Virchows Archiv B Cell Pathol (1992) 61:419-422

VirchowsArch&B CellPathology

br.t~,~ M o t ~ ecAot.av 9 Springer-Verlag 1992

Short communication

Endothelial cell protrusions in the rat aortic wall Immunocytochemical evidence for an alternative transendothelial passage of plasma proteins I. Londofio and M. Bendayan Department of Anatomy, University of Montreal, Montreal, Quebec, Canada Received June 16, 1990 / Accepted May 15, 1991

Summary. Focal morphological changes in the endothelial lining were observed in the aortic wall of control rats. They consisted of endothelial cytoplasmic projections and vacuolar structures protruding towards the luminal space and containing electron-dense material. Some of these structures were observed to open into the subendothelial space. Endogenous albumin was detected in these c o m p a r t m e n t s by applying protein A-gold immunocytochemistry to thin tissue sections of glutaraldehydefixed, Lowicryl-embedded aortic segments. The labelling was mainly distributed along the plasma m e m b r a n e o f the projections as well as over the dense content of the endothelial protrusions. The presence o f endogenous albumin in these endothelial structures, together with their opening into the subendothelial space, suggests a role for these structures in an alternative transendothelial transport of albumin. Key words: Endothelium - Albumin - I m m u n o c y t o chemistry - Permeability - Aorta

Introduction In morphological studies of the normal aortic tissue, endothelial cell m e m b r a n e blebs and protrusions have been c o m m o n l y observed. Such morphological features have been described as spontaneously occurring events in the aortic endothelia of normal rats (Pease and Paule 1960; Gerrity and Cliff 1972; Schwartz and Benditt 1972). Aortic endothelial pseudo-vacuoles and myoherniae have also been described in control animals ( H o f f and G o t t l o b 1967; Stetz et al. 1979). They seem to derive from smooth muscle cells and concur almost exclusively with fenestrae of the internal elastic lamina. In the present study, we report endothelial vacuolar Offprint requests to: M. Bendayan, Department of Anatomy, Universit6 de Montreal, C.P. 6128, Succ. "A", Montr6al, Qu6bec, H3C 3J7, Canada

protrusions in the rat aortic wall enclosing an electrondense material. Immunocytochemical labelling has demonstrated the presence of endogenous albumin in these structures, which open to the interstitial space. These results suggest a role for these structures in the transendothelial transport of plasma proteins.

Materials and methods Specimen preparation. Normal Sprague-Dawley rats (400 g body weight) were used in this study. They were fed with a balanced standard diet (Prolab Chow; Agway, Syracuse, N.Y., USA) and tap water. The thoracic cavity was opened under ether anaesthesia and the fixative, 100 mM phosphate-buffered 1% glutaraldehyde, rapidly introduced in order to begin the fixation of the aortic tissue in situ. The aorta was then immediately dissected, removed from the animal and cross-sectioned in rings while immersed in the fixative. Fixation was further prolonged for an additional period of 2 h at 4~ C. The rings of aortic tissue were then washed in the same buffer, dehydrated in graded methanol solutions and embedded in Lowicryl K4M at - 2 0 ~ C, as previously described (Bendayan 1984). Thin sections of Lowicryl-embedded tissues were cut, mounted on nickel grids having a carbon-coated Parlodion film, and processed for immunocytochemistry. Immunocytochemistry. Endogenous albumin was localised at the electron-microscope level by applying the protein A-gold immunocytochemical technique (Bendayan 1984). Colloidal gold particles (15 nm in size) were prepared according to Frens (1973), using sodium citrate as the reducing agent. The protein A-gold complex was prepared as described by Bendayan (1984). Tissue sections were subsequently floated at room temperature, on: (a) a drop of a saturated aqueous solution of sodium metaperiodate, 60 min; (b) distilled water, three times, 3 min each; (c) a drop of 150 mM glycine in phosphate-buffered saline (PBS), 30 min; (d) phosphate-buffered saline (PBS) three times, 3 min each; (e) a drop of a 1:25 dilution of an anti-rat albumin antibody (IgG fraction, Cooper Medical, Malver, Pa., USA), 90 min; (f) PBS, three times, 3 min each; and (g) a drop of the protein A-gold complex (ODs2~ =0.5), 30 min. The grids were then washed with PBS and distilled water, dried, and finally stained with uranyl acetate and lead citrate before examination in the electron microscope (Siemens Elmiskop 101 or Phillips EM 410). Cytochemical controls. In order to assess the specificity of the labelling, several controls were performed: (a) adsorption of the anti-

420

body by its specificantigen (rat serum albumin; Sigma, St. Louis, Mo., USA) for 12 h at 4~ C, prior to its use for labelling; (b) omission of the antibody, i.e. incubation of the tissue section with protein A-gold complexalone; (c) substitution of the specificantibody by normal rabbit serum. Results

Several focally distributed endothelial structures, directly protruding towards the lumen, were frequently observed in the rat aortic endothelium (Figs. 1-3). Such structures consisted of thin projections of the endothelial cell cytoplasm enclosing a homogeneous electron-dense material. They formed large cytoplasmic vacuoles (Fig. 1), which, in some instances, communicated directly with the subendothelial extracellular space (Fig. 3). Endothelial flaps (Fig. 2) and blebs were also observed. The endothelial protrusions were frequently associated with the thinnest zones of endothelial cell, approaching the cell borders and, fact, they usually occurred near intercellular junctions (Figs 1, 3a). Due to the fact that the study was performed on fixed tissue, the chronological relationship between the different forms of these endothelial structures could not be established. Concerning the immunocytochemical labelling of endogenous albumin, gold particles, revealing albumin antigenic sites, were abundant over the plasma membrane of the endothelial cell flaps (Fig. 2) and protrusions (Fig. 3). The electron-dense content of the latter was intensely labelled (Fig. 3 a, b). Endothelial plasmalemmal vesicles in the proximity of these vacuoles also showed labelling for albumin (Fig. 3 a). The subendothelial space was moderately labelled though less intensely than the protrusions (Figs. 2, 3). Only a few particles were found in endothelial nuclei and mitochondria as well as over the inner elastic laminae. On the other hand, labelling was very intense in the lumen of the vessels in the adventitia (Fig. 4). Since fixation was performed by immersion, plasma proteins were retained in the capillary lumina as flocculent material; this explains the intense labelling for albumin observed in capillaries. In contrast, no labelling was found in the aortic lumen due to its exposure at the time of fixation which led to the washing out of its content. The intensity of labelling observed in the dense content of the endothelial protrusions was similar to that observed over the capillary lumina, suggesting that the concentration of albumin in these vacuoles is comparable to that found in plasma. The control experiments demonstrated the specificity of the results, since labelling was abolished when the antibody step was omitted in the protocol, when it was substituted by normal rabbit serum (Fig. 5) and when an excess of the corresponding antigen (rat albumin) was added to the antibody solution (not shown). Discussion

Blebs, flaps and other endothelium-derived structures extending towards the lumen have been reported as common features of normal endothelium (Pease and Paule

1960; Hoff and Gottlob 1967; Stetz et al. 1979), or have been attributed to pathological states (Wiener and Giacomelli 1973; Kjeldsen and Thomsen 1975; Smith and Heath 1977; Gordon etal. 1981). Hoff and Gottlob (1967) and Stetz et al. (1979) have proposed that smooth muscle cells could be partially responsible for the formation of pseudo-vacuoles and myoendothelial herniae frequently observed in normal rats. These authors have suggested that such herniae may result from the extension and subsequent detachment of smooth muscle cells through fenestrations in the inner elastic lamina, or after monocyte/lymphocyte diapedesis throughout the endothelial layer. More recently, Majno et al. (1985) have demonstrated an association between these herniated structures and the stomata and stigmata usually observed in aortic endothelial cells after silver nitrate perfusion. The possibility that endothelial protrusions and herniae might constitute an artefact as a consequence of manipulation of the tissue during processing for electron microscopy was rejected by Majno et al. (1985). Morphologically speaking, the endothelial protrusions described in the present study resemble the endothelial structures previously described by Stetz et al. (1979). They correspond to vacuolar structures present in the endothelial cell cytoplasm and containing electron-dense material. Their presence in the endothelial lining, however, does not necessarily correlate either with smooth muscle cell protrusions passing through a fragmented elastic lamina, such as proposed by Hoff and Gottlob (1967), Stetz et al. (1979) and Majno et al. (1985), or with the presence of "ghost bodies" accompanying the herniae at these sites, as described by Stetz et al. (1979). They thus do not correspond to the previously reported herniated structures, but rather to endothelial vacuoles. In addition, we have observed that some of these vacuoles open into the subendothelial space of the aortic intima and are associated with several plasmalemmal vesicles. Endogenous albumin, as revealed by immunocytochemistry, was found in the content of these vacuolar endothelial protrusions and over the endothelial surface surrounding them. The presence of albumin in these structures indicates that they accumulate plasma proteins. The endothelial plasmalemmal vesicles associated with these structures were also labelled for albumin. In addition, the fact that these vacuoles open and seem to release their albumin into the subendothelial space suggests a role for these structures in the transendothelial transfer of plasma proteins. Our results go along with the hypothesis proposed by Fawcett (1963), who suggested that droplets of plasma proteins can be nonselectively incorporated in endothelial cells by pinocytosis. On the other hand, it is well established that endothelial cells do transfer circulating proteins through plasmalemmal vesicles by either fluid-phase or adsorptive, receptor-mediated micropinocytosis. Specific receptors for the endothelial uptake of circulating molecules such as low-density lipoproteins and transferrin have been involved in the transcytotic passage of these proteins (Vasile et al. 1983; Tavassoli et al. 1986). In the specific case of albumin, Ghitescu et al. (1986) and Simionescu (1988)

421

Fig. 1. Aortic intima of control rat. An endothelial vacuole (*) is observed. It contains an electron-dense material and protrudes into the aortic lumen (AL). IC, Intimal cell. x 21000; bar= 1 lam

elastic lamina (IEL) is unlabelled. AL, Aortic lumen, x26000 (Fig. 2), x 19500 (Fig. 3a), • 17000 (Fig. 3b); bars=0.5 ~tm (Fig. 2) and 1 ~tm (Fig. 3 a, b)

Figs. 2, 3. Aortic intima: immunocytochemical labelling for albumin. Gold particles, revealing albumin antigenic determinants, are observed on the surface of the flap (Fig. 2) and of the endothelial protrusions (arrowheads) (Fig. 3 a, b). The electron-dense content of these structures (*) is intensely labelled. Figure 3 a and b shows openings of the vacuoles into the subendothelial space (SES), which is less intensely labelled. Plasmalemmal vesicles (arrows) neighbouring the endothelial vacuoles are also labelled. The inner

Fig. 4. Aortic adventitia: immunocytochemical labelling for albumin. Labelling is very intense in the capillary lumen (CL) and in the subendothelial space around it. x 15600; bar= 1 lam Fig. 5. Control of the immunoytochemical labelling. Labelling is completely abolished when the specific antibody is substituted by normal rabbit serum. CL, Capillary lumen, x 16600; bar= 1 I~m

422 have suggested the existence of a receptor-mediated, transendothelial transport in various continuous capillaries. They have shown, however, a slow, non-receptormediated transport of albumin in large-vessel endothelia (Simionescu and Simionescu 1987). The presence of albumin in endothelial vacuoles and its transfer to the subendothelial space, described in the present study, suggest the existence of an alternative pathway for the nonselective passage of plasma proteins across the endothelium. This supports the results o f previous studies demonstrating significant concentrations o f albumin in the intimal space o f normal rat aortas (Bratzler et al. 1977; Londofio and Bendayan 1989). Some authors (Packham et at. 1967; Bell et al. 1974) have demonstrated the presence of focal areas of enhanced endothelial permeability to 131I-albumin and Evans blue dye in the aortic arch of normal pigs. These areas exhibit endothelial and intimal alterations relative to areas with no dye accumulation (Packham et al. 1967; Gerrity et al. 1977). It is likely that cytoplasmic processes and endothelial vacuoles described in the present study correspond to those described in blue areas by K a t o r a and Hollis (1976) and Gerrity et al. (1977). These permeable areas have been considered to represent zones of the aortic wall susceptible to the development of pathological lesions (Fry 1973). Indeed, the transport of circulating proteins could to a certain extent lead to endothelial and subendothelial pathophysiological alterations of the aortic wall. Consistently, the frequency of endothelial processes such as flaps, blebs, and vacuoles and herniae has been found to be markedly increased in pathological situations such as hypoxia (Kjeldsen and T h o m p s e n 1975), hypertension (Wiener and Giacomelli 1973; Smith and Heath 1977), hypercholesterolaemia (Still and O'Neal 1962) or after a variety o f different stimuli in experimental animals ( H o f f and G o t t l o b 1967; G o r d o n et al. 1981). The current morphological and cytochemical observations indicate the presence of an alternative pathway for the passage of plasma proteins across the aortic endothelium. The existence of vacuolar protrusions in focal areas of the endothelium and their implication in the enhanced permeability of these areas seem plausible but remain to be elucidated.

Acknowledgements. The technical assistance of C~cile Venne, Diane Gingras and Jean Leveillre is gratefully acknowledged. This study was supported by grants from the Medical Research Council of Canada. M.B. is a Scientist of the MRC.

References Bell FP, Adamson IL, Schwartz CJ (1974) Aortic endothelial permeability to albumin: focal and regional patterns of uptake and transmural distribution of ~3q-albumin in the young pig. Exp Mol Pathol 20:57-68 Bendayan M (1984) Protein A-gold electron microscopic immunocytochemistry: methods, applications, and limitations. J Electron Microsc Tech 1 : 243-270 Bratzler RL, Chisolm GM, Colton CK, Smith KA, Zilversmith DB, Lees RS (1977) The distribution of labeled albumin across the rabbit thoracic aorta in vivo. Circ Res 40:182-190 Fawcett DW (1963) Comparative observations on the fine structure of blood capillaries. In: Robinson JL, Smith D (eds) The peripheral blood vessels.Williams and Wilkins, Baltimore, pp 17 24

Frens G (1973) Controlled nucleation for regulation of the particle size in monodisperse gold suspensions. Nature Phys Sci 242: 2022 Fry DL (1973) Responses of the arterial wall to certain physical factors. In: Porter R, Knight J (eds) Atherogenesis: initiating factors. Ciba Foundation Symposium 12. Elsevier, Amsterdam, p 93 Gerrity RG, Cliff WJ (1972) The aortic tunica intima in young and aging rats. Exp Mol Pathol 16:382M02 Gerrity RG, Richardson M, Bell FP, Somer JB, Scwartz CJ (1977) Endothelial cell morphology in areas of in vivo Evans blue uptake in the young pig aorta. II. Ultrastructure of the intima in areas of differing permeability to proteins. Am J Pathol 89: 313-334 Ghitescu L, Fixman A, Simionescu M, Simionescu N (1986) Specific binding sites for albumin restricted to plasmalemmal vesicles of continuous capillary endothelium: receptor-mediated transcytosis. J Cell Biol 102:1304--1311 Gordon D, Guyton JR, Karnowsky MJ (1981) Intimal alterations in rat aorta induced by stressful stimuli. Lab Invest 45 : 14-27 Hoff HF, Gottlob R (1967) A fine structure study of injury to the endothelial cells of the rabbit abdominal aorta by various stimuli. Angiology 18: 440-451 Katora ME, Hollis TM (1976) Regional variation in rat aortic endothelial surface morphology: relationship to regional aortic permeability. Exp Mol Pathol 24:23-34 Kjeldsen K, Thomsen HK (1975) The effect of hypoxia on the fine structure of the aortic intima in rabbits. Lab Invest 33: 533543 Londofio I, Bendayan M (1989) Distribution of endogenous albumin across the rat aortic wall as revealed by quantitative immunocytochemistry. Am J Anat 186: 407-416 Majno G, Underwood JM, Zand T, Joris I (1985) The significance of endothelial stomata and stigmate in the rat aorta. Virchows Arch [A] 408:75-91 Packman MA, Rowsell HC, Jorgensen L, Mustard JF (1967) Localized protein accumulation in the wall of the aorta. Exp Mol Pathol 7: 214-232 Pease DC, Paule WJ (1960) Electron microscopy of elastic arteries; the thoracic aorta of the rat. J Ultrastruct Res 3:469 483 Schwartz SM, Benditt EP (1972) Studies on aortic intima. I. Structure and permeability of rat thoracic aortic intima. Am J Pathol 66: 241-264 Simionescu M (1988) Receptor-mediated transcytosis of plasma molecules by vascular endothelium. In : Simionescu N, Simionescu M (eds) Endothelial cell biology in health and disease. Plenum Press, New York, pp 69-104 Simionescu N, Simionescu M (1987) Receptor-mediated transcytosis of albumin: identification of albumin binding proteins in the plasma membrane of capillary endothelium. In: Tsuchiya M, Asano M, Mishima Y, Oda M (eds) Microcirculation: an uptade, vol 1. Proceedings of the Fourth World Congress for Microcirculation. Excerpta Medica, Amsterdam, pp 67-82 Smith P, Heath D (1977) Ultrastructure of hypoxic hypertensive pulmonary vascular disease. J Pathol 121:93-100 Stetz EM, Majno G, Joris I (1979) Cellular pathology of the rat aorta. Pseudo-vacuoles and myo-herniae. Virchows Arch [A] 383:135-148 Still WJS, O'Neal RM (1962) Electron microscopic study of experimental atherosclerosis in the rat. Am J Pathol 40:21-35 Tavassoli M, Kishimoto T, Soda R, Kataoka M, Harjes K (1986) Liver endothelium mediates the uptake of iron-transferrin complex by hepatocytes. Exp Cell Res 165:369-379 Vasile E, Simionescu M, Simionescu N (1983) Visualization of the binding, endocytosis and transcytosis of low-density lipoprotein in the arterial endothelium in situ. J Cell Biol 96:1677 1689 Wiener J, Giacomelli F (1973) The cellular pathology of experimental hypertension. VII. Structure and permeability of the mesenteric vasculature in angiotensin-induced hypertension. Am J Pathol 72: 221-240