THE MIGRATION OF LYMPHOCYTES ACROSS SPECIALIZED VASCULAR ENDOTHELIUM

Lymphocytes were exposed in vitro to either trypsin or neuraminidase. The ability of the treated cells to migrate into tissues was measured (a) by i.v. injection into intact recipients and (b) by vascular perfusion through an isolated lymph-node preparation. The localization of trypsinized cells in the lymph-nodes of recipients was deficient when compared to untreated lymphocytes and there was a surplus of trypsinized cells in the blood. Trypsinized cells migrated into the isolated nodes in reduced numbers. By contrast, neuraminidase treated lymphocytes were markedly deficient in the blood of recipients early after injection; their localization in the spleen and lymph-nodes was also deficient but they were in surplus in the liver. Moreover they migrated into the isolated nodes in slightly increased numbers. By 24 hr after injection the perturbed localization pattern produced by either enzyme was partly restored to normal. In conclusion, trypsin interfered with the capacity of lymphocytes to migrate into lymph-nodes but neuraminidase did not; the latter promoted the hepatic sequestration of cells and the reduced localization in the blood and tissues was a consequence of this. The hypothesis that lymphocytes adhere to specialized endothelia in lymph-nodes because of specific glycoside sequences on their surface lacks experimental support.     

The migration of lymphocytes across specialized vascular endothelium. I. The entry of lymphocytes into the isolated mesenteric lymph-node of the rat.

The chain of lymph-nodes in the rat mesentery was isolated and the preparation was perfused via cannulae in the superior mesenteric vessels. The perfusate consisted of serum to which labelled lymphocytes had usually been added. The entry of radioactively labelled lymphocytes from the blood vessels into the lymph-nodes was studied by scintillation counting and autoradiography. It was observed that: (1) In the perfused node labelled lymphocytes localized in and around post-capillary venules in the paracortex as they do early after i.v. injection. (2) The number of lymphocytes which entered the node was directly proportional to the concentration in the perfusate over the range studied. The proportion of cells retained in the node varied considerably around a mean of 11% of lymphocytes reaching it. (3) The isolated lymph-node released few if any lymphocytes into the effluent (venous) perfusate. (4) Large lymphocytes migrated into isolated lymph-nodes slightly more readliy than did small lymphocytes. (5) Unlike intact cells isolated lymphocyte membranes did not adhere to specialized vascular endothelium.     

The influence of strong transplantation antigens (Ag-B) on lymphocyte migration in vivo.

The tissue localization of syngeneic thoracic duct lymphocytes was compared to that of allogeneic cells in four rat strain combinations differing at the Ag-B locus (HO → DA, DA → HO, AO → HO, HO → AO). Dual isotope labeling with [³H]uridine and [¹⁴C]uridine was applied in order so that the distribution of allogeneic and syngeneic cells could be followed in one recipient. During the first couple of hours after iv injection, allogeneic lymphocytes usually migrated as easily into the various tissues as did syngeneic cells. However, after 24 and 48 hr, a reduced amount of label associated with allogeneic cells was often measured in the tissues. This reduction differed in magnitude in the different strain combinations and was most pronounced in the lymph nodes. A reduced number of allogeneic cells also appeared in the thoracic duct. By contrast, no reduced localization of allogeneic lymphocytes was measured in the draining popliteal lymph nodes late after sc injection. In preimmunized animals allogeneic cells were rapidly removed from the blood and therefore failed to localize in the lymphoid tissues. Furthermore, the lymph node localization of allogeneic cells was more like that of syngeneic cells in splenectomized rats, as well as in irradiated recipients (when the irradiation was given shortly before cell transfer). It is concluded that transplantation antigens play no essential role in the interaction between recirculating lymphocytes and the venous endothelium at the sites where the large-scale physiological emigration of the cells takes place (the HEVS of the lymph nodes and the marginal zone vessels of the spleen). The elimination of allogeneic cells is found later; it probably takes place in the lymph nodes and spleen. Possible mechanisms responsible for this rapid removal of allogeneic lymphocytes in nonimmunized recipients are discussed.     

Further evidence for the existence of B-lymphocyte subpopulations.

We have studied the homing properties of B lymphocytes by using ⁵¹Cr-labeled lymphoid cells obtained from athymic, nu/nu mice, and animals made T-lymphocyte deficient by thymectomy and lethal irradiation followed by reconstitution with syngeneic bone marrow. Comparison was made to the patterns of distribution observed when cell preparations containing normal numbers of T and B lymphocytes were migrated. A small but significant percentage of labeled lymphocytes from lymph nodes, spleen, Peyer's Patches, and bone marrow of T-cell-deficient animals was shown to be lymph node seeking. Secondary transfers of lymph node cells from primary recipients caused enrichment of this lymph node-seeking population. Treatment of T-lymphocyte-deficient lymphoid cell preparations with neuraminidase reduced the percentages of cells homing to the lymph nodes. The data showed that B lymphocytes exhibit unique homing properties when injected into normal recipients. In addition, direct comparison of the homing patterns of B lymphocytes prepared from spleen and lymph nodes of athymic mice revealed differences suggesting that these lymphoid organs contained unique mixtures of at least two different kinds of B cell. The evidence supports the notion that the B-lymphocyte populations contain at least two subpopulations, one of which possesses the ability to home to lymph nodes.     

THE MIGRATION OF LYMPHOCYTES ACROSS SPECIALIZED VASCULAR ENDOTHELIUM

The chain of lymph-nodes in the rat mesentery was isolated and the preparation was perfused via cannulae in the superior mesenteric vessels. The perfusate consisted of serum to which labelled lymphocytes had usually been added. The entry of radioactively labelled lymphocytes from the blood vessels into the lymph-nodes was studied by scintillation counting and autoradiography. It was observed that: (1) In the perfused node labelled lymphocytes localized in and around post-capillary venules in the paracortex as they do early after i.v. injection. (2) The number of lymphocytes which entered the node was directly proportional to the concentration in the perfusate over the range studied. The proportion of cells retained in the node varied considerably around a mean of 11 % of lymphocytes reaching it. (3) The isolated lymph-node released few if any lymphocytes into the effluent (venous) perfusate. (4) Large lymphocytes migrated into isolated lymph-nodes slightly more readily than did small lymphocytes. (5) Unlike intact cells isolated lymphocyte membranes did not adhere to specialized vascular endothelium.     

Recruitment of effector lymphocytes by initiator lymphocytes. Recruited lymphocytes are immunospecific and include graft-versus-host-reactive lymphocytes.

We have shown previously that initiator T lymphocytes (ITL), sensitized in vitro against fibroblast antigens, recruit effector T cells in vivo. After injection into hind footpads of syngeneic recipients, sensitized ITL migrated to the draining popliteal lymph nodes (PLN) and activated a trapping mechanism by which circulating lymphocytes were recruited in the PLN. This paper reports experiments designed to test the immunospecificity of these recruited T lymphocytes (RTL). We found that immunospecific RTL were depleted from other lymphoid organs during recruitment in the PLN. However, immunospecific ITL were not depleted from spleens during PLN recruitment. Thus ITL and RTL are functionally distinguishable. We show that specific GVH reactive lymphocytes were also lost from spleens and distal lymph nodes during trapping of RTL in the PLN. Thus, the trapping phase of the recruitment response is immunospecific, as are the sensitization and effector phases. The trapped RTL are antigen-specific, and include the pool of GVH-reactive-lymphocytes committed to the same alloantigen. Thus, it appears that GVH-reactive cells respond to syngeneic ITL sensitized against allogeneic fibroblasts.     

Influenza A virus interaction with murine lymphocytes. I. The influence of influenza virus A/Japan 305 (H2N2) on the pattern of migration of recirculating lymphocytes.

The effect of influenza virus A/Japan 305 (H2N2) on the path of migration of recirculating lymphocytes has been studied. 51Cr-labeled rat thoracic duct lymphocytes (TDL) were incubated with virus at 37 degrees C for 1 hr and then infused i.v. into syngeneic recipients which were killed 1 hr later. Virus-treated TDL accumulated in the liver and their recovery in lymph nodes and spleen was severely reduced. Changes in lymphocytes induced by virus developed rapidly and were evident after incubation for only 15 min. UV-irradiated virus altered the pattern of lymphocyte localization but attachment of heat-inactivated virus to lymphocytes in vitro had no effect on their distribution in vivo. Evidence was obtained that some virus-treated TDL, initially sequestered in the liver, subsequently recovered their ability to circulate normally. Recovery was not complete and a population of cells failed to regain their ability to home into lymph nodes. Evidence is also presented demonstrating that influenza virus affected the homing properties of both T and B cells. It is suggested that aberrations in lymphocyte homing were mediated by the viral neuraminidase which induces changes in the cell membrane leading to their accumulation in the liver.     

Antigen transport II. The significance of the localization of antigen-laden cells in the spleen

Previous studies have demonstrated that macrophage-like cells transporting antigen, e.g., human serum albumin (HSA) appear in thoracic duct lymph and blood shortly after antigen injection. The in vivo migration of these antigen-laden (Ag-L) cells from the blood stream was examined systematically by transferring Ag-L cells bearing ¹²⁵I-labelled HSA into syngeneic rats. There was no evidence autoradiographically that Ag-L cells migrated into lymph nodes, but the localization in the spleen followed a defined pattern: within the first hours after transfer, a majority of radiolabelled cells were identified in the marginal zone; by 3 hr and up to 4 days later, 60–80% of labelled cells were resident in the red pulp; Ag-L cells failed to migrate into the white pulp in significant numbers. Ag-L cells which had localized to the spleen, when examined 3 and 18 hr after transfer using combined autoradiography and immunoperoxidase staining, did not express la determinants in situ. The ability of Ag-L cells to stimulate an adoptive secondary response was tested in splenectomized, irradiated recipients receiving HSA-specific memory cells. Removal of the spleen before transfer severely reduced the antibody response evoked by Ag-L cells transporting HSA, thus indicating the functional importance of antigen transport to the spleen. Since Ag-L cell migration was primarily into the red pulp, we have considered whether the red pulp may provide a relevant microenvironment for lymphocyte/ antigen interaction.     

Migration of recently divided B and T lymphocytes to peritoneum and lung

It is recognized that a population of newly divided (or young) cells migrate preferentially to inflamed foci. It has been shown that a large proportion of lymphocytes residing in the bronchoalveolar airspaces of rat are recently divided cells and that blood may be an important source of these cells. To further delineate how blood may contribute to lymphocyte subpopulations in inflamed peritoneum and lung, a comparison of the capacity of recently divided T and B cells to migrate from blood to inflamed peritoneum and lung was made. To label young lymphocytes, DA strain donor rats were given Initiated thymidine by vein in vivo for 7 days. After thoracic duct drainage, the following labeled cell populations were adoptively transferred by vein into syngeneic recipients: (i) unseparated thoracic duct lymphocytes (TDL), (ii) enriched T cells (>90%) or B cells (>80%) recovered after passage of TDL through nylon columns, and (iii) thoracic duct lymphocytes (> 99% B cells) obtained from “B rats” that were prepared by X irradiation, thymectomy, and bone marrow reconstitution. T and B cells were identified by specific heterologous antisera. The percentage recovery of labeled lymphocytes in the recipients with inflamed peritoneum or lung aspirates was determined from cell counts and autoradiographs. The studies indicated that (a) both labeled T and B cells migrated to inflamed peritoneum and lung; (b) labeled B cells migrated to peritoneum and lung better than did labeled TDL or T cells; and (c) labeled lymphocytes did not migrate to unstimulated peritoneum. The enhanced migration of newly divided B lymphocytes to inflamed peritoneum and normal lung (a site that is likely under chronic antigenic stimulation) was unexpected, but may provide additional information on the relative contribution of these subpopulations in the immune inflammatory response.     

Virus-induced alterations of lymphoid tissue. IV. The effect of Newcastle disease virus on the fate of transfused thoracic duct lymphocytes

⁵¹Cr-labeled thoracic duct lymphocytes were briefly incubated at 4 °C with Newcastle disease virus (NDV) and then infused into syngeneic rats. Virus diverted the homing of many donor cells from lymph nodes and spleen to the liver. Evidence was obtained suggesting that some NDV-treated lymphocytes initially trapped in the liver subsequently migrated into the lymph nodes. The results imply that NDV transiently interrupts the normal route of lymphocyte migration. Alterations in lymphocyte distribution were mediated by attachment of virus to the cell surface and were the same as those induced by incubating lymphocytes with V. cholera neuraminidase before infusion. It appears that reactions involving 2–3′ and/or2–8′ linked sialyl residues on the surface of recirculating lymphocytes can markedly affect their distribution in the body.     

The migration of lymphocytes across specialized vascular endothelium. VI. The migratory behaviour of thoracic duct lymphocytes retransferred from the lymph nodes, spleen, blood, or lymph of a primary recipient

Thoracic duct lymphocytes labelled with 51Cr were injected into a primary recipient and then were transferred for a second time from the lymph nodes (cervical and/or mesenteric), spleen, lymph, or blood into a series of final recipients. Measurement of the organ distribution of labelled lymphocytes in the final recipients enabled three main conclusions to be drawn. (1) Lymphocytes that had localized in the spleen, mesenteric lymph nodes (LN), or cervical LN of the first recipient showed no tendency to return in increased numbers to the same organ in the final recipient. (2) Lymphocytes that had recently entered the spleen or LN were temporarily impaired in their ability to reenter LN. This capacity was recharged when the cells returned to the lymph and the blood. (3) Lymphocytes that had been passaged from blood to lymph and collected for up to 4 hr at room temperature entered the LN of a recipient much faster than did nonpassaged thoracic duct lymphocytes collected overnight at 0 degree C. Supplementary experiments indicated that the different migratory behavior of thoracic duct lymphocytes under these two circumstances was mainly a consequence of their handling in vitro during the collecting and the labelling procedures. This functional impairment was not associated with a diminished ability to enter the spleen and bone marrow or to survive in recipients for up to 24 hr.     

The migration of lymphocytes across specialized vascular endothelium. IV. Prednisolone acts at several points on the recirculation pathways of lymphocytes

Radioactively labelled thoracic duct lymphocytes from syngeneic rat donors were injected iv into recipients which had been given a continuous iv infusion of prednisolone at 1 mg/hr for 15–18 hr previously. The tissue distribution and recirculation into lymph of the labelled lymphocytes were compared quantitatively in the prednisolone-treated and control recipients by scintillation counting and autoradiography. The most prominent effect of prednisolone was to retard recirculating lymphocytes within the tissues to which they are normally distributed by the blood, namely the bone marrow, the spleen, and the lymph nodes. Although lymphocyte traffic was almost completely frozen by prednisolone, recirculating lymphocytes were not killed. A second effect of prednisolone was to impair the influx of lymphocytes from the blood into lymph nodes. Different groups of lymph nodes varied in the extent to which prednisolone inhibited the entry of lymphocytes, and previous antigenic stimulation completely exempted lymph nodes from this inhibition. Lymphocytes took a longer time to cross the walls of high endothelial venules in the lymph nodes of prednisolone-treated rats. A third effect of prednisolone was to increase the rate at which lymphocytes entered the bone marrow from the blood by crossing sinusoidal endothelium.     

Lymphocyte traffic through granulomas: Differences in the recovery of indium-111-labeled lymphocytes in afferent and efferent lymph

Afferent lymphatics draining granulomas and efferent lymphatics from normal and stimulated lymph nodes were cannulated in sheep. There was a greatly increased output of cells in afferent lymph-draining chronic inflammatory sites or Freund's adjuvant-induced granulomas. Cells collected from these lymphatics were radiolabeled with ¹¹¹In and injected intravenously. The reappearance of these labeled cells in lymph at various sites was measured. Labeled afferent lymph cells migrated from blood through the granuloma back into afferent lymph in large numbers and with kinetics which were comparable to efferent lymphocytes recirculating through a lymph node. When labeled afferent lymph cells were injected the specific activity (cpm/10⁷ cells) in afferent lymph was five times higher than that in efferent lymph from a normal node. When efferent lymph cells were labeled the afferent lymph specific activity was one-half that in efferent lymph. It is suggested that the cells in afferent lymph migrate preferentially from blood through the granuloma and constitute a unique population of cells.     

The migration of lymphocytes into rat popliteal lymph nodes during antigenic promotion or competition between heterologous erythrocytes

The injection of chicken and sheep red blood cells (CRBC and SRBC) into rat popliteal lymph nodes either together or sequentially 2, 4, 6, or 8 days apart resulted in an enhanced immune response when the second antigen was injected 2 or 4 days after the injection of the first antigen (antigenic promotion) or a suppressed immune response when the second antigen was injected 6 days after the injection of the first antigen (antigenic competition). The immune response to either antigen was dependent upon the time of administration of the second antigen with respect to the first antigen. Lymphocyte migration into antigenically stimulated lymph nodes was greater when the two antigens were injected sequentially rather than together. Further, the migration of lymphocytes into the lymph node was enhanced when the second antigen was injected during the inductive or suppressive phase of the immune response to the first antigen (CRBC) regardless of whether the same (CRBC) or an antigenically unrelated antigen (SRBC) was used as the second antigen. While antigenic promotion may in part be explained by the increased rate at which lymphocytes migrate into lymph nodes, lymphocyte migration is also enhanced during antigenic competition. This suggests that while suppressor cells/factors may regulate the effector phase of an immune response they do not directly modulate the migration of blood-borne lymphocytes into the lymph node.     

The entry of T and B lymphocytes into rat popliteal lymph nodes undergoing a graft-versus-host reaction

The entry of radiolabeled blood-borne T and B lymphocytes into resting popliteal lymph nodes and popliteal lymph nodes stimulated with semiallogeneic lymphocytes was investigated in rats. Thoracic duct lymphocytes separated into T- and B-lymphocyte populations on nylon-wool columns were radiolabeled with 51chromium and equal numbers of T or B lymphocytes were injected intravenously. While the ratio of T and B lymphocytes in the blood is approximately 3:1 it was found that the ratio of T to B lymphocytes migrating into lymph nodes was approximately 9 T to 1 B lymphocyte in both resting and antigenically stimulated lymph nodes. Since the ratio of T to B lymphocytes in thoracic duct lymph is similar to that of blood, there is a disparity between the number of T cells entering and leaving lymph nodes. These results suggest that some T lymphocytes may return to the blood directly and/or there is increased T lymphocyte death in lymph nodes.     

Migration of lymphocytes through endothelial cell monolayers: augmentation by interferon-gamma

During normal lymphocyte recirculation and in chronic inflammation, lymphocytes emigrate from blood into the perivascular tissue. The mechanism of lymphocyte migration through the endothelial cell (EC) layer of blood vessels is poorly understood. To identify factors that control lymphocyte emigration, a method has been developed to measure human peripheral blood lymphocyte migration through monolayers of human umbilical vein EC and into nitrocellulose (NC) filters located below the EC monolayer. Counts were made of lymphocytes that had migrated into the NC filter using a particle counter. T lymphocytes attached to and migrated through EC monolayers in a T-cell-number- and time-dependent fashion. Migration required viable EC since lymphocytes failed to migrate through formaldehyde-fixed EC monolayers or monolayers of dermal fibroblasts. Interferon-gamma (IFN-gamma) markedly augmented the migration in a dose- and time-dependent manner when preincubated with the EC. When T lymphocytes were pretreated with IFN-gamma, no increase in migration was observed. Finally, IFN-gamma augmented the migration of T cells prebound to the EC, indicating that the IFN-gamma-enhanced migration was not due to increased binding of T cells to the EC, but rather to an action on the EC to facilitate subsequent migration.     

Lymphocyte traffic through lymph nodes and Peyer's patches of the rat: B- and T-cell-specific migration patterns within the tissue,and their dependence on splenic tissue

The migration routes of lymphocyte subsets through organ compartments are of importance when trying to understand the local events taking place during immune responses. We have therefore studied the traffic of B, T, CD4⁺, and CD8⁺ lymphocytes through lymph nodes and Peyer's patches. At various time points after injection into the rat, labeled lymphocytes were localized, and their phenotype characterized in cryostat sections using immunohistochemistry. Morphometry was also performed, and the recovery of ⁵¹Cr-labeled lymphocytes in these organs was determined. B and T lymphocytes entered the lymph nodes via the high endothelial venules in similar numbers. Most B lymphocytes migrated via the paracortex (T cell area) into the cortex (B cell area), and then back in substantial numbers into the paracortex. In contrast, T lymphocytes predominantly migrated into the paracortex and were rarely seen in the cortex. No obvious differences were seen between various lymph nodes and Peyer's patches and the routes of CD4⁺ and CD8⁺ lymphocytes. After injection of lymphocytes into animals with autotransplanted splenic tissue, the number of B lymphocytes that had migrated into the B cell area of lymph nodes and of Peyer's patches was significantly decreased, whereas CD4⁺ lymphocytes migrated in larger numbers into the T cell area of both organs.This study was supported by the Deutsche Forschungsgemein-schaft (SFB 244, A7).     

A monoclonal antibody specific for rat intestinal lymphocytes

A monoclonal antibody, RGL-1, was produced by fusion of NSI myeloma cells with spleen cells of a mouse immunized with isolated rat intraepithelial lymphocytes (IEL). SDS-PAGE analysis revealed that RGL-1 precipitated two major noncovalently bound chains of about m.w. 100,000 and 125,000, and a minor component of m.w. 200,000. Examination of both tissue sections and isolated cells indicated that RGL-1 stained the majority of the lamina propria lymphocytes and IEL but only very few cells (less than 2%) in the lymphoid organs and small numbers of lymphocytes in other mucosae. In the small intestine, RGL-1 stained lymphocytes with the helper (W3/25) as well as the cytotoxic/suppressor (OX8) phenotype. The antibody reacted with 95% of the granular IEL but with less than 0.1% of the blood large granular lymphocytes. Although mature IgA plasma cells in the lamina propria were RGL-1-, some large IgA-containing cells were weakly positive. In the gut-associated lymphoid tissues (GALT), studies combining immunofluorescence and autoradiography indicated that 56 and 73% of rapidly dividing cells of mesenteric lymph nodes and of thoracic duct lymph (TDL) stained with RGL-1, respectively. In addition, 90 to 100% of the IgA-containing blasts of MLN and 75% of those of TDL were labeled by RGL-1. In contrast, rapidly dividing cells of spleen and of peripheral lymph nodes did not stain with RGL-1. Because RGL-1 can be demonstrated on both intestinal lymphocytes and their immediate precursors in the GALT, its expression may be related to the homing of lymphocytes into the gut mucosa.     

Homing of intravenously and intralymphatically injected human dendritic cells generated in vitro from CD34+ hematopoietic progenitor cells

Dendritic cells (DC) are professional antigen-presenting cells that can be generated in vitro from CD34⁺ peripheral blood progenitor cells by recombinant cytokines. These cells have potential implications for immunotherapeutic approaches in the treatment of cancer and other diseases. Physiologically, immature DC in the periphery capture and process antigens, then mature to interdigitating DC and migrate to lymphoid organs, where they activate lymphocytes. However, it is not known if DC generated in vitro have the capacity to traffic in vivo to the lymphoid tissues, such as spleen and lymph nodes. We have investigated whether human radiolabeled DC differentiated in vitro migrate and localize to lymphoid tissues after intravenous and intralymphatic injection. The distribution and localization of the DC were evaluated in five patients with malignant melanoma using serial whole-body gamma camera imaging. Intravenously infused DC demonstrated transient lung uptake followed by localization in the spleen and liver for at least 7 days. DC injected into a lymphatic vessel at the dorsal foot were rapidly detected in the draining lymph nodes where they remained for more than 24 h. These data suggest that DC differentiated in vitro localize preferentially to lymphoid tissue, where they could induce specific immune responses. Received: 28 January 1999 / Accepted: 4 March 1999     

Recruitment of effector lymphocytes by initiator lymphocytes. Circulating lymphocytes are trapped in the reacting lymph node.

In previous studies, we have shown that initiator T lymphocytes (ITL) sensitized in vitro recruit effector T lymphocytes in vivo. When ITL were sensitized against fibroblast antigens in vitro and injected into footpads of syngeneic recipients, they induced enlargement of the draining popliteal lymph node (PLN) and the development there of specific effector lymphocytes of recipient origin. To study the basis of this lymph node response in recruitment, we injected 51Cr-labeled spleen cells i.v. into recipients of sensitized ITL and found that the labeled circulating lymphocytes were trapped in the reacting PLN. The trapping depended on surface properties of the labeled circulating lymphocytes, as revealed by various enzymatic treatments. The trapping process was radiosensitive, both on the part of the trapped lymphocytes and the lymph node-trapping mechanism. Thus, sensitized ITL injected into the hind footpads migrate to the PLN and induce the trapping of circulating recruitable lymphocytes, which either differentiate into or regulate the differentiation of effector T lymphocytes.