J. Anthony Firth

Fine structure of the paracellular junctions of terminal villous capillaries in the perfused human placeta

SummarySelected lobules of human term placentae were extracorporeally perfused for a recovery period of 20 min, fixed by perfusion and mordanted with ferrocyanide prior to processing for transmission electron microscopy. The lateral membranes of the endothelial cells of the terminal villous capillaries were found to be separated by paracellular clefts of mean width 15.6 nm. At tight junctional regions (1–4 sites per cleft) the two membranes approached each other more closely and frequently appeared to fuse. However, tilting of the sections in the electron microscope stage showed that the membranes were separated by a gap of mean width 4.1 nm in at least 94% of tight junctional profiles. When individual tight junctions were studied by a combination of serial sectioning and goniometric tilting, they were seen to widen abruptly within a distance of three to seven consecutive thin sections, indicating they were not continuous throughout the axial length of the capillaries. The wide regions of the clefts usually showed linkers, strands of glycocalyx-like material spanning the gap. Linkers may contribute to cell adhesion and possibly form part of a filter within the tortuous paracellular pathway provided by the discontinuous network of tight junctional strands. Human term placental capillaries appear to resemble closely other continuous non-brain capillaries.

Uptake and intracellular routing of peroxidase-conjugated immunoglobulin-G by the perfused human placenta

SummarySelected lobules of term human placenta were extracorporeally perfused and human immunoglobulin-G complexed to horseradish peroxidase (IgG-HRP) was added to the maternal perfusate. After different durations of perfusion IgG-HRP was visualised by use of diamino-benzidine cytochemistry. Within the first 10 min of perfusion IgG-HRP was found bound to microvilli and coated pits of the syncytiotrophoblast; internalisation into coated vesicles and tubulo-vesicular bodies was also observed. Subsequently, IgG-HRP was found in multivesicular bodies and by 30 min appeared in basal vesicles, the frequency of the latter event increasing with time. No routing of IgG-HRP into Golgi regions or lysosomes could be detected. By 60 min IgG-HRP was found in a few caveolae of fetal endothelium of both terminal and intermediate villi. IgG-HRP was not found in intercellular clefts of the endothelium. The pattern of uptake and routing observed suggests a receptor-mediated transcytosis of IgG-HRP across the syncytiotrophoblast and a transcellular pathway through the endothelium.

Cadherin-10 is a novel blood–brain barrier adhesion molecule in human and mouse

Maintenance of the specialised environment of the central nervous system requires barriers provided by the endothelium of brain microvessels (the blood-brain barrier (BBB)) or the epithelium lining the ventricles (CSF-brain barrier) or the choroid plexus (blood-CSF barrier). Inter-endothelial junctions are more extensive in the BBB than in other tissues, with elaborate tight junctions. However, few differences in the molecular composition of these junctions have been described. Here, we show, in both human and mouse brain, that the type II classical cadherin, cadherin-10, is expressed in BBB and retinal endothelia, but not in the leaky microvessels of brain circumventricular organs (CVO), or in those of non-CNS tissues. This expression pattern is distinct from, and reciprocal to, VE-cadherin, which is reduced or absent in tight cortical microvessels, but present in leaky CVO vessels. In CVO, the barrier function is switched from the microvasculature to the adjacent ventricular epithelium, which we also find to express cadherin-10. In the vessels of gliobastoma multiforme tumours, where BBB is lost, cadherin-10 is not detected. This demonstration of a distinctive expression pattern of cadherin-10 suggests that it has a pivotal role in the development and maintenance of brain barriers.

Immunogold localisation of endogenous immunoglobulin-G in ultrathin frozen sections of the human placenta

SummaryEndogenous immunoglobulin-G was localised in ultrathin frozen sections of human term placenta by use of an indirect immuno electron-histochemical methodology. Immunoreactivity of endogenous IgG to rabbit anti-human immunoglobulin-G antibody was visualised by use of protein-A — colloidal gold complex. Gold marked the syncytiotrophoblast in both coated and uncoated regions of the apical plasmalemma, in vesicles and multivesicular bodies, and in vesicles near the basal plasmalemma. Immunoreactivity was also seen in the interstitial space between the trophoblast and the fetal endothelial layer as well as in various types of vesicles within the endothelial cells. No immunoreactivity was seen in the intercellular clefts of the endothelium. The pattern of localisation observed is consistent with receptor-mediated uptake of immunoglobulin-G into the syncytiotrophoblast of the human placenta followed by release into the interstitial space and then vesciular transport through the endothelium.

Structure and permeability of human placental microvasculature

Endothelial paracellular junctions are important structures for the regulation of vascular permeability, junctional organisation being systematically related to the functional properties of the endothelium. Electron microscopic studies, immunocytochemistry, and single‐passage permeability measurements have established that the placental microvessels resemble the fairly tight continuous microvessels of skeletal muscle both in structure and permeability. The endothelial paracellular clefts of these microvessels contain two distinct junctional entities which may influence permeability: the tight junction and the adherens junction. These clefts impose a substantial restriction to molecules above RMM 1000 Da and large haemproteins cannot cross the clefts. The 18 nm‐wide zones of the clefts possess the transmembrane adhesion molecules PECAM‐1 and VE‐cadherin, which have been implicated in junctional assembly and permeability. Inflammatory mediators such as histamine and tumour necrosis factor cause a redistribution of these adhesion molecules to non‐junctional regions, and histamine (100 μM) causes a rapid and sustained rise in extraction of radio‐labeled tracers. Electron microscopy has also revealed possible first indications of tight junctional disassembly. Both the endothelia of larger placental vessels and isolated placental microvascular endothelial cells express cadherins and PECAM‐1 and contain an extensive F‐actin cytoskeleton, which is implicated in changes of cell shape and junctional assembly/disassembly. Thus, the human placental endothelium, using perfusion techniques and in vitro experiments, offers a valuable model for vascular permeability studies. Microsc. Res. Tech. 38:137–144, 1997.

Immunocytochemical and labelled tracer approaches to uptake and intracellular routing of Immunoglobulin-G (IgG) in the human placenta

SummaryIsolated lobules of freshly delivered human term placenta were (a) subjected to an indirect immunoelectron ultracryo method in which the immunoreactivity of endogenous Immunoglobulin-G (IgG) to rabbit anti-human IgG antibody was localized with protein-A-colloidal gold and (b) extracorporeally perfused and human IgG molecules complexed to horseradish peroxidase (HRP) added to the maternal perfusate and the uptake of IgG-HRP over different perfusion durations visualized ultrastructurally by using diaminobenzidine cytochemistry. Immunoreactivity to anti-human IgG antibody was localized all along the apical plasmalemma, in apical coated and uncoated vesicles, in apical and juxtanuclear multivesicular bodies, and in basal vesicles of the syncytiotrophoblast layer of the placenta. The stroma separating the syncytiotrophoblast from the foetal endothelium as well as vesicles within the endothelium were immunoreactive. No immunoreactivity was localized in paracellular clefts of endothelia. A similar distribution of exogenous IgG-HRP was observed for the perfused placentae. When bovine IgG-HRP or HRP alone were used as control tracers no uptake was seen for the former whilst the latter was observed only in early endosomal vesicles of the syncytiotrophoblast. The pattern of localization visualized in both studies is consistent with receptor-mediated uptake of IgG by the syncytiotrophoblast and a vesicular transport of IgG across the foetal endothelium.

Actin cytoskeletal isoforms in human endothelial cellsin vitro: Alteration with cell passage

SummaryThe microfilamentous actin component of the cytoskeleton is crucial to endothelial angiogenesis and vascular permeability. Differences in actin cytoskeletal profiles in cultured human endothelial cells were explored: when first isolated, both primary human umbilical vein endothelial cells (HUVEC) and primary human placental microvascular endothelial cells (HPMEC) expressed F-actin, but notβ-actin orα-smooth muscle actin. A similar endothelial actin profile was observed in cryo-sections of freshly delivered term umbilical cord and placenta. In subsequent cell culture, although the actin cytoskeleton of HUVEC remained unchanged, the actin profiles of HPMEC altered after the second passage with the induction ofα-smooth muscle actin expression, which was intercellularly heterogeneous and increased to 20% at P4. This behavior occurred in HPMEC monolayers cultured on a variety of extracellular matrices. Comparisons with a spontaneously immortalized human microvascular cell-line, HGTEN 21, revealed that inprolonged passage, bothα-smooth muscle actin andβ-actin were expressed, whereas HPMEC at P4 showed a lower level ofβ-actin expression. Therefore, in comparison with large vessels, microvascular cells are more likely to dedifferentiate. This may reflect the ability of microvascular cells to remodel according to changing requirement for new vessel formation. In conclusion, passage of human microvascular endothelial cells, but not of larger vessel endothelial cells, alters the expression of actin isoforms. This may be important in relation to comparisons ofin vitro andin vivo vascular permeability; higher passage microvascular endothelial cells should thus be used with caution in such studies.

Adenosine triphosphate-lead histochemical reactions in ependymal epithelia of murine brains do not represent calcium transport adenosine triphosphatase

The strong enzyme histochemical reactions for adenosine triphosphatase (ATPase) seen in ependymal tanycytes after incubation in calcium-containing media have previously been reported as calcium transport ATPase. Investigation of these reactions showed that: (1) any nucleoside triphosphate can serve as a substrate; (2) diphosphates and monophosphates cannot replace triphosphates; this includes p-nitrophenyl phosphate which is readily hydrolysed by plasma membrane transport ATPases; (3) strong localization occurs in the presence of millimolar concentrations of either calcium or magnesium ions; there is no absolute requirement for calcium ions; (4) they are not inhibited by sulphydryl inhibitors or calmodulin antagonists; (5) lead phosphate precipitates are localized almost entirely on the external face of tanycyte plasma membranes. In addition, the technique gives strong localization to vessels in the choroid plexus but not to the choroidal epithelium. Immunohistochemistry with a primary antibody raised against Ca2+,Mg2+-ATPase stains the choroidal epithelium but not the vessels or the ependymal tanycytes. These results are inconsistent with identification of the reaction as calcium transport ATPase but support characterization as an ecto-ATPase.SummaryThe strong enzyme histochemical reactions for adenosine triphosphatase (ATPase) seen in ependymal tanycytes after incubation in calcium-containing media have previously been reported as calcium transport ATPase. Investigation of these reactions showed that: (1) any nucleoside triphosphate can serve as a substrate; (2) diphosphates and monophosphates cannot replace triphosphates; this includes p-nitrophenyl phosphate which is readily hydrolysed by plasma membrane transport ATPases; (3) strong localization occurs in the presence of millimolar concentrations of either calcium or magnesium ions; there is no absolute requirement for calcium ions; (4) they are not inhibited by sulphydryl inhibitors or calmodulin antagonists; (5) lead phosphate precipitates are localized almost entirely on the external face of tanycyte plasma membranes. In addition, the technique gives strong localization to vessels in the choroid plexus but not to the choroidal epithelium. Immunohistochemistry with a primary antibody raised against Ca2+,Mg2+-ATPase stains the choroidal epithelium but not the vessels or the ependymal tanycytes. These results are inconsistent with identification of the reaction as calcium transport ATPase but support characterization as an ecto-ATPase.

Junctions between pericytes and the endothelium in rat myocardial capillaries: A morphometric and immunogold study

SummaryPericytes are cells of mesodermal origin which are closely associated with the microvasculature. Despite numerous studies little is known about their function. We have studied the relationship between pericytes and the endothelium in rat myocardial capillaries employing ultrastructural and immunogold techniques. 14% of the subendothelial cell membrane is covered by comparatively small pericytic cell processes. About half of these processes are completely embedded in baseement membrane material, whereas the remaining half forms closer contacts with the endothelium. These contacts are devoid of anti-laminin immunogold label, a marker for basement membranes. A small fraction of these contacts has been identified as tight junctions resembling those seen between endothelial cells in capillaries of the same tissue. The remaining majority of junctions reveals a cleft of approximately 18 nm between the apposed membranes in which a succession of cleft-spanning structures can often bedetected. It was also found that pericytic processes are preferentially located close to interendothelial junctions. We suggest that the high frequency of intimate junctions between pericytes and the endothelium and the preferential localisation near paracellular clefts may have functional significance.

Structure and permeability in human placental capillaries

Summary Capillaries in the human and guinea pig placenta closely resemble those of cardiac and skeletal muscle both in their continuous endothelium and in their relatively high restrictiveness to permeation by macromolecules. Their detailed paracellular junctional architecture not only closely resembles other continuous capillaries, but has served as an important model for the investigation of major questions about the subcellular and molecular basis of capillary permeability. In such capillaries the main site for transendothelial water and solute flows and for imposition of molecular size restriction appears to be the paracellular clefts of the endothelium. Within these clefts the discontinuous tight junctions, which show extensive molecular homology with epithelial junctions, may be rate-limiting for hydraulic conductivity but are not size-limiting for permeability of hydrophilic solutes. It is likely that the main sites of size restriction are adhesion belts of the paracellular cleft, where a likely filtering structure is the linker array which is rich in the endothelium-specific adhesion molecule VE-cadherin. An important area for experiments in the next few years will be the identification of the cellular signals involved in the regulation of endothelial permeability which lead to repositioning of the adhesion molecules of the paracellular cleft.