Intercellular contacts of lymphocytes during migration across high-endothelial venules of lymph nodes. An electron microscopic studyкод для вставкиСкачать
THE ANATOMICAL RECORD 207:643-652 (1983) Intercellular Contacts of Lymphocytes During Migration Across High-Endothelial Venules of Lymph Nodes. An Electron Microscopic Study FERRELL R. CAMPBELL Department ofAnutomy, Health Sciences Center, University of Louisville, Louisville, K Y 40292 ABSTRACT The migration of lymphocytes across the wall of high-endothelial venules was studied by electron microscopic examination of niurine lymph nodes fixed with glutaraldehyde and tannic acid. Regions of close membrane apposition, referred to in the present study as “intercellular contacts,” were observed between migrating lymphocytes and endothelial cells of the vessel wall. At high magnification the intercellular contacts resolve into pentalaminar structures resembling gap junctions. However, the location of these contacts suggests that they are regions of membrane adherence utilized for locomotion of the lymphocytes across the endothelium. At present, it is unclear whether these intercellular contacts are communicating junctions or sites of membrane adherence. In lymph nodes and other lymphatic tissues, lymphocytes migrate from the vascular bed across the endothelium of venules to enter the lymphatic stroma (Gowans and Knight, 1964; Howard et al., 1972; Sedgely and Ford, 1976).The venules involved in this cellular migration have a n unusually thick endothelium and are called high-endothelial venules (HEV) or postcapillary venules (PCV). Electron microscopic studies on HEV have described the normal structure of these vessels (e.g., Anderson, 19761, the effects of various substances upon HEV (Anderson et al., 1975; Henry and Beverley, 1976; Kittas and Henry, 1980), and the route of lymphocytic migration across the endothelium, whether intercellular (Schoefl, 1972; Wenk et al., 1974) or transcellular (Farr and DeBruyn, 1975; Cho and DeBruyn, 1979). The migration of lymphocytes across the walls of HEV seems to be a highly specific phenomenon involving attachment of lymphocytes by means of receptors to the endothelial cells (Stamper and Woodruff, 1976; Chin et al., 1980a,b; Carey et al., 1981). However, only a few studies have shown morphological evidence for cellular attachment or interaction between lymphocytes and endothelial cells. Anderson and Anderson (1976) and Ohmann (1980) described regions of decreased intercellular space between lymphocytes and endothelial cells, but according to these authors, 0 1983 ALAN R. LISS, INC such regions do not resemble any of the usual intercellular junctions. Migration of blood cells across the sinusoidal wall of bone marrow is also believed to be a highly specific phenomenon. Recent studies (Campbell, 1982) have demonstrated “intercellular contacts” resembling gap junctions between migrating blood cells and cells of the sinusoidal wall of bone marrow. In view of these findings, the present study was undertaken to determine whether similar intercellular contacts are present between migrating lymphocytes and cells of HEY Tannic acid-glutaraldehyde fixation was employed, because this fixative enhances membrane staining (van Deurs, 1975; Wagner, 1976)and preserves surface components of the membranes (Chew, 1980). MATERIALS AND METHODS Mesenteric and inguinal lymph nodes from six female Wistar rats weighing 150-200 gm were used in this study. The thoracic aorta of anesthetized animals was cannulated, the vascular system was flushed with isotonic Krebs’ buffer, pH 6.8, containing 1 unit of heparin per 100 ml, and then fixative was perfused into the vascular bed. The perfusion apparatus has been described previously (Campbell, 1972). The fixative consisted of Received March 24, 1983; accepted July 18, 1983. 644 F.R. CAMPBELL 2% glutaraldehyde and 2% tannic acid in 0.1 M cacodylate buffer, pH 6.8. Following perfusion of 100 ml of fixative through the vascular system, lymph nodes were excised, cut into small blocks, and fixed a n additional 1 hour in 4% glutaraldehyde and 4% tannic acid in 0.1 M cacodylate buffer, pH 6.5. Tissues were rinsed overnight in 0.1 M cacodylate, osmicated in the same buffer, dehydrated in a graded series of ethanol, stained with 10%uranyl acetate in methanol, and embedded in Maraglas. Thick sections were stained with azure I1 and examined to locate highendothelial venules for electron microscopic study. Thin sections were cut on a PorterBlum MT-2 Ultramicrotome using a diamond knife, stained with lead citrate, and examined at 60 kv in a Philips EM 200 electron microscope. During the course of the study, sections through about 400 lymphocytes in various stages of migration across the endothelium were photographed and the micrographs of these cells are the basis of the following observations. RESULTS Following fixation with glutaraldehyde and tannic acid, the high-endothelial venules appear much as they do with aldehydes alone. Readily recognized are the thick endothelial cells and the numerous small lymphocytes in the process of migrating across the vessel wall. Various organelles of the endothelial cells, described in detail by others (e.g., Anderson, 1976),appear well preserved, and the basal lamina of the endothelial cells is intensely stained (Figs. 3, 6). Tannic acid enhances staining of cellular membranes and, consequently, accentuates the intercellular space between lymphocytes and endothelial cells. Examination of this space shows a variable but appreciable amount of extracellular material which on occasion is rather electron dense and largely fills the intercellular space (arrows, Fig. 6). After perfusion of fixative through the vascular system, only a few blood cells, mostly small lymphocytes, remain in the lumina of HEV. Some of these lymphocytes appear unattached in a given section, but others are closely apposed to the surface of endothelial cells (Fig. 1).The intercellular space between these apparently adherent lymphocytes and the endothelial cells is variable in width but a t its narrowest measures 20-30 nm across. A few lymphocytes lying in the vessel lumina appear to be in a n early stage of endo- thelial penetration. Figure 2 illustrates this point; a process of the lymphocyte has invaginated the endothelium. These penetrating lymphocytes typically have one or more sites, often associated with villous projections of the lymphocytes, where the intercellular space is lacking and the adjacent plasmalemmas are closely apposed (arrowhead, Fig. 2). At higher magnification these regions of membrane apposition resolve into pentalaminar structures consisting of a central dense layer continuous with the outer leaflets of the membranes and flanked successively by electron lucent regions and by the dense inner leaflets of the membranes (Fig. 3, see inset). In the present study, these and similar pentalaminar regions of close membrane apposition are referred to as “intercellular contacts.” Figure 3 shows a lymphocyte (LY1)located largely within the endothelium of a HEV; only the blunt uropod of this migrating cell still lies in the lumen. An intercellular space of the usual width can be seen around most of the lymphocyte, located between the lymphocyte and endothelial cells or between this cell and another lymphocyte (LY2). However, near the luminal surface of the endothelium, at the site where the lymphocyte apparently first penetrated the endothelium, there are regions where a n intercellular space is lacking (arrowheads). In the inset of Figure 3, the membranes of the two cells can be observed coming together to form a pentalaminar region identical to that described above. The second lymphocyte (LY2) of Figure 3 lacks intercellular contacts with the endothelial cell, a t least in this section, even though it is in a similar stage of endothelial penetration. This observation is interpreted to indicate that these intercellular contacts are limited in extent and do not completely surround the migrating cell a t the site of initial penetration. Figure 3 also shows a n intercellular junction between two endothelial cells (arrow) which a t higher magnification (inset) has a n ultrastructure identical to that of the intercellular contacts associated with the miFig. l. A lymphocyte (LY) located in the lumen of a HEV abuts against an endothelial cell (El. The lymphocyte appears to be attached to the endothelium, but no intercellular contacts are observed. x 26,400. Fig. 2. A lymphocyte (LY) beginning to penetrate the endothelium (E) of a HEV. The lymphocyte has a process (P) invaginating the endothelium and an intercellular contact (arrowhead) with the endothelial cell. x 20,500. INTERCELLULAR CONTACTS OF HEV 645 646 F.R. CAMPBELL 647 INTERCELLULAR CONTACTS OF HEV grating lymphocyte. In view of the morphology of this junction and the cell types involved, the junction is believed to be a gap junction. The similarity of this junction to the intercellular contact between the lymphocyte and endothelial cell is remarkable. Figure 4 illustrates a lymphocyte that has penetrated the endothelium and now lies within this layer. An intercellular space of the usual width is observed around most of the lymphocyte, but a t one site (arrow) the membranes lie close together and a n intercellular space is lacking. Higher magnification (Fig. 4,inset) shows that this region is a n intercellular contact. Intercellular contacts were sometimes seen near the luminal surface of the endothelium as shown here, but they were also observed near the abluminal surface of the endothelium and a t intermediate levels. Sections through some lymphocytes showed no intercellular contacts, while sections through others showed more than one. Of the lymphocytes photographed and closely examined, about one quarter had distinct intercellular contacts or, when the membranes were not cut perpendicularly, regions that were presumed to be intercellular contacts. Since serial sections were not examined, it is not known whether all lymphocytes had intercellular contacts, but lack of contacts could be due to the plane of section. Many lymphocytes in passage across the endothelium appear to be solitary cells surrounded by endothelial cells as shown in Figure 4, but other lymphocytes form clusters of two or more cells among the endothelial cells. In such cases intercellular contacts were often present between adjacent lymphocytes (Fig. 5, arrows). In any one section, lymphocytes may have intercellular contacts only with other lymphocytes, only with endothelial cells, with both endothelial cells and lymphocytes, or no intercellular contacts at all. Again, because of plane of sectioning effects, Fig. 3. A lymphocyte (LY1) has largely penetrated the endothelium (El of a HEV. Intercellular contacts (arrowheads) are seen between lymphocyte and endothelial cell. A second lymphocyte (LY2) in a similar state of penetration lacks intercellular contacts, at least in this section. An intercellular junction between endothelial cells (arrow), the basal lamina @I) of the endothelium, and a pericyte (PI are also labeled. X21,300. Inset shows high magnification of the intercellular contact at the right between lymphocyte CLY) and endothelial cell (El and the intercellular junction between the endothelial cells. x 121,500. the frequency with which any of these combinations of intercellular contacts occur is uncertain. As lymphocytes pass through the endothehum, they penetrate the basal lamina of the endothelium, come into contact with pericytic cells of the vessel wall, and thereafter lie among cells of the lymphatic tissue. Figure 6 shows a lymphocyte that has penetrated the basal lamina of the endothelium and rests against processes of pericytic cells. Intercellular contacts were not observed during the course of this study between lymphocytes and pericytes or between lymphocytes and cells of the lymphatic tissue, despite the fact that many lymphocytes in this stage of migration were closely examined. Figure 6 also illustrates a mass of filamentous material (0 in the cytoplasm of the migrating lymphocyte. Similar filamentous material believed to be microfilaments was frequently observed in migrating lymphocytes, but its relationship, if any, to the intercellular contacts is not known. DISCUSSION The direction of lymphocytic migration across high-endothelial venules has been shown, using labeled lymphocytes, to be from the vascular bed into the surrounding lymphatic tissue (Gowans and Knight, 1964; Howard et al., 1972). The direction of migration can also be surmised in many instances from the morphology of fixed lymphocytes. Migrating lymphocytes display a typical configuration, with the nucleus located in the leading part of the cell and a blunt uropod containing most of the organelles trailing behind (Lewis, 1931; Norberg et al., 1973; Haston et al., 1982). In the present study, the expected lymphocytic polarity was evident among cells that were penetrating the endothelium (Fig. 3). Polarity of lymphocytes was also apparent among cells that were migrating out of the endothelium, but polarity was frequently lacking in lymphocytes that were located within the endothelial layer. This observation and the fact that most of the lymphocytes associated with HEV lie within the endothelium would seem to indicate that lymphocytes rapidly penetrate the endothelium, remain within it for a period of time, and then migrate into the surrounding tissue. A continuing controversy concerning lymphocytic migration across HEV relates to whether the lymphocytes penetrate the endothelial cells (Marchesi and Gowans, 1964; Farr and DeBruyn, 1975) or whether they 648 F.R. CAMPBELL INTERCELLULAR CONTACTS OF HEV pass through the intercellular space between endothelial cells (Schoefl, 1972; Wenk et al., 1974). The scanning electron microscopic studies of Cho and DeBruyn (1979) argue strongly in favor of the former view. This is a n interesting and possibly important point, because transcellular migration of lymphocytes in HEV may require cellular interactions between lymphocytes and endothelial cells. In a similar system, blood cells of bone marrow have been shown to pass through pores that penetrate individual endothelial cells (DeBruyn et al., 1971; Campbell, 1972; Becker and DeBruyn, 1976). Transcellular migration is believed by these authors to be of importance in the screening of blood cells thought to occur at the sinusoidal wall. In 649 view of these findings, one might expect transcellular migration and cellular interactions to be involved in the selection that occurs in HEV, i.e., only small lymphocytes pass in large numbers across these vessels. The present observations do not resolve this controversy, since the intercellular contacts are of unknown function and could presumably be involved in either route of migration. From the present study on fixed cells, it seems possible to reconstruct some of the events that occur as lymphocytes migrate from the vascular bed across the wall of HEV. Lymphocytes in the lumen of the vessels come into contact with the endothelial cells and adhere to them. The lymphocytes of the present study, apparently attached to the en- Fig. 6. A lymphocyte CLY) has a large process that passes through the basal lamina (bl) of the endothelium (E) and lies adjacent to the pericytes (P) of the vessel wall. Filamentous material (0 can be seen in the lymphocyte, and the dense intercellular material often seen between lymphocytes and endothelial cells is shown (arrows). x 24,100. Fig. 4. A lymphocyte (LY) surrounded by endothelial cells (E) of a HEV. An intercellular contact between lymphocyte and endothelial cell is shown at the arrow. X 19,800. Inset shows detail of membranes at the intercellular contact. x 146.000. Fig. 5. Two adjacent lymphocytes (LY) lying in the endothelium (E) of a HEV. Intercellular contacts (arrows) are shown between the lymphocytes. X 17,200. 650 F.R. CAMPBELL dothelium, were separated from the endothelial cells by a n intercellular space measuring at least 20 nm across. As penetration begins, intercellular contacts, often associated with villous projections of the lymphocytes, are formed with the endothelial cells. As the lymphocytes penetrate deeper into the endothelium, intercellular contacts are observed near the site of initial penetration. Once inside the endothelium, intercellular contacts are present between lymphocytes and endothelial cells and between adjacent lymphocytes. Finally, the basal lamina of the endothelium is penetrated, and the lymphocytes pass into adjacent tissue. Intercellular contacts have not been seen between migrating lymphocytes and pericytes or other nearby cells. The attachment of lymphocytes to endothelial cells of HEV has been investigated by several workers. Stamper and Woodruff (1976) and Butcher et al. (1979) have demonstrated selective binding of small lymphocytes to the surface of endothelial cells of HEV. More recently Chin et al. (1980a,b) and Carey et al. (1981)have presented evidence that lymphocytes possess a surface substance that mediates their binding to endothelial cells of HEV. This substance is trypsin-sensitive and appears to be a glycoprotein (Chin et al., 198Ob). Chin et al. (1980a) feel that the migration of lymphocytes across HEV consists of two distinct and separate phases: selective binding of lymphocytes to endothelial cells via their receptors, and then passage of lymphocytes across the endothelium. The present observations seem compatible with a two-phase process of lymphocytic migration across HEV. Furthermore, it is of interest and possible importance that attachment does not appear to involve intercellular contacts, but penetration and movement across the endothelium do. One of the most interesting questions raised by the present study concerns whether or not the intercellular contacts are gap junctions. The pentalaminar structure of the intercellular contacts appear to be identical to that of gap junctions of other cells following fixation with glutaraldehyde and tannic acid (see van Deurs, 1975; Kalderon et al., 1977; Gilula et al., 1978).Furthermore, in the present study, the morphology of the intercellular contacts is identical to that of junctions between endothelial cells. Since endothelial cells in general (Simionescu et al., 1976) and endothelial cells of HEV (van Deurs et al., 1975) may have gap junctions, and since the morphology seems correct, it would appear that the junctions described here between endothelial cells are indeed gap junctions. That being the case, it would seem to follow that the intercellular contacts associated with migrating lymphocytes are likewise gap junctions. Assuming this is true, then what are plausible functions of gap junctions a t these sites? It might be supposed that communicating junctions between endothelial cells and lymphocytes help determine which cells migrate across the endothelium. However, the studies of Chin et al. (1980a,b) and Carey et al. (1981) indicate that binding of lymphocytes to endothelial cells via receptors is the crucial aspect of specificity, and the present study shows that intercellular contacts are not associated with attachment. Communicating junctions between lymphocytes and endothelial cells or between lymphocytes within the wall of HEV could have many other functions, but our present knowledge concerning events that may occur during lymphocyte migration across HEV leaves one at a loss as to what there functions might be. With the present techniques, the structural similarity of the intercellular contacts to gap junctions is striking, but beyond that little can be stated. Examination of the particle distribution on the apposed membranes by means of freeze-fracture replicas would be of assistance in determining how closely these contacts resemble gap junctions, but to the author’s knowledge such studies have not been conducted. Another possibility is that the intercellular contacts of the present study are sites of adhesion between lymphocytes and endothelial cell membranes utilized for locomotion of the lymphocytes across the vessel wall. One might suppose that, as the lymphocytes begin to penetrate the endothelium, they form intercellular contacts with the endothelial cells and use these anchoring sites to invaginate the endothelium. Formation of additional contacts would occur as the lymphocytes penetrate deeper into the endothelium and these contacts would allow the lymphocytes to eventually pass through the endothelium. The observations of the present study seem to fit fairly well into such a schema. To begin with, during initial stages of penetration, the intercellular contacts are often associated with villous projections of the lymphocytes. Villous projections of attached lymphocytes (Anderson and Ander- INTERCELLULAR CONTACTS OF HEV son, 1976) and of lymphocytes during early stages of endothelial penetration (Cho and DeBruyn, 1979) have been reported previously. An interesting correlate of these observations is that villous projections of lymphocytes have been shown to contain high concentrations of actin (Sundqvist et al., 1980), a substance involved in adhesion and locomotion of several cell types, including lymphocytes (reviewed by Parrott and Wilkinson, 1981).If the intercellular contacts are adherent sites for locomotion of the lymphocytes, one would expect concentration of actin near the contacts. The present study shows microfilaments in the migrating lymphocytes but does not shown a n association of microfilaments and intercellular contacts. However, other techniques such as immunochemistry are better suited to demonstrate any association of actin and intercellular contacts. Another interesting observation of the present study is that the contacts that form as penetration of the endothelium begins appear to be maintained as penetration continues, suggesting that the contact sites may move along the surface of the cell. Capping of membrane components and their translocation along the membrane during locomotion of lymphocytes has been discussed by Jarvis et al. (1976) and by Parrott and Wilkinson (1981). All these observations suggest that the intercellular contacts described here are involved in attachment and locomotion of the lymphocytes, but their ultrastructure does not correlate well with cell to cell contacts described by others. Cell to cell contacts have been studied extensively in tissue culture systems. These contacts (Chen and Singer, 1982), largely on the basis of intermembranous spacing, are currently classified as 1) focal adhesions (10-20 nm), 2) close contacts (30-50 nm), and 3) extracellular matrix contacts (100 nm or more). Focal adhesions and close contacts have submembranous densities associated with them. The intercellular contacts of the present study do not resemble any of these contact types; the membrane spacing is too close (2-4 nm), and submembranous densities are not present. The possibility that gap junctions function as adherent sites for cellular locomotion has not, to the author’s knowledge, been reported. Intercellular contacts identical to those of the present study have been reported in bone marrow following fixation with tannic acid and glutaraldehyde, but their function in 651 marrow is likewise unresolved. In marrow of mice, these contacts are observed between developing blood cells and fixed cells of the marrow (Campbell, 1980) and between migrating blood cells and cells of the sinusoidal wall (Campbell, 1982). In marrow of chickens, these contacts are observed between the intravascular erythroblasts and endothelial cells of the sinusoidal wall (Sorrel1and Weiss, 1982). Possible functions of these contacts in marrow are discussed by these authors. It need only be emphasized here that on the basis of the present study, it is apparent that identical contacts are involved in the delivery of blood cells from the marrow into the circulation, on the one hand, and with the transfer of lymphocytes from the circulation to lymph nodes, on the other. Both processes have in common a high degree of cell selectivity and the passage of cells from one compartment to another. Further studies would seem to be needed to elucidate the function of these intercellular contacts. At present one can only state that they are present between lymphocytes and endothelial cells and between lymphocytes and that they are likely of considerable importance to the migration of lymphocytes across the wall of HEV. LITERATURE CITED Anderson, A.O., and N.D. Anderson (1976) Lymphocyte emigration from high endothelial venules in rat lymph nodes. Immunology, 31:731-748. Anderson, N.D. (1976) Specialized structure and metabolic activities of high endothelial venules in rat lymphatic tissue. Immunolgy, 31:455-473. Anderson, N.D., A.O. Anderson, and R.G. Wyllie (1975) Microvascular changes in lymph nodes draining skin allografts. Am. J. Pathol., 81:131-153. Becker, R.P., and P.P.H. DeBruyn (1976) The transmural passage of blood cells into myeloid sinusoids and the entry of platelets into the sinusoidal circulation; A scanning electron microscopic investigation. Am. J. Anat., 145:183-206. Butcher, E.C., R.G. Scollay, and I.L. Weissman (1979) Lymphocyte adherence to high endothelial venules: Characterization of a modified in vitro assay, and examination of the binding of syngeneic and allogeneic lymphocyte populations. J. Immunol., 123:1996-2003. Campbell, F. (1972) Ultrastructural studies of transmural migration of blood cells in the bone marrow of rats, mice and guinea pigs. Am. J. Anat., 135.521-536. Campbell, F.R. (1980) Gap junctions between cells of bone marrow: An ultrastructural study using tannic acid. Anat. Rec., 196:lOl-117. Campbell, F.R. (1982) Intercellular contacts between migrating blood cells and cells of the sinusoidal wall of bone marrow: An ultrastructural study using tannic acid. Anat. Rec., 203:365-374. Carey, G.D., Y.-H. Chin, and J.J. Woodruff (1981) Lymphocyte recognition of lymph node high endothelium. 111. Enhancement by a component of thoracic duct 652 F.R. CAMPBELL lymph. J. Immunol., 127r976-979. Chen, W.-T., and S.L. Singer (1982) Immunoelectron microscopic studies of the sites of cell-substrate and cellcell contacts in cultured fibroblasts. J. Cell Biol., 95t205-222. Chew, E.C. (1980) On tannic acid fixation and staining. Acta Anat., 108t156-159. Chin, Y.-H., G.D. Carey, and J.J. Woodruff (1980a)Lymphocyte recognition of lymph node high endothelium. I. Inhibition of in vitro binding by a component of thoracic duct lymph. J. Immunol., 125t1764-1769. Chin, Y.-H., G.D. Carey, and J.J.Woodruff (1980b) Lymphocyte recognition of lymph node high endothelium. 11. Characterization of a n in vitro inhibitorv factor isolated by antibody affinity chromatography. J. Immunol., 125t1770-1774. Cho, Y., and P.P.H. DeBruyn (1979) The endothelial structure of the postcapillary venules of the lymph node and the passage of lymphocytes across the venule wall. J. Ultrastruct. Res., 69:13-21. DeBruyn, P.P.H., S. Michelson, and T.B. Thomas (1971) The migration of blood cells of the bone marrow through the sinusoidal wall. J. Morphol., 133r417-428. Farr, A.G., and P.P.H. DeBruyn (1975) The mode of lymphocyte migration through postcapillary venule endothelium in lymph node. Am. J. Anat., 143:59-92. Gilula, N.B., M.L. Epstein, and W.H. Beers (1978) Cellto-cell communication and ovulation. A study of the cumulus-oocyte complex. J. Cell Biol., 78t58-75. Gowans, J.L., and E. Knight (1964) The route of recirculation of lymphocytes in the rat. Proc. R. SOC. London, Ser. B., 159t257-282. Haston, W.S., J.W. Shields, and P.C. Wilkinson (1982) Lymphocyte locomotion and attachment on 2-dimensional surfaces and 3-dimensional matrices. J. Cell Biol., 92t747-752. Henry, L., and JK.A. Beverley (1976) Age and sex differences in the response of lymph node post-capillary venules in mice infected with Toxoplasma gondii. Br. J. Exp. Pathol., 57t274-280. Howard, J.C., S.V. Hunt, and J.L. Gowans (1972) Identification of marrow-derived and thymus-derived small lymphocytes in the lymphoid tissue and thoracic duct lymph of normal rats. J. Exp. Med., 135t200-219. Jarvis, S.C., R. Snyderman, and H.J. Cohen (1976) Human lymphocyte motility: Normal characteristics and anomalous behavior of chronic lymphocytic leukemia cells. Blood, 48t717-729. Kalderon, N., M.L. Epstein, and N.B. Gilula (1977) Cellto-cell communication and myogenesis. J. Cell Biol., 75t788-806. Kittas, C., and L. Henry (1980) Ultrastructural effects of sex hormones and infection on lymph node post-capil- lary venules. J. Pathol., 132t121-131. Lewis, W.H. (1931) Locomotion of lymphocytes. Bull. Johns Hopkins Hosp., 49t29-36. Marchesi, V.T., and J.L. Gowans (1964) The migration of lymphocytes through the endothelium of venules in lymph nodes: An electron microscope study. Proc. R. SOC.London, Ser. B., 159:283-290. Norberg, B., L. Rydgren, and N. Soderstrom (1973) Lymphocyte locomotion. Morphological criteria of the direction of lymphocyte movement. Lymphology, 6t187-191. Ohmann, H.B. (1980) Electron microscopic study of the paracortical postcapillary “high endothelial venules” in lymph nodes of the normal calf. Cell Tissue Res., 212t465-474. Parrott, D.M.V., and P.C. Wilkinson (1981)Lymphocyte locomotion and migration. Prog. Allergy, 28t193-248. Schoefl, G.I. (1972) The migration of lymphocytes across the vascular endothelium in lymphoid tissue. J. Exp. Med., 136568-584. Sedgley, M., and W.L. Ford (1976) The migration of lymphocytes across specialized vascular endothelium. I. The entry of lymphocytes into the isolated mesenteric lymph node of the rat. Cell Tissue Kinet., 9t231-243. Simionescu, M., N. Simionescu, and G.E. Palade (1976) Segmental differentiation of cell junctions in the vascular endothelium: Arteries and veins. J. Cell Biol., 68t705-723. Sorrell, J.M., and L. Weiss (1982) Intercellular junctions in the hematopoietic compartments of embryonic chick bone marrow. Am. J. Anat., 164t57-66. Stamper, H.B. Jr., and J.J. Woodruff (1976) Lymphocyte homing into lymph nodes: In vitro demonstration of the selective affinity of recirculating lymphocytes for high endothelial venules. J. Exp. Med., 144328-833. Sundqvist, K.G., P. Otteskog, L. Wanger, R. Thorstensson, and G. Utter (1980) Morphology and microfilament organization in human blood lymphocytes. Effects of substratum and mitogen exposure. Exp. Cell Res., 130t327-337. van Deurs, B. (1975) The use of tannic acid-glutaraldehyde fixatives to visualize gap and tight junctions. J. Ultrastruct. Res., 50t185-192. van Deurs, B., C. Ropke, and E. Westergaard (1975) Permeability properties of the postcapillary high-endothelial venules in lymph nodes of the mouse. Lab. Invest., 32t201-208. Wagner, R.C. (1976) The effect of tannic acid on electron images of capillary endothelial cell membranes. J. U1trastruct. Res., 57t132-139. Wenk, E.J., D. Orlic, E.J. Reith, and J.A.G. Rhodin (1974) The ultrastructure of mouse lymph node venules and the passage of lymphocytes across their walls. J. U1trastruct. Res., 47r214-241.