 mM CaCl₂, 1.0 mM MgSO₄, 4.4 mM glucose (80 mg/100 ml), and 29.8 mEq/l NaHCO₃. The placenta was placed in a water-jacketed perfusion chamber warmed to 37°C by a Haake constant-temperature circulating water bath (Haake model D1L, Fisher Scientific Co., Pittsburgh, PA). To continuously monitor the perfusion pressure, the fetal arterial catheter was connected to a pressure transducer connected to a Transbridge TBM 4 transducer amplifier connected to a MP100WS data acquisition workstation (World Precision Instruments, Inc., Sarasota, FL). A Macintosh computer was used with AcqKnowledge wave form data analysis software (World Precision Instruments, Inc., Sarasota, FL). The fetal side of the cotyledon was perfused at a rate of 3 to 4 ml/min to adjust the starting fetal side perfusion pressure to approximately 30 mm Hg. The intervillous space on the maternal side of the cotyledon was perfused at a rate of 6 to 8 ml/min by placing a butterfly needle attached to a catheter underneath the basal plate. Two Masterflex multichannel pumps were used for perfusion (Cole-Parmer Instrument Co., Chicago, IL).

Experimental Design. The experimental designs were based on a previous study demonstrating that t-butyl hydroperoxide (Px) induces vasoconstriction in the human placenta specifically by stimulating thromboxane (Walsh et al., 1993). Peroxides stimulate the synthesis of thromboxane by stimulating the activity of the cyclooxygenase enzyme (Hemler et al., 1979; Kulmacz and Lands, 1983). A vasoconstrictive state can therefore be simulated by perfusing the human placental cotyledon with Px. Px was obtained from Sigma Chemical Co. (St. Louis, MO) and AA-2414 was obtained from TAP Holdings, Inc. (Deerfield, IL). AA-2414 was synthesized at Takeda Chemical Industries, Ltd., Osaka, Japan.

Study 1: Dose Response of AA-2414. The first study evaluated whether the effects of AA-2414 were dose dependent. Placental cotyledons (n = 5) were perfused serially for 20-min intervals with control KRB buffer, Px (100 µM), KRB buffer, and KRB buffer containing Px (100 µM) to which progressively increasing concentrations of AA-2414 were added (1 × 10⁸, 1 × 10⁷, 1 × 10⁶, 1 × 10⁵, and 1 × 10⁴ mol/l). To validate the specificity of any inhibitory effects observed for AA-2414, four additional term placentas were studied in which the Px (100 µM) perfusions were repeated without AA-2414.

Study 2: Single Dose of AA-2414. A second study was done using a dose of AA-2414 of 1 × 10⁵ mol/l that is equivalent to plasma levels achieved in clinical studies after oral administration of AA-2414 (Hussein et al., 1994). In the second study we evaluated the effects of AA-2414 on fetal, as well as maternal, secretion rates of lipid peroxides, thromboxane, and prostacyclin, and we evaluated whether its inhibitory effects persisted after its perfusion is discontinued. Placental cotyledons (n = 6) were perfused serially for 20-min intervals with the following treatments: control KRB buffer, Px (100 µM), control KRB buffer, AA-2414 (1 × 10⁵ mol/l), AA-2414 plus Px, and Px alone. Both maternal and fetal effluent samples were collected and processed as described above for study 1.

Sample Analysis. The samples were analyzed for thromboxane and prostacyclin by radioimmunoassay of their stable metabolites, thromboxane B₂ (TXB₂) and 6-keto prostaglandin F₁ (6-keto PGF₁). Both assays were validated for placental perfusion samples as previously described (Walsh et al., 1993). TXB₂ and 6-keto PGF₁ standards and antibodies were purchased from PerSeptive Diagnostics, Inc. (Cambridge, MA). Tritiated TXB₂ and 6-keto PGF₁ were purchased from New England Nuclear (Dupont Research, Wilmington, DE). Analysis of various volumes of perfusate samples resulted in parallelism with the standard curve for both assays. Recovery of exogenously added known amounts of TXB₂ or 6-keto PGF₁ to 5 ml of KRB buffer followed by the concentration procedure was 83 to 98%. KRB buffer blanks resulted in zero dose responses. Within- and between-assay variations were <10% for both assays.

Calculations. Placental secretion rates were calculated by multiplying the concentrations in either the fetal or maternal effluents by their respective effluent perfusion flow rates. Placental vascular resistance was calculated by dividing the chorionic plate arteriovenous pressure difference by the fetal effluent flow rate.

Statistical Analysis. Data were analyzed by ANOVA using the randomized complete block design to account for variation between placentas. Duncan's new multiple range post hoc test was used to determine statistical differences between treatment means. A statistical computer software program was used (SuperANOVA, Abacus Concepts, Inc., Berkeley, CA). Log (X + 1) transformation was used when the variances were not equal. A probability level of P < .05 was considered significant. Data are presented as the mean ± S.E.

	Results

Top Abstract Introduction Materials and Methods Results Discussion References

Study 1: Dose Response of AA-2414. Figure 1 shows a representative tracing of the changes in perfusion pressure in response to peroxide perfusion and peroxide perfusion plus increasing concentrations of AA-2414. Perfusion pressure was substantially increased by the peroxide (Px) alone, but the increase in pressure was inhibited in a dose-response manner by AA-2414. Endothelin (ET; 40 nM), a compound that vasoconstricts independent of thromboxane, was perfused at the end of the experiment to demonstrate that the tissue was viable and capable of vasoconstriction.

View larger version (25K): [in this window] [in a new window]   Fig. 1.   A representative recording of changes in placental perfusion pressure in response to perfusion with Px, and Px + varying doses of AA-2414 (AA), a potent antioxidant and thromboxane receptor blocker. Perfusion with peroxide alone (Px-1) significantly increased the perfusion pressure. Doses of AA-2414 between 1 × 108 mol/l (AA-8) to 1 × 106 mol/l (AA-6) produced a dose-response inhibition of Px-induced increases in perfusion pressure. Doses of AA-2414 between 1 × 106 mol/l to 1 × 104 mol/l (AA-4) completely inhibited Px-induced increases in perfusion pressure. To demonstrate that the effects of AA-2414 were specific and not due to vascular fatigue, ET, 40 nM, a hormone that vasoconstricts independent of thromboxane receptors, was perfused at the end of the experiment. ET resulted in an immediate and substantial increase in perfusion pressure.

8

View larger version (19K): [in this window] [in a new window]   Fig. 2.   Perfusion with peroxide alone (Px-1) significantly increased placental perfusion pressure and vascular resistance. Perfusion of Px in combination with progressively increasing concentrations of AA-2414 (AA) resulted in a dose-dependent decrease in peroxide-induced increases in perfusion pressure and vascular resistance. C represents KRB buffer control perfusions. AAx represents log concentration of AA-2414. a = significantly higher than controls; b = significantly higher than controls but significantly lower than Px-1; c = not significantly different from controls; P < .05. Data represent mean ± S.E.

View larger version (21K): [in this window] [in a new window]   Fig. 3.   To evaluate whether attenuation in perfusion pressure was a specific effect of AA-2414 and not due to vascular fatigue or some other nonspecific factor, we conducted additional experiments (n = 4) in which we repeatedly challenged placental cotyledons with Px without AA-2414. The repeated challenges without AA-2414 resulted in increases in perfusion pressure and vascular resistance comparable with the first challenge demonstrating that the inhibitory effect observed for AA-2414 was specific. a = significantly higher than controls; P < .05. Reprinted from Holles SM et al. (1997) courtesy of Marcel Dekker, Inc.

View larger version (33K): [in this window] [in a new window]   Fig. 4.   Recording of changes in placental perfusion pressure for a placenta repeatedly challenged with Px. Note that the second response in this placenta is greater than the first response and that subsequent challenges with Px result in faster rates of rise in pressure than what occurs with the first challenge. C, Control KRB buffer perfusion. Reprinted from Holles SM et al. (1997) courtesy of Marcel Dekker, Inc.

7

8

View larger version (20K): [in this window] [in a new window]   Fig. 5.   Perfusion with Px alone significantly increased the maternal secretion rates of lipid peroxides and thromboxane. When AA-2414 was perfused in conjunction with Px, it resulted in a dose-dependent inhibition of the Px-induced increases that paralleled the declines in perfusion pressure and vascular resistance shown in Fig. 2. a = significantly higher than controls; c = significantly lower than Px-1, not significantly different from controls; P < .05. Abbreviations are the same as for Figs. 1 and 2.

Study 2: Single Dose of AA-2414. Consistent with the first study, compared with the control perfusion, perfusion with Px significantly increased perfusion pressure (30.8 ± 0.8 versus 60.3 ± 5.6 mm Hg, P < .01) and vascular resistance (11.5 ± 1.3 versus 23.4 ± 1.9 mm Hg · min/ml, respectively, Fig. 6). Perfusion with AA-2414 alone had no effect, but when the placenta was again challenged with Px, AA-2414 completely blocked the ability of peroxide to increase perfusion pressure (25.7 ± 1.1 mm Hg) and vascular resistance (9.4 ± 0.8 mm Hg · min/ml).

View larger version (17K): [in this window] [in a new window]   Fig. 6.   Perfusion with AA-2414 at a concentration of 1 × 105 mol/l completely inhibited the ability of Px to induce vasoconstriction. a = significantly higher than controls; c = not significantly different from controls; P < .05. Abbreviations are the same as for Figs. 1 and 2.

View larger version (12K): [in this window] [in a new window]   Fig. 7.   AA-2414 at a concentration of 1 × 105 mol/l completely inhibited the Px-induced increases in maternal and fetal lipid peroxide secretion. a = significantly higher than controls; c = not significantly different from controls; P < .05. Abbreviations are the same as for Figs. 1 and 2.

View larger version (12K): [in this window] [in a new window]   Fig. 8.   AA-2414 at a concentration of 1 × 105 mol/l significantly inhibited the Px-induced increases in maternal and fetal TXB2 secretion. Perfusion with AA-2414 alone (AA) resulted in declines in the maternal and fetal basal secretion rates of TXB2 that reached statistical significance on the fetal side. a = significantly higher than controls; c = not significantly different from controls; d = significantly lower than controls; P < .05. Abbreviations are the same as for Figs. 1 and 2.

View larger version (13K): [in this window] [in a new window]   Fig. 9.   Px significantly increased fetal secretion of prostacyclin. In contrast to results obtained for lipid peroxides and thromboxane, AA-2414 only partially inhibited ability of Px to increase fetal secretion of prostacyclin. Although the fetal concentration for AA + Px was lower than peroxide alone (Px-1), it was nevertheless significantly higher than controls and AA alone. a = significantly higher than controls; b = significantly higher than controls but significantly lower than Px-1; c = not significantly different from controls; P < .05. Abbreviations are the same as for Figs. 1 and 2.

5

	Discussion

Top Abstract Introduction Materials and Methods Results Discussion References

This study demonstrated that AA-2414, a compound that is both a potent antioxidant and thromboxane receptor antagonist, caused dose-response inhibition of peroxide-induced vasoconstriction in the isolated perfused human placental cotyledon. The inhibition was a specific effect of AA-2414 because the isolated placental cotyledon will repeatedly vasoconstrict to repeated challenges with Px.

AA-2414 also inhibited in a dose-response manner the maternal secretion rates of lipid peroxides and thromboxane. At a dose of AA-2414 of 1 × 10⁵ mol/l that approximates plasma levels achieved in clinical studies (Hussein et al., 1994), there was essentially complete inhibition of peroxide-induced vasoconstriction, as well as inhibition of peroxide-stimulated increases in the maternal secretion rates of lipid peroxides and thromboxane. Interestingly, AA-2414 only partially inhibited peroxide-stimulated increases in the fetal secretion rate of prostacyclin. The inhibitory effects of AA-2414 persisted after perfusion of the drug was discontinued.

The maternal and fetal secretion rates of both lipid peroxides and thromboxane were highly correlated with changes in perfusion pressure, suggesting that peroxide-induced changes in vascular tone were dependent on its ability to stimulate both lipid peroxidation and synthesis of thromboxane. These data also suggest that the inhibitory effects of AA-2414 were manifest both through its antioxidant effect and its thromboxane receptor blocking effect.

When AA-2414 was perfused alone for 20 min, basal maternal and fetal secretion rates of lipid peroxides and thromboxane started to decline with the decline in fetal thromboxane reaching statistical significance. A 20-min perfusion period is a rather short time to test the effects of AA-2414 by itself, and was used primarily as a pretreatment for the combined perfusion of AA-2414 plus peroxide. It is impressive that within such a short time AA-2414 was able to significantly inhibit basal secretion of thromboxane on the fetal side of the placenta and it is possible that the declines in the secretion rates of lipid peroxides and maternal thromboxane would have reached statistical significance if the duration of the perfusion with AA-2414 would have been longer.

The ability of AA-2414 to block peroxide-induced secretion of thromboxane is a rather interesting finding with regard to the mechanism of action of AA-2414. In other studies using cyclooxygenase purified from bovine vesicular glands, AA-2414 had only weak inhibitory effects on cyclooxygenase activity. In our study, it is likely that the inhibitory effects on thromboxane secretion were indirect and related to the inhibition of lipid peroxidation. The amount of peroxide tone in a tissue is an important factor in determining the activity of cyclooxygenase (Hemler et al., 1979; Kulmacz and Lands, 1983). When the level of peroxide tone increases, the activity of cyclooxygenase increases. When the level of peroxide tone decreases, the activity of cyclooxygenase decreases. In our study, AA-2414 completely inhibited the ability of exogenous peroxide to stimulate an increase in lipid peroxide secretion, indicating inhibition of endogenous lipid peroxide formation. Therefore, in the presence of AA-2414, the level of endogenous peroxide tone was not increased by exogenous peroxide and the secretion rates of thromboxane did not increase over control.

AA-2414 differentially affected the placental secretion rates of eicosanoids. AA-2414 completely blocked the ability of peroxide to increase the secretion of thromboxane above control, but it only partially inhibited the ability of peroxide to increase prostacyclin secretion. Although the fetal prostacyclin secretion rate for AA-2414 plus peroxide was significantly lower than for peroxide alone, it was significantly higher than control. A lesser blocking effect on the fetal side would be a favorable effect because the increase in prostacyclin would promote vasodilatation of the placental vasculature, whereas the vasoconstrictive effects of thromboxane would be blocked by the thromboxane receptor antagonistic properties of AA-2414.

The reason AA-2414 differentially affects placental eicosanoid secretion is not known, but it may relate to the compartmentalization of thromboxane and prostacyclin synthesis within the human placenta. Thromboxane is primarily synthesized by the trophoblast cells on the maternal side of the placenta, whereas prostacyclin is primarily synthesized by the endothelial cells on the fetal side of the placenta (Thorp et al., 1988; Nelson and Walsh, 1989; Shellhaas et al., 1997). Lipid peroxides are also primarily synthesized by trophoblast cells as opposed to the vasculature (Walsh and Wang, 1995), so inhibition of lipid peroxidation by AA-2414 would conceivably have a greater impact on thromboxane synthesis in the trophoblast cells than on prostacyclin synthesis in the endothelial cells because peroxide-induced activity of cyclooxygenase would be inhibited to a greater extent in the trophoblast cells than in the endothelial cells.

AA-2414 clearly prevents peroxide-induced vasoconstriction and placental secretion of lipid peroxides demonstrating both its antioxidant and thromboxane receptor blocking effects. It also inhibits peroxide-induced increases in thromboxane secretion. Given the abnormal increases of lipid peroxides and thromboxane in preeclampsia, AA-2414 has pharmacologic properties that would make it a candidate to consider for the treatment of women with preeclampsia. Its actions as a thromboxane receptor blocker should prevent thromboxane-induced vasoconstriction and platelet aggregation, and its action as an antioxidant should decrease lipid peroxide production.