Basement membrane thickening of postcapillary venules and capillaries in rheumatoid synovium. Immunoelectron microscopic and electron microscopic morphometric analysis

Pathologic changes in the basement membrane (BM) of postcapillary venules (PCV) and capillaries in rheumatoid arthritis (RA) synovium were studied by immunoelectron microscopy, using a monoclonal antibody against human type IV collagen, C(IV)22, and by electron microscopic morphometric analysis. The sublining region of RA synovium was classified into lymphocyte-rich areas, transitional areas, and interstitial areas, according to their pattern of cellular infiltration. In lymphocyte-rich areas, the BM of the PCV and capillaries were minimally thickened; disruption of the lamina densa was seldom seen. Transitional areas, which contained macrophages, lymphocytes, and plasma cells, had numerous PCV and capillaries. The BM was markedly thickened and partially multilamellated, and there were many disruptions in the lamina densa. The BM contained degenerated endothelial cells and cell debris. On immunostaining of this BM with monoclonal antibody C(IV)22, type IV collagen stained heavily, mainly in the disrupted lamina densa; this indicates that the thickening was, at least in part, the result of an increase in BM collagen. In uninfiltrated interstitial areas, BM were moderately thickened and multilamellated, and showed few disruptions of the lamina densa; there were similar increases in type IV collagen, but cell debris was seldom observed. Measurement of the BM width, the ratio of BM width to vessel diameter, and the fraction of vascular cross-sectional area occupied by BM demonstrated that the thickness of the BM of both PCV and capillaries was greatest in transitional areas and was smallest in lymphocyte-rich areas (P less than 0.01). Since macrophages and macrophage-derived factors have been found to promote synthesis of BM collagen type IV, and since transitional and interstitial areas are rich in macrophages and histiocytes, respectively, it is suggested that these mononuclear cells and the factors secreted by them play a significant role in the thickening of the BM of PCV and capillaries in RA synovitis.     

The thickening of basement membrane in synovial capillaries in rheumatoid arthritis

Summary Synovial tissues from seven rheumatoid arthritis (RA) patients were used for the ultrastructural investigation of capillary cellular components and basement membranes (BM). Attention has been specially paid to the mechanism of BM thickening of the capillaries in the inflammatory sites.The capillary BM were multilamellated in the inflammatory sites. The multilamellation was characteristic not only in the BM surrounding the endothelial cells and pericytes but also in the BM between these two types of cells. Cell debris was frequently encountered between the multilamellated BM. The hyperplasia and various stages of degeneration of the endothelial cells were observed in these regions. Some endothelial cells were activated and occasionally located in capillaries containing degenerated endothelial cells. The high incidence of these findings indicates the following hypotheses.The accelerated rate of death and replenishment of capillary cellular components may play a role in BM thickening in the inflammatory sites of RA synovium. These cells may not only produce one layer of BM in their life-time but may also be activated to produce excessive amounts of BM components to make several layers.     

Neutrophil-derived oxidants promote leukocyte adherence in postcapillary venules.

The objective of this study was to determine whether hydrogen peroxide (H2O2), hypochlorous acid (HOCl), and monochloramine (NH2Cl), at concentrations produced by activated neutrophils, promote leukocyte adherence to microvascular endothelium in post-capillary venules. Cat mesenteric venules (30-45 microns diameter) were examined using intravital video microscopy. Red blood cell velocity (VRBC), venular diameter (DV), and the number of adherent leukocytes (NWBC) were measured in postcapillary venules. Venular blood flow and wall shear rate (tau) were calculated from the measured values of VRBC and DV. Different concentrations (0.01-1.0 mM) of H2O2, HOCl, or NH2Cl were superfused on the mesentery. In some experiments, the contributions of the leukocyte adhesive glycoprotein CD11/CD18 and platelet-activating factor (PAF) in the oxidant-induced leukocyte adherence were determined using a CD18-specific antibody (IB4) and a PAF-receptor antagonist (WEB 2086), respectively. The results of our in vivo experiments indicate that H2O2 and NH2Cl, but not HOCl, promote leukocyte adhesion to venular endothelium. Incubation of isolated cat neutrophils with either NH2Cl or H2O2 resulted in activation of CD11/CD18, as assessed by flow cytometry. Although the leukocyte adhesion induced by both H2O2 and NH2Cl was associated with a reduction in venular wall shear rate, corresponding decrements in shear rate induced by partial occlusion of the mesenteric artery did not lead to similar levels of leukocyte adherence. The leukocyte adherence induced by H2O2 and NH2Cl was largely prevented by monoclonal antibody IB4, indicating that both oxidants promote leukocyte adherence via activation of CD11/CD18. The H2O2-induced, CD18-mediated leukocyte adherence appears to be elicited by PAF and by a direct effect of the oxidant on CD11/CD18 expression. The mechanism underlying the NH2Cl-induced leukocyte adherence remain unclear.     

Hypercholesterolemia promotes leukocyte-dependent platelet adhesion in murine postcapillary venules

OBJECTIVE: Leukocyte-platelet aggregates form in blood during the development of cardiovascular diseases, including atherosclerosis. The study determined whether leukocytes contribute to the platelet adhesion induced by hypercholesterolemia in postcapillary venules. METHODS: Wild-type (WT) C57Bl/6 or CD18-deficient (CD18-/-) mice were placed on a normal (ND) or high-cholesterol (HC) diet for 2 weeks. Platelets isolated from ND, HC, or CD18-/- mice were fluorescently labeled with CFSE, and administered to either ND, HC, HC-CD18-/- or anti-neutrophil serum (HC-ANS)-treated mice. Rhodamine 6G was administered to label and visualize leukocytes. Intravital fluorescence microscopy was used to quantify leukocyte and platelet adhesion in intestinal postcapillary venules. RESULTS: HC increased both leukocyte and platelet adhesion (relative to ND). Greater than 50% of adherent platelets in HC mice were bound to adherent leukocytes. When HC platelets were examined in HC-ANS-treated or HC-CD18-/- mice, leukocyte-dependent platelet adhesion was significantly attenuated. Conversely, when HC-CD18-/- platelets were observed in HC recipients both leukocyte-dependent and endothelium-dependent platelet adhesion was comparable to HC mice receiving WT platelets. CONCLUSIONS: The findings demonstrate that the pro-thrombogenic phenotype assumed in the microvasculature during hypercholesterolemia is largely attributed to leukocyte-dependent platelet adhesion.     

Effect of fibrinogen on leukocyte margination and adhesion in postcapillary venules

OBJECTIVE: To quantitatively evaluate the role of fibrinogen (Fb) as a determinant of leukocyte (WBC) margination in postcapillary venules in light of its ability to induce red blood cell (RBC) aggregation with reductions in shear rate (gamma) and increase adhesiveness of WBCs to endothelium (EC). METHODS: Red cell aggregation (RCA), WBC margination (flux at the EC), rolling velocity, and adhesion to the EC were measured in rat mesenteric postcapillary venules upon reducing gamma, prior to and following systemic infusion of Fb. Proximal occlusion of feeding microvessels with a blunted probe facilitated reductions in gamma from 600 to 50 s(-1). An index of aggregation (G) was derived from light-scattering properties of RBCs, where G was proportional to the number of RBCs per aggregate. WBC margination was measured as the percentage of total luminal WBC flux that rolled on the EC, F*(WBC). RESULTS: For normal levels of Fb (0.07 g%), reductions in gamma resulted in a 4-fold rise in F*(WBC) and no change in G as gamma was reduced to 50 s(-1). Infusion of Fb to achieve a plasma concentration to 0.7 g% caused a modest 20% increase in G and a 2.5-fold increase in F*(WBC) at gamma = 50 s(-1). WBC-EC adhesion appeared to increase significantly, but much less than with infusion of high molecular weight dextran (Dx). With Dx, G increased 3-fold, with reductions in gamma, but F*(WBC) increased only half the amount incurred with Fb at low shear. The greater margination in the presence of Fb results from RBC rouleaux that promote radial migration of WBCs. In contrast, clumps of RBCs resulting from high molecular weight Dx entrain WBCs within plasma gaps along the vessel centerline. CONCLUSIONS: In the presence of Fb, margination of WBCs increases dramatically at low shear due to rouleaux formation, which enhances radial migration of WBCs. This effect is much greater than with Dx because disruption of the much weaker Fb induced rouleaux precludes reductions in H(MICRO), whereas clumping aggregates induced by Dx form plasma gaps. Thus, modest levels of RCA caused by increased Fb may greatly enhance margination and with an enhancement of adhesiveness synergistically promote firm WBC-EC adhesion in the low flow state.     

Dendritic cells and high endothelial venules in the rheumatoid synovial membrane

Since dendritic cells are believed to play a crucial role in the pathogenesis of rheumatoid arthritis (RA) we studied the microenvironmental relationship of these cells with endothelial cells, lymphocytes and macrophages in the rheumatoid synovial membrane. With the monoclonal antibodies OKIa (MHC Class II determinants), RFD1 and L25 (both specific for "active" human dendritic cells) we identified large numbers of dendritic cells. With the monoclonal antibody HECA 452 [specific for a putative adhesion molecule notably present on high endothelial venules (HEV)], a subset of dendritic cells could be detected. HECA-452 positive dendritic cells were found in 2 basic patterns: (1) associated with small lymphoid cell clusters in the neighborhood of vessels with flat, HECA-452 negative endothelium, (2) at the periphery of dense organoid lymphoid infiltrates, surrounding HECA-452 positive HEV-like vessels. Our data suggest that the influx of HECA-452, L25, RFD1 and MHC Class II positive dendritic cells is an early event in the development of the inflammatory infiltrate found in the rheumatoid synovial membrane. The formation of organoid lymphoplasmacellular infiltrates with high endothelial venules would be secondary to this event.     

Desmopressin induces endothelial P-selectin expression and leukocyte rolling in postcapillary venules

Desmopressin, (DDAVP; 1-desamino-8-D-arginine vasopressin) increases the release and activity of von Willebrand factor (vWF); however, its effects on the other major constituent of endothelial Weibel-Palade bodies, P-selectin, has not been investigated. DDAVP-induced P-selectin expression may explain DDAVP's efficacy in bleeding disorders in which vWF levels are normal. Therefore, the objective of this study is to assess the effect of DDAVP on P-selectin expression on endothelial cells of postcapillary venules in vivo and on human umbilical vein endothelium in vitro, and to determine whether DDAVP has direct effects on leukocyte behavior in postcapillary venules. DDAVP (0.1 and 1.0 microgram/mL) induced a significant but transient increase in P- selectin expression on human umbilical vein endothelial cells as well as on rat and human platelets. Immunohistochemical analysis of rat postcapillary venules showed that in contrast to saline, DDAVP injection (1 microgram/kg, intravenous) induced significant endothelial P-selectin expression. DDAVP administration also induced a rapid and significant increase in leukocyte rolling in rat mesenteric venules in vivo. This response was entirely dependent on P-selectin, as an anti-P- selectin antibody rapidly reversed the DDAVP-induced increase in leukocyte rolling. DDAVP induced leukocyte rolling in medium (20 to 40 microns) and large (> 40 microns), but not small (< 20 microns), postcapillary venules. In animals that were treated with DDAVP, there was a steady and significant increase in leukocyte adhesion. This study shows that DDAVP can directly induce P-selectin expression on endothelium in vitro and in vivo and that the latter response is capable of supporting prolonged leukocyte rolling in rat postcapillary venules.     

Basement membrane of central nervous system capillaries lacks ruthenium red-staining sites

We used the cationic dye ruthenium red to examine the distribution of anionic macro-molecules (presumably proteoglycans) in the basement membranes of the fenestrated capillaries and epithelium of the choroid plexus, and of the continuous capillaries forming the blood-brain barrier. Both the endothelial and epithelial basement membranes of choroid plexus displayed discrete, 10- to 20-nm-diameter, electron-dense sites after exposure to ruthenium red. These sites were similar in size, appearance, and distribution to those found in the vascular and epithelial basement membranes of a variety of tissues outside the central nervous system. In contrast, the basement membranes of continuous, blood-brain barrier capillaries did not display electron-dense sites following exposure to ruthenium red, even after measures had been taken to enhance penetration of the dye across the endothelial cells. The lack of discrete ruthenium red staining in the basement membrane of continuous blood-brain barrier capillaries could be due to a relative paucity of anionic macromolecules, or may be the result of the compact architecture of this particular basement membrane. Regardless of the final explanation, these findings suggest that the basement membrane of the blood-brain barrier, like its endothelium, is structurally (and perhaps functionally) unique.     

Leukocyte adherence initiation in skeletal muscle capillaries and venules

In vivo leukocyte adherence has many physiological implications but the effect of vessel size and hemodynamic parameters on adherence initiation remains unclarified. The early phases of adherence, defined to include attachment of previously freely moving leukocytes to endothelium or leukocytes which roll along the vessel wall, are the focus of this report. Blood velocity and diameter in 121 capillaries and venules in the cremaster of 13 rats were measured. The hemodynamic and vessel size difference characterizing vessels in which either initial leukocyte wall attachment or leukocyte rolling occurred (termed "adherent vessels"), as distinguished from vessels in which neither transient stoppage nor cell rolling was observed (termed "nonadherent vessels"), was then determined. Fifty-seven percent of the vessels observed were adherent vessels. These vessels were characterized by a larger diameter, a smaller blood velocity, and a smaller calculated wall shear rate than the nonadherent vessels. The results show that leukocyte adherence can be initiated in capillaries and postcapillary vessels with diameters that are equal to and less than the leukocyte size. Within vessels less than 11 micron, adherence or nonadherence appears dependent on local hemodynamics with the possibility of a critical shear rate threshold of about 400 sec-1. In vessels with shear rates greater than this value only 7% were observed to have adherence. In vessels greater than 11 micron the absence of hemodynamic differences between adherent and nonadherent vessels suggests the presence of other adherence-initiating mechanisms. It is thought that in these larger size vessels both local erythrocyte effects and adherence initiation within upstream capillaries affect observed adherence.     

Immunohistology of rheumatoid nodules and rheumatoid synovium.

The immunohistological features of rheumatoid nodules and rheumatoid synovium were examined using monoclonal and polyclonal antibodies raised against macrophages, HLA-DR, leucocyte common antigen, and immunoglobulin components. The palisading cells surrounding the necrotic centre of the rheumatoid nodule were shown to be HLA-DR positive leucocytes, mostly histiocytes. The inflammatory infiltrate associated with rheumatoid nodules showed many immunohistochemical similarities to that of rheumatoid synovium, including a preponderance of IgG positive plasma cells, and a similar number and microanatomical pattern of distribution of HLA-DR positive cells. The significance of these findings for the cellular immunopathology and aetiology of the rheumatoid lesion is discussed.     

Basement membrane proteins in synovial membrane: Distribution in rheumatoid arthritis and synthesis by fibroblast-like cells

Summary Rheumatoid arthritis is a complex disease of unknown origin. In consequence of some immunological reactions, proliferative invading synovial tissue leads to destruction of normal joint architecture. The aim of this study was to investigate qualitative changes in extracellular matrix distribution of proliferating rheumatoid synovium and their cellular origin. Synovial tissues from 57 clinically indicated arthrotomies were investigated with immunofluorescence, using specific antibodies against extracellular matrix proteins in tissue slides and cultured cells, which were also studied for collagen biosynthesis. Results indicated that synovial fibroblast-like cells synthesize and secrete basement membrane proteins laminin and collagen type IV as e.g. endothelial cells or organogenic fibroblasts. Laminin and collagen type IV were specifically demonstrated pericellularly in the hyperplastic lining layer of active rheumatoid synovitis. These findings are discussed with respect to the possible implication of altered cell-matrix interactions in rheumatoid synovial proliferation.     

The distensibility of single capillaries and venules in the cat mesentery

The distensibility of single capillaries and venules of the cat mesentery has been determined using a simple technique in vivo. Selected vessels were occluded under microscopic control and red cell oscillation in the occluded segment was monitored using cinemicrography. The cells were seen to oscillate in synchrony with local vascular pressure which was recorded simultaneously using the servo-null technique. Less than 10% of the oscillation could be accounted for by filtration/reabsorption so the bulk of the movement was thought to be a consequence of distension of the vessel during systole and relaxation during diastole. Changes in radius would be amplified by the factor 2¢l/r (where r is radius and ¢l the length of the occluded vessel) to give the observed movement of red cells. Mean pulse pressure was 3.2 mm Hg in capillaries and 2.3 mm Hg in venules which gave rise to increases in radius of 0.03 and 0.028 μm, respectively. The nature of the structure supporting the vessels was considered and it seemed likely to be the basement membrane which would require a Young's modulus of 1.8 × 10⁶ N m⁻² for the capillary and 5.4 × 10⁶ N m⁻² for the venule although pericytes might contribute to support of the latter. It is considered likely that some of the higher values quoted in the literature are due to measurements being made at different ends of a nonlinear stress-strain curve, the present data being obtained at the low stresses normally found in mammalian microvessels.     

Topology of pulmonary arterioles,capillaries, and venules in the cat

It is shown in this paper that each alveolus is not a unit of microcirculation: It is not supplied by one arteriole and one venule; each arteriole supplies several alveoli, which are connected to other alveoli before draining into venules. The spatial relationship among the arterioles, venules, and capillaries in the lung is not clear in the literature. To clarify the topology, we developed a special method of particulate polymer casting of the vascular tree, and four criteria to decide whether a noncapillary vessel is arterial or venous. Then on each histological cross section of the lung parenchyma, we marked all arteries with white dots, and all veins with red dots, and covered the area of white dots with one color, and the area of red dots with another. A two-colored map emerged. A two-colored cartographic map can only show islands in an ocean. Our histological maps of the cat lung show that the arterial zones are islands, and the venous zone is the ocean. From these maps quantitative information about the arterial and venous zones in the lung is obtained by stereological methods, and the data for the cat are presented. A morphological definition of the average length of capillary blood vessels connecting arterioles and venules is proposed. The average morphological length of capillary blood vessels in cat lung was found to be 556 ± 286 (SD) μm when the transpulmonary pressure was 10 cm H₂O. The average length of capillaries (L) is an important determinant of pulmonary blood flow: for given arterial and venous pressures, the flow rate is inversely proportional to L², the square of the average length; the transit time (time available for diffusion and oxygenation) varies directly with L²; the regional difference of blood flow also depends on L.