Okajimas Folia Anat. Jpn., 57(2-3) : 145-158, August 1980 Junction between Kupffer Cells and Hepatic Sinusoidal Endothelium. A Review By TOSHIO ITO, YUTAKA TANUMA and SUSUMU SHIBASAKI Department of Anatomy, Teikyo University School of Medicine, Tokyo 173 and Department of Anatomy, Gunma University School of Medicine, Maebashi 371, Japan -Received for Publication, January 24, 1980- Key Words : Kupffer-endothelial junction, Human and bat liver, TEM and SEM. Summary. In this review, the junctions between Kupffer and sinusoidal endothelial cells previously revealed by the present authors under the transmission electron microscope in both normal human and bat livers has been discussed and compared with stereo- images of the connections between the two cell types reported by other authors in scanning electron microscopic studies. It is concluded that Kupffer cells lie on the gap or opening of the endothelial lining of the sinusoid, that part of the perisinusoidal surface of the cell body which is applied to the opening directly faces the perisinusoidal space, and that the margin of this part of the perisinusoidal surface of the Kupffer cell and the margin of the endothelial opening are circularly bound by means of a peculiar close junctional area identical in structure and appearance with the so-called "junctional complex" (Wisse) between endothelial cells . The two junctional areas between the Kupffer cell and the endothelial sheet which are found in a pair in trans- mission electron microscopic preparations on two diametrical parts of the perisinusoidal surface of the Kupffer cell, should be derived from the perpendicular section of such a circular margin of the endothelial opening attached to the Kupffer cell body. Thus, the integrity of the lining of the sinusoid is maintained and Kupffer cells are immo- vably fixed to the sinusoidal lining resisting the drag induced by the blood flow in the sinusoid. This short review was prepared on the Recently, using the scanning electron basis of transmisson electron microscopic microscope, many investigators have findings for the junction between Kupffer observed the sinusoidal surface of the and sinusoidal endothelial cells in normal hepatic sinusoidal wall in an attempt to human livers (Ito and Shibasaki, 1968 ; clarify the stereostructures of Kupffer Ito, 1973) and in normal as well as excess and endothelial cells. As a result, several vitamin A-administered bat livers (Tanu- topographical relationships between these ma and Ito, 1978), and gives a critical morphologically distinct cell types have appraisal of scanning electron microscopic been elucidated to some extent (Itoshima stereoimages of Kupffer cells and their et al., 1974 ; Motta and Porter, 1974 ; connections with sinusoidal endothelial Muto, 1975; Motta, 1977; Vonnahme, cells. 1977 ; etc.). Some of the scanning electron 145 146 T. Ito, Y. Tanuma and S. Shibasaki microscopic findings reported by the must be bound to the sinusoidal endothelial above authors, especially concerning the lining by means of a definite junctional relations between Kupffer and endothelial structure. This was first demonstrated b cells, seemed to require further precise Ito and Shibasaki (1968) with the trans-- reexamination with reference to those mission electron microscope in normal obtained by transmission electron micro- human livers and then confirmed by scopy. The scanning electron microscopic Tanuma and Ito (1978) in bat livers. images often indicated that filopodia and According to Ito and Shibasaki, a finger- lamellipodia projecting from the surface shaped end part of the cytoplasmic sheet of the Kupffer cell body might spread of the endothelial cell came in contact towards the endothelial wall to be with the marginal portion of the perisi- attached to or overlap the latter, or that nusoidal surface of the Kupffer cell. At in some cases they might communicate the contact site, a minute slit about 250 A with adjacent Kupffer cells across the in width was recognized between juxta- sinusoidal lumen (Muto, 1975 ; Vonnahme, posed plasma membranes of the two cell 1977; Motta, 1977). The above findings, types (Figs. 3, 4). In these junctional however, were only rarely confirmed in areas, a density increase in the plasma transmission electron microscopic images membranes and adjacent cytoplasm was (Fig. 1). It was also the case in trans- observed, and in the slit a slightly electron mission electron microscopy that filopodia dense material was demonstrated. How- penetrated through endothelial gaps (Fig. ever, no fusion line of the outer leaflets 2). However, filopodial communication of the plasma membranes as seen in the between adjacent Kupffer cells has never tight junction was noted, so that the been demonstrated by transmission electron Kupffer-endothelial junctions were identi- microscopy. fied with neither the tight junction nor The Kupffer cells are fixed macrophages desmosome. They represented a peculiar of the liver ; they are anchored on the junctional pattern different from a simple hepatic sinusoidal wall bulging into the membrane apposition. As shown in sinusoidal lumen and are provided with an Figure 4, two junctional areas were often irregular and changeable rough sinusoidal found in a pair on two diametrical surface which projects temporal ruffles, marginal regions of the perisinusoidal microvillous processes, filo- and lamelli- surface of the Kupffer cell. podia as seen .under the scanning electron The above transmission electron micro- microscope. Recently, Carr (1977) humor- scopic findings for junctional areas ously conjectured on the immobility of between the Kupffer cell body and Kupffer cells as follows : "they don't endothelial sheet were confirmed later by move and if they don't move, the drag Tanuma and Ito in their bat liver study. of the sinusoidal blood on their irregular In the bat livers examined, junctional surfaces must be considerable. Their areas between Kupffer and endothelial contacts with adjacent cells do not involve cells were frequently encountered, often the formation of elaborate desmosome. in a pair on two diametrical parts of the So we don't really understand how they perisinusoidal surface of the Kupffer cell resist the temptation to move, sitting like body with a more of less wide gap in the spinsters continually saying no". To be endothelial sheet in between (Figs. 5, 6), immovably anchored to the hepatic although it remained unknown whether sinusoidal wall resisting the blood flow such a gap was intracellular or intercel- in the sinusoid, the Kupffer cell body lular. ,High power views of such junctional Junction between Kupper Cells and Hepatic Sinusoidal Endothelium 147 areas between the Kupffer cell body as proposed by Wisse (1974). and "cytoplasmic process" of the endo- In his transmission electron microscopic thelial sheet are shown in Figures 7 and 8. studies on the rat hepatic sinusoidal wall, Density increases in the juxtaposed plasma Wisse (1970, 1972, 1974) noted that Kupffer membranes and adjacent cytoplasm as well cell bodies were incorporated into the as slightly electron opaque material filling sinusoidal wall, which remained intact as up the minute slits of aboUt 200 A in width a whole, because of the contact between between apposed plasma membranes, were endothelial cell processes and the surface apparent. As Tanuma and Ito (1978) of the Kupffer cells and because the pointed out, the junctions between the endothelial lining might be replaced by Kupffer cell body and cytoplasmic process insertions of Kupffer cell cytoplasm of of the sinusoidal endothelium were identi- varying dimensions. cal in structure and appearance with the In his scanning microscopic observations so-called "junctional complex" (Wisse, on Kupffer cells in monkey liver, 1970, 1972) between the margins of the Vonnahme (1977) noted that they gene- "cytoplasmic processes" of neighboring rally lay on the endothelial lining, over- endothelial cells. It is of great interest lapping it, whereas in rat sinusoids their that two cell types which composed the cell body filled the gaps between ad jacent hepatic sinusoidal wall were bound by endothelial cells. The latter description means of an identical peculiar close agrees with the scanning electron micro- junction different in fine structure from scopic studies of Muto (1975) on the rat the tight junction and desmosome. Wisse liver sinusoid in which he reported that (1974) considered the contacts between the Kupffer cell generally lay on a large Kupffer and endothelial cells as simple endothelial opening and was anchored to appositions of cell membranes with some its margin by filopodia to be incorporated electron dense material present in which, in the formation of the sinusoidal wall. according to his own description, he could As mentioned above, the transmission not detect the exact fine structure, pro- electron microscopic studies on the normal bably because he failed to demonstrate human and bat sinusoidal wall often the trilaminar aspect of plasma membranes. revealed two junctional areas between Kurtz (1964) had already described the the Kupffer cell body and endothelial presence, between the endothelial cyto- processes in a pair on two diametrical plasmic process and Kupffer cell body, of margins of the perisinusoidal surface of a similar membrane apposition across a the Kupffer cell. This suggests that the 300 A slit containing electron dense Kupffer cell body might have lain on a material. large endothelial gap or opening as shown Concerning the relationship