Nitric oxide and human umbilical vessels: Pharmacological and immunohistochemical studies

Placenta

(1995),

16, 277-288

Nitric Oxide and Human Umbilical Vessels: Pharmacological and Immunohistochemical Studies A. J. SEXTON, A. LOESCH, M. TURMAINE, S. MIAH & G. BURNSTOCK” Department of Anatomy and Developmental Biology and Centre for Neuroscience, University College London, Cower Street, London, WCIE 6BT, UK “To whom correspondence should be addressed Paper accepted 4. I. I 995

SUMMARY Human umbilical vessels are devoid of nerves and therefore endothelial cells may play an important role in the control offetoplacental bloodftow. In this study we examined the pharmacological eflects of various substances, known to produce endothelialmediated vasodilatation in many blood vessels, on the human umbilical artery and vein from legal terminations [mean gestational age, 15 (8-17) weeks; n=12] and normal term vaginal deliveries [mean gestational age, 39 (38-41) weeks; n=12]. Acetylcholine, adenosine S-triphosphate, the calcium ionophore A23187 and substance P had no effect on raised vascular tone, whereas sodium nitroprusside relaxed Fhydroxytyptamine (5-HT) preconstricted, umbilical artery and vein from both early and late pregnancy. L-fl-Nitroarginine methyl ester (L-NAME) had no efect on basal tone or on high tone, after it was raised by S-HT. Localization of nitric oxide synthase [NOS, type I (neuronal)] was examined in the same umbilical vessels using electron immunocytochemistry. No NOS-immunoreactive endothelial cells were observed in the umbilical vessels taken during early pregnancy. However, the percentage of NOS-immunoreactive endothelial cells in umbilical artery and vein from late pregnancy was 3 and 10 per cent, respectively. These results suggest that nitric oxide contributes little, if any, to the local control of umbilical blood flow throughout pregnancy, despite the presence of NOS-immunoreactivity in a subpopulation of endothelial cells in late pregnancy.

INTRODUCTION

The human umbilical and placental circulation both lack innervation (Spivak, 1943; Reilly and Russell, 1977; Fox and Khong, 1990). Therefore the endothelium may play an important role in the regulation of vascular tone. Endothelium-derived relaxing factor (EDRF) was first 0143-4004/95/030277

+ 12 $08.00/O

0

1995 W. B. Saunders

Company

Ltd

278

Placenta

(1995),

Vol. 16

demonstrated in 1980 by Furchgott and Zawadzki. It has since been identified as nitric oxide (NO) (Palmer, Ferrige and Moncada, 1987), and is produced within endothelial cells, from the amino acid L-arginine by the enzyme nitric oxide synthase (NOS) (Palmer, Ashton and Moncada, 1988). NO appears to be an autocrine/paracrine signalling agent synthesized in many tissues including the endothelium, cerebellum, platelets, hepatocytes and macrophages, where it exhibits diverse actions including vasodilatation (Palmer, Ashton and Moncada, 1988). In 1987, Van de Voorde and colleagues, demonstrated that histamine, the calcium ionophore A23187 and adenosine 5’-triphosphate (ATP) [but not acetylcholine (ACh) or 5-hydroxytryptamine (S-HT)] released a substance from the umbilical artery and vein with an intact endothelium which caused relaxation in endothelium denuded pre-contracted rat aortic rings. This substance was identified as NO. Later, Myatt et al (1991) also demonstrated the presence of EDRF in the human preconstricted dually perfused cotyledon, and Tsukahara, Gordienko and Goligorsky (1993) have shown that human umbilical vein endothelial cells (HUVECs) release NO on stimulation to several agonists. The aim of this study was to investigate endothelium-dependent relaxation in umbilical vesselsfrom different stages of gestation, and to observe whether the endothelial cells of these umbilical vessels contain NOS.

METHODS Collection

of tissue

Human umbilical cords from legal terminations of pregnancy [early pregnancy group, gestational age, 15 (8-17) weeks, n=12] were obtained from the MRC Tissue Bank, Royal Marsden Hospital and from normal term vaginal deliveries at University College Hospital [late pregnancy group, gestational age, 39 (38-41) weeks, n=12]. The cords were placed into a modified Kreb’s solution at 4°C of the following composition (mM): NaCl 118; KC1 4.7; NaH2P0, 25; Mg,SO, 1.2; CaCl, 2.5; glucose 11.1. Segments of the same cords were used within 12 h for both pharmacological and immunocytochemical studies.

Isolated

vessel pharmacology

Ring preparations of umbilical artery and vein, approximately 5 mm in length, were dissected free of surrounding tissue and mounted horizontally in 5 ml organ baths containing a modified Kreb’s solution which was continually aerated with 95 per cent 0,/S per cent CO, and maintained at 37°C. Preparations were mounted horizontally in the organ baths by inserting two tungsten wires through the lumen of the vessel, attached to a rigid support and a Grass force displacement transducer FT03C. Mechanical activity was displayed on a Grass ink-writing oscillograph. An initial load of 2 g was applied and the tissues were allowed to equilibrate for 2 h, resulting in an approximate tension of 1.5 g. Concentration-response curves (10 nM-30 PM) were constructed for .5-HT in order to obtain an EC,, (the concentration which caused 75 per cent of the maximal contraction) for each segment. This concentration of S-HT was used to constrict the vesselsin order to observe relaxations to a variety of known vasodilators. The effect of L-NG-nitroarginine methyl ester (L-NAME) (100 PM) was studied on basal tone and tone raised by 5-HT. The presence of an intact endothelium was checked after each experiment using silver nitrate staining. Vasodilator responses are expressed as mean per cent contraction of 5-HT f s.e. and statistical

Sexton et al: Nitric

Oxide and Human

Umbzlical

Vesseb

significance was determined using an unpaired Student’s t-test. A probability of taken as significant.

Pre-embedding

279

PcO.05was

immunocytochemistry

Segments of umbilical cords, from the mid to placental region, were immersed in fixative containing 4 per cent paraformaldehyde and 0.1 per cent glutaraldehyde in 0.1 M phosphate buffer at pH 7.4 and stored in the fixative at 4°C overnight. The following day the umbilical arteries and veins were dissected free of surrounding tissue, washed in phosphate buffer and cut longitudinally to produce about 3-5 mm long strips. These were then processed for immunocytochemistry according to the peroxidase-antiperoxidase (PAP) method of Sternberger (1979) using the same steps of the procedure as previously reported (Loesch and Burnstock, 1993; Loesch, Belai and Burnstock, 1993, 1994). Initially, however, the strips were exposed to 0.3 per cent H,O, in 30-50 per cent methanol for 30 min (for blocking of endogenous peroxidases). The rabbit NOS antiserum was used in this study at 1: 1000 dilution. After immunoprocedure, the specimens were post-fixed in 1 per cent OsO,, dehydrated in a graded series of ethanol and embedded in Araldite. The ultrathin circumferential sections were stained with uranyl acetate and lead citrate and subsequently examined with a JEM-1010 transmission electron microscope.

Controls

The rabbit polyclonal NOS antiserum, prepared by Abbott Laboratories, Chicago, USA, was kindly provided by Dr U. Forstermann (Abbott Laboratories, USA). It was raised against soluble NOS Type I (neuronal) purified from rat cerebellum and tested for specificity as described previously in detail (Schmidt et al, 1992). Anti-NOS I antibody reacted with a single 160 kDa protein band in rat cerebella 105 000 g supernatant fraction. In all tissues examined (rat: cerebellum, kidney, pancreas, lung, uterus) the relative intensity of the cross-reacting protein band (1.50-l 65 kDa) coincided with NOS I activity (Schmidt et al, 1992). Pre-absorption (22 h at 4°C) of the NOS antiserum (diluted 1:lOOO)with soluble NOS Type I purified from rat cerebellum resulted in elimination of positive labelling in endothelial cells and perivascular nerves in the pre-embedding PAP immunoprocedure (Loesch and Burnstock, 1993; Loesch, Belai and Burnstock, 1993, 1994). In the present study, the specificity of the immunolabelling was tested by omission of the NOS antiserum and anti-rabbit immunoglobulin G serum independently, as well as by using non-immune normal goat serum or non-immune normal rabbit serum (Nordic Immunology, Tilberg, The Netherlands) at a dilution of 1:lOOOinstead of NOS antiserum. No labelling was observed in these control experiments.

RESULTS Preconstriction

by S-HT

5-HT (10 nM-30 PM) induced concentration-dependent constriction of the human umbilical artery and umbilical vein with pD, values of 6.9 f 0.1 in the artery, 6.9 + 0.2 in the vein from early pregnancy and 7.0 & 0.1 in the artery and 6.9 f 0.1 in the vein from late pregnancy. Vasoconstriction was completely reversible in both artery and vein after washout to remove 5-HT.

Placenta

280

(1995),

Vol. 16

(a) Control

(4 SNP 2g

LU

A 0.1

I 0.3

5-HT c

I 1

2 min

I 3

I 10 PM

Figure 1. Original trace from experiments showing the effect of (a) control, (b) acetylcholine (ACh) (0.1-300 ELM), (c) adenosine 5’-triphosphate (ATP) (0.1-300 PM), and (d) sodium nitroprusside (SNP) (0.1-310 PM) on human umbilical artery (a) and (b) and vein (c) and (d) from late pregnancy. S-HT, S-hydroxytryptamine.

Endothelium-dependent

relaxation

in the human

umbilical

artery

and vein

Umbilical artery and umbilical vein preparations from early (~~12) and late pregnancy (~~12) with an intact endothelium constricted by S-HT did not relax in response to acetylcholine (ACh, 0.1-300 PM), adenosine 5’-triphosphate (ATP, 0.1-300 PM), substance P (SP, O.Ol30 nM) and calcium ionophore A23187 (0.01-30 PM) (see Figures 1 and 2). Sodium nitroprusside (SNP, 0.1-300 PM) was able to relax preconstricted artery and vein preparations from both early and late pregnancy (see Figures 1 and 2). This was 87.9 f 3.8 per cent (~6) and 95.0 f 6.8 per cent (~5) of the established tone, for artery and vein, respectively, from early pregnancy and 78.8 f 6.9 per cent (n=6) and 81.6 f 8.7 per cent (~6) for the artery and vein, respectively, of the established tone for late pregnancy (Figure 3). Effect

of L-NAME

No change in basal or raised vessel tone was observed on incubation for up to 4 h of umbilical artery and vein preparations from both early and late pregnancy, with 100 PM of L-NAME. Electron

immunocytochemistry

By using the technique of pre-embedding immunocytochemistry, endothelial cells displaying NOS (type I)-immunoreactivity were observed in both umbilical artery and vein of late

Sexton et al: Nitric

Oxide and Human

Umbilical

Vessels

(a) ACh I 0.5 g

Cd) SNP lg

L

1 min

Figwe 2. Original trace from experiments showing the effect of (a) acetylcholine lonophore A23187 (0.01-30 5’-triphosphate (ATP) (0.1-300 FM), (c) Ca” (SNP) (0.1-300 PM) on human umbilical artery (a), (b) and (d) and vein 5-hydroxytryptamine.

(ACh) (0.1-300 p.~), (b) adenosine and (d) sodium nitroprusside (c) from early pregnancy. 5HT,

FM),

pregnancy (Figure 4). Examination of numerous sections that were taken at different levels of the specimens showed a different percentage of immunoreactivity in the umbilical artery and umbilical vein. Analysis of 490 arterial endothelial cells revealed that only 3 per cent of these cells were positive for NOS (15 positive cells out of the 490 cells examined), whereas analysis of 566 venous endothelial cells showed NOS-immunoreactivity in 10 per cent of cells (56 positive cells out of the 566 cells examined). In the labelled cells the immunoprecipitate was distributed throughout the cell cytoplasm, although the density of immunostaining varied from cell to cell. Labelled cells contained a large, unlabelled nucleus and numerous organelles, especially swollen mitochondria and endoplasmic reticulum. Labelling to NOS was also observed outside the endothelium (extracellularly). In the umbilical vein, immunoprecipitate was seen in structures difficult to identify as illustrated in Figure 5(a). In the umbilical artery NOS-immunoreactivity was associated with material covering the endothelial cells from the luminal side which showed NOS-immunoreactivity, as shown in Figure 5(b). Labelling to NOS was not observed in the control preparations of umbilical arteries and veins from early and late pregnancy [Figure S(c)], nor was any NOS-immunoreactivity observed in endothelial cells from umbilical arteries and veins from early pregnancy (Figure 6). DISCUSSION Our pharmacological studies have failed to demonstrate any significant evidence of endothelial-dependent relaxation in the human umbilical artery and vein, either at early or late

Placenta

282

(199.5),

Vol. 16

loo- (4 go8070g ‘~3 60d 3 50& $ 40B g 30d

Jl

20-

-Log Figure 3. Cumulative (EC,, concentration) from early pregnancy. represent percentage

[SNPI

-Log

[SNPI

concentration-response curve to sodium nitroprusside (SNP) on 5-hydroxytryptamine (5-HT) preconstricted umbilical vessels. (a) ( l ), Umbilical artery (~6); and ( W), umbilical vein (~5) (b) ( l ), Umbilical artery (~6); and ( W), umbilical vein (~6) from late pregnancy. Symbols relaxation f s.e. (unless masked by symbol).

pregnancy. This is in accordance with the reports of other authors; Monuszko et al (1990) did not see any relaxation to ACh or A23187. Griggs and McLaughin (1989) demonstrated this lack of endothelial-dependent relaxation in the chorionic plate arteries of the term placenta, and proposed that this was due to the insufficient release of EDRF/NO to have an effect on the underlying smooth muscle. According to Hull, White and Pearce (1993) human umbilical vesselsdo not relax to ACh, adenosine diphosphate (ADP), bradykinin (Bk), histamine, SP, or A23187. It has been suggested that this lack of relaxation results from a near maximal production of EDRF in the placental arteries, leaving no capacity for increased release to secondary stimulation. However, Van de Voorde, Vanderstichele and Leusen (1987) and Chaudhuri et al (1991) like ourselves, have demonstrated the lack of relaxation due to basal release of NO. Klockenbusch and co-workers (1992) h ave also seen a lack of relaxation due to basal NO release, despite a significant basal NO generation, as seen from cyclic GMP generation. He also rules out the possibility that this could be due to a deficiency of endogenous substrate, i.e. L-arginine, since histamine induced an increase in cyclic GMP accumulation, which was suppressed by nitro-L-arginine. It has been shown that HUVECs are capable of releasing NO when stimulated by the agonists, thrombin, Bk, L-arginine and ionomycin (Tsukahara, Gordienko and Goligorsky, 1993). Cascade experiments with human umbilical vessels, have shown that they release a substance capable of dilating the endothelium denuded, ring preparations of rat aorta and bovine pulmonary artery (Van de Voorde, Vanderstichele and Leusen, 1987; Chaudhuri et al, 1991). Using the same techniques Izumi et al (1994) suggest that the amount of EDRF and the sensitivity of smooth muscle to EDRF decreases with gestation. NO synthesized from L-arginine acts through the stimulation of the soluble guanylate cyclase to increase cyclic

Sexton et al: Nitric

Oxide and Human

Umbilical

Vessels

283

Figure 4. Electron micrographs of umbilical vessels in late pregnancy immunolabelled for nitric oxide synthase (NOS): (a) Umbilical artery and (b, c) umbilical vein immunolabelled for NOS. (a) A fragment of artery displays the presence of one endothelial cell with positive cytoplasmic immunoreactivity for NOS (white star). Note the unlabelled nucleus (N). The neighbouring cell profiles (“) are NOS-negative. lo, lumen ( x 3800). (b) A fragment of vein shows at least two NOS-positive neighbouring endothelial cells. el, elastic lamina; sm, smooth muscle ( x 4550). (c) Note the two NOS-positive endothelial cells showing a difference in the intensity of cytoplasmic immunolabelling. Endoplasmic reticulum (er), mitochondria (m) and filaments (f) are seen in intense labelled cell. no, nucleus ( x 6770).

284

Placenta

(1995),

Vol. 16

Figure 5. Electron micrographs of (a) umbilical vein (term) and (b) umbilical artery (term) immunolabelled for nitric oxide synthase (NOS), and (c) an example of control specimen (term). (a) Note the intense NOS-immunolabelling in a membrane bound profile (large white star) localized below the NO.5negative endothelial cell (“). The neighbouring endothelial cell (small white star) is NOS-positive. N, cell nucleus; er, endoplasmic reticulum; m, mitochondria; el, elastic lamina; sm, smooth muscle; lu, lumen ( x 6150). (b) Note the intense NOS-positive material (arrows) associated with the luminal surface of unlabelled endothelial cells ( x 4900). (c) Control preparation (omission of NOS-antiserum) showing absence of NOS-immunolabelling ( x 3800).

Sexton et al: Nitric

Oxide and Human

Umbilical

Vessels

Figure 6. Electron micrograph of the (a) umbilical artery and (b) umbilical vein, from early pregnancy following immunolabelling for nitric oxide synthase (NOS). No NOS-positive endothelial cells were seen. N, cell nucleus; er, endoplasmic reticulum; m, mitochondria; sub, subendothelial zone; sm, smooth muscle; lu, lumen. (a) X 5170; (b) x 6460.

GMP levels in vascular smooth muscle to produce relaxation (Moncada and Palmer, 1991; Nathan, 1992). Renowden, Edwards and Griffith (1992), have shown that the umbilical artery is unable to relax significantly to atria1 natriuretic peptide (ANP), SNP, and A23187, which activate cyclic GMP and forskolin which activates cyclic AMP, when using maximal concentrations of S-HT. These results imply that the umbilical artery and vein smooth muscle are not responsive to endothelial-derived NO, although it is curious that SNP, which is usually taken to mimic the actions of NO, was shown to produce relaxation of these vessels. However, it has been demonstrated that changes in flow and shear stress may be a stimulus for NO release (Hecker et al, 1993). The present study shows that no NOS (type I, neuronal)-immunoreactive endothelial cells are present in human umbilical artery and vein endothelial cells from early pregnancy. By contrast, endothelial cells from umbilical artery and vein in late pregnancy are a heterogeneous

286

Placenta (199.i), Vol. 16

population for NOS-immunoreactivity, with, although still small, a greater percentage of NOS-immunoreactive endothelial cells in the umbilical vein. A low percentage of NOSimmunoreactive endothelial cells has also been observed in other vessels, for example, 6 per cent of endothelial cells were immunoreactive for NOS in the rabbit aorta (Loesch and Burnstock, 1993) and 7 per cent of endothelial cells were immunoreactive for NOS in the rat basilar artery (Loesch, Belai and Burnstock, 1994). Both these vessels are known to exhibit strong endothelium-dependent relaxations. Present results are in contrast to the relatively high percentage of immunoreactive cells for the peptides ANP and neuropeptide Y (NPY), 32 and 44 per cent, respectively, as observed by Cai et al (1993) in h uman umbilical vessels.This appearance of NOS-immunoreactivity within endothelial cells may be a physiological process through which an increase in fetoplacental blood flow is facilitated as gestation advances (Macara, Kingdom and Kaufmann, 1993). Our results on neuronal NOS-immunoreactivity differ from those of Myatt et al (1993) and Pollock et al (1993), where both groups found a strong immunoreactivity to NOS in umbilical artery and vein using a monoclonal antibody ‘H32’ to an endothelial isoform of NOS (NOS type III). The results of these studies, however, indicate that 50 per cent of veins examined with light microscopy did not display any NOS-immunoreactivity (Myatt et al, 1993), whereas Pollock et al (1993) found ‘strong immunoreactivity’ to NOS in both umbilical artery and vein. Springall et al (1993) have demonstrated that polyclonal antibodies to rat brain NOS (type I) enzyme react with the human equivalent and also with other isoforms in human endothelium, and have found positive immunolabelling in the human umbilicus and trophoblast of the human placenta. Although a 57 per cent sequence homology has been identified between NOS type I (neuronal isoform) and NOS type III (endothelial isoform) (Bredt et al, 1991; Lamas et al, 1992), it cannot be ruled out that the polyclonal antibody used in this study has not recognized all isoforms of NOS that may be present in human umbilical endothelium, labelling for example only those cells which contained neuronal NOS type I. It is striking, however, that our results differ from those of Dikranian et al (1994), showing that in the human umbilical vein at the light microscopic level, high proportions of endothelial cells (97 per cent) immunoreact with an antibody to neuronal NOS (the same antibody as that used in our study). The localization of neuronal NOS in vascular endothelium may not be surprising, since it has already been found that endothelial isoform of NOS and neuronal isoform of NOS may occur in the same neuronal cell populations in the brain (Dinerman et al, 1994). Our data shows that the primary antibody can be adsorbed onto the luminal surface of the umbilical artery and vein, despite the use of strong blockers to inhibit endogenous peroxidases. This may produce misleading immunocytochemical results at the light microscopic level, as at this level, such material can easily account for the endothelium. The possibility cannot be excluded that this phenomenon was partially responsible for high proportions of immunolabelled cells at the light microscopic level as seen by Dikranian et al (1994). Luminal labelling appears to be only in the term vessels, suggesting the physical properties of the endothelial luminal surface may be different in late pregnancy. At this stage it is difficult to explain this phenomenon, but in situ hybridization studies would clarify this issue. We can only suggest that it is possibly a material from the blood trapped between the endothelial cells, when the vesselsconstrict/collapse at birth, or possibly a material secreted from the endothelial cells, which is rich in NOS and/or peroxidases. This study demonstrates the inability of the umbilical artery and vein to relax to known endothelium-dependent vasodilators. This is despite demonstrating a small subpopulation of NOS-immunoreactive endothelial cells, and a functional cyclic GMP system in the smooth

Sexton

et al: Nitrsc

Oxide and Human

Umbilical

Vessels

287

muscle, which is responsible for the action of NO. Previous studies have demonstrated the release of NO from human umbilical endothelial cells from artery and vein, and shown its effect on the vascular smooth muscle from other animals. This could imply that the amount of NO released may be insufficient to affect smooth muscle locally, or that the mechanism coupling with NO with soluble cyclic GMP may be impaired. Therefore it is unlikely that NO is involved in the local regulation of umbilical blood flow throughout pregnancy. ACKNOWLEDGEMENTS The authors wish to thank the MRC Tissue Bank, the Royal supported by a studentship from Heart Foundation. The authors UCH, for his useful comments

nursing staff of the labour suite, University College London, and Dr Marsden Hospital, for their help and assistance in the collection of the Anatomical Society of Great Britain and Ireland. AL is supported also wish to thank Dr J. C. P. Kingdom, Department of Obstetrics and and discussion during preparation of this manuscript.

Wong of the tissue. AJS is by the British Gynaecology,

REFERENCES Bredt, D., Hwang, P., Glatt, C., Lowenstein, C., Reed, R. & Synders, S. (1991) Cloned and expressed nitric oxide synthase structurally resembles cytochrome P450 reductase. Nature, 351, 714-718. Cai, W. Q., Bodin, P., Sexton, A., Loesch, A. & Burnstock, G. (1993) Localization of neuropeptide Y and atria1 natriuretic peptide in endothelial cells of human blood vessels. Cell and Tissue Research, 272, 175-181. Chaudhuri, G., Buga, G. M., Gold, M. E., Wood, K. S. & Ignarro, L. J. (1991) Ch aracterization and actions of human umbilical endothelium-derived relaxing factor. British 3ournal of Pharmacology, 102, 331-336. Dikranian, K., Trosheva, M., Nikolov, S. & Bodin, P. (1994) Nitric oxide synthase (NOS) in the human umbilical cord vessels, An immunohistochemical study. Acta Histochemica, 96, 145-153. Dinerman, J. L., Dawson, T. M., Schell, M. J., Snowman, A. & Snyder, S. H. (1994) Endothelial nitric oxide synthase localized to hypocampal pyramidal cells: implications for synaptic plasticity. Proceedings of the -National Academy of Sciences, USA, 91, 4214-4218. Fox, S. B. & Khong, T. Y. (1990) Lack of innervation of human umbilical cord. An immunohistochemical and histochemical study. Placenta, 11, 59-62. Furchgott, R. F. & Zawadzki, J. V. (1980) The obligatory role of endothelial cells in the relaxation of arterial smooth muscle by acetylcholine. Nature, 288, 3733376. Griggs, K. C. & McLaughin, M. K. (1989) Human placental chorionic plate arteries lack an endotheliumdependent relaxation response. 36th Annual Meeting of the Society for Gynecologic Investigation. San Diego, California, 162, 162. Hecker, M., Mulsch, A., Bassenge, E. & Busse, R. (1993) Vasoconstrictors and increased flow: two principal mechanisms of shear stress-dependent endothelial autocoid release. American Journal of Physiology, 265, H8233H833. Hull, A. D., White, C. R. & Pearce, W. J. (1993) Human placental arteries have a high basal rate of EDRF synthesis. FASEB Journal, 7, A783. Izumi, H., Garfield, R. E., Makino, Y., Shirakawa, K. & Itoh, T. (1994) Gestational changes in endothelialdependent vasorelaxation in human umbilical artery. American Journal of Obstetrics and Gynecology, 170, 236245. Klockenbusch, W., Braun, M. S., Schriider, H., Heckenberger, R. E., Strobach, H. & SchrGr, K. (1992) Prostacyclin rather than nitric oxide lowers human umbilical artery tone in vitro. European3ournal of Obstetrics and Gynecology and Reproductive Biology, 47, 109-l 15. Lamas, S., Marsden, P., Li, G. K., Tempst, P. & Michel, T. (1992) Endothelial nitric oxide synthase molecular cloning and characterisation of a distinct constitutive enzyme isoform. Proceedings of the National Academy of Sciences, USA, 89, 6348-6352. Loesch, A. & Burnstock, G. (1993) Ultrastructural localization of nitric oxide synthase in intima of rabbit aorta. Endothelium, 1, 23-29. Loesch, A., Belai, A. & Burnstock, G. (1993) Ultrastructural localization of NADPH-diaphorase and colocalization of nitric oxide synthase in endothelial cells of the rabbit aorta. Cell and Tissue Research, 274, 539-545. Loesch, A., Belai, A. & Burnstock, G. (1994) An ultrastructural study of NADPH-diaphorase and nitric oxide synthase in the perivascular nerves and vascular endothelium of the rabbit basilar artery. &wnal oJNeurocytology, 23, 49959. Macara, L. M., Kingdom, J. C. P. & Kaufmann, P. (1993) Control of the fetoplacental circulation. Fetal and Maternal Medicine Review, 5, 167-179. Moncada, S. &Palmer, R. M. J. (1991) Biosynthesis and actions of nitric oxide. Seminars in Perinatology, 15, 16-19.

288

Placenta

(1995),

Vol. 16

Monuszko, E., Halevy, S., Freese, K. J., Shih, H. J., Parikh, N. S., Jelveh, Z., Liu-Barnett, M. & Cybulska, J. (1990) Umbilical vessels endothelium and vascular reactivity. Micvocirculation, Endothelium and Lymph&s, 6, 183-208. Myatt, L., Brewer, A. & Brockman, D. E. (1991) The action of nitric oxide in the perfused human fete-placental circulation. American 3ofoumal of Obstetrics and Gynecology, 164, 687-692. Myatt, L., Brockman, D. E., Eis, A. L. W. & Pollock, J. S. (1993) Immunohistochemical localization of nitric oxide synthase in the human placenta. Placenta, 14, 487-495. Nathan, C. (1992) Nitric oxide as a secretory product of mammalian cells. Fj4.5’EB Jfbuvnal, 6, 3051-3064. Palmer, R. M. J., Ferrige, A. G. & Moncada, S. (1987) Nitric oxide release accounts for the biological activity of endothelium-derived relaxing factor. Nature, 327, 524-526. Palmer, R. M. J., Ashton, D. S. & Moncada, S. (1988) Vascular endothelial cells synthesise nitric oxide from L-arginine. Nature, 333, 664-666. Pollock, J. S., Nakane, M., Buttery, L. D. K., Martinez, A., Springall, D., Polak, J. M., Fiirstermann, U. & Murad, F. (1993) Characterization and localization of endothelial nitric oxide synthase using specific monoclonal antibodies. Ame%an Journal of Physiology, 265, C1379-C1387. Reilly, F. D. & Russell, P. T. (1977) Neurohistochemical evidence supporting an absence of adrenergic and cholinergic innervation in the human placenta and umbilical cord. Anatomical Record, 188, 277-286. Renowden, S., Edwards, D. H. & Griffith, T. M. (1992) Impaired cyclic nucleotide-mediated vasorelaxation may contribute to closure of the human umbilical artery after birth. British Journal of Pharmacology, 106, 348-353. Schmidt, H. H. H. W., Gagne, G. D., Nakane, M., Pollock, J. S., Miller, M. F. & Murad, F. (1992) Mapping of neural nitric oxide synthase in the rat suggests frequent co-localization with NADPH-diaphorase but not with soluble guanylate cyclase, and novel paraneural functions for nitrinergic signal transduction. The Journal of Histochemistry and Cytochemisty, 40, 1439-1456. Spivak, M. (1943) On the presence and absence of nerves in the umbilical blood vessels of man and guinea pig. Anatomical Record, 85, 85-109. Springall, D. R., Riveros-Moreno, V., Buttery, L., Suburo, A., Bishop, A. E., Merrett, M. & Moncada, S. (1993) Immunological detection of nitric oxide synthase(s) in human tissues using heterologous antibodies suggesting different isoforms. Histochemistry, 98, 259-266. Sternberger, L. A. (1979) Immonocytochemistry. 2nd Edn. New York: John Wiley. Tsukahara, H., Gordienko, D. V. & Goligorsky, M. S. (1993) Continuous monitoring of nitric oxide release from human umbilical vein endothelial cells. Biochemical and Biophysical Research Communications, 193, 722-729. Van de Voorde, J., Vanderstichele, H. & Leusen, I. (1987) Release of endothelium-derived relaxing factor from human umbilical vessels. Circulation Research, 60, 517-522.