Placental function in maternal disease : ex vivo assessment of foetoplacental vascular function and transport in diabetes and preeclampsia

PDF hosted at the Radboud Repository of the Radboud University Nijmegen The following full text is a publisher's version. For additional information about this publication click this link. http://hdl.handle.net/2066/19535 Please be advised that this information was generated on 2021-11-26 and may be subject to change. PLACENTAL FUNCTION IN MATERNAL DISEASE EX VIVO ASSESSMENT OF FOETOPLACENTAL VASCULAR FUNCTION AND TRANSPORT IN DIABETES AND PREECLAMPSIA Bisseling, Tanya Maria – Placental function in maternal disease – 2004 Thesis University of Nijmegen – with ref. – with summary in Dutch – 160 p. ISBN: 90-9018565-8 Print: Print Partners Ipskamp, Enschede Graphic Design: Bas van Abel No part of this book may be reproduced in any form without permission of the author. This research project was financially supported by Praeventiefonds (thans: Zorg Onderzoek Nederland), and the Dutch Diabetes Research Foundation Publication of this thesis was financially supported by Dutch Diabetes Research Foundation Novo Nordisk B.V. Ferring Nederland B.V. Organon Nederland B.V. PLACENTAL FUNCTION IN MATERNAL DISEASE EX VIVO ASSESSMENT OF FOETOPLACENTAL VASCULAR FUNCTION AND TRANSPORT IN DIABETES AND PREECLAMPSIA Een wetenschappelijke proeve op het gebied van de Medische Wetenschappen Proefschrift ter verkrijging van de graad van doctor aan de Radboud Universiteit Nijmegen op gezag van de Rector Magnificus, prof. Dr. C.W.P.M. Blom, volgens besluit van het College van Decanen in het openbaar te verdedigen op woensdag 24 november 2004 des namiddags om 1:30 uur precies door Tanya Maria Bisseling geboren op 24 mei 1971 te Nijmegen Promotores: Prof. dr. P. Smits Prof. dr. F.G.M. Russel Prof. dr. E.A.P. Steegers Co-promotor: Dr. L.D. Elving Manuscriptcommissie: Prof. dr. F.K. Lotgering Prof. dr. A.H.J. Danser Dr. C.J.J. Tack Paranymfen: Else Bisseling Maike Bisseling JE KUNT TEGENWOORDIG ZOVEEL WORDEN DAT IK MAAR GEWOON BLIJF WIE IK BEN Loesje voor papa voor Rianne CONTENTS List of abbreviations 9 Chapter 1 11 General introduction 13 Objective and Outline of the thesis 24 PART I: Foetoplacental endothelial function in maternal type 1 diabetes mellitus Chapter 2 Nitric oxide-mediated vascular tone in the foetal placental circulation from patients with type 1 diabetes mellitus 35 Placenta 2003; 24: 974 – 978 Chapter 3 Impaired KATP channel function in the foetoplacental circulation from patients with type 1 diabetes mellitus 47 American Journal of Obstetrics and Gynecology [in press] Chapter 4 Dysfunction of the cyclooxygenase pathway in the foetoplacental circulation in type 1 diabetes mellitus Diabetic Medicine [in press] 65 PART II: Effects of antioxidants on the NO pathway in the foetoplacental circulation Chapter 5 Vitamin C improves the disturbed NO- pathway in the human foetoplacental circulation in preeclampsia 77 Part of this chapter was published as a letter in Lancet 2000; 355: 65 Chapter 6 N-acetylcysteine restores NO-mediated effects in the foetoplacental circulation of preeclamptic patients 87 American Journal of Obstetrics and Gynecology 2004; 191: 328-333 PART III: Placental folate uptake in pregnancies with foetal growth retardation Chapter 7 Placental folate transport and binding are not impaired in pregnancies complicated by foetal growth restriction 103 Placenta 2004; 25:588-593 Chapter 8 Summary and discussion 117 General conclusion 124 Samenvatting 127 Bibliography 135 Extra 139 Dankwoord 153 Curriculum Vitae 159 LIST OF ABBREVIATIONS BK big conductance potassium channel cAMP cyclic adenosine monophosphate cGMP cyclic guanosine monophosphate COX cycloogygenase DM diabetes mellitus type 1 EC endothelial cell EDHF endothelium derived hyperpolarizing factor EDRF endothelium derived relaxing factor eNOS endothelial nitric oxide synthase FRα folate receptor-α RFC reduced folate carrier IK intermediate conductance potassium channel KATP adenosine triphosphate dependent potassium channel KV voltage dependent potassium channel L-NAME N –nitro-L-arginine methyl ester 5MTF 5-methyltetrahydrofolic acid NAC N-acetyl-L-cysteine NO nitric oxide PGI2 prostacyclin ROS reactive oxygen species SK small conductance potassium channel TXA thromboxane VSMC vascular smooth muscle cell 9 CHAPTER 1 GENERAL INTRODUCTION OBJECTIVE AND OUTLINE OF THE THESIS CHAPTER 1 GENERAL INTRODUCTION Embryology and anatomy of the human placenta The human placenta is divided into 20 to 40 independent functional units, called cotyledons. An isolated cotyledon is an ideal target for studies on foetoplacental vascular, and transplacental transport function. For a better understanding of these placental units, and the experimental methods used in the studies described, this first paragraph deals with the development and anatomy of the human placenta. About three days after ovulation and fertilization, the zygote reaches the uterine cavity as the so-called morula. In the uterine cavity the morula develops into a blastocyst; a cavity with fluid inside, surrounded by cells called blastomeres. In one pole of the blastocyst, a compact cell mass comes into existence – the inner cell mass - which is designed to produce the embryo. The outer cell mass is destined to become nonembryonic cells, the trophoblasts. The blastocyst causes an irritative reaction in the maternal decidua to the effect that it can fully nidate into the maternal mucous tissue. It is surprising that the maternal body does not reject the embryonic tissue. This is probably due to the fact that the extravillous trophoblastic cells, which are the only cells in direct contact with maternal tissue, produce HLA antibodies, which prevent such rejection [1]. Morphologically, the trophoblasts are either cellular or syncythial, and may appear as (syncytiotrophoblasts). uninuclear The (cytotrophoblasts) cytotrophoblast is a or multinuclear cellular giant progenitor of cells the syncytiotrophoblast. Soon after nidation, the blastocysts and trophoblasts grow and expand rapidly. One pole of this mass extends towards the uterine cavity; the other remains buried in the decidua. As the embryo enlarges, more decidua basalis is invaded and the walls of the superficial decidual capillaries are eroded. As a consequence, maternal blood leaks out into lacunae. From throphoblasts primitive villi develop, which traverse these lacunae. Most villi disappear again, except those located at the placental pole. Cytotrophoblastic cells invade the spiral arteries inducing degenerative changes in these vessels, causing the vascular smooth muscle in these vessels to become non-recognizable. The cytotrophoblastic cells bathe in the maternal blood. Primary villi can be first distinguished in the human placenta on about the 12th day after fertilization. In secondary villi the trophoblast column is invaded by mesenchymal cells. From these mesenchymal cells angiogenesis occurs to form the foetoplacental vascular bed. The resulting villi are called tertiary villi. On about the 17th day the foetal vessels are functional and the first preliminary placental circulation is established. A number of villi extend to the decidua to serve as anchoring villi, but most villi project into the intervillous space without reaching the decidua. As the placenta matures, the early 13 GENERAL INTRODUCTION, OBJECTIVE AND OUTLINE stem villi branch and subdivide into increasingly small villi. These terminal villi are the most important functional unit in a placenta. Each main stem villus and its finer subdivisions form a placental cotyledon, each with its own branch of the chorionic artery and vein (figure 1). A full-grown placenta is a discoid flat organ with a diameter of 15 – 20 cm, a thickness of 2.0 – 3.5 cm and an average weight of 450 – 600 gr. The human placenta is, like that of the guinea pig, of the haemochorioendothelial type; at all sites of direct cell-tocell contact, maternal tissues (decidua and blood) are juxtaposed to extra-embryonic cells (syncytiotrophoblasts) and not to embryonic cells. The apical (microvillous) membrane of the syncytiotrophoblast faces towards the maternal tissues, whereas the basolateral membrane is directed towards the foetal tissues. Foetoplacental circulation Deoxygenated blood flows from the foetus towards the placenta through two umbilical arteries. At the chorionic plate they branch into the chorionic arteries. After three to four divisions the chorionic arteries perforate the chorionic plate and are now called truncal arteries. Each truncal artery supplies one cotyledon. From this division on there is a decrease of smooth muscle cells in the vessel wall. A further division leads to villous vessels, finally resulting in the tertiary villous network. Oxygenated blood returns from the placenta to the foetus through one umbilical vein. Maternal placental circulation Maternal blood enters through the basal plate and reaches the chorionic plate driven by the maternal blood pressure. After bathing the microvillous surface of the chorionic villi in the intervillous space, the blood returns to the uterine veins. Where the uterine spiral arteries are perpendicular to the uterine wall, the veins are parallel. This means that during a uterine contraction the veins are closed and no maternal blood can disappear from the intervillous space, and so it creates an optimal environment for continuous maternal-to-foetal and vice versa exchange of gas and (micro)nutrients. Foetoplacental vascular function Because the foetoplacental vascular system lacks autonomic innervation [2], its vascular function depends entirely on biophysical, and humoral/paracrine/autocrine stimuli. In the foetoplacental vascular bed a variety of vasoactive mediators and their effects have been recognized (table 1). Most of these effects are identical in the 14 CHAPTER 1 Figure 1 Anatomical presentation of a human placental cotyledon From: Human Microscopic Anatomy ©Prof R.J. Krstic / Springer-Verlag GmbH Berlin Heidelberg 1991 A = umbilical artery, V = umbilical vein, UC = umbilical cord, PV = placental villus, Sy = syncytiotrophoblast, Cy = cytotrophoblast, DC = decidual cell, IS = intervillous space, CP = chorionic plate, DB = decidua basalis, FP = foetal placenta, MP = maternal placenta, UA = uteroplacental arteries, MV = maternal veins, UG = uterine glands 15 GENERAL INTRODUCTION, OBJECTIVE AND OUTLINE foetoplacental circulation and other vascular beds. Nitric oxide (NO), endothelium derived hyperpolarizing factor (EDHF), and prostacyclin (PGI2) seem to be the most relevant endothelium derived relaxing factors (EDRF). Endothelin and thromboxane A2 (TXA) are common vasoconstrictors. Although bradykinin, serotonin, and hypoxia result in vasoconstriction in the foetoplacental vascular bed, these agents induce vasodilation in other vascular beds. Figure 2 Schematic presentation of the NO pathway in the foetoplacental vascular bed, and the effect of L-NAME on this pathway. VSMC = vascular smooth muscle cell. In the past decade, NO has been identified as an important endogenous regulator of vascular tone. Endothelial release of NO causes a significant dilation of the foetoplacental vascular bed in the placental circulation, as well [3;4]. The endothelial NO-pathway is schematically presented in figure 2. Apart from the NO pathway, EDHF seems to be an important factor for the maintenance of baseline vascular tone. Different mediators are mentioned to act as an EDHF [5-8]. Although the identity of EDHF has not fully been elucidated, the activation of potassium channels is a final common pathway. Smooth muscle potassium channels, which in open state cause a shift of potassium out of the cell, play a major role in the hyperpolarization of the vascular smooth muscle cell wall. This triggers the closure of 16 CHAPTER 1 voltage-dependent calcium channels, resulting in a lower level of cytosolic calcium and subsequent relaxation of the vascular smooth muscle cell (figure 3). Figure 3 Schematic presentation of the EDHF pathway in the foetoplacental vascular bed, and the influence of potassium channel blockers on this pathway in the vascular smooth muscle (left), and endothelial cell (right). Em = membrane potential. 17 GENERAL INTRODUCTION, OBJECTIVE AND OUTLINE A third EDRF, which is involved in the maintenance of baseline foetoplacental vascular tone, is the prostanoid PGI2. PGI2 acts as a vasodilator in the foetoplacental vascular bed [9;10], and has a role in the maintenance of baseline vascular tone [11;12]. It has a close relationship with the prostanoid TXA, because the enzyme cycloogygenase (COX) is involved in the biosynthesis of both prostanoids (figure 4), but TXA acts as a vasoconstrictor [9;13]. Figure 4 Schematic reproduction of the COX pathway in the foetoplacental vascular bed, and the point of application of indomethacin on this pathway As the foetoplacental vascular bed is characterized by low resistance, an impaired availability of vasoactive mediators might result in a disturbed foetoplacental vascular function, and subsequent disturbed foetal development. Foetoplacental vascular function in diabetes mellitus Despite excellent metabolic control of women with Type 1 diabetes mellitus even long before conception, obstetrical complications, such as perinatal mortality, remain more prevalent in diabetic mothers [14, 15]. Unfortunatly, for most complications it remains unknown why these complications are more prevalent in diabetic mothers. Next to intrauterine periods of hypo- or hyperglycemia, a disturbance in the foetoplacental 18 CHAPTER 1 circulation may contribute to this increased complication rate. Diabetes is known to have a functional effect on both macro- and microcirculation. Especially endothelial dysfunction is a common finding in diabetes [16-19]. Although exposure of the placenta to the diabetic environment is limited to the duration of pregnancy, previous experiments have shown that this is long enough to develop functional changes in vasculature [20]. The expression of endothelial NO-synthase (eNOS) is increased in the human placenta in diabetes [21]. This suggests an increased placental production of NO. The vascular reactivity to the NO-donor glyceryl trinitrate, however, is decreased when compared to healthy controls [22]. This may be due to reactive oxygen species (ROS), which inactivate NO by formation of peroxynitrite. Indeed, there is evidence for an increased formation of peroxynitrite in the foetoplacental circulation [23]. Additionally, hyperglycaemia has been reported to impair endothelium-dependent vasodilation [24], in particular by reducing the bioavailability of NO. The foetoplacental circulation of diabetic patients is not only exposed to high glucose concentrations, but also to high insulin concentrations, as hyperglycaemia will elicit insulin release from the foetal pancreas. Interestingly, high insulin levels have been reported to increase the endothelial release of NO [25]. Hyperglycaemia and hyperinsulinaemia occur simultaneously in the foetoplacental circulation, and therefore the activity of the NO pathway is thought to be different in this circulation from women with diabetes. Additionally, recent research points towards an impaired release of EDHF in diabetes [26;27]. The baseline release of EDHF may be impaired in the foetoplacental circulation as well, and therefore contribute to the aetiology of perinatal complications in diabetes. Finally, in diabetes a disturbed TXA to PGI2 ratio has been shown in the foetoplacental and other vascular beds [28-30]. This is more profound in the foetoplacental vascular bed, than in the umbilical arteries [31]. The foetoplacental response to PGI2 as well as to the TXA-analogue U-46619 is attenuated in diabetes [22]. Moreover, the affinity of TXA for its receptor is reduced in placentae from diabetic women [20]. As such, the functions of PGI2 as well as TXA seem to be impaired in diabetes. Therefore, the underlying defect may be an impairment of the function of the enzyme COX in this particular vascular bed in diabetes. To our knowledge no data on the contribution of COX to baseline vascular tone in the foetoplacental vascular bed in diabetes are available. 19 GENERAL INTRODUCTION, OBJECTIVE AND OUTLINE Table 1 Vasoactive mediators / stimuli and their effects in the human placental vascular bed from uncomplicated nondiabetic pregnancies. Mediator Technique Effect Ref. acetylcholine cotyl. perf. none [32;33] patch clamp hyperpolarization of VSMC [34] adenosine / ATP cotyl. perf. foetoplacental vasodilation [35] angiotensin II myography foetoplacental vasoconstriction [36] cotyl. perf. foetoplacental vasoconstriction [35;37] myography constriction of foetoplacental arterial strips [36] cotyl. perf. increased foetoplacental vascular resistance [37] cotyl. perf. < 10-14 M; foetoplacental vasodilation bradykinin higher doses; foetoplacental vasoconstriction [33] cotyl. perf. foetoplacental vasoconstriction [39;40] myography constriction of foetoplacental arterial strips [41] cotyl. perf. foetoplacental vasoconstriction [42;43] VSMC patch clamp hyperpolarization of VSMC [34] EC patch clamp hyperpolarization of EC [44] patch clamp hyperpolarization of VSMC [34] EC patch clamp hyperpolarization of EC [44] KATP channel VSMC patch clamp hyperpolarization of VSMC [34] EC patch clamp none [44] cotyl. perf. foetoplacental vasodilation [3;4;33] patch clamp hyperpolarization VSMC [34] myography dilation of foetoplacental arterial strips [36] cotyl. perf. foetoplacental vasodilation [9;10] serotonin myography constriction chorionic and stem villous [45] thromboxane A2 cotyl. perf. foetoplacental vasoconstriction [9;13] endothelins hypoxia KV channel KCa2+ channel VSMC nitric oxide prostacyclin cotyl. perf. = isolated perfused foetoplacental cotyledon VSMC = vascular smooth muscle cell EC = endothelial cell 20 CHAPTER 1 Baseline vascular tone is not only dependent on specific functional mediators. Nonspecific factors, like morphological characteristics and elasticity of resistance vessels may contribute to baseline vascular tone, as well. Morphological and human in-vivo studies point towards alterations in the structure and function of placental vessels in diabetes [46-48]. Additionally, the elasticity of resistance vessels (compliance) is reduced in diabetes [49]. However, no data exist on vascular compliance of foetoplacental vessels. Foetoplacental vascular function in preeclampsia Preeclampsia is a major complication of pregnancy. To date, the exact aetiology of the origin of preeclampsia remains uncertain. Like diabetes, preeclampsia is associated with endothelial dysfunction [50]. An impairment of endothelium-derived vasoconstrictor [51-54] as well as vasodilator [53;55;56] functions has been described in the foetoplacental vascular bed from patients with preeclampsia. Partly, oxidative stress may be an explanation for the origin of this endothelial dysfunction [22;57]. NO, the most important endothelium derived vasodilator in this vascular bed, is inactivated by reactive oxygen species (ROS), resulting in the formation of peroxynitrite. Peroxynitrite affects foetoplacental vascular function [22]. Administration of vitamin C and E to patients who ran the risk of developing preeclampsia seemed to be beneficial in the prevention of this disorder [58]. The effect of anti-oxidants on the foetoplacental circulation, however, is unknown. In different human vascular beds the antioxidants gluthation and its pharmacological precursor N-acetylcysteine (NAC), can improve endothelial function [59;60]. Such effect has been observed after short-term [59] as well as long-term administration [60]. In theory, administration of NAC into the foetoplacental vascular bed could improve endothelial function, especially the NO pathway, in women with preeclampsia. Placental transport function For the foetal well-being, the placenta has a role in respiration, nutrition, excretion, protection, storage, and hormone production. Exchange of gas, (micro)nutrients, hormones and excretion-products between maternal and foetal blood occurs over the so-called placental barrier. As mentioned, the human placenta is of the haemomonochorial type. This means that the placental barrier consists of a unilayer of syncytiotrophoblastic cells, a basal membrane and a unilayer of endothelial cells. From a pharmacological viewpoint, transport across the placental barrier can occur in four different ways; simple diffusion, facilitated diffusion, active transport, or pinocytosis. Glucose for example can pass the placental barrier freely through fascilitated diffusion. 21 GENERAL INTRODUCTION, OBJECTIVE AND OUTLINE Insulin, however, cannot pass the placental barrier at all. A disturbed placental uptake and transfer of nutrients, hormones and other products might result in an impaired placental as well as foetal growth and development. Folate transfer in foetal growth restriction An optimal transplacental transfer of (micro)nutrients is essential for development and growth of both the placenta and the foetus. Folate is an essential micronutrient in human beings. It plays an important role in a number of intracellular processes resulting in cell growth, such as carbon unit transfer, amino acid metabolism, and DNA synthesis. So, folate is involved in the development of the placental vascular bed, as well. A maternal folate deficiency is associated with several complications during pregnancy, including low birth weight [61-67]. Additionally, in vitro, folate deficiency induces an increase in apoptosis of the syncytiotrophoblast [68], probably due to a disturbed DNA synthesis and repair. Increased placental apoptosis during pregnancy is known to be associated with foetal growth restriction [69]. Additionally, folate is able to restore an impaired endothelial function. Folate does so, by increasing the activity of NO-synthase [70], and by reducing the catabolism of NO [71]. A such, apart from its effect on cellular growth and development, folate deficiency could result in an impaired vascular endothelial function. In theory, this endothelial dysfunction could be a second explanation for the increase in perinatal complications as observed in folate deficiency. The predominant form of folate in the human body is 5-methyltetrahydrofolic acid (5MTF). Cellular uptake of 5MTF is mediated by folate receptors in the microvillous plasma membrane. Three isotypes of the folate receptor (FR) have been described, of which only FRα is expressed in placental tissue [72]. FRα is a glycosylphosphatidylinositol-anchored protein located at the microvillous membrane. This receptor has direct contact with the maternal blood and binds folates with high affinity [73] with a dissociation constant for folic acid (pteroylglutamic acid) in the low nanomolar range [74-76]. After receptor-mediated uptake, folate efflux into the foetal blood is facilitated by the pH-dependent reduced folate carrier (RFC) at the basolateral side of the syncytiotrophoblast [72;77] (figure 5). To our knowledge only one study on the kinetics of human placental folate uptake has been published [78]. Placental uptake of 5MTF appeared not to be saturable at folate concentrations up to 1000 nM. During pregnancy, reduced availability of folate for the placenta or foetus can impair normal cellular growth and replication. This may arise from low folate intake or deficient placental uptake and subsequent transfer to the 22 CHAPTER 1 foetal blood. No study has been performed yet on folate transport in placentae of pregnancies complicated by foetal growth restriction. syncytiotrophoblast 1 6 2 5 3 4 microvillous membrane basolateral membrane 5MTF FRα RFC Figure 5 Schematic presentation of human syncytiotrophoblast; 5MTF binds to placental transfer of 5 MTF across the Folate Receptor-α (FRα) on the microvillous membrane (1), the FRα-5MTF complex is internalised into the syncytiotrophoblast via endocytosis (2, 3), 5MTF is released from FRα, probably due to a decrease of pH in the vesicles (4), 5MTF is transferred across the basolateral membrane by the Reduced Folate Carrier (RFC) (5,6) 23 GENERAL INTRODUCTION, OBJECTIVE AND OUTLINE OBJECTIVE AND OUTLINE OF THE THESIS The placenta has a key role in human reproduction. Moreover, it is a very exceptional organ; it combines functions which are performed by separate organs after birth. It transfers (micro)nutrients, oxygen/carbondioxide, hormones and waste products between mother and the developing foetus without actually mixing maternal and foetal blood. The low-resistance foetoplacental vascular bed plays a key role in placental physiology. An undisturbed function of this vascular bed is important for an optimal development and growth of the foetus. Although some data are available on placental vascular function in healthy circumstances, the involvement of disturbed placental physiology in several complications of pregnancy remains under-exposed. The placental vascular bed lacks autonomic innervation [2], therefore local mediators, like endothelium derived vasoactive factors, are important for the regulation of foetoplacental vascular tone. Diabetes mellitus type 1 and preeclampsia are associated with vascular endothelial dysfunction [16;50]. Moreover, both diseases are associated with an increased perinatal morbidity and mortality. In theory, in both diseases endothelial dysfunction in the foetoplacental vascular bed could contribute to the occurrence of perinatal complications. The main objective of this thesis was to investigate the effect of diabetes mellitus type 1 on foetoplacental vascular function (part I). Since maintenance of low-resistance is essential for an optimal placental function, it were mainly vasodilator mechanisms which were investigated. The effect of administration of the anti-oxidants vitamin C and NAC on the NO pathway in the foetoplacental circulation from preeclampsia (NAC) and healthy controls (vitamin C and NAC) was studied, as well (part II). Finally, placental transfer and binding of folate were investigated in placentae from pregnancies with foetal growth retardation as compared to controls (part III). Part I In chapter 2 the contribution of NO to the regulation of the baseline vascular tone in the foetoplacental circulation in DM was quantified. Additionally, the effect of insulin on the NO pathway in the foetoplacental vascular bed was investigated. These investigations were performed by means of ex vivo dual perfusion of the isolated human placental cotyledon. Chapter 3 describes a study on the contribution of different potassium channels to baseline vascular tone in the foetoplacental vascular bed in DM versus matched controls. Also, in this chapter the results of the study of the non-specific morphometry 24 CHAPTER 1 and vascular compliance are presented. In part, this study was performed by use of ex vivo cotyledon perfusion as described in chapter 2. Additionally, the morphometric analysis was performed using the VIDASplus analysis system, and elasticity of foetoplacental resistance vessels was assessed by use of a pressure myograph. The contribution of cyclooxygenase products to baseline vascular tone in the foetoplacental vascular bed in DM as compared to controls was studied by ex vivo cotyledon perfusion in chapter 4. Part II In chapter 5 the contribution of the NO pathway to baseline vascular tone in this vascular bed in preeclampsia as compared to controls is described. Additionally, the effect of acute administration of vitamin C in the foetoplacental circulation in placentae from uncomplicated pregnancies was assessed. Chapter 6 describes the results of a study in which we investigated the effect of acute administration of NAC on the NO pathway in the foetoplacental vascular bed in placentae from women with mild preeclampsia as compared to placentae from healthy controls. In both these chapters, the studies were performed by ex vivo cotyledon perfusion. Part III In chapter 7 a comparison was made of the transplacental folate transfer, binding of folate to the FRα, and expression of FRα and RFC between placentae from controls, and those from pregnancies complicated with foetal growth restriction. Folate transfer was investigated by ex vivo cotyledon perfusion. Binding of folate to the FRα was studied with a radioligand binding assay on isolated microvillous membranes from syncytiotrophoblasts. The expression of FRα and RFC was studied by Western Blot analysis on the same microvillous membrane fraction as well as a crude mince from each placenta. 25 GENERAL INTRODUCTION, OBJECTIVE AND OUTLINE REFERENCES [1] Billingham RE. Transplantation immunity and the maternal-foetal relation. N Engl J Med 1964; 270:667-672. [2] Reilly RD, Russell PT. Neurohistochemical evidence supporting an absence of adrenergic and cholinergic innervation in the human placenta and umbilical cord. Anat Rec 1977; 188(3):277-286. [3] Myatt L, Brewer A, Brockman DE. The action of nitric oxide in the perfused human foetal-placental circulation. Am J Obstet Gynecol 1991; 164(2):687-692. [4] Gude NM, King RG, Brennecke SP. Role of endothelium-derived nitric oxide in maintenance of low foetal vascular resistance in placenta [letter; comment]. Lancet 1990; 336(8730):1589-1590. [5] Hecker M, Bara AT, Bauersachs J, Busse R. Characterization of endothelium-derived hyperpolarizing factor as a cytochrome P450-derived arachidonic acid metabolite in mammals. J Physiol 1994; 481 ( Pt 2):407-414. [6] Beny JL, Schaad O. An evaluation of potassium ions as endothelium-derived hyperpolarizing factor in porcine coronary arteries. Br J Pharmacol 2000; 131(5):965973. [7] Matoba T, Shimokawa H, Nakashima M, Hirakawa Y, Mukai Y, Hirano K et al. Hydrogen peroxide is an endothelium-derived hyperpolarizing factor in mice. J Clin Invest 2000; 106(12):1521-1530. [8] Chaytor AT, Martin PE, Edwards DH, Griffith TM. Gap junctional communication underpins EDHF-type relaxations evoked by ACh in the rat hepatic artery. Am J Physiol Heart Circ Physiol 2001; 280(6):H2441-H2450. [9] Howard RB, Hosokawa T, Maguire MH. Pressor and depressor actions of prostanoids in the intact human foetoplacental vascular bed. Prostaglandins Leukot Med 1986; 21(3):323-330. [10] Soares DM. Effect of prostacyclin on the perfusion pressure and on the vasoconstrictor response of angiotensin II in the human isolated foetal placental circulation. Br J Clin Pharmacol 1987; 23(6):765-768. [11] Klockenbusch W, Braun MS, Schroder H, Heckenberger RE, Strobach H, Schror K. Prostacyclin rather than nitric oxide lowers human umbilical artery tone in vitro. Eur J Obstet Gynecol Reprod Biol 1992; 47(2):109-115. [12] Glance DG, Elder MG, Myatt L. The actions of prostaglandins and their interactions with angiotensin II in the isolated perfused human placental cotyledon. Br J Obstet Gynaecol 1986; 93(5):488-494. [13] Hedberg A, Mento PF, Liu EC, Hollander AM, Wilkes BM. Evidence for functional thromboxane A2-prostaglandin H2 receptors in human placenta. Am J Physiol 1989; 256(2 Pt 1):E256-E263. 26 CHAPTER 1 [14] von Kries R, Kimmerle R, Schmidt JE, Hachmeister A, Bohm O, Wolf HG. Pregnancy outcomes in mothers with pregestational diabetes: a population-based study in North Rhine (Germany) from 1988 to 1993. Eur J Pediatr 1997; 156(12):963-967. [15] Evers IM, Bos AM, Aalders AL, van Ballegooie E, de Valk HW, van Doormaal JJ, ter Brugge HG, Visser GH. Pregnancy in women with daibets mellitus type I: maternal and perinatal complications, in spite of good blood glucose control. Ned Tijdschr Geneeskd 2000; 144(17): 804-809 [16] De Vriese AS, Verbeuren TJ, Van D, V, Lameire NH, Vanhoutte PM. Endothelial dysfunction in diabetes. Br J Pharmacol 2000; 130(5):963-974. [17] Poston L, Taylor PD. Glaxo/MRS Young Investigator Prize. Endothelium-mediated vascular function in insulin-dependent diabetes mellitus [see comments]. Clin Sci (Colch ) 1995; 88(3):245-255. [18] Williams SB, Goldfine AB, Timimi FK, Ting HH, Roddy MA, Simonson DC et al. Acute hyperglycemia attenuates endothelium-dependent vasodilation in humans in vivo. Circulation 1998; 97(17):1695-1701. [19] Scherrer U, Randin D, Vollenweider P, Vollenweider L, Nicod P. Nitric oxide release accounts for insulin's vascular effects in humans. J Clin Invest 1994; 94(6):2511-2515. [20] Wilkes BM, Mento PF, Hollander AM. Reduced thromboxane receptor affinity and vasoconstrictor responses in placentae from diabetic pregnancies. Placenta 1994; 15(8):845-855. [21] Rossmanith WG, Hoffmeister U, Wolfahrt S, Kleine B, McLean M, Jacobs RA et al. Expression and functional analysis of endothelial nitric oxide synthase (eNOS) in human placenta. Mol Hum Reprod 1999; 5(5):487-494. [22] Kossenjans W, Eis A, Sahay R, Brockman D, Myatt L. Role of peroxynitrite in altered foetal-placental vascular reactivity in diabetes or preeclampsia. Am J Physiol Heart Circ Physiol 2000; 278(4):H1311-H1319. [23] Myatt L, Rosenfield RB, Eis AL, Brockman DE, Greer I, Lyall F. Nitrotyrosine residues in placenta. Evidence of peroxynitrite formation and action. Hypertension 1996; 28(3):488493. [24] Lash JM, Nase GP, Bohlen HG. Acute hyperglycemia depresses arteriolar NO formation in skeletal muscle. Am J Physiol 1999; 277(4 Pt 2):H1513-H1520. [25] Steinberg HO, Brechtel G, Johnson A, Fineberg N, Baron AD. Insulin-mediated skeletal muscle vasodilation is nitric oxide dependent. A novel action of insulin to increase nitric oxide release. J Clin Invest 1994; 94(3):1172-1179. [26] Wigg SJ, Tare M, Tonta MA, O'Brien RC, Meredith IT, Parkington HC. Comparison of effects of diabetes mellitus on an EDHF-dependent and an EDHF-independent artery. Am J Physiol Heart Circ Physiol 2001; 281(1):H232-H240. [27] Fukao M, Hattori Y, Kanno M, Sakuma I, Kitabatake A. Alterations in endotheliumdependent hyperpolarization and relaxation in mesenteric arteries from streptozotocininduced diabetic rats. Br J Pharmacol 1997; 121(7):1383-1391. 27 GENERAL INTRODUCTION, OBJECTIVE AND OUTLINE [28] Hishinuma T, Tsukamoto H, Suzuki K, Mizugaki M. Relationship between thromboxane/prostacyclin ratio and diabetic vascular complications. Prostaglandins Leukot Essent Fatty Acids 2001; 65(4):191-196. [29] Saldeen P, Olofsson P, Parhar RS, al Sedairy S. Prostanoid production in the umbilicoplacental arterial tree relative to impaired glucose tolerance. Early Hum Dev 1998; 50(2):175-183. [30] Kuhn DC, Botti JJ, Cherouny PH, Demers LM. Eicosanoid production and transfer in the placenta of the diabetic pregnancy. Prostaglandins 1990; 40(2):205-215. [31] Saldeen P, Olofsson P, Parhar RS, al Sedairy S. Prostanoid production in umbilical vessels and its relation to glucose tolerance and umbilical artery flow resistance. Eur J Obstet Gynecol Reprod Biol 1996; 68(1-2):35-41. [32] Boura AL, Gude NM, King RG, Walters WA. Acetylcholine output and foetal vascular resistance of human perfused placental cotyleda. Br J Pharmacol 1986; 88(2):301-306. [33] Amarnani S, Sangrat B, Chaudhuri G. Effects of selected endothelium-dependent vasodilators on foetoplacental vasculature: physiological implications. Am J Physiol 1999; 277(2 Pt 2):H842-H847. [34] Guiet-Bara A, Ibrahim B, Leveteau J, Bara M. Calcium channels, potassium channels and membrane potential of smooth muscle cells of human allantochorial placental vessels. Bioelectrochem Bioenerg 1999; 48(2):407-413. [35] Read MA, Boura AL, Walters WA. Vascular actions of purines in the foetal circulation of the human placenta. Br J Pharmacol 1993; 110(1):454-460. [36] Tulenko TN. The actions of prostaglandins and cyclo-oxygenase inhibition on the resistance vessels 21(6):1033-1043. supplying the human foetal placenta. Prostaglandins 1981; [37] Kisch ES, Jaffe A, Knoll E, Stern N. Role of the lipoxygenase pathway in angiotensin IIinduced vasoconstriction in the human placenta. Hypertension 1997; 29(3):796-801. [38] Groom TM, Gautieri RF. Influence of a stable prostacyclin analogue (iloprost) and cyclooxygenase inhibition on angiotensin-II in the perfused human placenta. Res Commun Chem Pathol Pharmacol 1989; 66(1):21-32. [39] Gude NM, King RG, Brennecke SP. Endothelin: release by and potent constrictor effect on the foetal vessels of human perfused placental lobules. Reprod Fertil Dev 1991; 3(4):495-500. [40] Myatt L, Brewer AS, Brockman DE. The comparative effects of big endothelin-1, endothelin-1, and endothelin-3 in the human foetal-placental circulation. Am J Obstet Gynecol 1992; 167(6):1651-1656. [41] Sand AE, Ostlund E, Andersson E, Fried G. Endothelin-induced contractions in placental arteries is mediated by both ETA- and ETB-receptors. Acta Physiol Scand 1998; 163(3):227-234. [42] Wilkes BM, Mento PF. Bradykinin-induced vasoconstriction and thromboxane release in perfused human placenta. Am J Physiol 1988; 254(6 Pt 1):E681-E686. 28 CHAPTER 1 [43] Hampl V, Bibova J, Stranak Z, Wu X, Michelakis ED, Hashimoto K et al. Hypoxic foetoplacental vasoconstriction in humans is mediated by potassium channel inhibition. Am J Physiol Heart Circ Physiol 2002; 283(6):H2440-H2449. [44] Guiet-Bara A, Bara M. Evidence of K and Ca channels in endothelial cells of human allantochorial placental vessels. Cell Mol Biol (Noisy -le-grand) 2002; 48 Online Pub:OL317-OL322. [45] Maigaard S, Forman A, Andersson KE. Relaxant and contractile effects of some amines and prostanoids in myometrial and vascular smooth muscle within the human uteroplacental unit. Acta Physiol Scand 1986; 128(1):33-40. [46] Teasdale F. Histomorphometry of the human placenta in Class B diabetes mellitus. Placenta 1983; 4(1):1-12. [47] Teasdale F. Histomorphometry of the human placenta in Class C diabetes mellitus. Placenta 1985; 6(1):69-81. [48] Mayhew TM, Jairam IC. Stereological comparison of 3D spatial relationships involving villi and intervillous pores in human placentae from control and diabetic pregnancies. J Anat 2000; 197 ( Pt 2):263-274. [49] Romney JS, Lewanczuk RZ. Vascular compliance is reduced in the early stages of type 1 diabetes. Diabetes Care 2001; 24(12):2102-2106. [50] Roberts JM, Taylor RN, Musci TJ, Rodgers GM, Hubel CA, McLaughlin MK. Preeclampsia: an endothelial cell disorder. Am J Obstet Gynecol 1989; 161(5):12001204. [51] Singh HJ, Rahman A, Larmie ET, Nila A. Endothelin-l in foeto-placental tissues from normotensive pregnant women and women with pre-eclampsia. Acta Obstet Gynecol Scand 2001; 80(2):99-103. [52] Walsh SW, Wang Y. Trophoblast and placental villous core production of lipid peroxides, thromboxane, and prostacyclin in preeclampsia. J Clin Endocrinol Metab 1995; 80(6):1888-1893. [53] Walsh SW. Preeclampsia: an imbalance in placental prostacyclin and thromboxane production. Am J Obstet Gynecol 1985; 152(3):335-340. [54] Cruz MA, Gonzalez C, Gallardo V, Lagos M, Albornoz J. Venous placental reactivity to serotonin in normal and preeclamptic gestants. Gynecol Obstet Invest 1993; 36(3):148152. [55] Myatt L, Eis AL, Brockman DE, Greer IA, Lyall F. Endothelial nitric oxide synthase in placental villous tissue from normal, pre-eclamptic and intrauterine growth restricted pregnancies. Hum Reprod 1997; 12(1):167-172. [56] King RG, Gude NM, Di Iulio JL, Brennecke SP. Regulation of human placental foetal vessel tone: role of nitric oxide. Reprod Fertil Dev 1995; 7(6):1407-1411. [57] Walsh SW. Maternal-placental interactions of oxidative stress and antioxidants in preeclampsia. Semin Reprod Endocrinol 1998; 16(1):93-104. 29 GENERAL INTRODUCTION, OBJECTIVE AND OUTLINE [58] Chappell LC, Seed PT, Briley AL, Kelly FJ, Lee R, Hunt BJ et al. Effect of antioxidants on the occurrence of pre-eclampsia in women at increased risk: a randomised trial [see comments]. Lancet 1999; 354(9181):810-816. [59] Prasad A, Andrews NP, Padder FA, Husain M, Quyyumi AA. Glutathione reverses endothelial dysfunction and improves nitric oxide bioavailability. J Am Coll Cardiol 1999; 34(2):507-514. [60] De Mattia G, Bravi MC, Laurenti O, Cassone-Faldetta M, Proietti A, De Luca O et al. Reduction of oxidative stress by oral N-acetyl-L-cysteine treatment decreases plasma soluble vascular cell adhesion molecule-1 concentrations in non-obese, non- dyslipidaemic, normotensive, patients with non-insulin-dependent diabetes. Diabetologia 1998; 41(11):1392-1396. [61] Hibbard BM. The role of folic acid in pregnancy. With particular referance to anemia, abruption and abortion. Journal of Obstetrics and Gynecology , 529-542. 1964. [62] Steegers-Theunissen RP, Boers GH, Trijbels FJ, Eskes TK. Neural-tube defects and derangement of homocysteine metabolism [letter]. N Engl J Med 1991; 324(3):199-200. [63] Steegers-Theunissen RP, Boers GH, Blom HJ, Trijbels FJ, Eskes TK. Hyperhomocysteinaemia and recurrent spontaneous abortion or abruptio placentae [letter]. Lancet 1992; 339(8801):1122-1123. [64] Goldenberg RL, Tamura T, Cliver SP, Cutter GR, Hoffman HJ, Copper RL. Serum folate and foetal growth retardation: a matter of compliance? Obstet Gynecol 1992; 79(5 ( Pt 1)):719-722. [65] Scholl TO, Hediger ML, Schall JI, Khoo CS, Fischer RL. Dietary and serum folate: their influence on the outcome of pregnancy. Am J Clin Nutr 1996; 63(4):520-525. [66] Scholl TO, Johnson WG. Folic acid: influence on the outcome of pregnancy. Am J Clin Nutr 2000; 71(5 Suppl):1295S-1303S. [67] George L, Mills JL, Johansson AL, Nordmark A, Olander B, Granath F et al. Plasma folate levels and risk of spontaneous abortion. JAMA 2002; 288(15):1867-1873. [68] Steegers-Theunissen RP, Smith SC, Steegers EA, Guilbert LJ, Baker PN. Folate affects apoptosis in human trofoblastic cells. Br.J.Obstet.Gynaecol. 107, 1513-1515. 2000. [69] Smith SC, Baker PN, Symonds EM. Increased placental apoptosis in intrauterine growth restriction. Am J Obstet Gynecol 1997; 177(6):1395-1401. [70] Stroes ES, van Faassen EE, Yo M, Martasek P, Boer P, Govers R et al. Folic acid reverts dysfunction of endothelial nitric oxide synthase. Circ Res 2000; 86(11):1129-1134. [71] Verhaar MC, Wever RM, Kastelein JJ, van Dam T, Koomans HA, Rabelink TJ. 5methyltetrahydrofolate, the active form of folic acid, restores endothelial function in familial hypercholesterolemia. Circulation 1998; 97(3):237-241. [72] Prasad PD, Ramamoorthy S, Leibach FH, Ganapathy V. Molecular cloning of the human placental folate transporter. Biochem Biophys Res Commun 1995; 206(2):681-687. [73] Antony AC. Folate receptors. Annu Rev Nutr 1996; 16:501-521. 30 CHAPTER 1 [74] Henriques C, Trugo NM. Partial characterization of folate uptake in microvillous membrane vesicles isolated from human placenta. Braz J Med Biol Res 1996; 29(12):1583-1591. [75] Green T, Ford HC. Human placental microvilli contain high-affinity binding sites for folate. Biochem J 1984; 218(1):75-80. [76] Antony AC, Utley C, Van Horne KC, Kolhouse JF. Isolation and characterization of a folate receptor from human placenta. J Biol Chem 1981; 256(18):9684-9692. [77] Dudeja PK, Torania SA, Said HM. Evidence for the existence of a carrier-mediated folate uptake mechanism in human colonic luminal membranes. Am J Physiol 1997; 272(6 Pt 1):G1408-G1415. [78] Henderson GI, Perez T, Schenker S, Mackins J, Antony AC. Maternal-to-foetal transfer of 5-methyltetrahydrofolate by the perfused human placental cotyledon: evidence for a concentrative role by placental folate receptors in foetal folate delivery. J Lab Clin Med 1995; 126(2):184-203. 31 PART I FOETOPLACENTAL ENDOTHELIAL FUNCTION IN MATERNAL TYPE 1 DIABETES MELLITUS CHAPTER 2 NITRIC OXIDE-MEDIATED VASCULAR TONE IN THE FOETAL PLACENTAL CIRCULATION FROM PATIENTS WITH TYPE 1 DIABETES MELLITUS Tanya M. Bisseling, Alfons C. Wouterse, Eric A. Steegers, Lammy Elving, Frans G. Russel , Paul Smits Placenta 2003; 24:974-978 © Elsevier Ltd NITRIC OXIDE IN THE FOETAL PLACENTAL VASCULAR BED IN DIABETES SUMMARY Endothelium-derived nitric oxide (NO) plays a key role in the regulation of vascular tone in health and disease. The present study addresses the contribution of NO to the baseline vascular tone in the foetoplacental circulation of type 1 diabetic women. To this end, we performed ex-vivo dual perfusions of isolated cotyledons from 7 women with type 1 diabetes mellitus and 24 healthy women. The foetoplacental arterial pressure was considered to be a measure of foetal vascular resistance. The contribution of NO to the baseline vascular tone of the foetoplacental circulation was quantified by addition of the NO-synthase inhibitor NG-nitro-arginine-methylester (L-NAME). Apart from the diabetic state, we studied the influence of exogenous insulin on the response to L-NAME. Mean (± SEM) baseline foetoplacental arterial pressure was higher in diabetes (25.7 ± 3.4 mm Hg versus 18.0 ± 1.7 mm Hg, P<0.05). Maximum perfusion pressure after LNAME was 87.9 ± 7.1 mm Hg in diabetes versus 58.9 ± 4.5 mm Hg in controls (P<0.01). The net L-NAME-induced increase in foetoplacental arterial pressure was higher in diabetes (62.2 ± 9.1 mm Hg versus 40.9 ± 3.5 mm Hg, P<0.05). Although insulin induced a shift to the left of the L-NAME-curve, the net L-NAME-induced increase in foetoplacental arterial pressure was not affected. We conclude that diabetes is associated with an increased baseline vascular tone of the foetoplacental vascular bed. This cannot be explained by attenuated NO-mediated effects. In contrast, the activity of the NO pathway seems to be increased in diabetes. The latter observation seems not to be caused by high insulin levels. 36 CHAPTER 2 INTRODUCTION In women with diabetes mellitus type 1 (DM), pregnancy is associated with an increased risk for perinatal complications. In part, this concerns complications like intrauterine growth retardation and foetal death [1], which may be explained by changes in the foetoplacental circulation. Indeed, DM is known to affect vascular function [2], and morphological as well as human in-vivo studies point towards alterations in the structure and function of placental vessels [3-5]. In the past decade, endothelium-derived nitric oxide (NO) has been identified as an important endogenous regulator of vascular tone. Also in the placental circulation, the endothelial release of NO significantly dilates the foetoplacental vascular bed [6;7]. In DM, hyperglycaemia has been reported to impair endothelium-dependent vasodilation [8], in particular by reducing the bioavailability of NO. The foetoplacental circulation of diabetic patients is not only exposed to high glucose concentrations, but also to high insulin, as hyperglycaemia will elicit insulin release from the foetal pancreas. Interestingly, high insulin levels have been reported to increase the endothelial release of NO [9]. Finally, structural diabetes-induced changes of the vascular wall may contribute to alterations of vessel function. All these effects have been observed in different vascular beds including the foetoplacental circulation [10]. However, up to now no data on the baseline release of NO from placental vessels are available. Realising that hyperglycaemia and hyperinsulinemia occur simultaneously in the foetoplacental circulation, the activity of the NO pathway is thought to be changed in diabetic women. Because the net effect of these different factors is not known, we aimed to quantify the contribution of NO to the regulation of the baseline vascular tone in the foetoplacental circulation of women with type 1 diabetes. METHODS Study population All pregnant women with type 1 diabetes mellitus were eligible to participate in this study. Controls were healthy pregnant women. Exclusion criteria for both diabetic women and healthy controls were multiple pregnancy, premature birth (<37 weeks gestation), retained placenta, pregnancy-induced hypertension (diastolic pressure > 90 mmHg on two following occasions), preeclampsia and HELLP syndrome (hemolysis, elevated liver enzymes and low platelets). Women with diabetes mellitus type 2 or 37 NITRIC OXIDE IN THE FOETAL PLACENTAL VASCULAR BED IN DIABETES gestational diabetes mellitus were excluded from this study as well. All women gave written informed consent. The local medical ethics committee approved this study. Maternal and umbilical venous blood samples for insulin and C-peptide assay were taken within 15 minutes after delivery. Placenta perfusion Placentae were obtained immediately following delivery. Within 15 minutes, a suitable cotyledon was selected for ex-vivo dual perfusion [11]. The third or fourth order artery and vein were cannulated just before passage through the chorionic plate. Foetal inflow was gradually increased to 6 ml/min, at which the baseline foetoplacental arterial pressure equilibrates to 15 to 40 mmHg. The cotyledon was placed in a chamber with the maternal side facing upward. Maternal inflow was kept constant at 12 ml/min. The maternal outflow was collected and returned into the maternal reservoir. The perfusion fluid (150 ml for both sides) was 37.0 – 37.5°C, and was oxygenated with 95% O2/ 5% CO2 (pH 7,4). Because the foetal arterial inflow was kept constant, the foetoplacental arterial pressure was considered to be a reflection of the foetal arterial resistance. After 30 minutes of equilibration, the NO pathway in the foetal circulation was investigated by adding increasing concentrations of the NO-synthase blocker L-NAME. To study the effect of insulin in this model, half of the control placentae (n=12) were pre-treated with exogenous insulin (150 mE/ml) at the foetal side, 15 minutes prior to L-NAME. Materials Perfusion fluid was a Krebs Henseleit buffer containing 121 mM NaCl, 4 mM KCl, 0.95 mM KH2PO4, 1.2 mM MgSO4-7H2O, 22 mM NaHCO3, 11.1 mM glucose-H2O, 2 mM CaCl2. Heparin was used in a concentration that does not affect vascular tone (2500 IE/L) [12]. L-NAME was obtained from Sigma (St. Louis, USA), human insulin (Actrapid) was obtained from Novo Nordisk Pharma B.V. (Alphen aan de Rijn, the Netherlands). Statistical analysis Comparison of the clinical characteristics was performed by a Mann-Whitney U test. Data on perfusion pressures were analysed using Prism 3.0 (Graphpad Software) by fitting individual concentration-response curves for each experiment. For differences between groups, these data were tested by the Kruskall-Wallis test (SPSS). For correlations, Spearman correlation coefficients were calculated. 38 CHAPTER 2 ∆ Foetoplacental arterial pressure (mm Hg) 60 40 20 0 -6 -5 -4 -3 L-NAME (logM) Figure 1 The mean (± SEM) change from baseline of the foetoplacental arterial pressure in response to increasing concentrations of the NO-synthase inhibitor L-NAME in diabetic women (DM, n=7, dashed line) versus healthy controls (n=12, solid line). RESULTS Clinical characteristics Seven placentae originated from women with DM and 24 placentae from healthy controls. Table 1 presents all relevant clinical characteristics of the participants. Besides a slightly lower maternal diastolic blood pressure and a shorter gestational age in DM, clinical characteristics between the groups were comparable. As expected maternal venous C-peptide concentration was decreased in DM (median and range: 0.20 nmol/l; 0 - 0.48) compared to the control group (0.88 nmol/l; 0.16 2.31, p < 0.05), whereas umbilical insulin concentration (132 mE/ml; 28 - 242 versus 19mE/ml; 6 - 54, p < 0.001) and umbilical C-peptide concentration (2.38 nmol/ml; 0.47 – 4.16 versus 1.01 nmol/ml; 0.21 – 6.00, p< 0.05) were increased in DM. There was no difference in maternal venous insulin concentration between DM and controls. The pO2 in the foetal perfusion fluid observed during these experiments was 471 (392 548) mm Hg. 39 NITRIC OXIDE IN THE FOETAL PLACENTAL VASCULAR BED IN DIABETES ∆ Foetoplacental arterial pressure (mm Hg) 60 40 20 0 -6 -5 -4 -3 L-NAME (logM) Figure 2 The mean (± SEM) change from baseline of the foetoplacental arterial pressure in response to increasing concentrations of the NO-synthase inhibitor L-NAME in healthy controls without insulin (n=12, solid line) versus healthy controls with insulin (n=12, dashed line). Type 1 diabetes mellitus versus healthy controls Baseline foetoplacental arterial pressure was increased in DM (mean ± SEM: 25.7 ± 3.4 mm Hg versus 18.0 ± 1.7 mm Hg, P<0.05). Maximum foetoplacental arterial pressure at the highest dose of L-NAME was 87.9 ± 7.1 mm Hg in DM versus 58.9 ± 4.5 mm Hg in the control group (P<0.01). The L-NAME induced increase from baseline in foetoplacental arterial pressure was significantly higher in DM (62.2 ± 9.1 mm Hg versus 40.9 ± 3.5 mm Hg, p < 0.05) (figure 1). The logEC50 was similar in both groups (–4.3 ± 0.1 versus –4.5 ± 0.1). There was no correlation between baseline foetoplacental arterial pressure and the maximum response to L-NAME. Insulin effects on the NO pathway in the foetoplacental circulation of healthy controls Addition of insulin prior to L-NAME did neither affect baseline foetoplacental arterial pressure (17.8 ± 1.7 with versus 18.0 ± 1.7 mm Hg without insulin), nor maximal 40 CHAPTER 2 foetoplacental arterial pressure after L-NAME (57.9 ± 2.4 versus 58.9 ± 4.5 mm Hg). Also, the net L-NAME-induced increase in foetoplacental arterial pressure was not affected by insulin pre-treatment (40.2 ± 3.0 versus 40.9 ± 3.5 mm Hg) (fig 2). However, the logEC50 was lower when insulin was added (-4.9 ± 0.1 versus -4.5 ± 0.1, p < 0.01). DISCUSSION Although impaired endothelial function has been observed in all kind of diabetes models, little is known about foetoplacental vascular function in DM. In established DM, reduced NO-mediated effects have been put forward as a cause of vascular dysfunction [13;14], whereas the early stages of DM may show increased NO production [15;16]. We now observed that L-NAME-induced vasoconstriction was more pronounced in DM versus controls. This observation suggests that the contribution of NO to the baseline vascular tone is increased in the diabetic placentae. Since the newly formed vascular bed in the placenta is only exposed to the diabetic environment during the limited period of pregnancy, this particular vascular bed may show functional characteristics similar to other vascular beds in the early stages of diabetes. As such, the present observation of an upregulated NO pathway in our material may be in line with previous findings in other vascular beds. The observed increased response to L-NAME could be a non-specific response rather than a specific upregulation of the NO pathway in the diabetic placentae. However, in an identical study with another vasoconstrictor (glibenclamide, an ATP-dependent potassium-channel blocker which does not act via the NO pathway), we did not observe an increase in the maximal response in DM as compared to controls [17]. Other investigators even observed attenuated responses to vasoconstrictors like angiotensin II and the prostanoid analogue U46619 [10]. Together, these data strongly argue against a non-specific explanation for the upregulated NO pathway including the occurrence of structural vascular changes in the foetal 41 vascular bed of diabetic patients. NITRIC OXIDE IN THE FOETAL PLACENTAL VASCULAR BED IN DIABETES Table 1 Clinical characteristics of the participants (median, min - max) Diabetes Control Control Number 7 12 12 Maternal age (years) 34.2 (26.1 - 36-7) 35.3 (27.0 - 42.8) 31.6 (21.1 - 41.0) Parity (number) 0 (0 - 1) 1 (0 - 3) 0 (0 - 1) Gestation (weeks) 37.9 (37.0 - 38.9)∗ 39.9 (37.7 - 41.6) 39.8 (37.0 - 42.3) Birth weight (grams) 3610 (2780 - 4300) 3330 (2500 - 4560) 3622 (2760 - 4500) Placental weight (grams) 600 (500 - 740) 645 (400 - 815) 585 (400 - 1000) Body Mass Index (kg/m2) 26.9 (23.8 - 39.7) 25.2 (15.4 - 30.8) 23.5 (19.0 - 28.1) Diastolic BP (mm Hg) 75.0 (65.0 - 78.0) 82.0 (70.0 - 90.0) 80.0 (68.0 - 91.0) Smokers (number) 0 1 1 Medication Insulin(n=7) 0 0 White Class 3 C, 2 D, 1 F, 1 R - - 1 6.1 (5.8 - 7.2) - - 1 HbA1c 3rd trimester (%) 6.0 (5.7 - 7.3) - - Insulin M 28.6 (5.0 – 59.0) 28.2 (0.2 – 161.0) - U 131.4 (28.0 – 242.0)∗ 17.4 (5.0 – 54.0) - M 0.20 (0.00 – 0.48)∗ 0.88 (0.16 – 2.31) - U 2.38 (0.47 – 4.16)∗ 1.01 (0.21 – 6.00) - HbA1c 1st trimester (%) (mE/ml) C-peptide (nmol/l) ∗ P<0.05 versus healthy controls, M = maternal, U = umbilical 1 normal value for HbA1c in our laboratory is 4.2 – 6.3 % Our data show that the baseline vascular resistance of the foetoplacental vascular bed is slightly increased in DM. Obviously, this observation cannot be explained by reduced NO-mediated effects in diabetes, because the data on L-NAME point towards the opposite. Vasodilator substances like prostacyclin and Endothelium Derived Hyperpolarizing Factor (EDHF) have been reported to be decreased in DM [18;19], whereas the production of vasoconstrictors like endothelin [20] were increased in other vascular beds in DM. Although those mechanisms would be compatible with our observations, no data on these specific mediators are available as far as the foetoplacental vascular bed is concerned. 42 On the vasoconstrictor CHAPTER 2 prostanoid thromboxane, specific data in the foetoplacental vascular bed show a reduced response in diabetes [21]. However, this appeared not to result in an altered baseline vascular resistance of diabetic placentae as observed in our study. Previous studies suggest that insulin triggers the release of endothelium-derived NO [9]. Since the foetoplacental circulation is exposed to high insulin in diabetes, this phenomenon may be the explanation for the observed increase in L-NAME-induced vasoconstriction. In our study, addition of insulin showed a shift to the left of the concentration-response curve of L-NAME, but the maximum increase in foetoplacental arterial pressure appeared not to be affected by insulin. As such, our data on insulin argue against a role for insulin in the increased vasoconstrictor response to L-NAME in diabetes, although the long-standing exposure to insulin in patients may have different effects than the short-term in-vitro exposure of the present study. In conclusion, our study in the foetoplacental circulation shows an increased baseline vascular resistance in type 1 diabetes as compared to controls. This increased baseline tone is not caused by reduced NO-mediated effects. In contrast, the NO pathway seems to be upregulated in diabetes as demonstrated by an elevated vasoconstrictor response to NO-synthase blockade. Increased concentrations of insulin seem not to be responsible for the increased vasoconstrictor response to L-NAME in diabetes. 43 NITRIC OXIDE IN THE FOETAL PLACENTAL VASCULAR BED IN DIABETES REFERENCES [1] von Kries,R., Kimmerle,R., Schmidt,J.E., Hachmeister,A., Bohm,O., & Wolf,H.G. Pregnancy outcomes in mothers with pregestational diabetes: a population-based study in North Rhine (Germany) from 1988 to 1993. Eur J Pediatr 1997; 156:963-967. [2] De Vriese,A.S., Verbeuren,T.J., Van,D., V, Lameire,N.H., & Vanhoutte,P.M. Endothelial dysfunction in diabetes. Br J Pharmacol 2000; 130:963-974. [3] Teasdale,F. Histomorphometry of the human placenta in Class B diabetes mellitus. Placenta 1983; 4:1–12 [4] Teasdale,F. Histomorphometry of the human placenta in Class C diabetes mellitus. Placenta 1985; 6:69–81 [5] Mayhew,T.M. Enhanced foetoplacental angiogenesis in pre-gestational diabetes mellitus: the extra growth is exclusively longitudinal and not accompanied by microvascular remodelling. Diabetologia 2002; 45:1434-1439 [6] [7] Myatt,L., Brewer,A., & Brockman,D.E. The action of nitric oxide in the perfused human foetal-placental circulation. Am J Obstet Gynecol 1991; 164:687-692. Boccardo,P., Soregaroli,M., Aiello,S., Noris,M., Donadelli,R., Lojacono,A., & Benigni,A. Systemic and foetal-maternal nitric oxide synthesis in normal pregnancy and preeclampsia. Br J Obstet Gynaecol 1996; 103:879-886. [8] [9] Lash,J.M., Nase,G.P., & Bohlen,H.G. Acute hyperglycemia depresses arteriolar NO formation in skeletal muscle. Am J Physiol 1999; 277: H1513-H1520 Steinberg,H.O., Brechtel,G., Johnson,A., Fineberg,N., & Baron,A.D. Insulin-mediated skeletal muscle vasodilation is nitric oxide dependent. A novel action of insulin to increase nitric oxide release. J Clin Invest 1994; 94:1172-1179. [10] Kossenjans,W., Eis,A., Sahay,R., Brockman,D., & Myatt,L. Role of peroxynitrite in altered foetal-placental vascular reactivity in diabetes or preeclampsia. Am J Physiol Heart Circ Physiol 2000; 278: H1311-H1319. [11] [12] Schneider,H. & Dancis,J. Modified double-circuit in vitro perfusion of placenta. Am.J.Obstet.Gynecol 1984; 148:836. Tiefenbacher,C.P. & Chilian,W.M. Basic fibroblast growth factor and heparin influence coronary arteriolar tone by causing endothelium-dependent dilation. Cardiovasc Res 1997; 34: 411-417. [13] [14] Poston,L. & Taylor,P.D. Endothelium-mediated vascular function in insulin-dependent diabetes mellitus. Clin.Sci.(Colch.) 1995; 88:245-255. Williams,S.B., Goldfine,A.B., Timimi,F.K., Ting,H.H., Roddy,M.A., Simonson,D.C., & Creager,M.A. Acute hyperglycemia attenuates endothelium-dependent vasodilation in humans in vivo. Circulation 1998; 97:1695-1701. 44 CHAPTER 2 [15] Chiarelli,F., Cipollone,F., Romano,F., Tumini,S., Costantini,F., di Ricco,L., Pomilio,M., Pierdomenico,S.D., Marini,M., Cuccurullo,F., & Mezzetti,A. Increased circulating nitric oxide in young patients with type 1 diabetes and persistent microalbuminuria: relation to glomerular hyperfiltration. Diabetes 2000; 49:1258-1263.. [16] [17] Fitzgerald,S.M. & Brands,M.W. Nitric oxide may be required to prevent hypertension at the onset of diabetes. Am J Physiol Endocrinol Metab 2000; 279: E762-E768. Bisseling,T.M.:Steegers,E.A.P.:Wouterse,A.C.:Elving,L.:Russel,F.G.M.:Smits,P. Baseline function of placental vascular KATP-channels in healty and diabetic women. Br.J Clin.Pharmacol 2002; 53:543P-544P [18] [19] Silberbauer,K., Clopath,P., Sinzinger,H., & Schernthaner,G. Effect of experimentally induced diabetes on swine vascular prostacyclin (PGI2) synthesis. Artery 1980; 8:30-36. Wigg,S.J., Tare,M., Tonta,M.A., O'Brien,R.C., Meredith,I.T., & Parkington,H.C. Comparison of effects of diabetes mellitus on an EDHF-dependent and an EDHFindependent artery. Am J Physiol Heart Circ Physiol 2001; 281:H232-H240. [20] [21] Verma,S., Arikawa,E., & McNeill,J.H. Long-term endothelin receptor blockade improves cardiovascular function in diabetes. Am.J.Hypertens 2001; 14:679-687. Wilkes,B.M., Mento, P.F. & Hollander, A.M. Reduced thromboxane receptor affinity and vasoconstrictor responses in placentae from diabetic pregnancies. Placenta 1994; 15:845-855. 45 CHAPTER 3 IMPAIRED KATP CHANNEL FUNCTION IN THE FOETOPLACENTAL CIRCULATION FROM PATIENTS WITH TYPE 1 DIABETES MELLITUS Tanya M. Bisseling, Marieke G. Versteegen, Selina van der Wal, Jenny J.H. Copius Peereboom–Stegeman, Joop M.P.M. Borggreven, Eric A.P. Steegers, Jeroen A.W.M. van der Laak, Frans G.M. Russel , Paul Smits Am J Obstet Gynecol [in press] © Elsevier Ltd DM AFFECTS FOETOPLACENTAL VASCULAR KATP CHANNEL SUMMARY The increased perinatal morbidity in diabetes may be partly related to vascular dysfunction. Since potassium channels play an important role in the regulation of vascular tone, this study explores the impact of diabetes on potassium channel function in the foetoplacental vascular bed. Vascular potassium channel function was investigated by ex-vivo dual perfusion of isolated placental cotyledons (n=47). Appropriate control experiments were carried out to exclude non-specific effects. Glibenclamide (KATP channel blocker) increased perfusion pressure to a maximum foetoplacental arterial pressure of 37 ± 6 mmHg in controls versus 15 ± 6 mmHg in diabetes (P<0.05). 4-Aminopyridine (KV channel blocker) equally increased foetoplacental arterial pressure in controls, and in diabetes (21 ± 4 mmHg versus 22 ± 2 mmHg). Apamin and charybdotoxin (KCa channel blockers) caused a negligible rise in foetoplacental arterial pressure. In the foetoplacental circulation, KATP channels and KV channels significantly contribute to baseline vascular tone. In diabetes, vascular KATP channel function is impaired. 48 CHAPTER 3 INTRODUCTION In diabetes, pregnancy is associated with an increased risk for perinatal complications [1], including foetal death. Theoretically, these complications can be explained by changes in the foetoplacental circulation, as vascular dysfunction may compromise optimal oxygenation of the vital organs of the foetus. Diabetes is known to affect vascular function, in both macro- and microcirculation. In particular, endothelial dysfunction is a rather common finding in diabetes [2-4]. Although the exposure to the diabetic environment is limited to the duration of pregnancy, previous experiments have shown that this period of time is sufficient to induce functional changes in the foetoplacental vasculature [5-6]. The endothelium plays a pivotal role in the maintenance of baseline vascular tone. In diabetes, endothelial dysfunction, in particular with respect to the NO- and the cyclooxygenase pathways, have been observed in different vascular beds including foetoplacental arteries [5-6]. Apart from this, recent research points towards an impaired release of endothelium derived hyperpolarizing factor (EDHF) in diabetes [7]. Opening of the vascular smooth muscle KCa channel seems to be a final common pathway in the mechanism of action of EDHF. Independent from the EDHF-pathway, diabetes has been associated with dysfunction of the cardiovascular KATP channel, an ion channel that has an important role in the regulation of vascular tone during ischemia [8]. We hypothesize that vascular potassium channel function in the foetoplacental circulation is compromised in diabetes. To address this hypothesis, we assessed the vasoconstrictor response to potassium channel blockers in the foetoplacental vascular bed in diabetes versus controls. METHODS Study population Pregnant women with type 1 diabetes were eligible to participate. Controls were healthy pregnant women with uncomplicated pregnancies. Exclusion criteria for both groups were multiple pregnancy, premature birth (<37 weeks gestation), retained placenta, pregnancy-induced hypertension (diastolic pressure > 90 mmHg on two following occasions), and preeclampsia or HELLP syndrome (hemolysis, elevated liver enzymes and low platelets). 49 DM AFFECTS FOETOPLACENTAL VASCULAR KATP CHANNEL From all participants maternal and umbilical venous blood samples for insulin and Cpeptide assay were taken within 15 minutes after delivery. All women gave written informed consent. The local medical ethics committee approved this study. Classification of diabetes mellitus in pregnancy. In pregnancy diabetes mellitus is classified according to the White Classification. This classification consists of seven group; A, B, C, D, F, R, and H. Group A is the classification for gestational diabetes mellitus. Groups B, C, D, F, R, and H are classifications for pregestational diabetes mellitus. In groups B, C, and D the age of onset of diabetes mellitus in years (>20, 10 - 19, and <10 respectively), and duration of the disease in years (<10, 10 - 19, >20) determine the classification. Groups B and C are not associated with vascular complications; group D is associated with a benign nephropathy. In groups F, R, and H vascular complications determine the classification. Group F is associated with nephropathy, group R with retinopathy and group H with heart disease. Cotyledon perfusion Over a period of two years, all placentae from type 1 diabetic women were obtained immediately following delivery and transported to the laboratory within 10 minutes following delivery. Controls matched for mode of delivery, parity and maternal age, were acquired throughout this period. After arrival at the laboratory, a suitable cotyledon was selected from the placenta for ex-vivo dual perfusion as described extensively before [6]. With a constant foetal arterial inflow, the foetoplacental arterial pressure was considered to be a reflection of net downstream foetoplacental vascular resistance. The contribution of four different vascular potassium channels to baseline foetoplacental vascular tone was investigated by adding increasing concentrations of the selective blockers for these channels both in control and in diabetic placentae. These blockers were: glibenclamide for the ATP dependent K channel (KATP channel), apamine for the small conductance KCa2+ channel (SK channel); charybdotoxin for the intermediate conductance KCa2+ channel (IK channel), large conductance KCa2+ channel (BK channel), and some voltage dependent K channels (KV channel), and 4aminopyridine for the KV channel. These blockers were added in 6 to 9 cumulative logdose steps in concentration ranges from (log concentration [mol/L]): –8.0 to –3.5 (glibenclamide), -9.0 to –6.0 (apamin), -10.0 to –7.0 (charybdotoxin) and –7.5 to –3.5 (4-aminopyridin). 50 CHAPTER 3 Control experiments on vascular compliance of isolated resistance arteries In order to assess the elasticity of the resistance arteries (compliance) of control and diabetic placentae, foetoplacental resistance arteries were isolated from 4 healthy and 4 diabetic patients. On isolation, all arteries were transferred to the 10-ml pressure myograph organ bath, where they were immersed in Ca-free medium as described previously by Smits et al. [9]. The artery was gradually pressurized to 50 mmHg in a pressure myograph (Danish MyoTech P100) over a period of 5 minutes, and the arterial diameter was studied by stepwise increasing the intraluminal pressure from 1 to 60 mmHg for a period of 2 min at each pressure step. This was done twice in succession for each artery. The vessel diameter values for these two series were then averaged for each pressure. In this way two or three arteries were studied per placenta. These vessel diameter values were averaged to one representative value for each placenta. All preparations were gassed with 95% O2 / 5% CO2 in order to maintain pH at 7.4 throughout the experiment. Control experiments on morphometry of the foetoplacental arterial circulation In order to quantify the vascular diameters and wall thickness we used the computerized image analysis system (e.g. Vidas PLUS system, Carl Zeiss GmbH, Germany) [10]. To perform this analysis, the vascular endothelial CD34 antigen was detected by the monoclonal mouse-anti-CD34 antibody (QBEnd) [10]. The Vidas PLUS system is organized in such way that it is able to discriminate the CD34-stained endothelium from the surrounding tissue. In 25 images of each placenta, the following vascularisation variables were calculated by the computer system: a) per image: vascular area as a percentage, vascular perimeter, and vascular number, and b) per vascular element: the area, the perimeter, and the diameter. Materials Apamine, 4-aminopyridine, and H2O2 were obtained from Sigma (St. Louis, USA), charybdotoxin was obtained from Alomone Laboratories (Jerusalem, Israel), Glibenclamide was obtained from Hoechst Marion Roussel, USA. Formalin was obtained from JTBaker (Deventer, The Netherlands), QBEnd was obtained from Biogenex (San Ramon, CA, USA), and horse antimouse and the ABC Elitekit were obtained from Vector Laboratories (Burlingame, CA, USA). Before each experiment the blockers were dissolved in Krebs-Henseleit buffer to a solution of 0.8 mmol/L, except for 4aminopyridine, which was dissolved in water. 51 DM AFFECTS FOETOPLACENTAL VASCULAR KATP CHANNEL Statistical analysis All data were tested for normality by Shapiro-Wilk test. Comparison of the clinical characteristics of the diabetic women versus the healthy controls was performed by a Mann-Whitney U test. Data on cotyledon perfusion pressures were analyzed using Prism 3.0 (Graphpad Software) by fitting individual concentration-response curves for each experiment. Maximal percentage changes were calculated by use of the quotient: [max. – baseline foetoplacental arterial pressure] / baseline foetoplacental arterial pressure * 100. The resulting parameters did not show Gaussian distribution and were therefore tested by a Mann-Whitney U test. Differences were considered to be significant at P-values < 0.05. Data concerning vascular compliance were compared by paired t-test. Vidas PLUS -data were tested by an unpaired t-test. All statistics were performed in SPSS (SPSS 10.0, SPSS Inc., Chigaco, IL, USA). RESULTS In total, 37 placentae were included in this study (19 controls, 18 diabetes). When possible, we investigated two cotyledons of each placenta simultaneously. In those cases, the 2 cotyledons were used for different series. As such, 26 control cotyledons versus 21 diabetes cotyledons were measured. For the separate potassium channel blocker series, all cotyledons originated from different placentae. Table 1 summarizes the clinical characteristics of the participants. Women with diabetes had a shorter gestational age at time of delivery. As expected, maternal venous C-peptide concentration was lower in diabetic patients versus controls. There was no difference in maternal insulin concentration between controls and diabetes. In umbilical venous plasma, the C-peptide concentration was lower in diabetes as compared with controls, but as expected higher than the maternal C-peptide concentrations. In umbilical venous plasma, insulin concentration was increased in diabetes as compared with controls. Overall, baseline foetoplacental arterial pressure was comparable in controls (n=19) versus diabetes (n=18) (22 ± 1 mmHg versus 24 ± 1 mmHg, mean ± SEM). 52 CHAPTER 3 Table 1 Clinical characteristics of the participants (median and range) Controls Diabetic women Number of women/placentae 19 18 Number of cotyledons tested 26 21 Maternal age (years) 33.1 (25.0 – 40.1) 30.5 (21.0 – 38.2) Parity (number) 1 (0 - 3) 1 (0 - 3) Gestation (weeks) 40 (38 – 42) 38 (37 - 39) ∗ Vaginal delivery (number) 14 13 CS + locoregional anaesthetic (number) 4 5 CS + general anaesthetic (number) 1 0 Birth weight (grams) 3097 (2555 - 3970) 3525 (2435 - 4875) Placental weight (grams) 548 (370 - 700) 670 (400 - 950) ∗ Body Mass Index (kg/m2) 23.0 (17.7 – 26.8) 23.9 (17.2 – 45.9) 79 (60 - 90) 87 (55 - 90) ∗ 1 Diastolic BP (mm Hg) Smokers (number) 1 2 White Class - 9 B, 6 C, 1 D, 1 F, 1 R HbA1c 1st trimester (%)2 - 6.5 (5.8 – 8.8) 2 - 6.5 (4.8 - 7.1) HbA1c 3rd trimester (%) Insulin (mE/ml) M 15.5 (0.2 – 161.0) 22.0 (5.0 – 59.0) U 10.0 (5.0 – 54.0) 52.0 (5.0 – 242.0) ∗ C-peptide (nmol/L) M 1.02 (0.16 – 2.35) 0.12 (0.00 – 0.48) ∗ U 0.50 (0.21 – 6.00) 0.26 (0.28 – 4.16) ∗ ∗ P<0.05 versus healthy controls CS=caesarean section, M=maternal, U=umbilical 1 diastolic BP was measured 0 to 6 days before delivery 2 normal value for HbA1c in our laboratory is 4.2 – 6.3 %; in patients with diabetes is aimed at a value < 7.0 % 53 DM AFFECTS FOETOPLACENTAL VASCULAR KATP CHANNEL KATP channel blockade by glibenclamide Glibenclamide induced a concentration-dependent rise in foetoplacental arterial pressure to a maximum of 56 ± 6 mmHg in controls (n=7) versus 38 ± 4 mmHg in diabetes (n=6) (table 2). As a consequence, the absolute glibenclamide-induced increase in foetoplacental arterial pressure was significantly lower in diabetes (figure 1). Figure 2 shows the percentage increments in foetoplacental arterial pressure for all K channel blockers in both groups. This figure also shows the impaired vasoconstrictor response to glibenclamide in diabetes. The LogEC50 for glibenclamide was similar in both groups. KV channel blockade by 4-aminopyridine Addition of 4-aminopyridine caused a significant rise in foetoplacental arterial pressure, which was equal in controls (n=7) and in diabetes (n=7) (P<0.001 in controls as well as in diabetes) (figure 1, figure 2). LogEC50 was similar in both groups (table 2). SK channel blockade by apamine and BK/IK channel blockade by charybdotoxin Both apamine and charybdotoxin only caused minor increments in foetoplacental arterial blood pressure (table 2, figure 2), both in controls (n=6) and in diabetes (n=4). Comparison of foetoplacental vascular compliance between diabetes and control experiments In theory, the observed differences between controls and diabetes in the response to glibenclamide may relate to diabetes-induced structural changes in the vascular wall. Therefore, foetoplacental vascular compliance was studied. The diameter at an intraluminal pressure of 0 mmHg was 90 ± 20 µm in foetoplacental arteries from controls and 90 ± 3 µm (average ± SEM) in those from diabetes. The maximum diameter, measured at an intraluminal pressure of 60 mmHg, averaged 260 ± 35 µm (n=4) in controls and 270 ± 10 µm (n=4) in diabetes. The pressure-diameter relationship was similar between controls and diabetes. Figure 3 shows the absolute pressure-diameter relationship. The figure for the relative diameter as percentage of the diameter at maximal tested pressure of 60 mmHg is comparable (not shown). 54 CHAPTER 3 ∆ foetal arterial pressure (mm Hg) 40 30 20 10 0 -8 -7 -6 -5 -8 -4 -7 -6 glibenclamide (logM) -5 -4 -3 4-aminopyridine (logM) Figure 1 The increase in foetoplacental arterial pressure in response to glibenclamide (left panel) and 4-aminopyridine (right panel) in the placenta in controls (solid line) and in women with diabetes mellitus type 1 (dotted line) (mean ± SEM) Table 2. Mean ± SEM characteristics of the concentration-response curves for the four potassium channel blockers in the foetoplacental circulation Foetoplacental arterial pressure (mmHg) Blocker Group Baseline At maximum Max. increase LogEC50 Glibenclamide Control (n=7) 20 ± 1 56 ± 6 37 ± 6 -6.4 ± 0.2 Diabetes (n=6) 23 ± 3 38 ± 4 ∗ 15 ± 3 ∗ -6.4 ± 0.2 Control (n=7) 22 ± 2 43 ± 4 21 ± 4 -5.2 ± 0.2 Diabetes (n=7) 26 ± 2 48 ± 2 22 ± 2 -5.0 ± 0.2 Control (n=6) 27 ± 2 37 ± 3 10 ± 2 -7.4 ± 0.2 Diabetes (n=4) 30 ± 3 38 ± 5 9±3 -7.6 ± 0.3 Control (n=6) 23 ± 3 26 ± 3 4±1 -8.5 ± 0.2 Diabetes (n=4) 23 ± 3 24 ± 3 1±1 -10.2 ± 2.1 4-Aminopyridine Apamin Charybdotoxin ∗ P < 0.05 as compared to control 55 DM AFFECTS FOETOPLACENTAL VASCULAR KATP CHANNEL ∆ Foetoplacental arterial pressure (%) 200 100 ot ox in ch a ry bd ap am in ne id i py r in o 4am gl ib en cl am id e 0 Figure 2 Maximal percentage increase in foetoplacental arterial pressure induced by the highest dose of the different potassium channel blockers in controls (open bars) and women with diabetes mellitus type 1 (dotted bars) (mean ± SEM). The maximum concentrations given were 0.3 mM for glibenclamide, 1 µM for apamin, 0.1 µM for charybdotoxin and 0.3 mM for 4-aminopyridin. * = P < 0.05 as compared to controls Comparison of placental vascular morphometry between diabetes and control experiments The impaired vasoconstrictor response to glibenclamide in diabetes may also relate to anatomical differences in the vascular beds between controls and diabetes. Therefore we performed control experiments on morphometry in placentae from controls (n=5) and from women with diabetes (n=8). The diabetes placentae originated from patients with White Class B (n=2), Class C (n=4), and Class D (n=2). Consequently, none of the women with diabetes had a history of vascular disease. As shown in table 3, all values for the individual vessels were comparable between controls and diabetes. 56 CHAPTER 3 Inner diameter (µm) 400 300 200 100 0 1.0 3.0 5.0 10.0 20.0 40.0 60.0 Transmural pressure (mm Hg) Figure 3 Inner diameter of foetoplacental resistance vessels from healthy controls (open bars) and women with diabetes mellitus type 1 ( mean ± SD) Table 3. Mean (± SEM) of some foetoplacental vascular parameters as calculated by the Vidas PLUS image analysis system in controls and DM Healthy controls Diabetes mellitus type 1 P-value (n=8) (n=7) Vascular area (%)1 0.18 ± 0.01 0.18 ± 0.01 0.90 Perimeter (µm/mm2)2 0.064 ± 0.003 0.058 ± 0.005 0.32 Vascular count (N/mm2)3 960 ± 70 830 ± 130 0.39 Area (%) 207 ± 23 290 ± 74 0.28 Perimeter (µm) 69 ± 4 80 ± 11 0.33 Diameter (µm) 8.5 ± 0.4 9.6 ± 1.1 0.34 Per field Per vascular element 1 CD34 positively stained area/total chorionic area x 100 2 total perimeter of vascular elements/total chorionic area x 106 3 total number of vascular elements/mm2 chorionic area 57 DM AFFECTS FOETOPLACENTAL VASCULAR KATP CHANNEL DISCUSSION The key observation in this study is that the vascular KATP and KV channel function significantly contributes to the baseline vascular tone of the ex vivo human foetoplacental circulation. In diabetes, this effect appears to be impaired for the vascular KATP channel, but not for the vascular KV channel. Under resting conditions, the KCa channels hardly contribute to baseline foetoplacental vascular tone, neither in women with diabetes nor in healthy women. Opening of potassium channels in vascular smooth muscle cells induces vasodilation by hyperpolarizing the cell membrane. By this hyperpolarization, voltage sensitive calcium channels will close resulting in a fall in calcium influx, a fall in intracellular calcium concentration and subsequent vasorelaxation. In resting conditions, vascular potassium channels may be open or closed depending on the type of tissue [11,12]. In our set up, the foetoplacental pharmacological vascular blockade of bed the shows a clear KATP channel vasoconstrictor response to by glibenclamide. From these observations, we conclude that the KATP channel is opened in the ex vivo perfused foetoplacental vascular bed, and contributes to baseline vascular tone. A similar line of reasoning concerns the KV channel since we observed a relevant vasoconstrictor response to 4-aminopyridine. Our observation on an impaired function of the vascular KATP channel in diabetes mellitus is unique as far as the foetoplacental vascular bed is concerned. However, other investigators have observed similar data in other vascular beds in diabetes [6,12]. In theory, the attenuated response to glibenclamide in diabetes may be related to dysfunction of the vascular KATP channel itself, for example as a result of glycosylation. The ATP/ADP ratio is a major determinant of the open state probability of KATP channels. As such, hypoxia or metabolic stress is a well-known trigger for the opening of KATP channels. From a theoretical point of view, the vasoconstrictor response to glibenclamide may be explained by the fact that the cotyledon is relatively hypoxic in our perfusion model. An argument against this mechanism is that the perfusion fluid is oxygenated intensively resulting in a pO2 in the inflow perfusion fluid between 400 and 550 mmHg. However, the tissue oxygenation may be different from normal values, since the perfusion fluid does not contain an oxygen carrier, and we were not able to measure the tissue oxygen content during perfusion. Although previous experiments from our laboratory have shown a constant lactate production in the perfused cotyledon model, this does not imply a hypoxic state because the placenta has been shown to use the glycolysis pathway independent of the oxygen status [13]. Recently, the KV channel has been shown to play a role in the vasoconstrictor response to hypoxia in the human foetoplacental vascular bed [14]. These investigators observed a 58 CHAPTER 3 hypoxia-induced increase in the perfusion pressure from 27 to 33 mmHg. In our test with the KV channel blocker 4-aminopyridine, the perfusion pressure rose from 22 to 43 mmHg. Because the baseline perfusion pressure was low in our model, and the response to 4-aminopyridine was more pronounced than the reported response to hypoxia, we think that hypoxia could not have played an important role in our set up. During pregnancy the foetoplacental vascular bed may be exposed to ischemic insults, which can harm foetal development or even result in foetal death [15]. In theory, the effects of these ischemic periods on the vascular bed might be comparable to those in cardiac and brain tissue. In the latter organs, powerful endogenous mechanisms have been described against ischemic injury, like for example hypoxic vasodilation and ischemic preconditioning [16,17], both mechanisms that contribute to the optimal match between oxygen use and metabolic demand. The KATP channel has been shown to play a crucial role in these protective mechanisms [18]. Interestingly, diabetes mellitus has been associated with impaired ischemic preconditioning [19]. Our observation of an impaired vascular KATP channel function in the foetoplacental vascular bed in diabetes implies that the matching between oxygen supply and demand may be less optimal in these patients. Such a defect may contribute to a poor outcome of ischemic insults in the foetoplacental circulation in diabetes. Along this line of reasoning, the vascular KATP channel may be an interesting pharmacological target to improve perinatal morbidity and mortality in type 1 diabetes mellitus. Apart from effects on the vascular smooth muscle cell, potassium channel blockers may affect the endothelium. Studies on the contribution of potassium channels to endothelium-dependent vascular responses in human vessels show that this contribution is small or does not play a role at all [20,21]. As such, the primary site of action of potassium channel blockers is more likely to be the vascular smooth muscle cell than the endothelial cell. In theory, the impaired vasoconstrictor response to glibenclamide in diabetes might reflect a more non-specific defect in vasoconstrictor capacity. However, this seems not to be the case because the vasoconstrictor response to other stimuli (for example 4aminopyridine) is similar in diabetes versus controls. For the NO-synthase inhibitor LNAME, we previously observed an even more pronounced vasoconstrictor response in diabetes as compared to controls [6]. As such, our observation on an impaired vasoconstrictor response to glibenclamide in diabetes seems to be specific. This conclusion is supported by the fact that our compliance and morphometric studies did not reveal structural differences of the foetoplacental vessels between diabetes and controls. In 2001 Langer et al presented a study, which concluded that oral treatment with glibenclamide is a clinically effective alternative to insulin for the treatment of gestational diabetes [22]. Our study shows that vascular KATP channels 59 do have a DM AFFECTS FOETOPLACENTAL VASCULAR KATP CHANNEL substantial role in the regulation of baseline foetoplacental vascular tone. Since glibenclamide blocks these KATP channels, this treatment could compromise placental vascular function. Nevertheless, the transfer of glibenclamide across the human placenta seems to be minimal [23]. Oral treatment with glibenclamide may therefore be expected to only minimally affect foetoplacental vascular function, and as such may be preferred over other sulfonylurea derivatives, which may cross the placenta. In conclusion, our human ex vivo study in the foetoplacental vascular bed shows that both the vascular KATP channel and the KV channel significantly contribute to the regulation of baseline vascular tone. In type 1 diabetes, the function of the foetoplacental vascular KATP channel appears to be impaired. 60 CHAPTER 3 REFERENCE LIST [1] von Kries R, Kimmerle R, Schmidt JE, et al. Pregnancy outcomes in mothers with pregestational diabetes: a population-based study in North Rhine (Germany) from 1988 to 1993. Eur J Pediatr. 1997;156:963-967. [2] Poston L, Taylor PD. Glaxo/MRS Young Investigator Prize. Endothelium-mediated vascular function in insulin-dependent diabetes mellitus. Clin Sci (Colch ). 1995;88:245255. [3] Williams SB, Goldfine AB, Timimi FK, et al. Acute hyperglycemia attenuates endotheliumdependent vasodilation in humans in vivo. Circulation. 1998;97:1695-1701. [4] Scherrer U, Randin D, Vollenweider P, et al. Nitric oxide release accounts for insulin's vascular effects in humans. J Clin Invest. 1994;94:2511-2515. [5] Wilkes BM, Mento PF, Hollander AM. Reduced thromboxane receptor affinity and vasoconstrictor responses in placentae from diabetic pregnancies. Placenta. 1994;15:845-855. [6] Bisseling TM, Wouterse AC, Steegers EA, et al. Nitric oxide-mediated vascular tone in the foetal placental circulation of patients with type 1 diabetes mellitus. Placenta. 2003;24:974-978. [7] Wigg SJ, Tare M, Tonta MA, et al. Comparison of effects of diabetes mellitus on an EDHF-dependent and an EDHF-independent artery. Am J Physiol Heart Circ Physiol. 2001;281:H232-H240. [8] Weintraub NL. Impaired hypoxic coronary vasodilation and ATP-sensitive potassium channel function: a manifestation of diabetic microangiopathy in humans? Circ Res. 2003;92:127-129. [9]. Smits BW, Siero HL, Ellenbroek BA, et al. Stress susceptibility as a determinant of the response to adrenergic stimuli in mesenteric resistance arteries of the rat. J Cardiovasc Pharmacol. 2002;40:678-683. [10] Nelen WL, Bulten J, Steegers EA, et al. Maternal homocysteine and chorionic vascularization in recurrent early pregnancy loss. Hum Reprod. 2000;15:954-960. [11] Nielsen JJ, Kristensen M, Hellsten Y, et al. Localization and function of ATP-sensitive potassium channels in human skeletal muscle. Am J Physiol Regul Integr Comp Physiol. 2003;284:R558-R563. [12] Farouque HM, Worthley SG, Meredith IT, et al J. Effect of ATP-sensitive potassium channel inhibition on resting coronary vascular responses in humans. Circ Res. 2002;90:231-236. 61 DM AFFECTS FOETOPLACENTAL VASCULAR KATP CHANNEL [13] Van der Aa, E.A. Drug transport and drug-nutrient interactions in the human placenta. Thesis, Chapter 7, 113-124. Nijmegen: University Medical Centre, the Netherlands. 1995 (Thesis). [14] Hampl V, Bibova J, Stranak Z, et al. Hypoxic foetoplacental vasoconstriction in humans is mediated by potassium channel inhibition. Am J Physiol Heart Circ Physiol. 2002;283:H2440-H2449. [15] Burke CJ, Tannenberg AE, Payton DJ. Ischaemic cerebral injury, intrauterine growth retardation, and placental infarction. Dev Med Child Neur 1997; 39:726-730 [16] Cleveland JC, Jr., Wollmering MM, Meldrum DR, et al. Ischemic preconditioning in human and rat ventricle. Am J Physiol 1996; 271:H1786-H1794. [17] Weih M, Kallenberg K, Bergk A, et al. Attenuated stroke severity after prodromal TIA: a role for ischemic tolerance in the brain? Stroke 1999; 30(9):1851-1854. [18] Speechly-Dick ME, Grover GJ, Yellon DM. Does ischemic preconditioning in the human involve protein kinase C and the ATP-dependent K+ channel? Studies of contractile function after simulated ischemia in an atrial in vitro model. Circ Res 1995; 77(5):10301035. [19] Ishihara M, Inoue I, Kawagoe T, et al. Diabetes mellitus prevents ischemic preconditioning in patients with a first acute anterior wall myocardial infarction. J Am Coll Cardiol 2001; 38(4):1007-1011. [20] Mc Auley D, McGurk C, Nugent AG, Hanratty C, Maguire S, Johnston GD. Forearm endothelium-dependent vascular responses and the potassium-ATP channel. Br J Clin Pharmacol 1997;44:292-294 [21] Hamilton CA, Berg G, Mc Arthur K, Reid JL, Dominiczak AF. Does potassium channel opening contribute to endothelium-dependent relaxation in human internal thoracic artery? Cli Sci (Lond) 1999;96:631-638 [22] Langer O, Conway DL, Berkus MD, et al. A comparison of glyburide and insulin in women with gestational diabetes mellitus. N Engl J Med. 2000;343:1134-1138. [23] Elliott BD, Langer O, Schenker S, et al. Insignificant transfer of glyburide occurs across the human placenta. Am J Obstet Gynecol. 1991;165:807-812. 62 CHAPTER 4 DYSFUNCTION OF THE CYCLOOXYGENASE PATHWAY IN THE FOETOPLACENTAL CIRCULATION IN TYPE 1 DIABETES MELLITUS T.M. Bisseling, A.C. Wouterse, E.A.P. Steegers, L. Elving, F.G.M. Russel, P. Smits Diabetic Medicine 2004 [in press] © Blackwell Publishing Ltd CYCLOOXYGENASE DYSFUNCTION IN THE FOETOPLACENTAL CIRCULATION IN DIABETES SUMMARY In diabetes, perinatal morbidity is significantly increased. This may partly be related to functional changes in the foetoplacental vascular bed. In diabetes models, a defect in the cyclooxygenase pathway is a common observation. Therefore, we hypothesise that the human foetoplacental circulation of diabetic patients is characterised by dysfunction of the cyclooxygenase pathway, as well. We performed ex-vivo perfusions of isolated placental cotyledons from healthy women (n=14) and from patients with type 1 diabetes (n=9). The contribution of cyclooxygenase products to foetoplacental vascular tone was quantified by measuring the response to the cyclooxygenase inhibitor indomethacin. Baseline foetoplacental arterial pressure was comparable between controls and diabetic women (mean ± SEM, 21.7 ± 1.2 mmHg versus 24.4 ± 2.0 mmHg). Maximum foetoplacental arterial pressure at highest dose of indomethacin was 32.8 ± 3.0 mmHg in controls versus 27.3 ± 2.3 mmHg in diabetic women. The indomethacin-induced increase in pressure was reduced in diabetes (2.9 ± 0.7 mmHg versus 11.2 ± 2.4 mmHg in controls, P=0.01). Under baseline conditions, the net effect of all cyclooxygenase products in the foetoplacental vascular bed is vasodilation. In diabetes, this vasodilator effect seems significantly impaired. 66 CHAPTER 4 INTRODUCTION In diabetes mellitus type 1, pregnancies are more frequently complicated by intrauterine growth retardation and foetal death [1]. Since diabetes is known to affect vascular function [2], these complications may in part be explained by functional changes in the foetoplacental circulation. Under physiological conditions, the baseline vascular tone of the foetoplacental circulation is low. The endothelial release of prostanoids is expected to contribute significantly to the low-pressure state of the foetoplacental circulation. In diabetes, dysfunction of the cyclo-oxygenase pathway has been shown in several vascular beds, even in patients without micro- or macrovascular complications [3]. Therefore, we hypothesize that the exposure of the foetoplacental vascular bed to the diabetic environment throughout pregnancy may result in a dysfunction of the cyclooxygenase pathway in this specific vascular bed. To address our hypothesis, we quantified the vasoconstrictor response to the cyclo-oxygenase blocker indomethacin by ex-vivo dual perfusion of isolated cotyledons from human placentae, both in healthy women and in patients with type 1 diabetes mellitus. PATIENTS AND METHODS Study population The experimental protocol was approved by the Medical Ethical Review Committee, and all participants gave written informed consent. All pregnant women with type 1 diabetes were eligible to participate in this study. Controls were healthy women with uncomplicated pregnancies. Exclusion criteria were multiple pregnancy, premature birth (<37 weeks gestation), retained placenta, pregnancy-induced hypertension (diastolic pressure > 90 mmHg on two following occasions), preeclampsia, and type 2 or gestational diabetes mellitus. In all cases, maternal and umbilical venous blood samples for insulin and C-peptide determination were taken within 15 minutes after delivery. Placenta perfusion Within 15 minutes after delivery of the placenta, a suitable cotyledon was selected for ex-vivo dual perfusion [4], as previously described by our laboratory [5]. The third or fourth order artery and vein were cannulated just before passage through the chorionic 67 CYCLOOXYGENASE DYSFUNCTION IN THE FOETOPLACENTAL CIRCULATION IN DIABETES plate. Foetal inflow was gradually increased to 6 ml/min, at which the baseline foetal arterial pressure equilibrates to 15 to 40 mmHg. The cotyledon was placed in a chamber with the maternal side facing upward. Maternal inflow was kept constant at 12 ml/min. The maternal outflow was collected and returned into the maternal reservoir. The perfusion fluid (150 ml for both sides) was 37.0 – 37.5°C, and was oxygenated with 95% O2/ 5% CO2 (pH 7,4). Because the foetal arterial inflow was kept constant, the foetal arterial pressure was considered to be a reflection of the foetal arterial resistance. After equilibration, the response of the foetoplacental arterial pressure to increasing concentrations of the cyclooxygenase-inhibitor indomethacin was investigated.. Indomethacin was added as a bolus injection into the foetal circulation in seven cumulative dosages (1 nM to 1 µM). After each dose, stabilization of the foetoplacental arterial pressure was awaited, which took maximally 30 min, before we added the next dose. Materials Krebs-Henseleit buffer was used for perfusion. Just before each experiment, indomethacin (Sigma, St. Louis, USA) was freshly prepared. Statistical analysis Clinical characteristics were compared by Mann-Whitney U test. Data on perfusion pressures were analysed using Prism 3.0 (Graphpad Software) by fitting individual concentration-response curves according to the sigmoid Emax model for each experiment. The indomethacin induced rise was calculated for each curve from the maximum foetoplacental pressure at highest dose of indomethacin, and the baseline foetoplacental pressure. Differences between groups were tested by the Kruskall-Wallis test (SPSS, SAS Institute, Cary, NC, USA). RESULTS Clinical characteristics of the participants Table 1 presents the clinical characteristics of the 14 healthy controls and the 9 patients with diabetes. Gestational age at delivery was shorter in diabetic patients because in this group labour is electively induced at 38 weeks in our hospital. Despite the exclusion of hypertension, mean diastolic blood pressure was higher in diabetic women as compared to controls. All women had a spontaneous onset of labour. In 4 diabetic women , a caesarean section had to be performed for the reason of prolonged labour. 68 CHAPTER 4 There was no difference in maternal venous insulin concentration between the groups. Maternal venous C-peptide concentration was lower in diabetes as compared with controls. Umbilical venous C-peptide and insulin concentrations were increased in diabetes (table 1). Table 1 Clinical characteristics of the participants (median and range) Control Type 1 Diabetes Number 14 9 Maternal age (years) 32.8 (22.8 – 37.2) 31.4 (27.5 – 38.0) Parity (number) 0 (0 - 2) 1 (0 - 2) Gestational age at delivery (weeks) 39.9 (38.6 - 41.3) 38.0 (37.0 - 38.6) ∗ Caesarian Section (number) 0 4 Birth weight (grams) 3115 (2505 – 3775) 3415 (2670 - 4875) 525 (370 - 680) 690 (435 - 950) ∗ Body Mass Index (kg/m ) 24.7 (18.6 – 27.6) 28.0 (20.6 – 35.9) Diastolic Blood Pressure (mmHg) 75.0 (60.0 - 90.0) 89.0 (55.0 - 90.0) ∗ Smokers (number) 1 1 Medication Paroxetine (n=1) Insulin (n=9) Induced labour with PGE2 (number (%) 1 (7) 9 (100) * Placental weight (grams) 2 White Class HbA1c 1st trimester (%)1 1,2 HbA1c 3rd trimester (%) Insulin (mE/ml) - 4 B, 2 C, 1 D, 1 F, 1 R - 6.2 (5.0 – 8.3) - 6.1 (4.8 - 7.1) M 20.0 (0.2 – 161.0) 28.0 (5.0 – 59.0) U 14.5 (6.0 – 49.0) 79.0 (28.0 – 216.0) ∗ C-peptide (nmol/l) M 0.97 (0.12 – 2.31) 0.12 (0. 00 – 0.27) ∗ U 0.55 (0.21 – 6.00) 1.61 (0.76 – 4.16) ∗ ∗ P<0.05 versus healthy controls 1 2 normal value for HbA1c in our laboratory is 4.2 – 6.3 %, determined between 32 and 37 weeks gestation; median 34 weeks 69 CYCLOOXYGENASE DYSFUNCTION IN THE FOETOPLACENTAL CIRCULATION IN DIABETES Ex vivo perfusion of the isolated cotyledon Baseline foetoplacental arterial pressure was comparable between controls and diabetic women (21.7 ± 1.2 mmHg versus 24.4 ± 2.0 mmHg, mean ± SEM). Maximum foetoplacental arterial pressure at the highest dose of indomethacin was 32.8 ± 3.0 mmHg in controls, versus 27.3 ± 2.3 mmHg in diabetic women. In diabetes, however, the indomethacin induced rise in foetoplacental arterial pressure was significantly reduced (2.9 ± 0.7 mm Hg versus 11.2 ± 2.4 mm Hg in controls, P = 0.01) (figure 1). LogEC50 was similar in both groups (–7.4 ± 0.2 mol/L in controls versus –7.8 ± 0.2 mol/L in diabetes). ∆ Fetoplacental arterial pressure (mmHg) 15 10 5 0 -9 -8 -7 -6 Indomethacin (logM) Figure 1 Mean (± SEM) indomethacin-induced increase in foetoplacental arterial pressure in type 1 diabetes (n=9, dashed line) versus controls (n=14, solid line) DISCUSSION This study shows that inhibition of cyclooxygenase by indomethacin induces vasoconstriction in the ex vivo perfused foetoplacental vascular bed from healthy women. This implies a more important role for vasodilator prostanoids like prostacyclin than for vasoconstrictor prostanoids like thromboxane A2 in this particular vascular bed. Moreover, the vasoconstrictor response to indomethacin is decreased in diabetes, 70 CHAPTER 4 suggesting a defect in the cyclooxygenase pathway in these patients. Since the lowpressure state of the foetoplacental circulation is largely dependent on the release of local vasodilator substances, our observation may be of relevance with respect to the increased perinatal mortality and morbidity in diabetes. A defect in the cyclooxygenase pathway is a rather common finding in diabetes models. However, we now report on a very specific vascular bed, which is exposed to the diabetic environment for the limited period of nine moths. Despite this short exposure, a clear difference was found between and diabetic and control patients. In previous studies, we already reported on alterations in vascular reactivity in the foetoplacental vascular bed in patients with diabetes [5]. Apparently, 38 weeks of pregnancy is long enough to induce clear changes in vascular function in the foetoplacental circulation. At baseline, perfusion pressure did not differ between groups. Based on the observed defect in the cyclooxygenase pathway, an increase in baseline vascular tone would be expected in diabetes. Apparently, other vasodilator mechanisms have compensated for this defect. The most probable candidate is the nitric oxide pathway, as we previously found an increased vasoconstrictor response to NO synthase-inhibition by L-NAME in diabetes [5]. Cyclooxygenase is involved in the formation of both prostacyclin and thromboxane A2, by stimulating the initial formation of prostaglandin endoperoxides from arachidonic acid. In a second step, PGI2 and TXA2 are formed by PGI-synthase and TXA-synthase respectively. Cyclooxygenase is involved in the formation of other prostanoids as well, although these are less relevant to vascular function. Although the total effect of the cyclooxygenase products on foetoplacental vascular tone is clear from our study, we are not informed on the respective contribution of PGI2 and TXA2. In theory, our data can be explained by both, an increased release of TXA2 and a reduced release of PGI2. In the literature, there is support for the latter mechanism since dysfunction of PGI2synthase may be caused by peroxynitrite [6,7], a radical oxygen species that is found to be increased in the foetoplacental vascular bed of diabetic patients [8]. Although increased synthesis of vasoconstrictor prostanoids has been suggested to occur in diabetes models, no such data are available concerning to the foetoplacental circulation. One study has reported on a reduced vasoconstrictor response to TXA2 as a result of a decrease in receptor affinity in placentae from patients with diabetes [9], but no data on the release of TXA2 were available. Of course, within the concept of receptor desensitisation, a reduced response would be compatible with an increased release of TXA2. The enzyme cyclooxygenase is known to exist in two isoforms; the cyclooxygenase-1 and cyclooxygenase-2. The cyclooxygenase-1 is expressed in most tissues and is responsible for the majority of the vascular effects of cyclooxygenase. Under baseline 71 CYCLOOXYGENASE DYSFUNCTION IN THE FOETOPLACENTAL CIRCULATION IN DIABETES conditions, the expression of cyclooxygenase-2 is low in almost all tissues. It is responsible for the response to tissue damage, and can be upregulated by (patho)physiological conditions, such as ischemia and infection. In our model we did not investigate the differences between selective inhibitors for both cyclooxygenase-1, and cyclooxygenase-2. Indomethacine is known as a combined cyclooxygenase-1/ cyclooxygenase-2 inhibitor. In theory, most of the vascular effects of indomethacin may be attributed to the effects of inhibition of cyclooxygenase-1. Studies on the role of cyclooxygenase-2 in vascular function are limited. We can not exclude that the use of selective cyclooxygenase-1 and cyclooxygenase-2 inhibitors would have resulted in different observations in our study. Insulin interacts with the cyclooxygenase pathway [10]. The most important interaction is the stimulation of prostacycline release by insulin [10]. If this mechanism was operative in our model, an increased vasoconstrictor response to cyclooxygenaseinhibition would have been expected. Actually, we observed the opposite. This discrepancy may be explained by the fact that the vasculature in our ex vivo model is not exposed to the in vivo concentrations of insulin during the experiment. As such, due to stimulation of prostacyclin production by high insulin levels, the in-vivo effects of cyclooxygenase inhibition on the foetoplacental vascular bed from women with diabetes may be different from our observations. Additionally, the use a buffer solution may be a limitation of our study because this excludes the interactions between prostanoids and blood cells. As such, the observations in this study are limited to the effects of prostanoids released from the vascular wall. The advantage of our approach is that we are sure that our observations refer to the vascular wall and not to indirect effects via blood cells. Prostaglandin E2 (PGE2) is applied (para)cervically to induce labour, and is known as a vasoconstrictor in the placental circulation [11]. As such, the use of PGE2 may have confounded our observations. However, a comparison of baseline foetoplacental arterial pressures from this and earlier studies, showed no effect of exogenous PGE2 in controls nor in diabetic women (results not shown). In the present study, no such comparison could be made, because in all diabetic women labour was induced with PGE2. The absence of an effect of PGE2 on baseline pressure suggests that the paracervically applied substance does not reach the foetoplacental circulation. Unfortunately, we cannot exclude an effect of PGE2 on the response to indomethacin. In conclusion, under baseline conditions, the net effect of all cyclo-oxygenase products is vasodilation in the foetoplacental vascular bed. In type 1 diabetes mellitus this vasodilator effect is significantly impaired. 72 CHAPTER 4 REFERENCES [1] von Kries R, Kimmerle R, Schmidt JE, Hachmeister A, Bohn O, Wollf HG. Pregnancyoutcomes in mothers with pregestational diabetes: a population-based study in North Rhine (Germany) from 1988 to 1993. Eur J Pediatr 1997; 156:963-967. [2] [3] De Vriese AS, Verbeuren TJ, Van D, V, Lameire NH, Vanhoutte PM. Endothelial dysfunction in diabetes. Br J Pharmacol 2000; 130:963-974. Hishinuma T, Tsukamoto H, Suzuki K, Mizugaki M. Relationship between thromboxane/prostacyclin ratio and diabetic vascular complications. Prostaglandins Leukot Essent Fatty Acids 2001; 65:191-196. [4] [5] Schneider H, Dancis J. Modified double-circuit in vitro perfusion of placenta (letter). Am J Obstet Gynecol 1984; 148:836. Bisseling TM, Wouterse AC, Steegers EA, Elving L, Russel FG, Smits P. Nitric oxidemediated vascular tone in the foetal placental circulation of patients with type 1 diabetes mellitus. Placenta 2003; 24:974-978. [6] Kossenjans W, Eis A, Sahay R, Brockman D, Myatt L. Role of peroxynitrite in altered foetal-placental vascular reactivity in diabetes or preeclampsia. Am J Physiol Heart Circ Physiol 2000; 278:H1311-H1319. [7] Cooke CL, Davidge ST. Peroxynitrite increases iNOS through NF-kappaB and decreases prostacyclin synthase in endothelial cells. Am J Physiol Cell Physiol 2002; 282:C395C402. [8] Lyall F, Gibson JL, Greer IA, Brockman DE, Eis AL, Myatt L. Increased nitrotyrosine in the diabetic placenta; evidence for oxidative stress. Diabetes Care 1998; 21: 17531758. [9] Wilkes BM, Mento PF, Hollander AM. Reduced thromboxane receptor affinity and vasoconstrictor responses in placentae from diabetic pregnancies. Placenta 1994; 15:845-855. [10] Kahn NN, Bauman WA, Hatcher VB, Sihna AK. Inhibition of platelet aggregation and the stimulation of prostacyclin 1993;265H:2160-2164 [11] synthesis by insulin in humans. Am J Physiol Rama Sastry BV, Hemontolor ME, Olenick M. Prostaglandin E2 in human placenta: its vascular effects and activation of prostaglandin E2 formation by nicotine and cotinine. Pharmacology 1999; 58:70-86. 73 PART II EFFECTS OF ANTIOXIDANTS ON THE NO PATHWAY IN THE FOETOPLACENTAL CIRCULATION CHAPTER 5 VITAMIN C IMPROVES THE DISTURBED NO-PATHWAY IN THE HUMAN FOETOPLACENTAL CIRCULATION IN PREECLAMPSIA Tanya M. Bisseling, Frans G.M. Russel, Simone Dekker, Eric A.P. Steegers, Paul Smits This work was published in a short version as: Anti-oxidants and preeclampsia. Lancet 2000; 355: 65 [LETTER] © Elsevier Ltd EFFECT OF VITAMIN C ON DISTURBED NO-PATHWAY IN PREECLAMPSIA SUMMARY Preeclampsia is associated with endothelial dysfunction, which is often related to oxidative stress. Oxidative stress might result in reduced effects of the endotheliumderived-relaxing-factor nitric oxide (NO). Antioxidants remove reactive oxygen species, resulting in an improvement of endothelial function. In this study, we investigated the contribution of the NO pathway to baseline vascular tone in the foetoplacental circulation from preeclampsia as compared to healthy controls. Additionally, we investigated if vitamin C, possessing antioxidant capacities, has a beneficial effect on the NO pathway in the human foetoplacental circulation in control pregnancies. We did so by use of the NO-synthase inhibitor L-NAME in an ex vivo cotyledon perfusion model. Cotyledons from 30 placentae were used in this study; 4 from women with preeclampsia, and 26 from healthy controls. Within the latter group 12 cotyledons were investigated after administration of 5 mmol/L vitamin C. At baseline, foetoplacental arterial pressure was comparable in all groups. The maximum L-NAME-induced rise in foetoplacental arterial pressure was attenuated in preeclampsia as compared to controls (39 ± 6.5 mmHg (mean ± SEM) versus 49 ± 2.1 mmHg, P < 0.05). Addition of vitamin C increased the maximum L-NAME-induced rise in foetoplacental arterial pressure to 77 ± 4.9 mmHg (P < 0.001). LogEC50 was comparable between the controls and preeclampsia (-4.6 ± 0.2 mol/L vs. –4.6 ± 0.2 mol/L). Addition of vitamin C, however, reduced the logEC50 to –5.1 ± 0.2 mol/L). In conclusion, the NO-mediated vasodilator component seems to be decreased in the foetoplacental vascular bed in preeclampsia as compared to controls. Administration of vitamin C increases the contribution of the NO pathway to baseline vascular tone in this vascular bed in controls. 78 CHAPTER 5 INTRODUCTION Preeclampsia is a major complication of pregnancy. Although the exact aetiology of the origin of preeclampsia remains uncertain, preeclampsia is associated with endothelial dysfunction [1]. Defective trophoblast invasion is one of the early pathophysiological features that occurs before the development of the maternal syndrome [2]. In the fullgrown foetoplacental vascular bed from women with preeclampsia an impairment in endothelium derived vasoconstrictor [3-6] as well as vasodilator [6-8] functions has been described. Lucy Chappell and colleagues [9] reason that endothelial dysfunction and reduced bioavailability of nitric oxide (NO) contribute to oxidative stress, in particular from the interaction between superoxide anion and NO-producing peroxynitrite radicals. Indeed, partly, oxidative stress might be an explanation for the origin of this endothelial dysfunction [10-11]. NO, the most important endothelium derived vasodilator in this vascular bed, is inactivated by reactive oxygen species (ROS), resulting in the formation of peroxynitrite, which affects foetoplacental vascular function [11]. In-vitro experiments have shown that vitamin C interferes with this reaction, thereby increasing the availability of NO [12]. Chappell and colleagues’ positive results on the ratio of plasminogen-activator inhibitor 1 and 2, and on clinical outcome, are comparable with a vitamin-C induced improvement of the endothelial NO pathway. As such, supplementation with the vitamin C and E to patients who run the risk of developing preeclampsia seemed to be beneficial to the prevention of this disorder [9]. The effect of anti-oxidants in the foetoplacental circulation, however, is unknown. We investigated the NO pathway in the human foetoplacental circulation from women with preeclampsia and healthy controls, using the perfused cotyledon technique to assess endothelial function. Additionally, we investigated the effect of acute administration of vitamin C on the contribution of the NO pathway to baseline foetoplacental tone in healthy controls. METHODS The Medical Ethical Review approved the experimental protocol. All patients gave their informed consent. Study population Pregnant women with preeclampsia were eligible to participate in this study. Preeclampsia was defined as a diastolic blood pressure ≥ 100 mmHg measured at least 79 EFFECT OF VITAMIN C ON DISTURBED NO-PATHWAY IN PREECLAMPSIA two times with an interval of 4 hours, in combination with proteinuria defined as a > 300 mg/24 h. Controls were healthy women after uncomplicated, normotensive pregnancy. Exclusion criteria were diabetes mellitus, multiple pregnancy, premature birth (<37 weeks gestation), foetal growth retardation, retained placenta, and HELLP syndrome (hemolysis, elevated liver enzymes and low platelets). Placenta perfusion Placentae were obtained immediately following delivery. Within 15 minutes, a suitable cotyledon was selected for ex-vivo dual perfusion [13]. The third or fourth order artery and vein were cannulated just before passage through the chorionic plate. Foetal inflow was gradually increased to 6 ml/min, at which the baseline foetoplacental arterial pressure equilibrates between 15 to 40 mmHg. The cotyledon was placed in a chamber with the maternal side facing upward. Maternal inflow was kept constant at 12 ml/min. The maternal outflow was collected and returned into the maternal reservoir. A recirculating system was used for both foetal and maternal side. The perfusion fluid (Krebs-Henseleit buffer without albumin, pH 7.4: 150 ml for both sides) was 37°C, and was oxygenated with 95% O2 and 5% CO2. Because the foetal arterial inflow was kept constant, the foetoplacental arterial pressure was considered to be a reflection of the foetoplacental arterial resistance. After 30 minutes of equilibration, when the remnant blood had been washed out and the foetoplacental arterial pressure was stabilised, the experiment was started. The NOmediated component of baseline vascular tone in the foetal circulation was investigated by addition of the specific NO-synthase blocker N(G)-nitro-arginine-methyl-ester (LNAME) over a concentration range from 1 to 500 µM in a cumulative way. L-NAME was added to the foetal circulation in six dosages. After each addition stabilization of the foetoplacental arterial pressure was awaited, which took maximally 30 min, before we added the next dose. In 12 of the cotyledons from placentae from uncomplicated pregnancies, fifteen minutes prior to the first dose of L-NAME, vitamin C was dissolved in the Krebs Henseleit buffer (KH buffer) in the foetal perfusion fluid, to a final circulating concentration of 5 mmol/L. Materials A KH buffer containing 121 mM NaCl, 4 mM KCl, 0.95 mM KH2PO4, 1.2 mM MgSO47H2O, 22 mM NaHCO3, 11.1 mM glucose-H2O, and 2 mM CaCl2 was used as perfusion fluid. Heparin was used in a concentration that does not affect vascular tone (2500 IE/L). L-NAME and vitamin C were obtained from Sigma (St. Louis, USA). 80 CHAPTER 5 Statistical analysis Comparison of the clinical characteristics was performed by a Mann Whitney U test. Data on perfusion pressures were analysed using Prism 3.0 (Graphpad Software) by fitting individual concentration-response curves to L-NAME for each experiment. For each experiment, the relevant curve characteristics were calculated. These were baseline foetoplacental arterial pressure, maximum L-NAME-induced foetoplacental arterial pressure, logEC50 [log of concentration at which 50% of maximum pressure occurred], and Hill slope. For differences between groups these curve characteristics were compared by one-way ANOVA, with post-hoc Tukey’s multiple comparison. RESULTS We studied 30 cotyledons, four from patients with preeclampsia, and 26 from normotensive women with pregnancies without complications. Within the latter group, 12 cotyledons were investigated after co-administration of 5 mmol/L vitamin C. Table 1 Clinical characteristics Control Preeclampsia Number 26 4 Maternal age (years) 32.1 (26.8 – 38.8) 29.5 (20.8 – 37.0) Parity (n) 2 (0 – 3) 0 (0 – 1) Gestational age (weeks) 40.0 (38.1 – 42.0) 37.1 (36.3 – 38.3) ∗ Birth weight (grams) 3795 (2550 – 4560) 2619 (1775 – 3700) ∗ 600 (400 – 800) 400 (345 – 500) ∗ Body Mass Index (kg/m ) 24.2 (18.8 – 30.4) 28.6 (21.0 – 32.7) Diastolic blood pressure (mmHg) 77.2 (70.0 – 90.0) 98.0 (90.0 – 110.0) ∗ Caesarean section (n) 12 (46 %) 2 (50%) Smokers (n) 5 (19%) 1 (25%) Placental weight (grams) 2 Data are given as medians (min-max) or numbers ∗ = P < 0.05 81 EFFECT OF VITAMIN C ON DISTURBED NO-PATHWAY IN PREECLAMPSIA Clinical characteristics The clinical characteristics from the women that participated in this study are presented in table 1. All women with preeclampsia had a proteinuria > 300 mg/24h, whereas all other women had no proteinurea. As expected diastolic blood pressure was increased in preeclampsia, and birth and placental weight were decreased as compared to controls. Effects of L-NAME in isolated cotyledons There were no differences in baseline perfusion pressures between groups. In cotyledons from preeclamptic women, L-NAME increased the perfusion pressure concentration dependently from mean 14 ± 4 mmHg (mean ± SEM) to 39 ± 6.5 mmHg (figure 1). This vasoconstrictor response was significantly higher in 14 cotyledons from healthy pregnant women (16 ± 3 mmHg to 49 ± 2.1 mmHg, P < 0.05 vs. controls). LogEC50 was the same in controls (-4.6 ± 0.2 mol/L) as in women with preeclampsia (-4.6 ± 0.2 mol/L). Hill slopes were similar in all three groups, and were not significantly different from 1.0. Fetal arterial pressure (mmHg) 100 75 50 25 0 -9-8 -6 -5 -4 -3 -2 L-NAM E (logM ) Figure 1 L-NAME-induced rise in foetoplacental arterial pressure in healthy controls without (○) and with ( ●) addition of vitamin C, and preeclampsia without (□). 82 CHAPTER 5 Effects of vitamin C In 12 cotyledons exposed to vitamin C the vasoconstrictor response to L-NAME was greatest (14 ± 9 mmHg to 77 ± 4.9 mmHg, P < 0.001 vs. controls without vitamin C). LogEC50 was significantly lower in the vitamin C group (-5.1 ± 0.2 mol/L), P < 0.001 vs. other groups). DISCUSSION These results suggest that the NO-mediated contribution to baseline foetoplacental vascular tone is increased in controls as compared to that in women with preeclampsia. Additionally, the contribution of the NO pathway to baseline tone was increased in the presence of vitamin C. NO has been identified as an important endogenous regulator of baseline vascular tone. Also in the placental circulation, the endothelial release of NO significantly dilates the foetoplacental vascular bed [14;15]. Our study suggests that the contribution of NO to baseline vascular tone is decreased in preeclampsia. In contrast with these results, an increased production of NO [16;17], and en increased expression of eNOS [18] have been reported in this vascular bed in preeclampsia. However, recently others did not observe a dysfunction of eNOS in preeclampsia [19]. As such, despite an increased production of NO in preeclampsia, its function seems to be disturbed. Reason for this may be that NO is rapidly inactivated by the superoxide-anion, resulting in the formation of peroxynitrite. Peroxynitrite is increased in the foetoplacental vascular bed in preeclampsia, and affects [20] foetoplacental vascular function [11]. Vitamin C is a potent water-soluble antioxidant [21]. The effect of vitamin C on vascular function has been beneficial in different human vascular beds under various pathological conditions [22;23], however, no data exist on the foetoplacental vascular bed. It is unlikely that this beneficial effect of vitamin C is a result of scavenging of the superoxide radical, because the interaction between the superoxide radical and NO has place rapidly with a high rate constant of about 1.9 1010 M/s [24]. As such, the mechanism of action of vitamin C remains elusive. In conclusion, the contribution of NO to baseline vascular tone is impaired in the foetalplacental circulation from patients with preeclampsia. Vitamin C is able to increase the NO-mediated vasodilator component in this vascular bed from healthy controls. 83 EFFECT OF VITAMIN C ON DISTURBED NO-PATHWAY IN PREECLAMPSIA REFERENCES [1] Roberts JM, Taylor RN, Musci TJ, Rodgers GM, Hubel CA, McLaughlin MK. Preeclampsia: an endothelial cell disorder. Am J Obstet Gynecol 1989; 161(5):12001204. [2] Meekins JW, Pijnenborg R, Hanssens M, McFadyen IR, van Asshe A. A study of placental bed spiral arteries and trophoblast invasion in normal and severe pre-eclamptic pregnancies. Br J Obstet Gynaecol 1994; 101(8):669-674. [3] Singh HJ, Rahman A, Larmie ET, Nila A. Endothelin-l in feto-placental tissues from normotensive pregnant women and women with pre-eclampsia. Acta Obstet Gynecol Scand 2001; 80(2):99-103. [4] Walsh SW, Wang Y. Trophoblast and placental villous core production of lipid peroxides, thromboxane, and prostacyclin in preeclampsia. J Clin Endocrinol Metab 1995; 80(6):1888-1893. [5] [6] Walsh SW. Preeclampsia: an imbalance in placental prostacyclin and thromboxane production. Am J Obstet Gynecol 1985; 152(3):335-340. Cruz MA, Gonzalez C, Gallardo V, Lagos M, Albornoz J. Venous placental reactivity to serotonin in normal and preeclamptic gestants. Gynecol Obstet Invest 1993; 36(3):148152. [7] Myatt L, Eis AL, Brockman DE, Greer IA, Lyall F. Endothelial nitric oxide synthase in placental villous tissue from normal, pre-eclamptic and intrauterine growth restricted pregnancies. Hum Reprod 1997; 12(1):167-172. [8] [9] King RG, Gude NM, Di Iulio JL, Brennecke SP. Regulation of human placental fetal vessel tone: role of nitric oxide. Reprod Fertil Dev 1995; 7(6):1407-1411. Chappell LC, Seed PT, Briley AL, Kelly FJ, Lee R, Hunt BJ et al. Effect of antioxidants on the occurrence of pre-eclampsia in women at increased risk: a randomised trial [see comments]. Lancet 1999; 354(9181):810-816. [10] [11] Walsh SW. Maternal-placental interactions of oxidative stress and antioxidants in preeclampsia. Semin Reprod Endocrinol 1998; 16(1):93-104. Kossenjans W, Eis A, Sahay R, Brockman D, Myatt L. Role of peroxynitrite in altered fetal-placental vascular reactivity in diabetes or preeclampsia. Am J Physiol Heart Circ Physiol 2000; 278(4):H1311-H1319. [12] van der Vliet A, V, Smith D, O'Neill CA, Kaur H, Darley-Usmar V, Cross CE et al. Interactions of peroxynitrite with human plasma and its constituents: oxidative damage and antioxidant depletion. Biochem J 1994; 303 ( Pt 1):295-301. [13] Schneider H, Dancis J. Modified double-circuit in vitro perfusion of placenta [letter]. Am J Obstet Gynecol 1984; 148(6):836. [14] Myatt L, Brewer A, Brockman DE. The action of nitric oxide in the perfused human fetalplacental circulation. Am J Obstet Gynecol 1991; 164(2):687-692. 84 CHAPTER 5 [15] Gude NM, King RG, Brennecke SP. Role of endothelium-derived nitric oxide in maintenance of low fetal vascular resistance in placenta [letter; comment]. Lancet 1990; 336(8730):1589-1590. [16] [17] Lyall F, Young A, Greer IA. Nitric oxide concentrations are increased in the fetoplacental circulation in preeclampsia. Am J Obstet Gynecol 1995; 173(3 Pt 1):714-718. Norris LA, Higgins JR, Darling MR, Walshe JJ, Bonnar J. Nitric oxide in the uteroplacental, fetoplacental, and peripheral circulations in preeclampsia. Obstet Gynecol 1999; 93(6):958-963. [18] Myatt L, Eis AL, Brockman DE, Kossenjans W, Greer I, Lyall F. Inducible (type II) nitric oxide synthase in human placental villous tissue of normotensive, pre-eclamptic and intrauterine growth- restricted pregnancies. Placenta 1997; 18(4):261-268. [19] Orange SJ, Painter D, Horvath J, Yu B, Trent R, Hennessy A. Placental endothelial nitric oxide synthase localization and expression in normal human pregnancy and preeclampsia. Clin Exp Pharmacol Physiol 2003; 30(5-6):376-381. [20] Myatt L, Rosenfield RB, Eis AL, Brockman DE, Greer I, Lyall F. Nitrotyrosine residues in placenta. Evidence of peroxynitrite formation and action. Hypertension 1996; 28(3):488493. [21] [22] Frei B, England L, Ames BN. Ascorbate is an outstanding antioxidant in human blood plasma. Proc Natl Acad Sci U S A 1989; 86(16):6377-6381. Taddei S, Virdis A, Ghiadoni L, Magagna A, Salvetti A. Vitamin C improves endotheliumdependent vasodilation by restoring nitric oxide activity in essential hypertension. Circulation 1998; 97(22):2222-2229. [23] Schindler TH, Nitzsche EU, Munzel T, Olschewski M, Brink I, Jeserich M et al. Coronary vasoregulation in patients with various risk factors in response to cold pressor testing: contrasting myocardial blood flow responses to short- and long-term vitamin C administration. J Am Coll Cardiol 2003; 42(5):814-822. [24] Kissner R, Nauser T, Bugnon P, Lye PG, Koppenol WH. Formation and properties of peroxynitrite as studied by laser flash photolysis, high-pressure stopped-flow technique, and pulse radiolysis. Chem Res Toxicol 1997; 10(11):1285-1292. 85 CHAPTER 6 N-ACETYLCYSTEINE RESTORES NO-MEDIATED EFFECTS IN THE FOETOPLACENTAL CIRCULATION FROM PREECLAMPTIC PATIENTS Tanya M. Bisseling, Eva Maria Roes, Maarten T.M. Raijmakers, Eric A.P. Steegers, Wilbert H.M.Peters, Paul Smits Am J Obstet Gynecol 2004; 191: 328-333 © Elsevier Ltd NAC AMELIORATES NO PATHWAY IN PREECLAMPSIA SUMMARY Preeclampsia is associated with an imbalance between oxidants and anti-oxidants resulting in reduced effects of the endothelium-derived-relaxing-factor nitric oxide (NO). Antioxidants, like N-acetylcysteine (NAC) remove reactive oxygen species, resulting in an improvement of endothelial function. We aimed to investigate the effect of NAC on the NO pathway in the human foetoplacental circulation in preeclampsia and control pregnancies. The NO pathway was investigated by use of the NO-synthase inhibitor L-NAME in an ex vivo cotyledon perfusion model. At baseline, foetoplacental arterial pressure was comparable in preeclampsia (n=8) and controls (n=8), and increased dose-dependently after L-NAME. The maximal L-NAMEinduced rise in foetoplacental arterial pressure was attenuated in preeclamptic pregnancies versus controls (20.8 ± 2.0 mmHg versus 36.7 ± 3.5 mmHg, P < 0.05). Addition of NAC increased the L-NAME-induced rise in foetoplacental arterial pressure to 36.4 ± 3.4 mmHg in preeclamptic pregnancies (P < 0.05) and to 49.2 ± 2.6 mmHg in controls (P<0.05) . Preeclampsia is associated with a dysfunction of the NO pathway. N-acetylcysteine increases NO-mediated effects in the foetoplacental circulation in preeclamptic placentae as well as in healthy controls. 88 CHAPTER 6 INTRODUCTION Preeclampsia, a major complication of pregnancy, is associated with endothelial dysfunction in the maternal and foetal circulation. In this disease, there is probably an increase in oxidants, such as reactive oxygen species (ROS) [1;2] and increased lipid peroxidation as well as a decrease in several antioxidants, such as α-tocopherol βcarotene, ascorbic acid and gluthation [3-5]. This dysbalance may be responsible for the observed endothelial dysfunction. Recently, vitamin C and E supplementation to women with an increased risk for preeclampsia was shown beneficial in prevention of this disorder [6]. An optimal placental vascular function is important for normal foetal growth and wellbeing. Since the placenta lacks autonomic innervation, locally produced mediators like the endothelium derived relaxing factor nitric oxide (NO) play an important role in the maintenance of normal foetoplacental blood flow. NO is synthesized by the enzyme NO-synthase (NOS) in the endothelial cell, after stimulation by various factors (like e.g. shear stress, serotonin, bradykinin). NO diffuses to the vascular smooth muscle cell, where it activates the guanylate cyclase to increase intracellular cGMP. This is followed by a closure of calcium channels, a fall in intracellular calcium and a subsequent relaxation of the vessel wall. NO is inactivated by ROS resulting in the formation of peroxynitrite. As such, an increase in ROS in preeclampsia might induce dysfunction of the NO pathway in the foetoplacental arterial circulation. In the present study we tried to investigate whether the antioxidant substance N-acetylsteine (NAC) is able to affect the NO pathway. NAC is a pharmacological substance, which is rapidly deacetylated into cysteine, a direct precursor of gluthation. Whole blood gluthation concentrations were shown to increase after NAC administration to healthy, non-pregnant women [7]. Short-term intra-arterially administered gluthation improves endothelial function in the femoral circulation in patients with arteriosclerosis [8]. Apart from the indirect protection against ROS, being a precursor of gluthation, N-acetylcysteine itself is able to scavenge ROS by nonenzymatic reductions [9]. Along this line of reasoning, we hypothesize that administration of NAC in the foetoplacental vascular bed might improve endothelial function, especially that mediated by the NO pathway, in the placenta from women with preeclampsia. To address this, we investigated the effect on the NO pathway of acute administration of NAC in the foetoplacental vascular bed of placentae from women with mild preeclampsia as compared to placentae from healthy pregnant control women. 89 NAC AMELIORATES NO PATHWAY IN PREECLAMPSIA METHODS The Medical Ethical Review Committee approved the experimental protocol, and each subject gave written informed consent. Study population Pregnant women with mild preeclampsia were eligible to participate in this study. Mild preeclampsia was defined as a diastolic blood pressure > 90 mmHg measured at least two times with an interval of 4 hours, in combination with proteinuria defined as a protein/creatinine ratio > 30 mg/mmol, according to the criteria of the International Society for the Study of Hypertension in Pregnancy. Controls were healthy women after uncomplicated, normotensive pregnancy. Exclusion criteria were diabetes mellitus, multiple pregnancy, premature birth (<37 weeks gestation), foetal growth retardation, retained placenta, and HELLP syndrome (hemolysis, elevated liver enzymes and low platelets). Placenta perfusion Placentae were obtained immediately following delivery. Within 15 minutes, two suitable cotyledons were selected for ex-vivo dual perfusion [10]. Perfusion of both these cotyledons was performed simultaneously on two identical perfusion devices. The third or fourth order artery and vein were cannulated just before passage through the chorionic plate. Foetal inflow was gradually increased to 6 ml/min, at which the baseline foetoplacental arterial pressure equilibrates between 15 to 40 mmHg. The cotyledon was placed in a chamber with the maternal side facing upward. Maternal inflow was kept constant at 12 ml/min. The maternal outflow was collected and returned into the maternal reservoir. A recirculating system was used for both foetal and maternal side. The perfusion fluid (Krebs-Henseleit buffer without albumin, pH 7.4: 150 ml for both sides) was 37°C, and was oxygenated with 95% O2 and 5% CO2. Because the foetal arterial inflow was kept constant, the foetoplacental arterial pressure was considered to be a reflection of the foetoplacental arterial resistance. After 30 minutes of equilibration, when the remnant blood had been washed out and the foetoplacental arterial pressure was stabilised, the experiment was started. The NOmediated component of baseline vascular tone in the foetal circulation was investigated by addition of the specific NO-synthase blocker L-NAME over a concentration range from 1 to 500 µM in a cumulative way. L-NAME was added to the foetal circulation in 90 CHAPTER 6 six dosages. After each addition stabilization of the foetoplacental arterial pressure was awaited, which took maximally 30 min, before we added the next dose. Fifteen minutes prior to the first dose of L-NAME, NAC was dissolved in the Krebs Henseleit buffer (KH buffer) in one of the perfusion devices, to a final circulating concentration of 50 µmol/L, whereas KH buffer with no additives (placebo) was added in the other device. Materials A KH buffer containing 121 mM NaCl, 4 mM KCl, 0.95 mM KH2PO4, 1.2 mM MgSO47H2O, 22 mM NaHCO3, 11.1 mM glucose-H2O, and 2 mM CaCl2 was used as perfusion fluid. Heparin was used in a concentration that does not affect vascular tone (2500 IE/L) [11]. L-NAME and N-acetylcysteine were obtained from Sigma (St. Louis, USA). The NAC concentration in the perfusion fluid (50 µmol/L) was based on the plasma levels in a clinical study of healthy volunteers who received 3 x 1800 mg Nacetylcysteine daily. The median plasma level of N-acetylcysteine after three gifts of NAC was 48 µmol/L (range 37 - 71 µmol/L). Statistical analysis Comparison of the clinical characteristics was performed by a Mann-Whitney U test. Data on perfusion pressures were analysed using Prism 3.0 (Graphpad Software) by fitting individual concentration-response curves for each experiment. For each experiment, the curve characteristics were calculated. These were baseline foetoplacental arterial pressure, maximum L-NAME-induced foetoplacental arterial pressure, net L-NAME-induced increase in foetoplacental arterial pressure, and logEC50. For differences between groups these curve characteristics were compared by two-way ANOVA, with group (control or preeclampsia) and medication (vehicle or NAC) as independent factors. Differences were considered to be statistically significant when zero was not included in the 95 % confidence interval. RESULTS Clinical characteristics Apart from the expected differences in diastolic blood pressure and proteinuria, there were no statistically significant differences in clinical characteristics between the two 91 NAC AMELIORATES NO PATHWAY IN PREECLAMPSIA groups (Table 1). In the control group one patient used vitamin B12, in the preeclampsia group one patient used the low molecular weight heparin nadroparine (thrombosis profylaxis) and one used methyldopa. Table 1 Clinical characteristics Control Preeclampsia Number 8 8 Maternal age (years) 31.7 (28.2 – 38.0) 27.9 (26.3 – 41.3) Parity (n) 1 (0 – 3) 0 (0 – 3) Gestational age (weeks) 39.5 (38.0 – 41.1) 39.1 (37.6 – 39.6) Birth weight (grams) 3757 (2610 – 3990) 3175 (2745 – 4035) 595 (410 – 755) 582 (390 – 790) Body Mass Index (kg/m ) 21.7 (18.9 – 28.8) 27.5 (19.4 – 43.6) Diastolic blood pressure (mmHg) 77.5 (75.0 – 85.0) 98.0 (90.0 – 110.0) ∗ Caesarean section (n) 1 1 Smokers (number) 1 0 Protein/creatinine ratio (g/10 mmol) n.m 1.15 (0.59 – 3.38) ∗ Placental weight (grams) 2 Data are given as medians (min-max) or numbers ∗ = P < 0.05 n.m.: dipstick negative in all women Effects of L-NAME in isolated cotyledons Two cotyledons of 8 placentae from women with uncomplicated pregnancies and of 8 from women with mild preeclampsia were perfused. The concentration-response curves for L-NAME are presented in Figure 1. Baseline foetoplacental arterial pressure was comparable in all groups (Table 2). LNAME elicited a concentration-dependent rise in foetoplacental arterial pressure. Maximum L-NAME-induced foetoplacental arterial pressure in the placentae from healthy controls was 60.5 ± 3.1 mmHg (mean ± SEM). The absolute L-NAME-induced rise in foetoplacental arterial pressure was 36.7 ± 3.5 mmHg. In preeclampsia, the maximum L-NAME-induced foetoplacental arterial pressure was 45.6 ± 4.5 mmHg. The absolute L-NAME-induced increase in foetoplacental arterial 92 CHAPTER 6 pressure was 20.8 ± 2.0 mmHg (P <0.05 as compared to controls). There was no difference in logEC50 values between the groups. ∆ fetal arterial pressure (mmHg) 50 50 healthy controls 40 40 30 30 20 20 10 10 0 0 -6 -5 -4 preeclamptic patients -6 -5 -4 L-NAME (logM) Figure1 L-NAME-induced rise in foetoplacental arterial pressure in healthy controls without (□) and with (■) addition of N-acetylcysteine, and preeclampsia without (○) and with (●) Nacetylcysteine. The horizontal dottet line shows that after NAC the net foetoplacental arterial pressure in preeclampsia is comparable to that in the placenta of healthy controls without NAC. Effects of NAC Addition of NAC did not significantly affect baseline foetoplacental arterial pressure. After addition of NAC, the maximum L-NAME-induced foetoplacental arterial pressure was 72.2 ± 5.6 mmHg in the placentae of healthy controls and 59.4 ± 6.7 mmHg in preeclampsia. The absolute L-NAME-induced increase in foetoplacental arterial pressure was 49.2 ± 2.6 mmHg in controls and 36.4 ± 3.4 mmHg in preeclampsia. This NACeffect reached statistical significance in healthy controls (P<0.05) as well as in preeclampsia (P<0.05). 93 NAC AMELIORATES NO PATHWAY IN PREECLAMPSIA DISCUSSION The two main observations of our study are that 1) the contribution of NO to baseline vascular tone is impaired in the foetoplacental circulation of preeclamptic patients, and 2) the NO-mediated effects can be ameliorated to normal levels by addition of the antioxidant N-acetylcysteine. Table 2 Characteristics of the fitted concentration-response curves of the NOS inhibitor LNAME Number of experiments Control Control Preeclampsia Preeclampsia 8 8 8 8 Baseline FAP (mmHg) 23.7 (2.8) 23.0 (3.6) 24.8 (2.7) 23.0 (2.5) Maximal FAP (mmHg) 60.5 (3.1) 72.2 (5.6) 45.6 (4.5) 59.4 (6.7) ∆ L-NAME-induced increase 36.7 (3.5) 49.2 (2.6) # 20.8 (2.0) ∗ 36.4 (3.4) # -4.7 (0.1) -4.9 (0.1) -4.5 (0.1) -4.5 (0.1) in FAP (mmHg) Log EC 50 (logM) Statistical comparison by two-way ANOVA, with group (control or preeclampsia) and medication (vehicle or NAC) as independent factors. Data are presented as mean ± SEM. FAP = Foetoplacental Arterial Pressure ∗ P < 0.05 versus controls without NAC # P < 0.05 versus corresponding group without addition of NAC Our conclusions on the NO pathway are based on the quantitive assessment of the vasoconstrictor response to the NO-synthase inhibitor L-NAME. This is an indirect measure of the functional vasodilator effect of NO in this specific vascular bed. An attenuated vasoconstrictor response to L-NAME can be interpreted as a reduced effect of functional NO in the circulation. In a previous study, we already reported on an impaired effect of NO in the placenta of preeclamptic patients [12]. This reduced NO pathway in the foetoplacental vascular bed in preeclampsia is expected to be associated with an increase in baseline foetoplacental arterial pressure. In our study, however, we did not observe a difference in baseline foetoplacental arterial pressure between patients with preeclampsia and the healthy controls. Nevertheless, we did observe a significantly decreased response to L-NAME in preeclampsia, which proves that the NO pathway is impaired in this vascular bed in these patients. Other vasodilator mechanisms might be compensating for the dysfunction of the NO pathway 94 CHAPTER 6 in preeclampsia. Candidates for these compensatory mechanisms are an increased release of prostacycline or endothelium-derived hyperpolarizing factor (EDHF) on the one hand, and a reduced release of endothelin or vasoconstrictor prostanoids on the other hand. This kind of compensation has also been observed in other vascular beds with dysfunction of the NO pathway [13;14]. Additional experiments with the combined blockade of the aforementioned mechanisms are necessary to be sure whether this line of reasoning is correct. Other investigators have shown a decrease of NO-products in the urine of preeclamptic patients [15]. Of course, urinary NO-products reflect tubular or renal NO-production and/or secretion rather than indicating levels of NO in the foetoplacental circulation. Interestingly, two independent research groups reported an increase rather than a reduction in NO-products in uteroplacental and foetoplacental plasma or serum of preeclamptic women as compared to control pregnancies [16;17]. This increase in NOproducts has been attributed to a compensatory mechanism to offset the impaired placental perfusion, as frequently observed in preeclampsia [17]. The observations in our study can still be compatible with the reports on increased levels of NO-products in the preeclamptic placenta, because we quantified the vascular effects of NO, and not the concentrationsof NO itself, whereas functional effects of NO can be impaired because of a lower responsiveness of the tissue, or due to mechanisms that inactivate NO, like oxidative stress. The increase in foetoplacental arterial pressure in response to L-NAME could also be attributed to an excessive production of a vasoconstricting mediator. L-NAME, however, is a specific blocker of the endothelial NO-synthase (eNOS). As such, the change in foetoplacental arterial pressure somehow must be mediated by the NO pathway. Of course, the observed increase in foetoplacental arterial pressure may be an indirect effect of NO, since there is much interaction between the NO pathway and other endothelium mediated mechanisms, like the production of endothelins, and prostanoids. This implies that the decreased vasoconstrictor response to L-NAME in the foetoplacental vascular bed in preeclampsia could also be attributed to diminished effects of vasoconstricting mediators. In addition, alterations of vascular structure could play a role in the observed functional differences between the foetoplacental vascular bed from patients with preeclampsia as compared to healthy controls. On average, however, we did not observe a difference in baseline pressure between preeclampsia and the control group. Since perfusion flow, vessel length, and viscosity of the perfusion flow were similar, we can conclude from Poisseuille’s Law that the diameters should be similar in both groups. As such, differences in structure of the vessels do not seem to be the underlying cause of the observed functional differences between preeclampsia and controls. 95 NAC AMELIORATES NO PATHWAY IN PREECLAMPSIA The addition of N-acetylcysteine positively affected NO-dependent vasodilation in our model. As such a decrease in baseline foetoplacental arterial pressure would have been expected after addition of NAC. In our study, however, baseline foetoplacental arterial pressure was not affected by NAC. Again, this lack of a vasodilator effect of NAC can be explained by the putative activation of compensatory mechanisms. Additionally, we have to assume that during the subsequent administration of L-NAME the reductions in the NO pathway are so large that these cannot be compensated for by the aforementioned mechanisms. The effect of NAC on the NO pathway could be explained in different ways. At first, NAC has antioxidant properties, and has the ability to scavenge reactive oxygen species [18]. In preeclampsia, oxidative stress is a mediator of endothelial cell dysfunction through the excess of superoxide anion and other free radicals. Superoxide may have a direct vasoconstrictor effect on the vasculature or it may react with NO to produce peroxynitrite. Myatt and co-workers demonstrated the presence of nitrotyrosine residues in placental tissue of preeclamptic pregnancies, indicating the presence of peroxynitrite [19]. Thus the availability of NO to act as a vasorelaxant may be reduced, directly resulting in vasoconstriction [20]. Addition of NAC to the perfusion fluid may result in scavenging of superoxide anions and possibly other reactive oxygen species, leading to an increased availability of NO, resulting in NO-mediated vasodilatation [21]. Secondly, it is important to realise that NO can easily react with albumin to form S-NOalbumin, which is the predominant form of NO in plasma and is an efficient NO transporter to the smooth muscle cells of the vessels. NAC, being an aminothiol, might compete for NO with albumin (by transnitrosation) and thus may cause decreased availability of S-NO-albumin [22]. However, in the KH buffer we used during the perfusion experiments, no albumin was added and therefore this protein was not available for the formation of S-NO-albumin. In addition, N-acetylcysteine is rapidly deacetylated to cysteine, which is the key substrate in gluthation synthesis. Gluthation, one of the most important intracellular thiols, enhances the bioavailability of NO by forming a stable, biologically active Snitrosoglutathion adduct, as described in a group of patients with arteriosclerosis or risk factors for arteriosclerosis [23]. Moreover, also the sulfhydryl groups of Nacetylcysteine and cysteine will bind to nitric oxide resulting in formation of Snitrosothiols, a stored form of nitric oxide. The S-nitrosothiol compounds activate the guanylate cyclase enzyme causing vasodilatation [24]. We do not think that the latter mechanism was operative in our model because we did not observe a vasodilator effect of NAC alone on baseline vascular tone in the foetoplacental circulation. In theory, the vasodilator effect could also be independent from NO. Due to addition of N-acetylcysteine in the perfusion fluid an excess of free sulfhydryl groups is present, 96 CHAPTER 6 which could result in their auto-oxidation, with subsequent formation of superoxide and hydrogen peroxide. This may directly lead to a vasoconstrictor response or indirectly through the formation of peroxynitrite. Again, this is not plausible because we did not observe an increase in foetoplacental arterial pressure during our perfusion experiments after addition of NAC. Our results do not necessarily mean that oral NAC-treatment of preeclamptic women will have comparable effects on the foetoplacental circulation as described here in the ex vivo system. First, in vivo there may be an equilibrium between free and bound NAC, due to disulphide bonds [9]. In this ex vivo study, the availability of free NAC (SH) probably is higher as compared to the in vivo situation, because the perfusion fluid does not contain proteins, although we realise that NAC may be auto-oxidied. Additionally, a dose-response study to estimate the optimal dose for clinical use has not been performed yet. In vivo, over 80 % of NAC is bound to plasma and tissue proteins by labile sulphide bonds [9], and is therefore not available for scavenging of ROS. Further, the present study describes the effects of short-term administration of NAC to the perfusion buffer, resulting in an increase in the vasoconstrictor response to LNAME. This was in concordance with other studies regarding vascular function after administration of an antioxidant [8;12]. Various studies showed anti-oxidant biochemical effects of NAC without investigating vascular function [6;7;21]. Nacetylcysteine is transported across the placenta, suggesting that maternal NAC supplementation might reach the foetoplacental endothelium [25]. However, the effect on the foetoplacental vascular function of long-term maternal NAC administration remains unknown. In conclusion, ex vivo the contribution of NO to the baseline vascular tone is impaired in the foetoplacental circulation from preeclamptic patients. Moreover, the NO pathway can be ameliorated by addition of the anti-oxidant N-acetylcysteine to the perfusion buffer. 97 NAC AMELIORATES NO PATHWAY IN PREECLAMPSIA REFERENCES [1] Walsh SW. Maternal-placental interactions of oxidative stress and antioxidants in preeclampsia. Semin Reprod Endocrinol 1998;16:93-104. [2] Kossenjans W, Eis A, Sahay R, Brockman D, Myatt L. Role of peroxynitrite in altered foetal-placental vascular reactivity in diabetes or preeclampsia. Am J Physiol Heart Circ Physiol 2000;278:H1311-H1319. [3] Knapen MFCM, Mulder TPJ, Van Rooij IA, Peters WHM, Steegers EAP. Low whole blood glutathione levels in pregnancies complicated by preeclampsia or the hemolysis, elevated liver enzymes, low platelets syndrome. Obstet Gynecol 1998;92:1012-15. [4] Uotila JT, Tuimala RJ, Aarnio TM, Pyykko KA, Ahotupa MO. Findings on lipid peroxidation and antioxidant function in hypertensive complications of pregnancy. Br J Obstet Gynaecol 1993;100:270-76. [5] Chen G, Wilson R, Cumming G, Walker JJ, Smith WE, McKillop JH. Intracellular and extracellular antioxidant buffering levels in erythrocytes from pregnancy-induced hypertension. J Hum Hypertens 1994;8:37-42. [6] Chappell LC, Seed PT, Briley AL, Kelly FJ, Lee R, Hunt BJ et al. Effect of antioxidants on the occurrence of pre-eclampsia in women at increased risk: a randomised trial. Lancet 1999;354:810-16. [7] Roes EM, Raijmakers MTM, Peters WHM, Steegers EAP. Effects of oral Nacetylcysteine on plasma homocysteine and whole blood glutathione levels in healthy, non-pregnant women. Clin Chem Lab Med 2002;40:496-98. [8] Prasad A, Andrews NP, Padder FA, Husain M, Quyyumi AA. Glutathione reverses endothelial dysfunction and improves nitric oxide bioavailability. J Am Coll Cardiol 1999;34:507-14. [9] Holdiness MR. Clinical pharmacokinetics of N-acetylcysteine. Clin Pharmacokinet 1991;20:123-34. [10] Schneider H, Dancis J. Modified double-circuit in vitro perfusion of placenta. Am J Obstet Gynecol 1984;148:836. [11] Tiefenbacher CP, Chilian WM. Basic fibroblast growth factor and heparin influence coronary arteriolar tone by causing endothelium-dependent dilation. Cardiovasc Res 1997;34:411-17. [12] Bisseling TM, Russel FG, Dekker S, Steegers EA, Smits P. Antioxidants and preeclampsia. Lancet 2000;355:65. [13] Thollon C, Fournet-Bourguignon MP, Saboureau D, Lesage L, Reure H, Vanhoutte PM, Vilaine JP. Consequences of reduced production of NO on vascular reactivity of porcine coronary arteries after angioplasty: importance of EDHF. Br J Pharmacol 2002; 136: 1153-61 98 CHAPTER 6 [14] Kenny LC, Baker PN, Kendall DA, Randall MD, Dunn WR. Differential mechanisms of endothelium-dependent vasodilator responses in human myometrial small arteries in normal pregnancy and preeclampsia. Clin Sci (Lond) 2002; 103: 67-73 [15] Davidge ST, Stranko CP, Roberts JM. Urine but not plasma nitric oxide metabolites are decreased in women with preeclampsia. Am J Obstet Gynecol 1996;174:1008-13. [16] Norris LA, Higgins JR, Darling MR, Walshe JJ, Bonnar J. Nitric oxide in the uteroplacental, foetoplacental, and peripheral circulations in preeclampsia. Obstet Gynecol 1999;93:958-63. [17] Lyall F, Young A, Greer IA. Nitric oxide concentrations are increased in the foetoplacental circulation in preeclampsia. Am J Obstet Gynecol 1995;173:714-18. [18] Kelly GS. Clinical applications of N-acetylcysteine. Altern Med Rev 1998;3:114-27. [19] Myatt L, Rosenfield RB, Eis AL, Brockman DE, Greer I, Lyall F. Nitrotyrosine residues in placenta. Evidence of peroxynitrite formation and action. Hypertension 1996;28:48893. [20] Davidge ST. Oxidative stress and altered endothelial cell function in preeclampsia. Semin Reprod Endocrinol 1998;16:65-73. [21] Raijmakers MTM, Schilders GW, Roes EM, van Tits LJH, Hak-Lemmers HLM, Steegers EAP et al. N-acetylcysteine improves the disturbed thiol redox balance after methionine loading. Clin Sci 2003;105:173-80. [22] Scharfstein JS, Keaney JF, Jr., Slivka A, Welch GN, Vita JA, Stamler JS et al. In vivo transfer of nitric oxide between a plasma protein-bound reservoir and low molecular weight thiols. J Clin Invest 1994;94:1432-39. [23] Prasad A, Andrews NP, Padder FA, Husain M, Quyyumi AA. Glutathione reverses endothelial dysfunction and improves nitric oxide bioavailability. J Am Coll Cardiol 1999;34:507-14. [24] Stamler JS, Slivka A. Biological Chemistry of thiols in the vasculature and in vascularrelated disease. Nutrition Rev 1996;54:1-30. [25] Horowitz RS, Dart RC, Jarvie DR, Bearer CF, Gupta U. Placental transfer of Nacetylcysteine following human maternal acetaminophen toxicity. J Toxicol Clin Toxicol 1997;35:447-51. 99 PART III PLACENTAL FOLATE UPTAKE IN PREGNANCIES WITH FOETAL GROWTH RETARDATION CHAPTER 7 PLACENTAL FOLATE TRANSPORT AND BINDING ARE NOT IMPAIRED IN PREGNANCIES COMPLICATED BY FOETAL GROWTH RESTRICTION T.M. Bisseling, E.A.P. Steegers, J.J.M. van den Heuvel, H.L.M. Siero, F.M. van de Water, A.J. Walker, R.P.M. Steegers-Theunissen, P. Smits, F.G.M. Russel Placenta 2004; 25(56):588-593 © Elsevier Ltd PLACENTAL UPTAKE AND BINDING OF 5MTF IN FOETAL GROWTH RESTRICTION SUMMARY Maternal folate deficiency is associated with foetal growth restriction, however, transfer of folate across placentae from pregnancies complicated by foetal growth restriction has never been investigated. We studied whether maternal to foetal 5methyltetrahydrofolate (5MTF) transport in the ex vivo dually perfused isolated cotyledon, binding of [3H] folate (PteGlu) to the syncytial microvillous membrane, and protein expression of folate receptor-α (FRα) and reduced folate carrier (RFC) in these placentae are disturbed. Placental clearance of 5MTF from the maternal perfusate appeared to be non-saturable over a range of 50 to 500 nM, independent of albumin and flow-independent. No statistically significant differences between placentae complicated with foetal growth restriction and uncomplicated pregnancies were observed. Binding characteristics of [3H-]PteGlu to microvillous membranes of foetal growth restriction versus control placentae were similar: Bmax of 3.9 ± 2.0 (mean ± SD) versus 4.0 ± 1.6 pmol/mg protein and a Kd of 0.037 ± 0.010 versus 0.040 ± 0.018 nM. Expression of FRα and RFC were not different in placentae of both groups studied. In conclusion, foetal growth restriction appears not to be associated with impaired maternal to foetoplacental folate transport, placental receptor binding, or expression of FRα and RFC. 104 CHAPTER 7 INTRODUCTION Folate is an essential micronutrient for human beings, because of its role in a number of intracellular processes resulting in cell growth. Especially during pregnancy, an optimal placental uptake of folate from the maternal circulation is critical for development and growth of the placenta and an adequate supply of folate to the developing foetus. Maternal folate deficiency is associated with various complications of pregnancy under which a low birth weight [1-6]. The exact mechanism of maternal-to-foetal folate transport is still not completely elucidated. The initial uptake step in placental folate transfer most likely involves endocytosis mediated by the folate receptor-α (FRα) at the microvillous membrane of the syncytiotrophoblast. FRα is a glycosylphosphatidylinositol-anchored protein located at the microvillous membrane. This receptor is in direct contact with the maternal blood and binds folates with high affinity in the low nanomolar range [7-10]. On the basolateral side of the syncytiotrophoblast, folate efflux into the foetal blood is facilitated by the reduced folate carrier (RFC), which is driven by the transmembrane H+-gradient [11;12]. To our knowledge only one study on the kinetics of folate uptake in isolated human perfused placenta has been published [13]. Despite the low dissociation constant, placental uptake appeared not to be saturable at folate concentrations up to 1000 nM. No study has been performed yet on folate transport in placentae from pregnancies complicated by foetal growth restriction. During pregnancy, marginal availability of folate to the placenta or foetus can impair normal cellular growth and replication. This may arise from low folate intake or deficient placental uptake and subsequent transfer to the foetal blood. The aim of this study was to investigate whether there is a difference in folate binding and transfer between placentae from healthy controls and those from foetal growth restriction. METHOD Study populations All pregnant women with uncomplicated pregnancies were eligible to participate in this study. Foetal growth restriction was 105 diagnosed if the foetal abdominal PLACENTAL UPTAKE AND BINDING OF 5MTF IN FOETAL GROWTH RESTRICTION Table 1 Patient characteristics Foetal growth restriction Control Number 10 50 Maternal age (years) 28.3 (22.3 – 37.9) 32.9 (20.8 – 42.1) Parity (number) 0 (0 - 2) 1 (0 - 4) Gestation (days amenorrhoea) 267 (257 - 287) 276 (259 - 297) Birth weight (grams) 2095 (749 - 2575) ∗ 3375 (2375 - 4650) Placental weight (grams) 400 (200 - 635) ∗ 593 (350 - 1110) Gender child (number of boys (%)) 5 (50%) 24 (48%) Body Mass Index (kg/m ) 20.7 (17.7 – 25.5) 22.9 (18.3 – 34.0) Diastolic BP (mm Hg) 75.0 (70.0 – 90.0) 70.0 (40.0 – 90.0) Caesarean section (number (%)) 4 (40%) 16 (32%) Smokers (number (%)) 3 (30%) 9 (18%) Alcohol (number) 0 0 Medication (number) Oxazepam (1) Iron (3) Periconceptual folic acid use (number (%)) 5 (50%) 24 (48%) Maternal serum folic acid (nmol/L) 18.0 (8.4 – 27.4) 12.4 (6.7 – 32.0) Umbilical venous serum folic acid (nmol/L) 45.7 (5.9 – 65.7) 28.8 (9.3 – 40.7) Maternal erythrocyte folic acid (nmol/L) 896 (311 - 1595) 844 (213 - 1474) Umbilical venous erythrocyte folic acid 1728 (311 - 3448) 1254 (566 - 1818) 2 Data presented as median and range, ∗ P < 0.05 versus healthy controls. Table 2 Kinetic parameters characterising placental 5-methylTHF disappearance from the maternal perfusate of the ex vivo isolated perfused cotyledon of foetal growth restriction versus control Control Foetal growth restriction t1/2α (min) 10.2 (4.2) 4.4 (1.2) t1/2β (min) 209 (47) 388 (95) Co (nM) 173 (17) 155 (23) CL (ml/min) 0.90 (0.44) 1.32 (0.79) Values are presented as mean (SEM) 106 CHAPTER 7 circumference was below P10 on two following occasions, and growth had followed a curve within the normal ranges for the Dutch population during a longer period of pregnancy, until this deviation had been observed. Exclusion criteria for both groups were an intrauterine infection, a congenital abnormality, an abnormal umbilical pulsatility index [14] and/or cardiotocogram abnormalities prior to labour, multiple pregnancy, premature birth (<37 weeks amenorrhoea), retained placenta, pregnancyinduced hypertension (diastolic pressure >90 mmHg on two following occasions), preeclampsia and HELLP syndrome (hemolysis, elevated liver enzymes and low platelets). Twenty-eight placentae from women with uncomplicated pregnancies were used to study the kinetics of 5MTF foetal to maternal transfer. Twenty-two placentae from women with uncomplicated pregnancies were used as a control in the comparison with 10 placentae from pregnancies complicated with foetal growth restriction. Fifteen minutes after delivery maternal and umbilical venous blood samples for serum and erythrocyte folate determination were taken. The clinical characteristics as presented in table 1 were obtained from medical records. All women gave their written informed consent. The medical ethics committee of the University Medical Centre approved this study. Placenta perfusion Ex vivo placental perfusion was performed as described before [15]. Antipyrine (0.43 mM) was used as a diffusion marker and was measured in perfusate according to Brodie et al. [16]. If more than 10 % perfusion fluid had disappeared, after correction for loss due to sampling, the experiment was discarded. To avoid photodecomposition of 5MTF, the experiments were performed in subdued daylight. Dose-dependent transport in control placentae was studied by administering one dose of 5MTF into the maternal perfusate at t = 0 min to a final concentration varying from 0 to 500 nM. Samples of both foetal and maternal fluid were collected in dark reaction tubes containing 1% (w/v) of ascorbic acid to stabilize the 5MTF, immediately frozen in liquid nitrogen, and stored at -80°C until further assay. In addition, the influence of increased maternal flow (20 ml/min) and albumin binding (1.8 g/L) on placental folate uptake were investigated a concentration of 100 nM 5MTF. 5MTF was measured in maternal and foetal perfusion fluid by HPLC with fluorescence detection (details available upon request). 107 PLACENTAL UPTAKE AND BINDING OF 5MTF IN FOETAL GROWTH RESTRICTION Isolation of syncytial microvillous membranes Syncytial microvillous membranes were isolated from human term placentae from uncomplicated pregnancies (n=8) and from placentae from pregnancies with foetal growth restriction (n=4) according to method 3 as described by Glazier, Jones and Sibley [17], with some modifications as described by Van der Aa et al. [18]. To determine the purity of the microvillous membrane suspension, alkaline phosphatase was analysed in the starting mince and final suspension according to Mircheff and Wright [19]. Alkaline phosphatase had to be enriched at least 9 times, otherwise the material was not used. Binding of [3H]PteGlu to MMV The microvillous membrane suspension was diluted in Krebs Henseleit phosphate buffer, pH 7.4 (KH buffer) to a protein concentration of 4 µg/ml. Each assay was run in triplicate at 37°C. A solution of 100 µl [3H]folate ([3H ]PteGlu) in KH buffer and 150 µl KH buffer were transferred into glass test tubes (4 ml) in a concentration range of [3H]PteGlu from 0.02 to 1.70 nM. At t=0, 250 µl membrane suspension was added to a final concentration of 2 µg/ml protein and a final concentration range of [3H]PteGlu from 0.01 to 0.85 nM. In a second set of tubes at the three highest concentrations of [3H] PteGlu 150 µl 60.0 µM PteGlu in KH instead of KH was transferred to analyse the non-specific binding. The test tubes were vortexed and placed in a water bath of 37°C and incubated for 90 minutes. Incubations were stopped by adding 2.0 ml ice-cold KH buffer, immediately followed by vacuum filtration over a Whatman GF/F glass fibre filter (Whatman, Kent, UK). Every filter was washed three times with 5.0 ml ice-cold KH buffer to remove unbound [3H]PteGlu. Filters were put in vials containing 4.0 ml Aqualuma scintillation fluid (Lumac, Groningen, the Netherlands) and radioactivity retained on the filters was counted in a Beckman LS 6000 LL scintillation counter. Immunoblot analysis of FRα and Reduced Folate Carrier Western Blot analysis was performed on microvillous membrane fractions and crude mince for FRα, and on crude mince for RFC. For RFC only crude mince was used, because we did not have enough material left to isolate a basal membrane fraction as well. The immunoblot analysis was performed as described before for RFC [20]. To detect FRα proteins the same procedure was used with MOv-18 as primary antibody (Alexis Biochemicals, Montreal, Canada) followed by horseradish peroxidase conjugated goat anti mouse IgG as secundary antibody (Sigma Immunochemicals, St. Louis MO, USA). Intensities of bands were measured using Scion Image Release Beta 3b (1998) 108 CHAPTER 7 (www.scorpioncorp.com). The average pixel intensity for each placenta was calculated and averaged for the 8 placentae from healthy controls and for the 4 placentae from foetal growth restriction. Materials Perfusion fluid consisted of a KH buffer containing 121 mM NaCl, 4 mM KCl, 0.95 mM, KH2PO4, 1.2 mM MgSO4, 20 mM NaHCO3, 11.1 mM glucose, 2 mM CaCl2 and 2500 IE/L heparin. Antipyrine, 5MTF, and PteGlu were obtained from Sigma (St. Louis, USA), ascorbic acid from Merck (Darmstadt), [3H]PteGlu from Amersham (Buckinghamshire, UK). Statistical analysis Data were analysed by curve fitting of individual experiments using Prism 3.0 (Graphpad Software), Patient and placenta characteristics were compared using Mann Whitney U test or Fisher’s Exact test when appropriate (SPSS 10.0). RESULTS Patient characteristics Placentae from 60 women were investigated. The clinical characteristics are presented in table 1. Apart from foetal and placental weight, patient characteristics, including serum and erythrocyte folate, were comparable. Kinetics of 5MTF disappearance from maternal perfusate The kinetics of 5MTF disappearance from maternal perfusate at an initial concentration of 100 nM was compared with doses of 50 and 500 nM, and the influence of albumin (1.8 g/L) and an increased maternal flow (20 ml/min) was studied. Data of each individual experiment were analysed best according to a two-phase exponential decay curve. Rate constants alpha (α, initial decay) and beta (β, slow decay) and initial concentration (C0) were determined by curve fitting. Kinetic parameters after 100 nM (n=5) were, alpha half-life (t1/2α) = 7.3 + 2.6 min, beta half-life (t1/2β) = 530 + 70 min, and clearance (Cl) = 1.39 + 0.15 ml/min, and were not significantly different from 50 (n=4) and 500 nM (n=3) or influenced by albumin (n=3) or an increased 109 PLACENTAL UPTAKE AND BINDING OF 5MTF IN FOETAL GROWTH RESTRICTION maternal inflow (n=5). Appearance of 5MTF in the foetal perfusate increased with the concentration in maternal perfusion fluid and was about 6 % of the initial maternal concentration. Foetal growth restriction versus uncomplicated pregnancies Perfusion of the isolated placental cotyledon Fourteen placentae from controls and 6 from foetal growth restriction were used for the perfusion experiments. In each experiment 5MTF was added to the maternal perfusion buffer in a concentration of 200 nM at t = 0 min. The kinetic parameters did not differ between the groups (table 2), and no differences were observed in the percentage of the initial maternal concentration that appeared in the foetal perfusate. spec. bound [3H]PteGlu (pmol/mg protein) 5.0 2.5 0.0 0.0 0.2 0.4 0.6 0.8 3 free [ H]PteGlu (nM) Figure 1 Binding of [3H]PteGlu to syncytial microvillous membranes of placentae from pregnancies complicated by foetal growth restriction (n=4, ∆) and uncomplicated pregnancies (n=8, □) Binding of [3H]PteGlu to syncytiotrophoblast microvillous membranes Binding to microvillous membranes of 8 placentae from uncomplicated pregnancies was compared with those from 4 pregnancies complicated by foetal growth restriction. The purity of the preparation was checked by enrichment of alkaline phosphatase, which was 10.6 ± 0.4 (mean ± SEM) for the uncomplicated group, versus 12.7 ± 1.7 for the foetal growth restriction group. Concentration-dependent binding of [3H]-PteGlu was 110 CHAPTER 7 saturable and could be analysed according to a one-site model with a Bmax of 3.9 ± 2.0 pmol/mg protein for controls and 4.1 ± 1.6 pmol/mg for foetal growth restriction. The Kd was 0.037 ± 0.010 nM and 0.040 ± 0.018 nM, respectively (fig. 1). Figure 2 Immunoblot analysis of reduced folate carrier (RFC, upper pannel) and folate receptor-α (FRα, lower panel). Proteins (30 µg) were separated on a 10% SDS polyacrylamide gel, transferred to a nitrocellulose membrane, and incubated with polyclonal GST-hRFC antibody (RFC) or monoclonal MOv-18 antibody (FRα). Proteins were visualized using enhanced chemiluminescence. Lanes with placental proteins from foetal growth retarded pregnancies are marked with an asterix (∗) Immunoblot analysis of FRα and RFC The MOv-18 antiserum detected a protein of ~ 44kDa in microvillous membranes of placentae from healthy controls as well as in placentae from foetal growth retardation (fig. 2), which is in concordance with the weight for FRα found by others [21]. Pixel intensity imaging resulted in similar intensities for both the RFC protein and the FR-α protein in placentae from foetal growth retardation and healthy controls. For RFC, the average intensities were 1.40 ± 0.06 (mean ± SEM) in placentae from foetal growth retardation versus 1.30 ± 0.05 in placentae from healthy controls. For FRα protein the pixel intensities were 1.58 ± 0.17 and 1.52 ± 0.12, respectively. Western Blot Analysis for FRα with MOv-18 antibody in the crude mince showed a protein of ~ 44 kDa as well, with again pixel intensities that were not different between placentae from foetal growth restriction and healthy controls (results not shown). 111 PLACENTAL UPTAKE AND BINDING OF 5MTF IN FOETAL GROWTH RESTRICTION DISCUSSION Radioligand binding studies showed no difference in binding of folate to the microvillous membranes isolated from these placentae, and on Western Blot we found no difference between expression of both FRα and RFC between both groups studied. This was confirmed at a functional level in the perfused cotyledon, in which no difference in transport of 5MTF could be observed between placentae from healthy controls and those from foetal growth restriction. We found a dissociation constant of 0.04 nM for binding of [3H]PteGlu to its receptor on isolated syncytial microvillous membranes. In comparable studies, other authors reported higher dissociation constants of 0.9 nM [22], 1.8 nM [8] and 3.0 nM [9]. The reason for the discrepancy with our results is not clear. Maybe these authors characterized a lower affinity binding site, as they used 30–40 times higher concentrations of [3H]PteGlu in their binding assay. For 5MTF, the folate we used during the perfusion experiments, the Kd value is much higher and approximately 30 nM, as derived from competitive inhibition of [3H]PteGlu binding [22]. This value is in the range of physiological 5MTF concentrations and indicate that in our perfusion experiments at 200 nM, which is about ten times the serum concentrations following optimal folate suppletion [23], virtually all available receptors must have been occupied. A disturbed availability of folate could play a role in foetal growth restriction through different pathways, including maladaption of the placenta in early pregnancy or vascular placental insufficiency. In our study group, vascular placental insufficiency was unlikely to be a cause for foetal growth reduction, because cases with an abnormal umbilical pulsatility index and/or abnormal cardiotocogram were excluded. Low folates in the foetus could be also caused by maternal folate depletion, and a decreased transport of folate to the foetus. Foetal folate insufficiency may cause metabolic problems leading to an impaired foetal growth. Despite the fact that folate deficiency is associated with foetal growth restriction in observational studies(4;5), our observations show that in case of pregnancies complicated by foetal growth restriction, appearance of 5MTF in the foetal circulation as well as microvillous folate receptor binding and expression of FRα and RFC appeared not to be different from that in uncomplicated pregnancies. Foetal growth restriction, if due to a folate deficiency, is therefore more likely a result of a low supply caused by maternal folate deficiency. Additionally, other factors that could influence the rate of folate transfer have been described, like increased NO-levels in the retina [24], and a rise in pH in the colonic basal membrane [25]. Foetal growth restriction is, however, associated with a decrease in NO in the placenta [26]. Furthermore, if the foetus is small for gestation due to a placental insufficiency, it is more likely that the pH is decreased rather than increased. Our 112 CHAPTER 7 findings are in line with earlier observations that uptake of 5MTF from the maternal circulation in an isolated perfused cotyledon of the human placenta is not saturable and not limited by binding to the folate receptor [13], and from the present study we conclude that foetal growth restriction is not associated with impaired placental folate uptake. ACKNOWLEDGEMENTS The authors gratefully acknowledge Dr. L.H. Matherly from the Barbara Ann Karmanos Cancer Institute, Wayne State University, Detroit, MI, USA for the kind disposal of the hRFC antibody, Y. Tan for his skilful assistance with the set-up of the HPLC assay and Dr. D. Swinkels and Dr. G. van de Wiel of the Central Clinical Chemical Laboratory of the University Medical Centre Nijmegen for the validation of this assay. 113 PLACENTAL UPTAKE AND BINDING OF 5MTF IN FOETAL GROWTH RESTRICTION REFERENCES [1] Hibbard BM. The role of folic acid in pregnancy. With particular reference to anemia, abruption and abortion. Journal of Obstetrics and Gynecology , 529-542. 1964. [2] Steegers-Theunissen RP, Boers GH, Trijbels FJ, Eskes TK. Neural-tube defects and derangement of homocysteine metabolism. N Engl J Med 1991; 324(3):199-200. [3] Steegers-Theunissen RP, Boers GH, Blom HJ, Trijbels FJ, Eskes TK. Hyperhomocysteinaemia and recurrent spontaneous abortion or abruptio placentae. Lancet 1992; 339(8801):1122-1123. [4] Goldenberg RL, Tamura T, Cliver SP, Cutter GR, Hoffman HJ, Copper RL. Serum folate and foetal growth retardation: a matter of compliance? Obstet Gynecol 1992; 79(5 ( Pt 1)):719-722. [5] Scholl TO, Hediger ML, Schall JI, Khoo CS, Fischer RL. Dietary and serum folate: their influence on the outcome of pregnancy. Am J Clin Nutr 1996; 63(4):520-525. [6] George L, Mills JL, Johansson AL, Nordmark A, Olander B, Granath F et al. Plasma folate levels and risk of spontaneous abortion. JAMA 2002; 288(15):1867-1873. [7] Antony AC. Folate receptors. Annu Rev Nutr 1996; 16:501-521. [8] Henriques C, Trugo NM. Partial characterization of folate uptake in microvillous membrane vesicles isolated from human placenta. Braz J Med Biol Res 1996; 29(12):1583-1591. [9] Green T, Ford HC. Human placental microvilli contain high-affinity binding sites for folate. Biochem J 1984; 218(1):75-80. [10] Antony AC, Utley C, Van Horne KC, Kolhouse JF. Isolation and characterization of a folate receptor from human placenta. J Biol Chem 1981; 256(18):9684-9692. [11] Prasad PD, Ramamoorthy S, Leibach FH, Ganapathy V. Molecular cloning of the human placental folate transporter. Biochem Biophys Res Commun 1995; 206(2):681-687. [12] Dudeja PK, Torania SA, Said HM. Evidence for the existence of a carrier-mediated folate uptake mechanism in human colonic luminal membranes. Am J Physiol 1997; 272(6 Pt 1):G1408-G1415. [13] Henderson GI, Perez T, Schenker S, Mackins J, Antony AC. Maternal-to-foetal transfer of 5-methyltetrahydrofolate by the perfused human placental cotyledon: evidence for a concentrative role by placental folate receptors in foetal folate delivery. J Lab Clin Med 1995; 126(2):184-203. [14] [15] Lees C, Albaiges G, Deane C, Parra M, Nicolaides KH. Assessment of umbilical arterial and venous flow using color Doppler. Ultrasound Obstet Gynecol 1999; 14(4):250-255. Bisseling TM, Wouterse AC, Steegers EA, Elving L, Russel FG, Smits P. Nitric oxidemediated vascular tone in the foetal placental circulation of patients with type 1 diabetes mellitus. Placenta 2003; 24(10):974-978. 114 CHAPTER 7 [16] [17] Brodie MM, Axelrod J, Soberman R, Levy BB. The estimation of antipyrine in biological materials. Journal of Biological Chemistry 1949; 179:25-31. Glazier JD, Jones CJ, Sibley CP. Purification and Na+ uptake by human placental microvillus membrane vesicles prepared by three different methods. Biochim Biophys Acta 1988; 945(2):127-134. [18] Van der AA EM, Copius Peereboom-Stegeman JH, Russel FGM. Isolation of syncytial microvillous membrane vesicles from human term placenta and their application in drugnutrient interaction studies. J Pharmacol Toxicol Methods 1995; 34(1):47-51. [19] Mircheff AK, Wright EM. Analytical isolation of plasma membranes of intestinal epithelial cells: identification of Na, K-ATPase rich membranes and the distribution of enzyme activities. J Membr Biol 1976; 28(4):309-333. [20] Liu XY, Matherly LH. Analysis of membrane topology of the human reduced folate carrier protein by hemagglutinin epitope insertion and scanning glycosylation insertion mutagenesis. Biochim Biophys Acta 2002; 1564(2):333-342. [21] Wong SC, Zhang L, Witt TL, Proefke SA, Bhushan A, Matherly LH. Impaired membrane transport in methotrexate-resistant CCRF-CEM cells involves early translation termination and increased turnover of a mutant reduced folate carrier. J Biol Chem 1999; 274(15):10388-10394. [22] Wang X, Shen F, Freisheim JH, Gentry LE, Ratnam M. Differential stereospecificities and affinities of folate receptor isoforms for folate compounds and antifolates. Biochem Pharmacol 1992; 44(9):1898-1901. [23] De Weerd S, Thomas CM, Cikot RJ, Steegers-Theunissen RP, de Boo TM, Steegers EA. Preconception counseling improves folate status of women planning pregnancy. Obstet Gynecol 2002; 99(1):45-50. [24] Smith SB, Huang W, Chancy C, Ganapathy V. Regulation of the reduced folatetransporter by nitric oxide in cultured human retinal pigment epithelial cells. Biophys Res Commun 1999; 257(2):779-783. [25] Biochem Dudeja PK, Kode A, Alnounou M, Tyagi S, Torania S, Subramanian VS et al. Mechanism of folate transport across the human colonic basolateral membrane. Am J Physiol Gastrointest Liver Physiol 2001; 281(1):G54-G60. [26] Morris NH, Sooranna SR, Learmont JG, Poston L, Ramsey B, Pearson JD et al. Nitric oxide synthase activities in placental tissue from normotensive, pre-eclamptic and growth retarded pregnancies. Br J Obstet Gynaecol 1995; 102(9):711-714. 115 CHAPTER 8 SUMMARY AND DISCUSSION GENERAL CONCLUSION CHAPTER 8 SUMMARY AND DISCUSSION In spite of an optimal control of mother and foetus during pregnancy, the incidence of perinatal complications is increased in women with diabetes mellitus. The cause of these perinatal complications often remains unclear. Diabetes mellitus type 1 and preeclampsia, as well as folate deficiency are pathological conditions that lead to vascular (endothelial) dysfunction. As such, foetoplacental vascular dysfunction may contribute to the increased prevalence of perinatal complications observed in relationship to these maternal diseases. Although the placenta plays a major role in an optimal development and growth of the foetus, little is known about foetoplacental function in maternal disease. In this thesis the possible relationship between maternal diabetes, preeclampsia (part I and II), and foetal growth restriction (part III), and placental vascular dysfunction were studied. Part I Diabetes is associated with endothelial dysfunction in different vascular beds. Regardless of its importance for the development and growth of the unborn child, only little is known about the possible changes in the foetoplacental vascular bed in women with diabetes. The foetoplacental vascular bed differs from most vascular beds, since the placenta lacks autonomic innervation. The vascular function of this organ depends on locally available vasoactive mediators. There is evidence from other vascular beds that the availability of these mediators may be impaired as a result of diabetes. We studied the availability of some of these mediators in the foetoplacental vascular bed from women with diabetes mellitus type 1 as compared to healthy controls. In baseline conditions the foetoplacental vascular bed is one of low resistance. Therefore, studies on mediators causing vasodilation, such as the nitric oxide (NO) pathway and endothelium derived hyperpolarizating factor (EDHF) pathway, were performed. In part I investigations are presented, which were performed on the NO pathway (chapter 2), potassium channels (chapter 3), and the prostanoid pathway (chapter 4). Additionally, foetoplacental morphometry, and foetoplacental arterial elasticity were investigated (chapter 3). Placentae from healthy controls (n=57) and from women with DM (n=30) were investigated by ex vivo dual perfusion of the isolated human placental cotyledon. The clinical data of these women are presented in table 1. In the total group of women with DM baseline foetoplacental arterial pressure was increased as compared to the healthy controls (24.5; 17.0 – 40.0 mmHg versus 20.0; 12.0 – 35.5 mmHg, P<0.001 (median; min - max). This observation, however, was not consistently found within the 119 SUMMARY, DISCUSSION AND GENERAL CONCLUSION separate subgroups studied in the chapters 2 to 4, probably due to a type 2 statistical error. In the nineties of the last century, NO has been identified as an important vasodilatory mediator in the placental, and other vascular beds. Diabetes has been associated with an impaired vascular endothelial function. Hyperglycaemia reduces the bioavailability of NO, whereas hyperinsulinemia increases the endothelial release of NO. In the foetoplacental circulation from women with diabetes, hyperglycaemia and hyperinsulinemia occur simultaneously. Therefore, the activity of the local NO pathway is thought to be different in diabetic women. Because the net effect of these different factors is not known, in chapter 2 the contribution of the NO pathway to the regulation of baseline vascular tone in the foetoplacental circulation from women with type 1 diabetes was quantified. Additionally, the influence of short-term administration of insulin was investigated in the foetoplacental vascular bed from healthy controls. In diabetes an increased contribution of the NO pathway to foetoplacental baseline vascular tone was observed, which could not be attributed to the effect of acute administration of insulin. Beside NO, recent studies point towards the EDHF-pathway as having a major contribution to the regulation of baseline vascular tone. There is evidence that this EDHF pathway is disturbed in diabetes. Unfortunately, the identity of EDHF remains unknown. This makes a direct investigation of EDHF difficult. Nevertheless, the openstate-probability of vascular smooth muscle potassium channels seems to be a final common pathway in the mechanism of action. The increased conductance of potassium through the potassium channels hyperpolarizes vascular smooth muscle cells, which then triggers the closure of voltage-dependent calcium channels resulting in a reduced cytosolic calcium concentration and subsequent vasorelaxation. In this thesis not the role of EDHF, but the role of different potassium channels in the maintenance of baseline foetoplacental arterial tone was studied. A disturbance in the functioning of vascular potassium channels may contribute to an impaired endothelial function. In chapter 3 the contribution of the KATP, KV, SK, and BK channel was investigated by using the same model of ex vivo dual perfusion of the isolated human placental cotyledon as described in chapter 2. The results of these studies suggest a contribution of both the KATP and the KV channel to baseline tone in the foetoplacental vascular bed in healthy controls. In DM, however, the contribution of the KATP channel was impaired. Apart from NO and potassium channels, the prostanoid pathway contributes to baseline vascular tone. Of all prostanoids, the vasodilator prostacyclin (PGI2), and vasoconstrictor thromboxane (TXA) seem to be most important in the regulation of vascular function. In diabetes, there is evidence of a disturbed availability and function of both these prostanoids. The enzyme cyclooxygenase is a common pathway in the 120 CHAPTER 8 formation of prostanoids. In chapter 4 the net contribution of vasodilator and vasoconstrictor prostanoids to baseline vascular tone was investigated by quantification of the contribution of the enzyme cycloogygenase to baseline foetoplacental vascular tone. Inhibition of cyclooxygenase by indomethacin induced vasoconstriction in the ex vivo perfused foetoplacental vascular bed from healthy women. This suggests a more important role for vasodilator prostanoids like prostacyclin, than for vasoconstrictor prostanoids like thromboxane A2 in this particular vascular bed. In diabetes, the vasoconstrictor response to indomethacin was decreased, suggesting a defect in the cyclooxygenase pathway in these patients. Table 1 Clinical characteristics of the participants Control Diabetes Number 57 30 Maternal age (years) 33.1; 21.1 – 42.8 31.3; 21.0 – 38.2 Parity (number) 0; 0 - 3 0; 0 - 3 Gestation (weeks) 39.9; 37.0 – 42.3 38.1; 36.6 – 38.9 ∗ Birth weight (grams) 3333; 2500 – 4560 3578; 2435 - 4875 Placental weight (grams) 570; 370 - 1000 660; 400 – 950 ∗ Body Mass Index (kg/m2) 24.2; 15.4 – 30.8 24.0; 17.2 – 51.0 Diastolic BP (mm Hg) 80.0; 60.0 - 90.0 85.0; 55.0 - 90.0 Smokers (number ;%) 4; 7.0 2; 6.7 White Class HbA1c 1st trimester (%)1 1 HbA1c3rd trimester (%) Insulin (mE/ml) - 14 B, 9 C, 3 D, 2 F, 2 R - 6.3; 5.0 – 8.8 - 6.3; 4.8 - 7.3 M 21.0; 0.2 – 161.0 28.5; 5.0 – 61.0 U 14.5; 5.0 – 54.0 75.5; 5.0 – 242.0 ∗ C-peptide (nmol/l) M 0.98; 0.12 – 2.35 0.13; 0.00 – 0.48 ∗ U 0.54; 0.21 – 6.00 1.55; 0.28 – 4.16 ∗ Baseline foetoplacental arterial pressure 20.0; 12.0 – 35.5 24.5; 17.0 – 40.0 ∗ Data are presented as median; min – max, M=maternal, U=umbilical ∗ P < 0.05 versus healthy controls. 1 normal value for HbA1c in our laboratory is 4.2 – 6.3 %, in pregnant women with DM the aim is a value < 7.0 % 121 SUMMARY, DISCUSSION AND GENERAL CONCLUSION Other factors, like morphological characteristics and vascular elasticity could contribute to baseline vascular tone. Vascular tone is dependent on the diameter, and elasticity of the vascular wall of the investigated vascular bed. Studies on morphology of the placenta suggest that this organ is morphologically different in diabetes mellitus as compared to controls. However, a morphometric analysis of the foetoplacental vascular bed has never been performed. Additionally, the effect of diabetes on the elasticity of the vascular wall from foetoplacental resistance vessels is unknown. Therefore, morphometric parameters as well as vascular elasticity were investigated in the foetoplacental vascular bed. No differences in morphometry or elasticity of resistance vessels were observed between foetoplacental vascular beds from controls and DM. In conclusion, baseline foetoplacental vascular tone is increased in diabetes. This could be attributed to impaired function of the KATP channel, and the cyclooxygenase pathway. The contribution of the NO pathway, although different from controls, is increased in diabetes. In theory, this might be attributed to a compensatory vasoactive mechanism in the foetoplacental vascular bed in diabetes. Part II Preeclampsia has been associated with endothelial dysfunction, and a subsequent impaired availability of NO for vasodilation. The underlying cause might be a disturbed balance between oxidants and anti-oxidants, resulting in an increased breakdown of NO by reactive oxygen species (ROS). Inactivation of NO by ROS results in the formation of peroxynitrite, which is increased in the foetoplacental vascular bed in preeclampsia. Supplementation of the antioxidant vitamins C and E to women with an increased risk for preeclampsia was shown to be beneficial in the prevention of this disorder. Another pharmacological substance with antioxidant capacities is N-acetylcysteine (NAC). NAC is rapidly deacetylated into cysteine, a direct precursor of gluthation. Short-term intraarterially administered gluthation improves endothelial function. Theoretically, the additions of radical scavengers like vitamin C or NAC to the foetoplacental vascular bed might be able to improve endothelial function. In chapter 5 the contribution of the NO pathway to baseline foetoplacental vascular tone was investigated in the foetoplacental vascular bed from women with PE as compared to that from controls. In addition, the effect of acute administration of 5 mM vitamin C into the foetoplacental circulation on the NO pathway was studied in controls. From the results of these studies it could be concluded that the contribution of the NO pathway to baseline vascular tone is impaired in the foetoplacental circulation from patients with preeclampsia. Vitamin C is able to increase the NOmediated vasodilator component in this vascular bed from healthy controls. 122 CHAPTER 8 Apart from vitamin C, the acute effect of N-acetylcysteine (NAC) was studied, as well. Chapter 6 describes the results of a study in which NAC was administered in the foetoplacental circulation from women with preeclampsia as well as in that from healthy controls. For each experiment a simultaneous control experiment was performed in a cotyledon from the same placenta, however, with addition of a solvent instead of the active substrate. The results from this study confirm the observation from chapter 5 that ex vivo the contribution of NO to the baseline vascular tone is impaired in the foetoplacental circulation from preeclamptic patients. Moreover, this study shows that the NO pathway can be ameliorated by addition of the anti-oxidant Nacetylcysteine to the perfusion buffer. In conclusion, the contribution of the NO pathway to foetoplacental baseline vascular tone is decreased in preeclampsia as compared to controls. Acute administration of the radical scavengers vitamin C (in controls) and N-acetylcysteine (in controls, and preeclamptic patients) into the foetoplacental circulation was beneficial for the contribution of the NO pathway to vascular tone in this circulation. Part III Folate (vitamin B11) is an essential micronutrient, which plays a major role in a number of intracellular processes such as the formation of DNA, and therefore cellular growth. An optimal availability of this vitamin is important for normal development and growth of the unborn child. Maternal folate deficiency has been associated with a number of complications in pregnancy, such as foetal growth restriction. To date, the role of the placenta, being the regulating organ for transfer of (micro)nutrients, in foetal growth restriction is unclear. Several factors contribute to the availability of folate to the foetus. These include the maternal folate supply to the placenta, the uptake of folate by the placenta, the amount of folate binding to the folate-receptor, and the expression of folate binding proteins in the placenta. In chapter 7 we investigated whether foetal growth restriction is associated with a disturbed placental folate uptake. Maternal folate uptake during ex vivo dual perfusion of an isolated cotyledon, binding of folate to the folate receptor-α (FRα), and expression of FRα as well as reduced folate carrier (RFC) were compared between placentae from pregnancies complicated with foetal growth restriction and those from healthy controls. Placental folate clearance was not different, but because of endogenous folate release we had to use supraphysiological concentrations. Most probably, in these experiments simple diffusion of folate was the major determinant for maternal folate uptake. This may have contributed to the fact that no differences could be observed between placentae from pregnancies complicated by foetal growth restriction and those from controls. However, the radioligand-binding assay, which was performed with 123 SUMMARY, DISCUSSION AND GENERAL CONCLUSION physiological concentrations of folate, did not show differences between those groups, either. Moreover, the expression of FRα and RFC was comparable in placentae from pregnancies complicated by foetal growth restriction and those from controls. In conclusion, foetal growth restriction does not seem to be associated with an impaired maternal placental folate uptake. Also, placental folate binding to the FRα, and expression of FRα and RFC are comparable between placentae from pregnancies complicated with foetal growth restriction and those from controls. GENERAL CONCLUSION In this thesis a number of associations were studied between maternal diabetes, preeclampsia, and foetal growth restriction on onde hand, and a disturbed foetoplacental function on the other hand. In diabetes, baseline foetoplacental vascular tone seemed to be increased. This could be the result of a disturbed function of the KATP channel, and the enzyme cyclooxygenase. In contrast, the contribution of NO to baseline foetoplacental vascular tone seemed to be increased in diabetes. In preeclampsia, the disturbance in the NO pathway could be diminished by administration of the antioxidants vitamin C and NAC. No relationships between maternal placental folate uptake, placental folate receptor binding, as well as expression of FRα and RFC and foetal growth restriction could be observed. However, the association between folate insufficiency and foetal growth restriction could be the result of other causes such as a placental malfunction or a very low maternal folate intake. Little information is available about the contribution of other vasoactive mediators such as vasoconstrictors and biophysical stimuli in the foetoplacental vascular bed in diabetes as well as in preeclampsia, and the effect of addition of antioxidants in diabetes. These are interesting questions for future studies. Also, it is interesting to investigate whether administration of an antioxidant before and/or during pregnancy in women with (a risk of) diseases such as diabetes and preeclampsia, ameliorates the foetoplacental vascular endothelial function, and decreases the incidence of perinatal complications. If that is the case, antioxidants may play a role in the secondary prevention of foetoplacental vascular dysfunction in pregnancies complicated with maternal diabetes or pregnancies in women at risk of preeclampsia. The newly formed placental vascular bed has a life span limited to a period of only nine months, including the period of its development. The conclusions that can be drawn from the studies presented in this thesis suggest that during this short period, pathologic conditions such as diabetes and preeclampsia can result in foetoplacental 124 CHAPTER 8 vascular dysfunction. Type 1 diabetes mellitus exists before pregnancy, and therefore the foetoplacental vascular bed is exposed to both hyperglycaemia and hyperinsulinaemia from the onset of its development. Although the foetoplacental circulation is different because of its lack of autonomic innervation, it may possess functional characteristics similar to other vascular beds in the early stages of diabetes. If so, the results as presented in part I suggest that endothelial dysfunction is present from the very early stage of diabetes. Therefore, the foetoplacental vascular bed from women with diabetes might be a good target for future studies of early vascular endothelial dysfunction in diabetes. 125 SAMENVATTING SAMENVATTING VERKLARENDE WOORDENLIJST diabetes mellitus Type 1 suikerziekte Type 1 endotheel cellen die de binnenwand van bloedvaten bekleden foetoplacentaire vaten vaten van het kind in de moederkoek macrosome kinderen kinderen met een te hoog geboortegewicht voor de duur van de zwangerschap perinataal rond de geboorte preeclampsia zwangerschapsvergiftiging prevalentie frequentie waarin een bepaalde aandoening voorkomt in de populatie vasoconstrictie vernauwing van de bloedvaten vasorelaxatie verwijding van de bloedvaten 128 SAMENVATTING SAMENVATTING EN DISCUSSIE Anno 2004 worden vrouwen met diabetes mellitus zowel voor als tijdens de zwangerschap intensief gecontroleerd en begeleid. Desondanks hebben vrouwen met diabetes mellitus type 1 een verhoogd risico op complicaties in de zwangerschap in vergelijking met gezonde vrouwen. De meest bekende complicaties zijn macrosome kinderen en aangeboren afwijkingen. Minder bekend is dat deze groep vrouwen ook een hoger risico heeft op het krijgen van een te klein of een doodgeboren kind. De oorzaak van deze laatstgenoemde nare complicaties blijft tot op heden meestal onopgehelderd. Diabetes mellitus type 1, maar ook preeclampsie en een foliumzuurdeficiëntie zijn afwijkingen die kunnen leiden tot abnormaal functioneren van bloedvaten, in het bijzonder van het endotheel. Een veranderde functie van de foetoplacentaire bloedvaten zou kunnen bijdragen aan de verhoogde prevalentie van perinatale complicaties. Hoewel bekend is dat de placenta een belangrijke rol speelt in de groei en ontwikkeling van de foetus, is er weinig bekend over de functie van de placenta, bij moeders met een onderliggend lijden. In dit proefschrift werden de relatie tussen maternale diabetes mellitus type 1 (deel I), preeclampsie (deelI) en foetale groeivertraging (deel III) enerzijds en een verstoorde functie van de placenta anderzijds bestudeerd. In het bijzonder werden de effecten van ziekten op het endotheel van het foetoplacentaire vaatbed onderzocht. Deel I Diabetes mellitus wordt in verband gebracht met een gestoorde endotheelfunctie. Er is weinig bekend over de mogelijke gevolgen van diabetes mellitus voor het foetale vaatbed van de placenta van vrouwen met deze ziekte. In de meeste vaatbedden wordt de vaattonus gereguleerd door een combinatie van de invloed van hormonen, biofysische factoren (bijvoorbeeld pH) en het autonome zenuwstelsel. Het vaatbed in de placenta bevat echter geen zenuwen en om die reden is dit vaatbed voor zijn regulatie afhankelijk van locale mediatoren. In diverse vaatbedden zijn aanwijzingen gevonden voor een gestoorde aanwezigheid van deze locale mediatoren onder de invloed van diabetes. Wij hebben de effecten van een aantal van deze mediatoren in het foetoplacentaire vaatbed van vrouwen met diabetes mellitus type 1 bestudeerd in vergelijking met die in gezonde controles. Onder normale omstandigheden heeft het foetoplacentaire vaatbed een lage vaatweerstand. In dit proefschrift werden met name de mediatoren die de weerstand in het vaatbed laag houden bestudeerd. Deze mediatoren zijn als groep bekend onder de naam “endothelium derived relaxing factors”, waarvan stikstofmonoxide (NO), endothelium derived hyperpolarizing factor (EDHF) en 129 SAMENVATTING prostacycline (PGI2) de meest bekende zijn. In het laatste decennium van de vorige eeuw is het NO geïdentificeerd als een belangrijke vaatverwijder, zowel in het placentaire als in andere vaatbedden. Al langer is bekend dat diabetes kan leiden tot een verstoring van de endotheelfunctie. Hyperglycemie verlaagt de beschikbaarheid van het NO, terwijl hyperinsulinemie de endotheliale afgifte van het NO verhoogt. Het foetoplacentaire vaatbed van vrouwen met diabetes wordt gelijktijdig blootgesteld aan deze beide condities. Om die reden wordt gedacht dat de locale NO-functie anders is in patiënten met diabetes mellitus. Het effect van deze factoren op de beschikbaarheid van het NO is onbekend. In hoofdstuk 2 werd de bijdrage van NO aan de regulatie van de basale vaatweerstand in het foetoplacentaire vaatbed van vrouwen met diabetes mellitus type 1 gekwantificeerd. Aanvullend werd dit vaatbed van gezonde vrouwen het effect van acute toediening van insuline bestudeerd. Uit deze studies bleek dat bij diabetes de bijdrage van het NO aan de basale vaatweerstand was verhoogd, hetgeen niet kon worden geweten aan het effect van acute toediening van insuline. Naast het NO, lijkt het EDHF een grote rol te spelen in de regulatie van de rusttonus van vaten. Omdat de identiteit van EDHF onzeker is, is het rechtstreeks bestuderen niet mogelijk. Wel is uit eerder onderzoek gebleken dat de openingstoestand van verschillende kaliumkanalen in de vasculaire gladde spiercel een rol speelt in het werkingsmechanisme van dit EDHF. Een verhoogde doorlaatbaarheid van deze kaliumkanalen zorgt voor een verhoogd spanningsverschil over de wand van de gladde spiercel. Hierdoor sluiten voltage-afhankelijke calciumkanalen, met als gevolg een verlaagde concentratie van calcium in de gladde spiercel met als resultaat vasorelaxatie. Met deze kennis werd in dit proefschrift niet de rol van het EDHF zelf, maar die van de verschillende kaliumkanalen onderzocht. In hoofdstuk 3 werd de bijdrage van het KATP, KV, SK, and BK kanaal bestudeerd in hetzelfde model dat werd gebruikt in de studie in hoofdstuk 2. De resultaten van deze studie suggereren dat het KATP en het KV kanaal een rol spelen in de handhaving van de basale vaattonus in het humane foetoplacentaire vaatbed in gezonde controles. In dit vaatbed van vrouwen met diabetes was de bijdrage van het KATP kanaal echter verminderd ten opzichte van de controlegroep. Een derde factor die bijdraagt aan de rusttonus van de vaten is het mechanisme via prostanoïden. Van alle prostanoïden lijken het vaatverwijdende prostacycline (PGI2) en het vaatvernauwende thromboxane A2 (TXA) het meest belangrijk voor de regulatie van de vaattonus. Er zijn aanwijzingen dat de bijdrage van deze prostanoïden gestoord is in aanwezigheid van diabetes. Het enzym cyclooxygenase is een enzym dat zowel van belang is voor de vorming van PGI2 als voor de vorming van TXA. In hoofdstuk 4 werd de bijdrage van dit enzym aan de foetoplacentaire vaattonus bestudeerd door gebruik te maken van het ex vivo perfusie model. Remming van het enzym cyclooxygenase door middel van de stof indomethacine resulteerde in een stijging van de vaattonus in gezonde controles. Dit suggereert dat er in het foetoplacentaire vaatbed een grotere rol 130 SAMENVATTING is voor vaatverwijdende prostanoïden, waaronder het prostacycline, dan voor vaatvernauwende prostanoïden als thromboxane. In vrouwen met diabetes trad minder vaatvernauwing op na toediening van indomethacine, hetgeen suggereert dat het mechanisme dat loopt via indomethacine is gestoord in de placenta van deze vrouwen. Concluderend kan worden gezegd dat de vaattonus in het foetoplacentaire vaatbed is verhoogd in vrouwen met diabetes. Mogelijk speelt een gestoorde functie van het KATP kanaal en een gestoord mechanisme van cycloogygenase een rol in het ontstaan van deze verhoogde basale vaattonus. Ook de bijdrage van het NO is veranderd in diabetes. De beschikbaarheid van het NO is echter toegenomen, hetgeen resulteert in vasorelaxatie. Theoretisch zou de verhoogde beschikbaarheid van het NO in het foetoplacentaire vaatbed in diabetes een compensatiemechanisme kunnen zijn. Andere, niet-specifieke, factoren zoals morfologische karakteristieken en elasticiteit van de vaatwand zijn ook van invloed op de rusttonus van de vaatwand. Er zijn studies die suggereren dat er morfologische verschillen bestaan tussen de placenta van vrouwen met diabetes en die van gezonde controles. Een morfometrische analyse werd tot op heden nog niet verricht in dit vaatbed. Ook werd het effect van diabetes op de elasticiteit van de foetale vaten in de placenta niet eerder bestudeerd. Om er zeker van te zijn dat de verschillen tussen diabetische vrouwen en controles niet werden veroorzaakt door deze non-specifieke factoren, werden morfometrische parameters en de elasticiteit van de foetoplacentaire weerstandsvaten bestudeerd. In deze analyses werden geen verschillen gevonden tussen placenta’s van gezonde vrouwen en die van vrouwen met diabetes mellitus. In totaal werden in deze drie studies 57 placenta’s van gezonde vrouwen en 30 van vrouwen met diabetes onderzocht. Voor deze studie werd voornamelijk gebruik gemaakt van het model van ex vivo tweezijdige perfusie van het geïsoleerde humane placentaire cotyledon. In deze totale groep blijkt de foetoplacentaire druk in rust verhoogd in de placenta’s van vrouwen met diabetes in vergelijking met die in gezonde controles (24.5; 17.0 – 40.0 mmHg versus 20.0; 12.0 – 35.5 mmHg, P<0.001 (mediaan; min - max). In de separate studies beschreven in hoofdstuk 3 en 4 bleek dit verschil echter niet significant. Mogelijk is dit te wijten aan een type 2 statistische fout. Deel II Ook preeclampsie word in verband gebracht met een verslechterde functie van het endohteel, met als gevolg een verminderde beschikbaarheid van NO voor handhaving van de basale vaattonus. De onderliggende oorzaak zou verhoogde afbraak van NO 131 SAMENVATTING kunnen zijn door reactieve zuurstofdeeltjes (oxidanten). Bij deze afbraak wordt peroxynitriet gevormd, een stof waarvan bewezen is dat die de endotheelfunctie aantast. De hoeveelheid peroxynitriet is verhoogd in de placenta van vrouwen met preeclampsie. Toediening van de vitamines C en E, beiden beschikken over een antioxidant werking, aan vrouwen met een verhoogd risico op het ontwikkelen van preeclampsie, is gunstig gebleken in de preventie van deze ziekte. Ook N-acetylcysteïne (NAC) is een stof met anti-oxidante eigenschappen. Deze stof wordt in het lichaam omgezet in cysteïne, een directe voorloper van het anti-oxidant gluthation. Kort-durende intra-arteriele toediening van gluthation verbetert de endtoheelfunctie. In theorie zou toediening van anti-oxidanten aan het foetoplacentaire vaatbed de endotheelfunctie moeten kunnen verbeteren. In hoofdstuk 5 werd de bijdrage van het NO aan de basale vaattonus vergeleken in het foetoplacentaire vaatbed van controles in vergelijking tot die in hetzelfde vaatbed van vrouwen met preeclampsie. Tevens werd het effect van toediening van 5 mM vitamine C bestudeerd in een controlegroep. Naast vitamine C, is ook het acute effect van NAC bestudeerd. Hoofdstuk 6 beschrijft een studie waarin NAC werd toegediend in het foetoplacentaire vaatbed van vrouwen met preeclampsie en dat van gezonde controles. Voor elk van deze experimenten werd een simultaan experiment uitgevoerd in een identieke opstelling, waarbij een placebo werd toegevoegd. Conclusie van deze studies is dat de bijdrage van het NO aan de basale vaattonus van het foetoplacentaire vaatbed is verminderd in preeclampsie. Acute toediening van antioxidanten als vitamine C (controles) en NAC (controles en preeclampsiepatiënten) was gunstig voor de bijdrage van het NO aan de vaattonus in dit vaatbed. Deel III Foliumzuur (vitamine B11) is een essentiële voedingsstof, die een grote rol speelt in een aantal intracellulaire processen waaronder de vorming van DNA, en dus celgroei. Een optimale beschikbaarheid van deze vitamine is van groot belang voor een normale groei en ontwikkeling van het ongeboren kind. Een foliumzuurgebrek bij de moeder wordt in verband gebracht met verschillende complicaties in de zwangerschap, waaronder een foetale groeivertraging. De placenta speelt een belangrijke rol in het transport van voedingsstoffen naar het kind. Tot op heden is het onduidelijk of de placenta een rol speelt in het verband tussen foliumzuurtekort bij de moeder en een groeivertraging bij het kind. Verschillende factoren dragen bij aan de opname van foliumzuur vanuit het moederlijke bloed in de placenta. Dit zijn onder andere de maternale aanvoer van foliumzuur, de hoeveelheid foliumzuur dat aan de foliumzuurreceptor bindt en de expressie van foliumzuurbindende eiwitten in de placenta. 132 SAMENVATTING In hoofdstuk 7 hebben werd onderzocht of een foetale groeivertraging in verband kan worden gebracht met een gestoorde foliumzuuropname in de placenta. De opname van foliumzuur tijdens ex vivo perfusie van een geïsoleerd cotyledon, de binding van foliumzuur aan de foliumzuurreceptor-α (FRα) en de expressie van zowel FRα als de reduced foliumzuur carrier (RFC) werden bestudeerd in placenta’s van zwangerschappen gecompliceerd door een foetale groeivertraging als in die van gezonde zwangerschappen. Conclusie van dit onderzoek is dat een foetale groeiachterstand niet lijkt samen te hangen met een gestoorde opname van foliumzuur aan de maternale zijde van de placenta. Ook de binding van foliumzuur aan de FRα en de expressie van FRα en RFC was vergelijkbaar tussen placenta’s van zwangerschappen gecompliceerd door een foetale groeivertraging en die van gezonde controles. SLOTCONCLUSIE In dit proefschrift werd een aantal verbanden bestudeerd tussen maternale diabetes mellitus type 1, preeclampsie en foetale groeivertraging enerzijds en een gestoorde functie van de placenta anderzijds. In vrouwen met diabetes mellitus lijkt de basale foetoplacentaire vaattonus verhoogd te zijn. Mogelijk is dit het gevolg van een gestoorde functie van het KATP kanaal of van het enzym cyclooxygenase. Anderzijds blijkt de bijdrage van NO aan de basale foetoplacentaire vaattonus verhoogd te zijn. In preeclampsie kon de verminderde bijdrage van het NO worden verbeterd door toediening van de anti-oxidanten vitamine C en NAC. Er werd geen relatie gevonden tussen foetale groeivertraging enerzijds, en de opname van foliumzuur aan de maternale zijde van de placenta, binding van foliumzuur aan de FRα , en de expressie van FRα , en RFC anderzijds. Deze bevinding sluit overigens niet uit dat een foetale groeivertraging het resultaat zou kunnen zijn van andere (placentaire) problemen zoals een algeheel slecht functioneren van de placenta of van een veel te lage inname van foliumzuur door de moeder. 133 BIBLIOGRAFIE BIBLIOGRAFIE PUBLICATIONS Collagen synthesis in rat skin and ileum fibroblasts is affected differently by diabetesrelated factors. Michiel H.J. Verhofstad, Tanya M. Bisseling, Eloy M.H. Haans and Thijs Hendriks. Int J Exp Path 1998; 79: 321–328 Het Boerhaave-syndroom bij hyperemesis gravidarum; een casus. A.A. van Ginkel, T.M. Bisseling, P.J.H.M. Reuwer. NTOG 1999; 112: 239-240 Het spastische bekkenbodemsyndroom en de rol van het bekkenbodemteam. J.A. Wegdam, T.M. Bisseling, E.J. Spillenaar Bilgen, W.F. Eggink, H.C.M. Defourny, T.G. Wiersma, C. Mulder, G. den Hartog, H. Ruesink, P.C. Weijerman, A. Huisman, E. Blokzijl. NTOG 1999; 112: 97 Antioxidants and pre-eclampsia. Tanya M. Bisseling, Frans G.M. Russel, Simone Dekker, Eric A.P. Steegers, Paul Smits. Lancet 2000; 355; 65 [letter] Nitric Oxide-mediated vascular tone in the foetal placental circulation in diabetes mellitus type 1. Tanya M. Bisseling, Alfons C. Wouterse, Eric A.P. Steegers, Lammy Elving, Frans G. Russel, Paul Smits. Placenta 2003; 24, 974-978 Placental folate transport and binding are not impaired in pregnancies complicated by foetal growth restriction. Tanya Bisseling, Eric Steegers, Leni Siero, Annabel Walker, Femke van de Water, Regine Steegers-Theunissen, Paul Smits, Frans Russel. Placenta 2004; 25: 588-593 N-acetylcysteine restores baseline NO release in the foetoplacental circulation of preeclamptic patients. T.M. Bisseling, E.M. Roes, M.T.M. Raijmakers, E.A.P. Steegers, W.H.M. Peters, P. Smits. Am J Obstet Gynecol 2004; 191; 328-333 Dysfunction of the cyclooxygenase pathway in the foetoplacental circulation in type 1 diabetes mellitus. T.M. Bisseling, A.C. Wouterse, E.A.P. Steegers, L. Elving, F.G.M. Russel, P. Smits. Diabetic Medicine [in press] Impaired KATP channel function in the foetoplacental circulation of patients with type 1 diabetes mellitus. Tanya M. Bisseling, Marieke G. Versteegen, Selina van der Wal, Joop Borggreven, Eric A. Steegers Ph.D., Jeroen van der Laak, Frans G. Russel Ph.D., Paul Smits Ph.D. Am J Obstet Gynecol [in press] 136 BIBLIOGRAFIE Prospective analysis of complications of Tension-free Vaginal Tape (TVT) from The Netherlands TVT Study. Steven E. Schraffordt Koops, Tanya Bisseling, A. Peter.M. Heintz, Harry A.M. Vervest. [Submitted] Quality of life before and after TVT, a prospective multi centre cohort study, results from the Netherlands TVT database. Steven E. Schraffordt Koops, Tanya Bisseling, A. Peter.M. Heintz, Harry A.M. Vervest. [Submitted] PUBLISHED ABSTRACTS Indications for and functional results after surgical correction for enterocele. T.M. Bisseling, E.J. Spillenaar Bilgen, I.M.C. Jansen, G. den Hartog, T.G. Wiersma, W.F. Eggink. Eur J Gastroenterol Hepatol 1998; 10: A51 Increased resistance of the foetal placental vascular bed in diabetes mellitus. Bisseling T.M., Steegers E.A., Elving L., Russel F.G.M. en Smits P. The Netherlands Journal of Medicine 2001; 59: A1 Baseline function of placental vascular KATP-channels in healthy and in diabetic women Bisseling T.M., Steegers E.A., Wouterse A.C., Elving L., Russel F.G.M. & Smits P. J Clin Pharmacol 2002; 53: 548P–549P The outcome of TVT analysed with disease-specific quality of life questionnaires: results from the Netherlands Multicenter TVT Study. S. Schraffordt, H. Vervest. T. Bisseling. Neurourology and Urodynamics 2003;22:404-405. Quality of life after combined genitourinary prolapse repair and TVT. S. Schraffordt, H. Vervest. T. Bisseling. Proceedings of the International Continence Society; 2003: 183184. 137 EXTRA A HISTORICAL AND CULTURAL VIEW ON THE HUMAN PLACENTA INTRODUCTION The placenta may be the most important organ for human reproduction. During pregnancy it combines functions which will be performed by different organs after birth. These include respiration, uptake of nutrients, elimination of waste products, and hormone production. The functional part of the placenta conists of an extensive vascular network surrounded by one layer of so-called syncytiotrophoblastic cells bathing in maternal blood. An optimal placental transport and vascular function is important for a normal intrauterine development and growth of the foetus. In spite of this important role of the placenta, only limited studies have been performed of the degree in which placental dysfunction contributes to the prevalence of perinatal morbidity and mortality. Therefore, this organ is an interesting and fascinating subject for scientific research. Only during the last century has more knowledge become available about placental morphology and placental function. In the past, when the function of the placenta was not fully understood by human beings, all kinds of mysteries, beliefs, habits and rituals have arisen around the placenta. These socio-cultural perceptions of the placenta are at least as fascinating as the basic scientific observations. Therefore in this chapter a concise survey is presented of these perceptions. Unfortunately, some words or sentences were not translatable from Dutch to English. THE MEANING AND ORIGIN OF THE WORD “PLACENTA” The meaning of the word placenta, and related words, are presented as follows in the “Dikke van Dale” [1]: Placenta 1. moederkoek, nageboorte 2. [biol.] zaadkoek, aanhechting van de zaadknop aan het vruchtbeginsel Placentair [biol.] De moederkoek betreffend Placentatie [biol.] De ontwikkeling van de moederkoek 141 EXTRA The use of the word “placenta” originates from 1559. In this year a man named Realdus Columbus (1516-1559) is thought to be the first person that used this word. It is a derivation from the modern Latin words placenta uterina, which mean uterine cake (ned.: uterine koek). Moreover, the word was known in ancient Latin (placenta: flat cake (ned.: platte koek), and ancient Greek (πλακουντα [plakoenta]: accusativus from πλακουις [plakoeis]; flat), as well. HISTORY OF THE STUDY OF THE PLACENTA The earliest known representation of the placenta is seen on the frequently reproduced palette of King Narmer, a well-preserved palette found at Hierankopolis at the end of the 19th century (fig 1), probably dating from between 3500 and 3100 before Christ. The uppermost panel on the reverse side shows Narmer in ceremonial procession preceded by royal standards; to the right are piled decapitated bodies of his enemies. The standard nearest the king has the shape of a bilobate disc with a streamer hanging down, and has been interpreted as the king’s placenta and umbilical cord [2], the prototype of flags carried into battle with their streamers in red, white and blue. The Figure 1 Palette of King Narmer 142 ancient Egypt people believed the placenta was a child’s secret helper or quasi twin, since they believed the child was formed in the womb from blood and that the unused blood formed the placenta, a reserve of vital material [3]. Only during the renaissance, in the early 16th century, was a detailed description given of the placenta. However, both Leonardo da Vinci and Vesalius based the placenta in their pictures on that of other mammals and therefore they misrepresented it. Leonardo da Vinci (1452-1515) showing it to be cotyledonary as in the sheep, Vesalius (15141564) showing it to be of the zonary variety as in the bitch. Vesalius, however, corrected this error in the second edition (1555) of his great book. Harvey knew that the two blood streams in the placenta must be separate, each stream flowing in an opposite direction to the placenta, but he knew nothing about capillary vessels and could not explain how the blood passed from arteries to veins through the tissue. It was not until the publication of Malpighi about capillaries (1660), that anatomical bases for the regional circulation became available. Hereafter John Mayow (1643–1679) more adequately described the placenta’s nature. Around these days the term cotyledon – originally used by the Greek and Romans – was re-introduced for the isolated functional unit of the placenta. William Hunter (1718-1783) was the first investigator to give a sophisticated description of the human placenta. He and his brother John Hunter (1728 – 1793) injected fluid wax into the uterine circulation, which did not appear in the foetal circulation. Their finding finally set aside the assumption that there existed anastomoses between the maternal and foetal circulation. William Hunter was the first to effectively describe the decidua (1774). John Hunter first mentioned the decidua basalis. The establishment of the knowledge of the development and nature of placental villi only took place over many years. Indeed, their structure could not adequately be appreciated until the development of the compound microscope. The first convincing account of the villi was given by Weber (1832), in an illustration of teased out villi under low magnification. By 1880 the basic knowledge of the circulation in the intervillous space had been established. Langhans (Bern) and Hubrecht (Utrecht) were important investigators, responsible for the analysis of the histological nature of the covering of the foetal villi. Hubrecht’s term “trophoblast” (1889) – a term he used for the layer of cells in the early embryo that he interpreted as a nutritive layer and that does not contribute to the cellular portion of the embryo – was applicable to the two layers found by Langhans as the covering of the chorionic villi. Various investigators demonstrated the superficial layer to be syncýtial in nature. Therefore this superficial layer is now generally syncytiotrophoblast. The other layer is known as the cytotrophoblast. 143 known as the EXTRA Grosser (1873-1951) introduced a histological classification based in part upon the assumed biological activity of the placenta, the principle of which depends on the number and kind of maternal and foetal tissue layers interposed between the blood of the mother and the conceptus. This classification (epithelio-chorial, syndesmo-chorial, endothelio-chorial, haemo-chorial, and haemochorio-endothelial) is nowadays often criticised. Electron microscopy has made it clear that the arrangement of the trophoblasts is more complicated than suspected from studies with light microscopy, on which Grosser’s scheme was based. However, his publications were of great importance for the investigators who studied the early development of the placenta. A lot of investigators have contributed to unravel this puzzle, but the most important basic work was performed by Hertig and Rock (1941). They charted the course for later explorers in the field of placentology. DIFFERENT RITES AND HABITS AROUND THE WORLD Various qualities and powers have been attributed to the placenta. For this reason all kinds of rituals are performed with this special organ after birth. Most of these rituals find their origin in one of the three categories burial, consumption, or healing power of the placenta. However, differences in these rituals exist all over the world. Lotus birth There is no written information dating from before 1974 about this ritual being performed by humans. Until 1974 only has it been described in chimpanzees. It seems to be a modern habit, named after Claire Lotus Day. She was a midwife, who challenged doctors reasons for hasty clamping. The ritual of Lotus Birth is the practice of leaving the umbilical cord uncut, so that the baby remains attached to his/her placenta until the cord naturally separates at the umbilicus, exactly as a cut cord does, at 3 to 10 days after birth. To make sure the placenta dries, it is covered with salt and/or spice and wrapped into cloths. This prolonged birth can be seen as a time of transition, allowing the baby to slowly and gently let go of his/her attachment to the mother’s body. 144 Burn and burry The most common rituals are burial and burning of the placenta. Burial or burying of the placenta has been performed all over the world, however, the reason for doing so may be different. In Alaska, some Indians cremate the placenta, preserving the ashes for ultimate burial with those of the child after its death [4]. In the past, in the Salzburg region in Austria it was customary to bury the placenta under a green tree to ensure the woman’s continuing fertility [4]. In Bolivia Aymara and Quecha people believe the placenta has its own spirit. It is to be washed and buried by the husband in a shady place. If this ritual is not performed correctly, the mother or the bay may become very sick or even die. In Dalmatia (Croatia) the placenta was buried under a rose bush, thus the child would always have rosy cheeks [4]. In Indonesia the rituals vary per isle and/or tribe. Often offers such as flowers (symbol for good smell and fame), a piece of paper, a pencil (so that the baby will be able to write in future) and a needle (symbols for intelligence, smartness and wisdom) acompagny the placenta. The placenta is buried next to the house, on the left hand side if it’s a girl, on the right hand side if it’s a boy. A fence is erected around the site, which is lit by a lamp. In some parts of Java the placenta is set afloat, decorated with flowers and little lights, to be eaten by crocodiles [5]. In Bali the Balinese society is founded on the Hindu caste system; life on earth is only one stage in the continuity of existence. For this reason, the umbilical cord is preserved and kept for life, whilst the placenta is placed in a coconut and buried as they do in Java. On Sulawesi the Torajan of Tana Toraja bury the placenta in a bag of woven bamboo on the eastern side of the house, because east is associated with life and the rising sun. It is believed that if any Torajan travels far away from home, the placenta, a kind of double, will lure him/her back to his/her place of birth [6]. The Maori in New Zealand bury the placenta ritually on the ancestral grounds, as well. Among the Indian tribes in North-America there is a belief that the soul lives in the birth organ; when the cord and the afterbirth have dried up and dropped away, they’re tenderly gathered up by the mother and arranged in a sacred fabric that she have woven during pregnancy. This bundle is then buried secretly, and retrieved whenever the tribe moves. This bundle is presented to the grown person when they appear to have children who will survive adulthood. In the Philipines mothers bury the placenta with books in hopes of a smart child. Burying the placenta is one of the Buddhist ceremonies of Thai (Thailand). This is performed to be sure good will befall in life and evil forces will be warded off. The placenta is preserved in an earthenware jar after it has been dried with salt, because a suitable site for the burial has to be determined according to a set of strict rules. The 145 EXTRA burial must take place in a special site, dependent on the month and year of birth of the baby. For example a child born in the year of the ox and pig are best matched to the sugar palm, or lotus. The Hmong (Vietnam) word for placenta means “jacket”, as it is considered an infant’s first and finest clothing. The Hmong baby’s placenta is buried inside the house of its birth. Hmong people believe that after death the soul must retrace the journeys undertaken in life until is reaches the burial place of its placenta jacket. The placenta must be retrieved after death to ensure physical integrity in next life. CONSUMPTION OF THE PLACENTA (PLACENTOPHAGY) Placenta is the organ through which the foetus receives nourishment. Many cultures consider it rich in nutrients. It is even said to relieve, or prevent a postnatal depression. In Uganda, the Baganda refer to the placenta as a “spirit-child”. The grandparents eat it, so that the child’s spirit will remain in the clan. Except for the placenta of the king, because this is considered to posses lethal power. It is always carefully dried and preserved. When the old king dies, the new king is considered to own the royal power only upon being given the old king’s royal placenta [7]. Consumption of the placenta by mothers is considered traditional among some Vietnamese and Chinese people [3]. Chinese believe a nursing mother should boil the placenta to make a broth (bouillon) and then drink it to improve her milk. Li Shih-chen recommended a mixture of milk and placental tissue for a disease known as “ch’I exhaustion”, an ill-defined syndrome characterized by anemia, weakness of the extremities and coldness of the sexual organs with involuntary ejection of semen [8]. The Chinese ch’I was a vital force (like the placenta in Egypt). Indians eat the placenta as a delicacy and a source of strength and [9]. Even the British cooking show “TV Dinner” had an edition on placentophagy, although it did not go well. After the broadcasting in 1998, the British Journal “The Independent” wrote: “The program showed a London couple celebrating the birth of their granddaughter by preparing and eating the baby’s placenta as a means of reflecting rituals around the world and symbolical sharing in the baby’s gene pool. They served it as a paté on bread. The father ate 17 portions, but the other guests were less enthusiastic”. More recently, the Dutch “Volkskrant Magazine” had an item on modern placentophagy [10]. The owner of a restaurant in Amsterdam (“Kantine West”) prepared a special 7- 146 course dinner for a couple and six friends to celebrate their newborn baby. Two of the courses were prepared of the placenta from this baby. The guests were surprised about the tastiness of these courses. Although it is not presented on the official menu to avoid provoking other guests who are not interested, the owner of the restaurant is prepared to cook a comparable dinner on request. For those who want to try to prepare a “placental meal” themselves, in September 1983, Mothering Magazine (USA) published some recipes with human placenta. You can find these recipes after the reference-section of this survey. HEALING POWERS All over the world there are people who believe(d) in a beneficial power of the placenta and/or its umbilical cord. In Argentina Araucan Indians dried the cord and ground it to powder to give a little to the child when it was sick [3]. A comparable ritual was performed in Haiti, where the Mirebalais preserved part of the cord. If the child became ill the cord was boiled and the broth (bouillon) was given as a medicine [3]. The people from the village of Topotzlan In Mexico used the umbilical cord as a remedy for eye trouble, as described in study of Robert Redfield from Chigaco 1930 and confirmed by Oscar Lewis in 1951 [3]. In Peru they let an ailing infant suck on the cord [3]. Jamaicans used the placental membranes to prevent irritation of the child by a ghost. Therefore the tissue was carefully parched over a hot stone and put into the infant’s tea. In the “modern” Western world special power was attributed to the placenta, as well. Namely, some believe placentophagy helps to prevent postnatal depression [11]. Especially in France placental extracts were used in commercially available (facial) creams until 1994. From that year on, this fell into disuse, for in the united Kingdom the collection of placentae from unknowing mothers was forbidden. The reason for this was that it has become obvious that each year 360.000 kilograms were shipped into France for cosmetical and pharmaceutical industries over there. Up to about ten years ago in many Dutch hospitals – under which our own institution – placentae were sold for one guilder each to the pharmaceutical industry, as well. 147 EXTRA Nowadays there is a new placenta-related therapeutical goal. After birth of the child and clamping of the umbilical cord, foetal blood remains in the umbilical and placental vascular bed. This blood contains many stem cells, which can be used for a bonemarrow transplantation [12]. However, the placenta and placental proteins have also been associated with the origin of diseases. For example preeclampsia does not disappear as long as the placenta is in the uterus. Recently, Oura et al. [13] showed that a human placental growth factor (PIGF-2) is linked with an increase in inflammatory response, resulting in vascular enlargement, edema and inflammatory cell response. In summary, in most rituals the placenta is believed to contribute to a fortunate and healthy future for the child it originates from. However, no recent proof of these rituals was found in literature, and probably most rituals presented here are seldom, if ever, carried out nowadays. Nevertheless, these historical and socio-cultural perspectives on the human placenta emphasize the importance of this special organ for human beings and for human being. 148 REFERENCES [1] Dikke van Dale, 11e ed. 1984, Van Dale Lexicografie B.V., Utrecht, the Netherlands [2] Smith, G.E. Human History, New York, Norton, 1929, 326-343 [3] William B. Ober, M.D. Notes on placentophagy. Bull. N.Y. Acad. Med, 1979; 55(6):5919. [4] Iconographia Gyniatrica, a pictorial history of Gynecology and Obstetrics, Harold Speert, MD, F.A. Davis Company, Philadelphia, 1973 [5] Wilken, G.A. Handeling voor de Vergelijkende Volkenkinde van Nederlandsch-Indië. Leyden, Brill, 1893, 205 [6] Cadwell I, Volkman TA & Grimes J, 1997, Sulawesi, The Celebes. Periplus Editions, Singapore [7] [8] DeMause L, 1981. Foundations of Psychohistory. The Psychohistory Press, institute for Psychohistory, 140 Riverside Drive, NY 10024 Cooper, W.C. and Sivin, N. Man as a Medicine: Pharmacological and Ritual Aspects of Traditional Therapy Using Drugs Derived from the Human body. In: Nakayama N. and Sivin N. Dhinese Science: Explorations of an Ancient Tradition. Cambridge, MIT Press, 1973, 221, 227-228 [9] Notes from grandfather of A.H.T. Robb-Smith, M.D., F.R.C.P. Woodstock, England Letter to the editor: Egerton Y. Davis, M.D., and placentophagy, Bull. N.Y. Acad. Med 1980; 56(2): 258-9 [10] Volkskrant Magazine, PCM Journals, Amsterdam, the Netherlands, 13 December 2003, page 8 [11] Mary Field, RGN, SCM. Placentophagy, Midwives Chronicle & Nursing Notes, November 1994 [12] Barker JN, Wagner JE. Umbilical-cord blood transplantation for the treatment of cancer. Nat Rev Cancer. 2003 Jul; 3(7): 526-32. Review. [13] Oura H, Bertoncini J, Velasco P, Brown LF, Carmeliet P, Detmar, P. A critical role of placental growth factor in the induction of inflammation and edema formation. Blood. 2003 Jan 15; 101(2): 560-7. 149 EXTRA RECIPES (Mothering Magazine, USA, 1983) placenta cocktail ¼ cup raw placenta 8oz V-8 juice 2 ice cubes ½ cup carrot blend at high speed for 10 seconds placenta lasagna Use your favorite Lasagna recipe and substitute this mixture for one layer of cheese: Quickly saute meat of ¾ placenta, ground or minced in 2 tablespoons (tbl) of olive oil, add 2 sliced cloves of garlic, ½ tsp. Oregano, and one whole tomato or 2 tbl tomato paste. placenta spaghetti Cut the meat of ¾ placenta into bite size pieces, then brown quickly in 1 tbl butter plus 1 tbl olive oil. Then add one can of tomato puree, 2 cans crushed pear tomatoes, 1 onion, 2 cloves of garlic, 1 teaspoon (tsp) molasses (“suikerstroop”), 1 bay leaf, 1 tbl rosemary, 1 tsp each of salt, honey, oregano, basil and fennel. Simmer 1½ hour. placenta stew Cut meat or ¾ placenta in bite size chunks, 1 potato (cubed), ¼ cup fresh parsley, 2 carrots, 3 ribs celery, 1 zucchini, 1 large tomato, and 1 small onion. Dredge meat in 1 tbl spoon flour mixed with 1 tsp salt, ½ tsp paprika, a pinch of cloves, pinch of pepper, 6-8 crushed coriander seeds. Saute meat in 2 tbl oil, then add vegetables (cut cup) and 4-5 water. Bring to full boil, then simmer for one hour. placenta pizza Grind placenta. Saute in 2 tbl olive oil with 4 garlic cloves, then add ¼ tbl fennel, ¼ tsp pepper, ¼ tsp paprika, ¼ tsp salt, ½ tsp oregano, ¼ tsp thyme, and ¼ cup of wine. Allow to stand for 30 minutes with your favorite homemade pizza recipe. Additionally, my colleague (AR) suggested to start a new Dutch tradition, perhaps useful during a baby-shower? “moederkoekhappen” 150 DANKWOORD DANKWOORD Het leven van een promovendus wordt vaak gezien als een eenzaam bestaan. Het feit dat er uiteindelijk één naam op de voorkant van het boekje staat, versterkt deze gedachte alleen maar. Niets is echter minder waar. Ondanks tegenslagen is het me gelukt de overschrijding van de vier jaar die ik oorspronkelijk had te beperken. Niet in de laatse plaats is dit te danken aan de goede en prettige samenwerking met anderen. Een aantal wil ik met naam noemen. Op de eerste plaats natuurlijk Prof. Smits, beste Paul; Bedankt voor het telkens opbouwende commentaar en immer positieve instelling. Met name dat laatste is voor mij erg belangrijk geweest. Het is bekend dat een promotieonderzoek zelden zonder valkuilen, struikelblokken en andere tegenslagen verloopt. Maar de hindernissen op mijn weg waren soms wel buitenproportioneel van aard. Fijn dat er altijd een doos tissues op jouw kamer klaarstond, zodat ik me weer voldoende kon ontladen voor een volgende stap. Prof. Russel, beste Frans; De dag dat ik begon als arts-onderzoeker, was voor jou de eerste dag als hoogleraar. Het placentaonderzoek was binnen de transportgroep een beetje een buitenbeentje. Dat bleek zeker na de verhuizing van een deel van de afdeling naar “de toren”, waar men zich voornamelijk met transport in de nieren bezig houdt. Gelukkig bleef jij gemakkelijk bereikbaar. Wel was het voor mij merkbaar dat mijn basis totaal anders is dan de jouwe, want zo diep als jij in het transport zit, dat was voor mij niet altijd gemakkelijk te volgen. Ach, een beetje uitleg en studiewerk doet wonderen… Ook jij bedankt. Prof. Steegers, beste Eric; Jij was de eerste die met mij over onderzoek begon. Ik weet niet of ik zonder jou ooit aan een promotieonderzoek zou zijn begonnen. Ik ben blij dat ik het heb gedaan, ik heb er veel van geleerd. Midden in mijn onderzoeksjaren heb jij Nijmegen verlaten. Met name in het begin was dat lastig. Maar net als met jouw carrière is het ook met dit proefschrift goedgekomen. Dr. Elving, beste Lammy; Bedankt voor je kritische blik op het onderzoek en vooral op de manuscripten. Jouw adviezen hebben een grote bijdrage geleverd aan de tekst die in dit proefschrift is gedrukt. Bedankt ook dat je ook aandacht had voor het leven naast dat van onderzoeker. Fons Wouterse; De belangrijkste persoon bij het aanleren van de experimentele technieken en het leren kennen van de ins en outs van de opstellingen. Buiten dat kwam jij ook altijd opdraven op de meest idiote tijdstippen als ik weer eens met zwangerschapsverlof of iets dergelijks was om ervoor te zorgen dat er geen placenta van een zwangere vrouw met diabetes mellitus werd gemist. Jouw praktische bijdrage is denk ik wel de belangrijkste 155 CURRICULUM VITAE geweest om ervoor te zorgen dat het onderzoek niet te veel vertraging opliep. Maar ik geloof dat je wel weet hoe dankbaar ik je ben. Jenny Copius Peereboom-Stegeman; Net als jij, mag ik mijzelf nu een beetje placenta-goeroe noemen. Bedankt voor het meedenken aan het morfologisch gedeelte van mijn studie. Bedankt dat je bereid was de begeleiding van Selina over te nemen tijdens mijn afwezigheid, mede namens Selina. Jeroen van der Laak: Bedankt voor het ontwikkelen van de VIDAS, het rekenwerk en de begeleiding van Selina, die onverwacht iets intensiever was dan tevoren was ingeschat. Zonder ook maar een seconde erover na te denken, ben jij samen met Selina verder gegaan. Geweldig! Leni Siero; Ook jij nam belangeloos een deel van het onderzoek over, omdat ik met een kind in mijn buik niet aan straling blootgesteld mocht worden. De experimenten die jij hebt voortgezet zijn van groot belang geweest voor het totstandkomen van hoofstuk 7. Bedankt. Joop Borggreven; Inmiddels heb jij de afdeling verlaten om van je verdiende vrije tijd te genieten. Voor je vertrok heb je voor mij de compliantie van de foetale vaten van de placenta bestudeerd. Wat was ik blij dat deze niet verschillend bleek tussen diabetespatiënten en controlepersonen!!! Jeanne Pertijs; Bedankt voor het prepareren van de vaten die Joop heft gebruikt in het onderzoek. Ik heb het later ook geprobeerd voor ander onderzoek; het is monnikenwerk. Yuen Tan; Sommigen zullen jou alleen nog maar kennen uit “verhalen over vroeger”. Voor mij was je belangrijk voor het opzetten en aanleren van de HPLC-methode om het foliumzuurgehalte in de perfusievloeistof te bepalen. Bedankt voor je eeuwige geduld met deze – in jouw woorden - “kwebbeltante”. Jeroen van den Heuvel; Van jou heb ik geleerd dat er bij een Western Blot meer komt kijken dan je denkt. Ik ben er trots op dat ik heb bijgedragen aan jouw eerste publicatie. Lisanne van den Berg, Chantal Bovee, Miranda Stringer, Alexander Rennings en Peter Pickkers; Bedankt voor de gezelligheid, attentheid, eerlijkheid en alle andere dingen die je van een goede collega/kamergenoot mag verwachten. De omschakeling van vrouwen naar mannen was even groot, maar viel zeker niet tegen. 156 DANKWOORD Alle collega arts-onderzoekers van de afdelingen Obstetrie-Gynaecologie, Epidemiologie en Biostatistiek en Interne Geneeskunde; Dit zijn er de afgelopen jaren zoveel geweest dat ik maar geen namen noem omdat ik bang ben iemand te vergeten. Bedankt voor de prettige samenwerking, goede nieuwe ideeën, de uitjes etc. Kortom, alles wat voor een arts-onderzoeker van belang is. Eva Maria Roes en Maarten Raaijmakers bedankt voor de goede samenwerking en prettige afronding van de NAC-studie in de placenta. Simone Dekker, Femke van de Water, Annabel Walker, Marieke Verstegen, Selina van der Wal en Eva Goula; Bijna geen enkel proefschrift komt tot stand zonder een bijdrage van studenten. Buiten werkelijke hulp brengt dat ook een hoop leven in de brouwerij. Daar houd ik wel van… Hilde Basten - Marjo Beuting - Caren Thijssing - Marja Brederveld - Margreet Centen Loes van Dalen - Marloes van Dalen - Hans van Dillen - Rianne van Gaalen - Jos van Gent - Mariёlle Fisscher - Vera van Haaren - Irma Buis - Hester de Haas - Marijke van der Heijden - Inigo Heijmans - Els Hendriks - Erik Hendriks - Lena Holtman - Lilian Janssen - Maartje de Kat - Marlies Knuiman - Patricia Koppelman - Petri Kusters - Iris van Mil - Monique Hoogeveen - Moniek Raemakers - Marjolein Scheltinga - Daniēlle van Hove - Marja Scholten - Marij Strijbosch - Sandra Timmer - Maria Vos - Dorrit Grosfeld Kim van de Water - Betty van der Weijden - Ali Wessel; Bedankt voor jullie hulp bij het verkrijgen van de placenta’s. Moeders; Bedankt voor het afstaan van de placenta en bloed voor dit onderzoek. Medewerkers van de laboratoria CKCL en ACE; Bedankt voor de hulp bij de benodigde bepalingen, in het bijzonder Dorine Swinkels, Annelies Mooren, G. van de Wiel, Fred Sweep en Tijn Seegers. Greet Goverde; Fijn dat je bereid was om het manuscript na te lezen op Engelse taal- en stijlfouten. Bas van Abel; Een zelfgebakken kaft om het proefschrift. Ik moet de eerste zijn die dat kan zeggen. Hij ziet er geweldig uit. Else en Maike; Omdat jullie achter mij (zijn blijven) staan. Mama; Jij was en bent nog altijd een belangrijke spil in ons gezin. Er is zoveel waarvoor ik je moet bedanken. Met name voor al die keren dat ik je onverwacht om hulp vroeg bij de opvang van de kinderen. Maar niet in de laatste plaats gewoon omdat je al ruim 33 jaar mijn moeder bent. 157 CURRICULUM VITAE Amber, Mees en Jasmijn; Bedankt voor al die keren dat jullie mij frietjes kwamen brengen op het lab als ik weer eens tot diep in de nacht doorging. Erik; Mijn man, vader van mijn kinderen. Volgens mij was het af en toe flink afzien als ik het weer onverwacht liet afweten aan het eind van de dag. Je bent een geweldige vader voor onze kinderen en een fantastische man voor mij. Bedankt voor jouw onvoorwaardelijke steun in de afgelopen jaren. 158 CURRICULUM VITAE CURRICULUM VITAE Tanya Maria Bisseling werd geboren op 24 mei 1971 als oudste van vier dochters van Jan Bisseling en Ria Bisseling-Hermsen. In 1989 deed zij eindexamen aan het Stedelijk Gymnasium te Nijmegen, waarna zij aan de studie Geneeskunde aan de Katholieke Universiteit te Nijmegen begon. In mei 1994 behaalde zij haar doctoraalexamen, gevolgd door het artsexamen in maart 1997. Van april 1997 tot februari 1998 werkte zij als AGNIO Heelkunde in het Rijnstate Ziekenhuis te Arnhem. Vervolgens werkte zij als AGNIO Obstetrie-Gynaecologie in het Elisabeth Ziekenhuis te Tilburg. In 1997 trouwde zij met Erik Olde. Tijdens de zwangerschap van haar oudste dochter Amber aanvaarde zij in 1999 een baan als arts-onderzoeker op de afdeling FarmacologieToxicologie. In februari 2000 begon zij onder leiding van Prof. dr. P. Smits in samenwerking met de afdeling Obstetrie-Gynaecologie aan het onderzoeksproject dat uiteindelijk heeft geleid tot dit proefschrift met de titel “Placental function in maternal disease; ex vivo assessment of foetoplacental vascular function and transport in diabetes and preeclampsia”. Tegelijkertijd was zij als (toekomstig) co-promotor betrokken bij een studie naar de kwaliteit van leven van vrouwen na een behandeling met een Tensionfree Vaginal Tape voor stress-incontinentie. In maart 2001 en november 2002 werden Mees en Jasmijn geboren. Thans is zij werkzaam op de afdeling Interne Geneeskunde in het Ziekenhuis Bernhoven te Veghel. 160