Reprod. Fertil. Dev., 1991, 3, 475-81
Regulation of Fetal Vascular Tone in the Human Placenta
W. A . W. WaltersA and A . L. A . ~ o u r a ~ ~ Discipline of Reproductive Medicine, Faculty of Medicine, University of Newcastle, Shortland, NSW 2308, Australia. Department of Pharmacology, Monash University, Clayton, Vic. 3168, Australia. Present address: Discipline of Reproductive Medicine, Faculty of Medicine, University of Newcastle, Shortland, NSW 2308, Australia.
A
Abstract Using an in vitro placental lobule perfusion technique, the human fetal placental vasculature has been found to respond vigorously with high sensitivity to various vasoconstrictor substances, including angiotensin 11, endothelins 1 and 3, prostaglandins Fz,, Ez and D2 and the thromboxane A2 agonist U46619. Thromboxane A2 receptors mediating vasoconstriction have been characterized in fetal placental vessels and appear to be identical to those on human platelets and pulmonary blood vessels. Although the isolated fetal placental vessels are largely unresponsive to exogenous vasodilatory stimuli, when preconstricted they respond by vasodilatation to several vasodilator substances, including arachidonate, prostacyclin, prostaglandin El, theophylline and nitroglycerine. The resistance offered to flow in vitro by the villous vasculature is therefore low, as it is in vivo. Both intrinsic and extrinsic mechanisms probably operate to cause relaxation of the vascular smooth muscle with the vasodilatatory effects of locally released autacoids dominating the effects of those having vasoconstrictor actions.
Introduction The umbilical cord and placenta are devoid of nerves (Fox and Khong 1990) so that only non-neural mechanisms can influence vascular tone in these tissues. Humoral factors originating in the fetoplacental unit or maternal tissues, intrinsic myogenic mechanisms, changes in blood rheology, and extrinsic physical factors may be involved; of these, the humoral factors are likely to be of prime importance. Fetal umbilical-placental vascular resistance is low (Stuart et a(. 1980; Trudinger et a]. 1985). Thus, during normal pregnancy vasodilator influences appear to predominate. However, in various pathological states associated with pregnancy, changes in the balance of opposing vasoconstrictor and vasodilator humoral autacoids may well make a significant contribution to the observed increased placental resistance to blood flow (Trudinger et al. 1985). Several substances that can directly or indirectly affect vascular tone have been detected in the umbilical circulation: noradrenaline and adrenaline (Falconer and Lake 1982), histamine (Mitchell and Porter 1970), 5-hydroxytryptamine (Jones and Rowsell 1973), acetylcholine (Boura et al. 1986), prostaglandins F2, (Craft et al. 1973), E2 (Mackenzie et al. 1980), D2 and I2 and thromboxane A2 (Tuvemo 1980), leukotrienes (Piper and Levine 1987), bradykinin (Melmon et al. 1968), angiotensin I1 (Broughton Pipkin and Symonds 1977), endogenous opioids (Goland et al. 1981) and endothelin (Haegerstrand et al. 1989). 1031-3613/91/040475$05.00
W. A. W. Walters and A. L. A. Boura
The normal human placenta exerts profound effects on the biological activities of some substrates appearing in the umbilical circulation, although not on others. For example, 70-90% of 5-hydroxytryptamine activity is destroyed in the placenta and 98% of bradykinin is inactivated during a single passage through the human placental lobule (Prentice et al. 1987), and the placenta rapidly converts angiotensin I to the active form angiotensin I1 (Maguire et a/. 1985) and also des-asp-angiotensin I to angiotensin I11 (Prentice et al. 1987). Prostaglandins E2 and FZa, however, pass through the fetal placental vasculature without being altered (Prentice et a/. 1987). Thus, the placenta may protect the fetus and umbilical vein from actions of some potent agents such as 5-hydroxytryptamine and bradykinin while either increasing or not affecting the activities of others. The origins of the humoral factors may vary, although the placenta itself contains many of the enzymes necessary for their biosynthesis, e.g. mechanisms for acetylcholine (Sastry and Sadavongvivad 1979), angiotensin I1 (Lenz et al. 1989) and eicosanoid production (Gude et a/. 1990b). Furthermore, the placenta contains many enzymes known to destroy the biological activities of autacoids and neurohormones, e.g. histaminase, catechol O-methylamine-transferase, monoamine oxidase (Barnea et a/. 1986) and prostaglandin-15-hydroxydehydrogenase, (Thaler-Dao et al. 1974) terminating the action of histamine, catecholamines, 5-hydroxytryptamine and prostaglandins E2 and F2a, respectively. Methods Most investigations into the role of autacoids in the control of vascular tone in the human placenta and umbilical cord have been conducted in vitro because of the technical and ethical difficulties associated with in vivo studies. Various methods have been used to investigate the control of vascular tone in the fetal extracorporeal circulatory system: they include perfusion of the umbilical artery (Eltherington et al. 1968), umbilical vein (Sandford et al. 1981), the entire placenta (Gautieri and Ciuchta 1962) or a single placental lobule (Panigel 1962; Penfold et al. 1981) and measurement of smooth muscle tone in helical strips of umbilical artery or vein (Tuvemo et al. 1976; Tulenko 1981) or in segments of umbilical artery or vein and placenta (Bjdro and Stray-Pedersen 1986; Maigaard 1987). Perfusion of a single placental lobule with Krebs solution has evolved as the most satisfactory method for investigating placental vascular tone and is now used extensively (Penfold et al. 1981; Abramovich et al. 1983; Boura et al. 1986; Dancis et al. 1986; Glance et al. 1986; Howard et al. 1987; Soares de Moura 1987).
Placental Lobule Perfusion Our method of placental lobule perfusion (Mak et al. 1984) is a modification of that described by Penfold et al. (1981), based upon the original work of Panigel(1962). The fetal membranes are removed from the placenta immediately after delivery. A fetal artery and vein pair of vessels is tied off close to the point of insertion of the umbilical cord, the artery distal to its ligation is cut, and a plastic cannula (external diameter 1.34 mm) is ihserted intravascularly to a point at which the artery disappears below the surface of the chorionic plate. A cannula is tied into the arery and connected to a perfusion pump. The lobule is then perfused with Krebs solution (pH 7.3), maintained at 37.5OC and equilibrated with 95% O2 and 5% COz. The vein of the pair of vessels is cut at a convenient site distal to the ligature to allow blood and perfusate to escape and is then cannulated with plastic tubing (external diameter 2.2 mm) which is tied in place. The flow rate of perfusing fluid is maintained at 10 mL min-'. The placenta is immersed in normal saline at 37.5'C, the maternal component being flushed out by inserting two or three plastic tubes into spiral arteries in the basal plate and pumping perfusion fluid through them at 8 mL min-I. Changes in placental resistance to flow are measured as changes in perfusion pressure, obtained from a transducer (Statham P23AC) connected to the fetal artery cannula and displayed on a chart recorder. Initially there is a pressure of about 100 mm Hg, probably caused by release of autacoids in the blood as it is washed out, but the pressure subsequently declines to 20-60 mm Hg and remains stable. Only after this are drugs administered. The perfusion apparatus is shown in Fig. 1, the flow of perfusing fluid being from right to left. A roller pump delivers Krebs solution bubbled with carbogen from a container at room temperature, whence it is passed successively through a flow dampener, to attenuate the changes in pressure due
Fetal Vascular Tone in the Human Placenta
chart recorder flow meter
m
-
37.5"Cwater bath
transducer 7, / h b u b b l e trael
I
flow meter
-
placenta with cannulated lobule
flow dampener
amniotic Krebs medium
I
constant infusion pump Fig. 1. Apparatus used for perfusion of a single placental lobule (Gude 1988).
to the pulsatile action of the roller pump, a flow meter and a heating coil to bring it to 37.5OC. A bubbletrap is positioned beyond the heating coil to remove gas bubbles which come out of solution when the Krebs solution is heated. A graduated cylinder is used for measurement of the perfusate. A hypodermic needle, to which is attached a short length of polyethylene tubing, is inserted through the circuit tubing just proximal to the perfusion chamber, for administration of drugs by a Sage constant infusion pump until the subsequent response of the tissue reaches a plateau. Rates of drug administration vary between 0.013 and 0.47 mL min-'. Using this methodology during the last decade, our group has investigated the actions of a number of autacoids that might be involved in the control of fetal vascular tone in the human placenta and umbilical cord (Mak et al. 1984; Boura et al. 1986; Prentice et al. 1987). The results of our studies, reported elsewhere, will be summarized here along with those of other workers to provide an overview of current knowledge in the field.
Discussion
The Human Fetal Extracorporeal Circulation All vessels of the human fetal extracorporeal circulation are capable of responding to a variety of both vasodilator and vasoconstrictor substances in vifro. The influence of vasodilator mechanisms predominates during normal pregnancy as indicated by the high umbilical bloodflow rate in vivo (Assali et al. 1960) and the low resistance of the vessels when they are perfused in vitro under pressure and flow conditions which approach the physiological norm (Walsh et at. 1974). Furthermore, in vivo Doppler ultrasound recordings during normal pregnancy demonstrate umbilical artery wave forms indicative of a low placental resistance to flow (Trudinger et al. 1985). Vasodilatation Because of their intrinsically low vascular tone, the fetal placental vessels in vifro are virtually unresponsive to exogenous vasodilatory stimuli unless preconstricted. Thus, the placental villous vessels, first constricted by infusion of the thromboxane A2 agonist U46619, respond by vasodilatation to nitroglycerine (Boura and Walters, unpublished observations), arachidonate and prostacyclin (Mak et al. 1984). Histamine (Maguire et al. 1985), isoprenaline, terbutaline (Soares de Moura 1987), vasoactive intestinal peptide (Maigaard 1987) and theophylline (Nielsen-Kudsk 1985) also cause vasodilatation after placental vascular preconstriction. Although acetylcholine is produced in large amounts by the placenta, there is little evidence to suggest that it contributes to the control of vascular tone and it has not been found to affect vasoconstrictor responses (Boura et a/. 1986). Acetylcholine may be more important in mechanisms con-
W. A. W. Walters and A. L. A. Boura
trolling transplacental passage of amino acids and electrolytes (Fant and Harbison 1981) and release of placental hormones (Sastry et al. 1981). The low tone in the smooth muscle of the fetal extracorporeal vasculature is probably maintained by the continuous output of vasodilator autacoids, particularly prostacyclin and perhaps the endothelial-cell-derived relaxing factor (EDRF) or nitric oxide (Vanhoutte 1988; Gude et al. 1990a).
Vasoconstriction The fetal placental vessels respond vigorously to vasoconstrictor stimuli, although there are substantial differences between various vessels in responsiveness to autacoids (Mak et al. 1984; Maigaard et al. 1986). Furthermore, large differences are found in the responses of the villous vessels to the various autacoids, which vary in their vasoconstrictor potencies as well as in the magnitude of the maximum responses they are capable of causing (Mak et al. 1984). Thus, in vitro angiotensin I1 (Abramovich et al. 1983), endothelins 1 and 3 (Boura and Walters, unpublished observations) and the thromboxane A2 agonist U46619 (Mak et al. 1984) show high potencies, causing effects in concentrations of 1 pmol to 100 pmol L-', but U46619, when perfused at a concentration (100 pmol L-') causing a maximal effect, is capable of producing intense vasoconstriction. The mean maximum increase in pressure to U46619, when perfused under constant flow conditions, was approximately 200 mm Hg, compared with an increase of approximately 35 mm Hg in the presence of a high concentration (30 pmol L-') of angiotensin I1 (Mak et al. 1984). Prostaglandins F2,, E2 and D2 are capable of causing constriction of the villous vessels of a magnitude comparable to that seen during thromboxane A2 receptor stimulation with U46619 but are at least 100 (F2,) and 3000 (E2 and D2) times less potent (Abramovich et al. 1984; Mak et al. 1984; Howard et al. 1987). There is evidence that responses to many of these eicosanoids can be affected by concomitant output of substances arising in tissues of the vessel walls including thromboxane A2, prostacyclin and EDRF in the placenta and umbilical cord (Mak et al. 1984; Bjdro 1986; Walters and Boura, unpublished observations). Thromboxane A2 receptors have been identified in the vasculature of both umbilical cord and placenta utilizing the thromboxane A2 agonist U46619, an analogue of prostaglandin H2, and its specific competitive antagonist AH22921 (Boura et al. 1986). In the umbilical vein and placental villous vessels, these receptors have been characterized and appear to be identical to those on human platelets and pulmonary blood vessels (Boura et al. 1986). Stimulation of umbilical-placental thromboxane A2 receptors in the sheep (Trudinger et al. 1989) causes analogous changes in systolic-diastolic flow patterns and vascular resistance to those seen in preeclampsia. Thus, during placental ischaemia such as occurs during maternal hypoxia (Las Heras et al. 1985) and preeclampsia (Chesley 1978), a vicious circle of intense vasospasm, reduction of blood flow and intravascular thrombosis could be established, with platelet aggregation leading to further thromboxane A2 release from platelets and progressively increasing vasospasm.
Conclusions
It can be concluded from results of work mentioned in this review that, under normal conditions, the human fetal extracorporeal vasculature is in a predominantly dilated state. It is, however, sensitive to a wide range of vasoconstrictor autacoids, some of which can cause intense vasoconstriction. It is likely that a balance is maintained between the actions of opposing vasodilator and vasoconstrictor autacoids, which may be readily tipped towards vasoconstriction by alterations in placental autacoid production, as demonstrated by the reduced prostacyclin and increased thromboxane A2 output from the placenta in preeclampsia (Remuzzi et al. 1980; Makila et al. 1984; Walsh et al. 1985).
Fetal Vascular Tone in the Human Placenta
Acknowledgment
Neil Gude has made a major contribution to much of the work in this field conducted in our laboratories, and reported in this article and elsewhere. References Abramovich, D. R., Page, K. R., and Wright, F. (1983). Effect of angiotensin I1 and 5-hydroxytryptamine on the vessels of the human foetal cotyledon. Br. J. Pharmacol. 79, 53-6. Abramovich, D. R., Page, K. R., and Parkin, A. M. L. (1984). The effect of prostaglandin D2 on the blood vessels of the perfused isolated cotyledon and the human placenta. Br. J. Pharmacol. 81, 19-21. Assali, N. S., Rauramo, L., and Peltonen, T. (1960). Measurement of uterine blood flow and uterine metabolism. Am. J. Obstet. Gynecol. 79, 86-98. Barnea, E. R., Fakih, H., and Oelsner, G., Walner, S., DeCherney, A. H., and Naftolin, F. (1986). Effect of antihypertensive drugs on catechol-0-methyltransferase and monoamine oxidase activity in human term placental explants. Gynecol. Obstet. Invest. 21, 124-30. Bjdro, K. (1986). Prostacyclin and thromboxane formation in human umbilical arteries following stimulation with vasoactive autacoids. Prostaglandins 31, 699-714. Bjdro, S., and Stray-Pedersen, S. (1986). Effects of vasoactive autacoids on different segments of human umbilicoplacental vessels. Gynecol. Obstet. Invest. 22, 1-6. Boura, A. L. A., Gude, N. M., King, R. G., and Walters, W. A. W. (1986). Acetylcholine output and fetal vascular resistance of human perfused placental cotyleda. Br. J. Pharmacol. 88, 301-6. Broughton Pipkin, F., and Symonds, E. M. (1977). Factors affecting angiotensin I1 concentrations in the human infant at birth. Clin. Sci. Mol. Med. 52, 449-56. Chesley, L. C. (1978). 'Hypertensive Disorders in Pregnancy.' pp. 455-66. (Appleton-Century-Crafts: New York.) Craft, I. L., Scrivener, R., and Dewhurst, C. J. (1973). Prostaglandin F2, levels in the maternal and fetal circulations in late pregnancy. J. Obstet. Gynaecol. Br. Commonw. 80, 616-18. Dancis, J . , Lehanka, J., Levitz, M., and Schneider, H. (1986). Establishment of gradients of riboflavin, L-lysine and a-aminoisobutyric acid across the perfused human placenta. J. Reprod. Med. 31, 293-6. Eltherington, L. G., Stoff, J., Hughes, T., and Melmon, K. L. (1968). Constriction of human umbilical arteries. Interaction between oxygen and bradykinin. Circ. Res. 22, 747-52. Falconer, A. D., and Lake, D. M. (1982). Circumstances influencing umbilical-cord plasmacatecholamines at delivery. Br. J. Obstet. Gynaecol. 89, 44-9. Fant, M. E., and Harbison, R. D. (1981). Syncytiotrophoblast membrane vesicles: a model for examining the human placental cholinergic system. Teratology 24, 187-99. Fox, S. B., and Khong, T. Y. (1990). Lack of innervation of human umbilical cord. An immunohistological and histochemical study. Placenta 11, 59-62. Gautieri, R. F., and Ciuchta, H. P. (1962). Effect of certain drugs on perfused human placenta: I. Narcotic analgesics, serotonin and relaxin. J. Pharm. Sci. 51, 55-8. Glance, D. G., Elder, M. G., and Myatt, L. (1986). The actions of prostaglandins and their interactions with angiotensin I1 in the isolated perfused human placental cotyledon. Br. J. Obstet. Gynaecol. 93, 488-94. Goland, R. S., Wardlaw, S. L., Stark, R. I., and Frantz, A. G. (1981). Human plasma @-endorphin during pregnancy, labor and delivery. J. Clin. Endocrinol. Metab. 52, 74-8. Gude, N. M. (1988). An investigation into the mechanisms controlling vascular tone of the fetal vessels of the human isolated perfused placenta. M.Sc. Thesis, Monash University, Melbourne. Gude, N. M., King, R. G., and Brennecke, S. P. (1990~).Role of endothelium-derived nitric oxide in maintenance of low fetal vascular resistance in placenta. Lancet 336, 1589-90. Gude, N. M., Rice, G. E., King, R. G., Boura, A. L. A., and Brennecke, S. P. (1990b). Analysis of the responses of the fetal vessels of human perfused placental lobules to acute infusions of arachidonic acid. Reprod. Fertil. Dev. 2, 591-6. Haegerstrand, A., Hemsen, A,, Gillis, C., Larsson, O., and Lundberg, J. M. (1989). Endothelin: presence in human umbilical vessels, high levels in fetal blood and potent constrictor effect. Acta Physiol. Scand. 137, 541-2. Howard, R. B., Hosokawa, T., and Maguire, M. H. (1987). Hypoxia-induced fetoplacental vasoconstriction in perfused human placental cotyledons. A m . J. Obstet. Gynecol. 157, 1261-6.
W. A. W. Walters and A. L. A. Boura
Jones, J . B., and Rowsell, A. (1973). Fetal 5-hydroxytryptamine levels in late pregnancy. J. Obstet. Gynaecol. Br. Commonw. 80, 687-9. Las Heras, J., Baskerville, J. C., Harding, P. G. R., and Daria Haust, M. (1985). Morphometric studies of fetal placental stem arteries in hypertensive disorders ('toxemia') of pregnancy. Placenta 6, 217-28. Lenz, T., Sealey, J. E., August, P., James, G. D., and Laragh, J. H. (1989). Tissue levels of active and total renin, angiotensinogen, human chorionic gonadotrophin, estradiol, and progesterone in human placentas from different methods of delivery. J. Clin. Endocrinol. Metab. 69, 31-7. Mackenzie, I. Z., MacLean, D. A., and Mitchell, M. D. (1980). Prostaglandins in the human fetal circulation in mid-trimester and term pregnancy. Prostaglandins 20, 649-54. Maguire, M. H., Howard, R. B., and Hosokawa, T. (1985). Autacoid receptors in the human fetoplacental vasculature. Contrib. Gynecol. Obstet. 13, 170-2. Maigaard, S. (1987). Contraction and relaxation of human uterine and placental smooth muscle. Endogenous control and calcium activation mechanisms. Acta Obstet. Gynecol. Scand. Suppl. 143, 9-39. Maigaard, S., Forman, A., and Andersson, K.-E. (1986). Differential effects of angiotensin, vasopressin and oxytocin on various smooth muscle tissues within the human uteroplacental unit. Acta Physiol. Scand. 128, 23-31. Mak, K. K.-W., Gude, N. M., Walters, W. A. W., and Boura, A. L. A. (1984). Effects of vasoactive autacoids on the human umbilical-fetal placental vasculature. Br. J. Obstet. Gynaecol. 91, 99-106. Makila, U.-M., Viinikka, L., and Ylikorkala, 0. (1984). Increased thromboxane Az production but normal prostacyclin by the placenta in hypertensive pregnancies. Prostaglandins 27, 87-95. Melmon, K. L., Cline, M. J., Hughes, T., and Nies, A. S. (1968). Kinins: possible mediators of neonatal circulatory changes in man. J. Clin. Invest. 47, 1295-302. Mitchell, R . G., and Porter, J. F. (1970). Histamine and granulocytes in the umbilical cord blood of infants at birth. Br. J. Pharmacol. 40, 310-16. Nielsen-Kudsk, J. E. (1985). Enprofylline and theophylline on small human placental arteries: in vitro effects and mode of action. Acta Pharm. Toxicol. 56, 176-82. Panigel, M. (1962). Placental perfusion experiments. Am. J. Obstet. Gynecol. 84, 1664-83. Penfold, P., Drury, L., Simmonds, R., and Hytten, F. E. (1981). Studies of a single placental cotyledon in vitro: I. The preparation and its viability. Placenta 2, 149-54. Piper, P . J., and Levene, S. (1987). Human umbilical arteries and veins: generation of leukotrienes and response to exogenous leukotrienes. Biol. Neonate 52, 9-15. Prentice, D. A., Boura, A. L. A., Gude, N. M., Walters, W. A. W., and King, R. G. (1987). Changes in the biological activity of autacoids during passage through the human perfused fetoplacental lobule. Eur. J. Pharmacol. 141, 79-86. Remuzzi, G., Marchesi, D., Zoja, C., Muratore, D., Mecca, G., Misiani, R., Rossi E., Barbato, M., Capetta, P., Donati, M. B., and De Gaetano, G. (1980). Reduced umbilical and placental vascular prostacyclin in severe pre-eclampsia. Prostaglandins 20, 105-10. Sandford, A. S. C., Walters, W. A. W., and Boura, A. L. A. (1981). On the mechanism underlying closure of the human umbilical vein induced by cooling. In 'Prostaglandins and Thromboxanes in the Cardiovascular System and in Gynaecology and Obstetrics'. (Ed. W. Forster.) pp. 415-19. (Fischer: Jena.) Sastry, B. V. R., and Sadavongvivad, C. (1979). Cholinergic systems in non-nervous tissues. Pharmacol. Rev. 30, 65-132. Sastry, B. V. R., Tayeb, 0. S., Barnwell, S. L., Janson, V. E., and Owens, L. K. (1981). Peptides from human placenta: methionine enkephalin and substance P. Placenta, Suppl. 3, 327-37. Soares de Moura, R. (1987). Effect of prostacyclin on the perfusion pressure and on the vasoconstrictor response of angiotensin I1 in the human isolated foetal placental circulation. Br. J. Clin. Pharmacol. 23, 765-8. Stuart, B., Drum, J., Fitzgerald, D. E., and Duignan, N. M. (1980). Fetal blood velocity waveforms in normal pregnancy. Br. J. Obstet. Gynaecol. 87, 780-5. Thaler-Dao, H., Saintot, M., Baudin, G., Descomps, B., and Crastes de Paulet, A. (1974). Purification of the human placental 15-hydroxy prostaglandin dehydrogenase: properties of the purified enzyme. FEBS Lett. 48, 204-8. Trudinger, B. J., Giles, W. B., Cook, C. M., Bombardieri, J., and Collins, L. (1985). Fetal umbilical artery flow velocity waveforms and placental resistance: clinical significance. Br. J. Obstet. Gynaecol. 92, 23-30.
Fetal Vascular Tone in the Human Placenta
Trudinger, B. J., Connelly, A. J., Giles, W. B., Hales, J. R., and Wilcox, G. R. (1989). The effects of prostacyclin and thromboxane analogue (U46619) on the fetal circulation and umbilical flow velocity waveforms. J. Dev. Physiol. 11, 179-84. Tulenko, T. N. (1981). The actions of prostaglandins and cyclo-oxygenase inhibition on the resistance vessels supplying the human fetal placenta. Prostaglandins 21, 1033-43. Tuvemo, T. (1980). Role of prostaglandins, prostacyclin, and thromboxanes in the control of the umbilical-placental circulation. Semin. Perinatol. NY 4, 91-5. Tuvemo, T., Strandberg, K., Hamberg, M., and Samuelsson, B. (1976). Maintenance of tone of the human umbilical artery by prostaglandin and thromboxane formation. In 'Advances in Prostaglandin and Thromboxane Research'. Vol. 1. (Eds B. Samuelsson and R. Paoletti.) pp. 425-8. (Raven: New York.) Vanhoutte, P. M. (1988). Relaxing and contracting factors. In 'Biological and Clinical Research'. (Humana: Clifton, NJ.) Walsh, S. W. (1985). Pre-eclampsia: an imbalance in placental prostacyclin and thromboxane production. Am. J. Obstet. Gynecol. 152, 335-40. Walsh, S. Z., Meyer, W. W., and Lind, J. (1974). The human fetal and neonatal circulation. (Ed. C. Thomas.) pp. 1-35 1. (Springfield: Ill.)
Manuscript received 4 November 1990; revised and accepted 4 March 1991
Regulation of Fetal Vascular Tone in the Human Placenta
W. A . W. WaltersA and A . L. A . ~ o u r a ~ ~ Discipline of Reproductive Medicine, Faculty of Medicine, University of Newcastle, Shortland, NSW 2308, Australia. Department of Pharmacology, Monash University, Clayton, Vic. 3168, Australia. Present address: Discipline of Reproductive Medicine, Faculty of Medicine, University of Newcastle, Shortland, NSW 2308, Australia.
A
Abstract Using an in vitro placental lobule perfusion technique, the human fetal placental vasculature has been found to respond vigorously with high sensitivity to various vasoconstrictor substances, including angiotensin 11, endothelins 1 and 3, prostaglandins Fz,, Ez and D2 and the thromboxane A2 agonist U46619. Thromboxane A2 receptors mediating vasoconstriction have been characterized in fetal placental vessels and appear to be identical to those on human platelets and pulmonary blood vessels. Although the isolated fetal placental vessels are largely unresponsive to exogenous vasodilatory stimuli, when preconstricted they respond by vasodilatation to several vasodilator substances, including arachidonate, prostacyclin, prostaglandin El, theophylline and nitroglycerine. The resistance offered to flow in vitro by the villous vasculature is therefore low, as it is in vivo. Both intrinsic and extrinsic mechanisms probably operate to cause relaxation of the vascular smooth muscle with the vasodilatatory effects of locally released autacoids dominating the effects of those having vasoconstrictor actions.
Introduction The umbilical cord and placenta are devoid of nerves (Fox and Khong 1990) so that only non-neural mechanisms can influence vascular tone in these tissues. Humoral factors originating in the fetoplacental unit or maternal tissues, intrinsic myogenic mechanisms, changes in blood rheology, and extrinsic physical factors may be involved; of these, the humoral factors are likely to be of prime importance. Fetal umbilical-placental vascular resistance is low (Stuart et a(. 1980; Trudinger et a]. 1985). Thus, during normal pregnancy vasodilator influences appear to predominate. However, in various pathological states associated with pregnancy, changes in the balance of opposing vasoconstrictor and vasodilator humoral autacoids may well make a significant contribution to the observed increased placental resistance to blood flow (Trudinger et al. 1985). Several substances that can directly or indirectly affect vascular tone have been detected in the umbilical circulation: noradrenaline and adrenaline (Falconer and Lake 1982), histamine (Mitchell and Porter 1970), 5-hydroxytryptamine (Jones and Rowsell 1973), acetylcholine (Boura et al. 1986), prostaglandins F2, (Craft et al. 1973), E2 (Mackenzie et al. 1980), D2 and I2 and thromboxane A2 (Tuvemo 1980), leukotrienes (Piper and Levine 1987), bradykinin (Melmon et al. 1968), angiotensin I1 (Broughton Pipkin and Symonds 1977), endogenous opioids (Goland et al. 1981) and endothelin (Haegerstrand et al. 1989). 1031-3613/91/040475$05.00
W. A. W. Walters and A. L. A. Boura
The normal human placenta exerts profound effects on the biological activities of some substrates appearing in the umbilical circulation, although not on others. For example, 70-90% of 5-hydroxytryptamine activity is destroyed in the placenta and 98% of bradykinin is inactivated during a single passage through the human placental lobule (Prentice et al. 1987), and the placenta rapidly converts angiotensin I to the active form angiotensin I1 (Maguire et a/. 1985) and also des-asp-angiotensin I to angiotensin I11 (Prentice et al. 1987). Prostaglandins E2 and FZa, however, pass through the fetal placental vasculature without being altered (Prentice et a/. 1987). Thus, the placenta may protect the fetus and umbilical vein from actions of some potent agents such as 5-hydroxytryptamine and bradykinin while either increasing or not affecting the activities of others. The origins of the humoral factors may vary, although the placenta itself contains many of the enzymes necessary for their biosynthesis, e.g. mechanisms for acetylcholine (Sastry and Sadavongvivad 1979), angiotensin I1 (Lenz et al. 1989) and eicosanoid production (Gude et a/. 1990b). Furthermore, the placenta contains many enzymes known to destroy the biological activities of autacoids and neurohormones, e.g. histaminase, catechol O-methylamine-transferase, monoamine oxidase (Barnea et a/. 1986) and prostaglandin-15-hydroxydehydrogenase, (Thaler-Dao et al. 1974) terminating the action of histamine, catecholamines, 5-hydroxytryptamine and prostaglandins E2 and F2a, respectively. Methods Most investigations into the role of autacoids in the control of vascular tone in the human placenta and umbilical cord have been conducted in vitro because of the technical and ethical difficulties associated with in vivo studies. Various methods have been used to investigate the control of vascular tone in the fetal extracorporeal circulatory system: they include perfusion of the umbilical artery (Eltherington et al. 1968), umbilical vein (Sandford et al. 1981), the entire placenta (Gautieri and Ciuchta 1962) or a single placental lobule (Panigel 1962; Penfold et al. 1981) and measurement of smooth muscle tone in helical strips of umbilical artery or vein (Tuvemo et al. 1976; Tulenko 1981) or in segments of umbilical artery or vein and placenta (Bjdro and Stray-Pedersen 1986; Maigaard 1987). Perfusion of a single placental lobule with Krebs solution has evolved as the most satisfactory method for investigating placental vascular tone and is now used extensively (Penfold et al. 1981; Abramovich et al. 1983; Boura et al. 1986; Dancis et al. 1986; Glance et al. 1986; Howard et al. 1987; Soares de Moura 1987).
Placental Lobule Perfusion Our method of placental lobule perfusion (Mak et al. 1984) is a modification of that described by Penfold et al. (1981), based upon the original work of Panigel(1962). The fetal membranes are removed from the placenta immediately after delivery. A fetal artery and vein pair of vessels is tied off close to the point of insertion of the umbilical cord, the artery distal to its ligation is cut, and a plastic cannula (external diameter 1.34 mm) is ihserted intravascularly to a point at which the artery disappears below the surface of the chorionic plate. A cannula is tied into the arery and connected to a perfusion pump. The lobule is then perfused with Krebs solution (pH 7.3), maintained at 37.5OC and equilibrated with 95% O2 and 5% COz. The vein of the pair of vessels is cut at a convenient site distal to the ligature to allow blood and perfusate to escape and is then cannulated with plastic tubing (external diameter 2.2 mm) which is tied in place. The flow rate of perfusing fluid is maintained at 10 mL min-'. The placenta is immersed in normal saline at 37.5'C, the maternal component being flushed out by inserting two or three plastic tubes into spiral arteries in the basal plate and pumping perfusion fluid through them at 8 mL min-I. Changes in placental resistance to flow are measured as changes in perfusion pressure, obtained from a transducer (Statham P23AC) connected to the fetal artery cannula and displayed on a chart recorder. Initially there is a pressure of about 100 mm Hg, probably caused by release of autacoids in the blood as it is washed out, but the pressure subsequently declines to 20-60 mm Hg and remains stable. Only after this are drugs administered. The perfusion apparatus is shown in Fig. 1, the flow of perfusing fluid being from right to left. A roller pump delivers Krebs solution bubbled with carbogen from a container at room temperature, whence it is passed successively through a flow dampener, to attenuate the changes in pressure due
Fetal Vascular Tone in the Human Placenta
chart recorder flow meter
m
-
37.5"Cwater bath
transducer 7, / h b u b b l e trael
I
flow meter
-
placenta with cannulated lobule
flow dampener
amniotic Krebs medium
I
constant infusion pump Fig. 1. Apparatus used for perfusion of a single placental lobule (Gude 1988).
to the pulsatile action of the roller pump, a flow meter and a heating coil to bring it to 37.5OC. A bubbletrap is positioned beyond the heating coil to remove gas bubbles which come out of solution when the Krebs solution is heated. A graduated cylinder is used for measurement of the perfusate. A hypodermic needle, to which is attached a short length of polyethylene tubing, is inserted through the circuit tubing just proximal to the perfusion chamber, for administration of drugs by a Sage constant infusion pump until the subsequent response of the tissue reaches a plateau. Rates of drug administration vary between 0.013 and 0.47 mL min-'. Using this methodology during the last decade, our group has investigated the actions of a number of autacoids that might be involved in the control of fetal vascular tone in the human placenta and umbilical cord (Mak et al. 1984; Boura et al. 1986; Prentice et al. 1987). The results of our studies, reported elsewhere, will be summarized here along with those of other workers to provide an overview of current knowledge in the field.
Discussion
The Human Fetal Extracorporeal Circulation All vessels of the human fetal extracorporeal circulation are capable of responding to a variety of both vasodilator and vasoconstrictor substances in vifro. The influence of vasodilator mechanisms predominates during normal pregnancy as indicated by the high umbilical bloodflow rate in vivo (Assali et al. 1960) and the low resistance of the vessels when they are perfused in vitro under pressure and flow conditions which approach the physiological norm (Walsh et at. 1974). Furthermore, in vivo Doppler ultrasound recordings during normal pregnancy demonstrate umbilical artery wave forms indicative of a low placental resistance to flow (Trudinger et al. 1985). Vasodilatation Because of their intrinsically low vascular tone, the fetal placental vessels in vifro are virtually unresponsive to exogenous vasodilatory stimuli unless preconstricted. Thus, the placental villous vessels, first constricted by infusion of the thromboxane A2 agonist U46619, respond by vasodilatation to nitroglycerine (Boura and Walters, unpublished observations), arachidonate and prostacyclin (Mak et al. 1984). Histamine (Maguire et al. 1985), isoprenaline, terbutaline (Soares de Moura 1987), vasoactive intestinal peptide (Maigaard 1987) and theophylline (Nielsen-Kudsk 1985) also cause vasodilatation after placental vascular preconstriction. Although acetylcholine is produced in large amounts by the placenta, there is little evidence to suggest that it contributes to the control of vascular tone and it has not been found to affect vasoconstrictor responses (Boura et a/. 1986). Acetylcholine may be more important in mechanisms con-
W. A. W. Walters and A. L. A. Boura
trolling transplacental passage of amino acids and electrolytes (Fant and Harbison 1981) and release of placental hormones (Sastry et al. 1981). The low tone in the smooth muscle of the fetal extracorporeal vasculature is probably maintained by the continuous output of vasodilator autacoids, particularly prostacyclin and perhaps the endothelial-cell-derived relaxing factor (EDRF) or nitric oxide (Vanhoutte 1988; Gude et al. 1990a).
Vasoconstriction The fetal placental vessels respond vigorously to vasoconstrictor stimuli, although there are substantial differences between various vessels in responsiveness to autacoids (Mak et al. 1984; Maigaard et al. 1986). Furthermore, large differences are found in the responses of the villous vessels to the various autacoids, which vary in their vasoconstrictor potencies as well as in the magnitude of the maximum responses they are capable of causing (Mak et al. 1984). Thus, in vitro angiotensin I1 (Abramovich et al. 1983), endothelins 1 and 3 (Boura and Walters, unpublished observations) and the thromboxane A2 agonist U46619 (Mak et al. 1984) show high potencies, causing effects in concentrations of 1 pmol to 100 pmol L-', but U46619, when perfused at a concentration (100 pmol L-') causing a maximal effect, is capable of producing intense vasoconstriction. The mean maximum increase in pressure to U46619, when perfused under constant flow conditions, was approximately 200 mm Hg, compared with an increase of approximately 35 mm Hg in the presence of a high concentration (30 pmol L-') of angiotensin I1 (Mak et al. 1984). Prostaglandins F2,, E2 and D2 are capable of causing constriction of the villous vessels of a magnitude comparable to that seen during thromboxane A2 receptor stimulation with U46619 but are at least 100 (F2,) and 3000 (E2 and D2) times less potent (Abramovich et al. 1984; Mak et al. 1984; Howard et al. 1987). There is evidence that responses to many of these eicosanoids can be affected by concomitant output of substances arising in tissues of the vessel walls including thromboxane A2, prostacyclin and EDRF in the placenta and umbilical cord (Mak et al. 1984; Bjdro 1986; Walters and Boura, unpublished observations). Thromboxane A2 receptors have been identified in the vasculature of both umbilical cord and placenta utilizing the thromboxane A2 agonist U46619, an analogue of prostaglandin H2, and its specific competitive antagonist AH22921 (Boura et al. 1986). In the umbilical vein and placental villous vessels, these receptors have been characterized and appear to be identical to those on human platelets and pulmonary blood vessels (Boura et al. 1986). Stimulation of umbilical-placental thromboxane A2 receptors in the sheep (Trudinger et al. 1989) causes analogous changes in systolic-diastolic flow patterns and vascular resistance to those seen in preeclampsia. Thus, during placental ischaemia such as occurs during maternal hypoxia (Las Heras et al. 1985) and preeclampsia (Chesley 1978), a vicious circle of intense vasospasm, reduction of blood flow and intravascular thrombosis could be established, with platelet aggregation leading to further thromboxane A2 release from platelets and progressively increasing vasospasm.
Conclusions
It can be concluded from results of work mentioned in this review that, under normal conditions, the human fetal extracorporeal vasculature is in a predominantly dilated state. It is, however, sensitive to a wide range of vasoconstrictor autacoids, some of which can cause intense vasoconstriction. It is likely that a balance is maintained between the actions of opposing vasodilator and vasoconstrictor autacoids, which may be readily tipped towards vasoconstriction by alterations in placental autacoid production, as demonstrated by the reduced prostacyclin and increased thromboxane A2 output from the placenta in preeclampsia (Remuzzi et al. 1980; Makila et al. 1984; Walsh et al. 1985).
Fetal Vascular Tone in the Human Placenta
Acknowledgment
Neil Gude has made a major contribution to much of the work in this field conducted in our laboratories, and reported in this article and elsewhere. References Abramovich, D. R., Page, K. R., and Wright, F. (1983). Effect of angiotensin I1 and 5-hydroxytryptamine on the vessels of the human foetal cotyledon. Br. J. Pharmacol. 79, 53-6. Abramovich, D. R., Page, K. R., and Parkin, A. M. L. (1984). The effect of prostaglandin D2 on the blood vessels of the perfused isolated cotyledon and the human placenta. Br. J. Pharmacol. 81, 19-21. Assali, N. S., Rauramo, L., and Peltonen, T. (1960). Measurement of uterine blood flow and uterine metabolism. Am. J. Obstet. Gynecol. 79, 86-98. Barnea, E. R., Fakih, H., and Oelsner, G., Walner, S., DeCherney, A. H., and Naftolin, F. (1986). Effect of antihypertensive drugs on catechol-0-methyltransferase and monoamine oxidase activity in human term placental explants. Gynecol. Obstet. Invest. 21, 124-30. Bjdro, K. (1986). Prostacyclin and thromboxane formation in human umbilical arteries following stimulation with vasoactive autacoids. Prostaglandins 31, 699-714. Bjdro, S., and Stray-Pedersen, S. (1986). Effects of vasoactive autacoids on different segments of human umbilicoplacental vessels. Gynecol. Obstet. Invest. 22, 1-6. Boura, A. L. A., Gude, N. M., King, R. G., and Walters, W. A. W. (1986). Acetylcholine output and fetal vascular resistance of human perfused placental cotyleda. Br. J. Pharmacol. 88, 301-6. Broughton Pipkin, F., and Symonds, E. M. (1977). Factors affecting angiotensin I1 concentrations in the human infant at birth. Clin. Sci. Mol. Med. 52, 449-56. Chesley, L. C. (1978). 'Hypertensive Disorders in Pregnancy.' pp. 455-66. (Appleton-Century-Crafts: New York.) Craft, I. L., Scrivener, R., and Dewhurst, C. J. (1973). Prostaglandin F2, levels in the maternal and fetal circulations in late pregnancy. J. Obstet. Gynaecol. Br. Commonw. 80, 616-18. Dancis, J . , Lehanka, J., Levitz, M., and Schneider, H. (1986). Establishment of gradients of riboflavin, L-lysine and a-aminoisobutyric acid across the perfused human placenta. J. Reprod. Med. 31, 293-6. Eltherington, L. G., Stoff, J., Hughes, T., and Melmon, K. L. (1968). Constriction of human umbilical arteries. Interaction between oxygen and bradykinin. Circ. Res. 22, 747-52. Falconer, A. D., and Lake, D. M. (1982). Circumstances influencing umbilical-cord plasmacatecholamines at delivery. Br. J. Obstet. Gynaecol. 89, 44-9. Fant, M. E., and Harbison, R. D. (1981). Syncytiotrophoblast membrane vesicles: a model for examining the human placental cholinergic system. Teratology 24, 187-99. Fox, S. B., and Khong, T. Y. (1990). Lack of innervation of human umbilical cord. An immunohistological and histochemical study. Placenta 11, 59-62. Gautieri, R. F., and Ciuchta, H. P. (1962). Effect of certain drugs on perfused human placenta: I. Narcotic analgesics, serotonin and relaxin. J. Pharm. Sci. 51, 55-8. Glance, D. G., Elder, M. G., and Myatt, L. (1986). The actions of prostaglandins and their interactions with angiotensin I1 in the isolated perfused human placental cotyledon. Br. J. Obstet. Gynaecol. 93, 488-94. Goland, R. S., Wardlaw, S. L., Stark, R. I., and Frantz, A. G. (1981). Human plasma @-endorphin during pregnancy, labor and delivery. J. Clin. Endocrinol. Metab. 52, 74-8. Gude, N. M. (1988). An investigation into the mechanisms controlling vascular tone of the fetal vessels of the human isolated perfused placenta. M.Sc. Thesis, Monash University, Melbourne. Gude, N. M., King, R. G., and Brennecke, S. P. (1990~).Role of endothelium-derived nitric oxide in maintenance of low fetal vascular resistance in placenta. Lancet 336, 1589-90. Gude, N. M., Rice, G. E., King, R. G., Boura, A. L. A., and Brennecke, S. P. (1990b). Analysis of the responses of the fetal vessels of human perfused placental lobules to acute infusions of arachidonic acid. Reprod. Fertil. Dev. 2, 591-6. Haegerstrand, A., Hemsen, A,, Gillis, C., Larsson, O., and Lundberg, J. M. (1989). Endothelin: presence in human umbilical vessels, high levels in fetal blood and potent constrictor effect. Acta Physiol. Scand. 137, 541-2. Howard, R. B., Hosokawa, T., and Maguire, M. H. (1987). Hypoxia-induced fetoplacental vasoconstriction in perfused human placental cotyledons. A m . J. Obstet. Gynecol. 157, 1261-6.
W. A. W. Walters and A. L. A. Boura
Jones, J . B., and Rowsell, A. (1973). Fetal 5-hydroxytryptamine levels in late pregnancy. J. Obstet. Gynaecol. Br. Commonw. 80, 687-9. Las Heras, J., Baskerville, J. C., Harding, P. G. R., and Daria Haust, M. (1985). Morphometric studies of fetal placental stem arteries in hypertensive disorders ('toxemia') of pregnancy. Placenta 6, 217-28. Lenz, T., Sealey, J. E., August, P., James, G. D., and Laragh, J. H. (1989). Tissue levels of active and total renin, angiotensinogen, human chorionic gonadotrophin, estradiol, and progesterone in human placentas from different methods of delivery. J. Clin. Endocrinol. Metab. 69, 31-7. Mackenzie, I. Z., MacLean, D. A., and Mitchell, M. D. (1980). Prostaglandins in the human fetal circulation in mid-trimester and term pregnancy. Prostaglandins 20, 649-54. Maguire, M. H., Howard, R. B., and Hosokawa, T. (1985). Autacoid receptors in the human fetoplacental vasculature. Contrib. Gynecol. Obstet. 13, 170-2. Maigaard, S. (1987). Contraction and relaxation of human uterine and placental smooth muscle. Endogenous control and calcium activation mechanisms. Acta Obstet. Gynecol. Scand. Suppl. 143, 9-39. Maigaard, S., Forman, A., and Andersson, K.-E. (1986). Differential effects of angiotensin, vasopressin and oxytocin on various smooth muscle tissues within the human uteroplacental unit. Acta Physiol. Scand. 128, 23-31. Mak, K. K.-W., Gude, N. M., Walters, W. A. W., and Boura, A. L. A. (1984). Effects of vasoactive autacoids on the human umbilical-fetal placental vasculature. Br. J. Obstet. Gynaecol. 91, 99-106. Makila, U.-M., Viinikka, L., and Ylikorkala, 0. (1984). Increased thromboxane Az production but normal prostacyclin by the placenta in hypertensive pregnancies. Prostaglandins 27, 87-95. Melmon, K. L., Cline, M. J., Hughes, T., and Nies, A. S. (1968). Kinins: possible mediators of neonatal circulatory changes in man. J. Clin. Invest. 47, 1295-302. Mitchell, R . G., and Porter, J. F. (1970). Histamine and granulocytes in the umbilical cord blood of infants at birth. Br. J. Pharmacol. 40, 310-16. Nielsen-Kudsk, J. E. (1985). Enprofylline and theophylline on small human placental arteries: in vitro effects and mode of action. Acta Pharm. Toxicol. 56, 176-82. Panigel, M. (1962). Placental perfusion experiments. Am. J. Obstet. Gynecol. 84, 1664-83. Penfold, P., Drury, L., Simmonds, R., and Hytten, F. E. (1981). Studies of a single placental cotyledon in vitro: I. The preparation and its viability. Placenta 2, 149-54. Piper, P . J., and Levene, S. (1987). Human umbilical arteries and veins: generation of leukotrienes and response to exogenous leukotrienes. Biol. Neonate 52, 9-15. Prentice, D. A., Boura, A. L. A., Gude, N. M., Walters, W. A. W., and King, R. G. (1987). Changes in the biological activity of autacoids during passage through the human perfused fetoplacental lobule. Eur. J. Pharmacol. 141, 79-86. Remuzzi, G., Marchesi, D., Zoja, C., Muratore, D., Mecca, G., Misiani, R., Rossi E., Barbato, M., Capetta, P., Donati, M. B., and De Gaetano, G. (1980). Reduced umbilical and placental vascular prostacyclin in severe pre-eclampsia. Prostaglandins 20, 105-10. Sandford, A. S. C., Walters, W. A. W., and Boura, A. L. A. (1981). On the mechanism underlying closure of the human umbilical vein induced by cooling. In 'Prostaglandins and Thromboxanes in the Cardiovascular System and in Gynaecology and Obstetrics'. (Ed. W. Forster.) pp. 415-19. (Fischer: Jena.) Sastry, B. V. R., and Sadavongvivad, C. (1979). Cholinergic systems in non-nervous tissues. Pharmacol. Rev. 30, 65-132. Sastry, B. V. R., Tayeb, 0. S., Barnwell, S. L., Janson, V. E., and Owens, L. K. (1981). Peptides from human placenta: methionine enkephalin and substance P. Placenta, Suppl. 3, 327-37. Soares de Moura, R. (1987). Effect of prostacyclin on the perfusion pressure and on the vasoconstrictor response of angiotensin I1 in the human isolated foetal placental circulation. Br. J. Clin. Pharmacol. 23, 765-8. Stuart, B., Drum, J., Fitzgerald, D. E., and Duignan, N. M. (1980). Fetal blood velocity waveforms in normal pregnancy. Br. J. Obstet. Gynaecol. 87, 780-5. Thaler-Dao, H., Saintot, M., Baudin, G., Descomps, B., and Crastes de Paulet, A. (1974). Purification of the human placental 15-hydroxy prostaglandin dehydrogenase: properties of the purified enzyme. FEBS Lett. 48, 204-8. Trudinger, B. J., Giles, W. B., Cook, C. M., Bombardieri, J., and Collins, L. (1985). Fetal umbilical artery flow velocity waveforms and placental resistance: clinical significance. Br. J. Obstet. Gynaecol. 92, 23-30.
Fetal Vascular Tone in the Human Placenta
Trudinger, B. J., Connelly, A. J., Giles, W. B., Hales, J. R., and Wilcox, G. R. (1989). The effects of prostacyclin and thromboxane analogue (U46619) on the fetal circulation and umbilical flow velocity waveforms. J. Dev. Physiol. 11, 179-84. Tulenko, T. N. (1981). The actions of prostaglandins and cyclo-oxygenase inhibition on the resistance vessels supplying the human fetal placenta. Prostaglandins 21, 1033-43. Tuvemo, T. (1980). Role of prostaglandins, prostacyclin, and thromboxanes in the control of the umbilical-placental circulation. Semin. Perinatol. NY 4, 91-5. Tuvemo, T., Strandberg, K., Hamberg, M., and Samuelsson, B. (1976). Maintenance of tone of the human umbilical artery by prostaglandin and thromboxane formation. In 'Advances in Prostaglandin and Thromboxane Research'. Vol. 1. (Eds B. Samuelsson and R. Paoletti.) pp. 425-8. (Raven: New York.) Vanhoutte, P. M. (1988). Relaxing and contracting factors. In 'Biological and Clinical Research'. (Humana: Clifton, NJ.) Walsh, S. W. (1985). Pre-eclampsia: an imbalance in placental prostacyclin and thromboxane production. Am. J. Obstet. Gynecol. 152, 335-40. Walsh, S. Z., Meyer, W. W., and Lind, J. (1974). The human fetal and neonatal circulation. (Ed. C. Thomas.) pp. 1-35 1. (Springfield: Ill.)
Manuscript received 4 November 1990; revised and accepted 4 March 1991
