Expression of MHC class II molecules on human monocytes is regulated independently from each other after phagocytosis of bacteria

Reactive arthritis is usually self-limiting polyarthritis which develops in HLA-B27 positive individuals after certain gastrointestinal or urogenital infections. The pathogenesis of reactive arthritis is unknown but T cells seem to have a crucial role. Most of the antigen-specific T cells isolated from the synovial fluid have been MHC class II restricted. The role of antigen presentation in the pathogenesis of reactive arthritis has been studied relatively little. In this work the authors studied the effect of arthritis-triggering bacterium (Yersinia enterocolitica O:3) on the expression of MHC class II molecules on human monocytes and found that the expression of different MHC class II molecules was regulated independently from each other in half of the individuals after certain incubation periods. In these cases the expression of HLA-DP was parallel to the expression of HLA-DQ, while HLA-DR expression went to the opposite direction or did not change at all. No difference between HLA-B27 negative and HLA-B27 positive healthy individuals was seen. The authors conclude that independent regulation of the expression of different MHC class II antigens on antigen-presenting cells is a more common phenomenon than usually thought.

CD73 is involved in lymphocyte binding to the endothelium: characterization of lymphocyte-vascular adhesion protein 2 identifies it as CD73

We have recently described a monoclonal antibody (mAb) 4G4 recognizing a 70-kD molecule constitutively expressed on human endothelial cells and on subpopulations of lymphocytes. We showed that this molecule, which we named lymphocyte-vascular adhesion protein 2 (L-VAP-2), mediates lymphocyte adhesion to cultured endothelial cells. Protein sequencing of tryptic peptides from immunoaffinity-purified L-VAP-2 revealed sequence identity between L-VAP-2 and CD73 (ecto-5'-nucleotidase, E.C.3.1.3.5), and COS cells transfected with a CD73 cDNA were positively stained with the mAb 4G4, which recognizes L-VAP-2. mAb 4G4 was also able to partially inhibit the ecto-5'-nucleotidase activity of peripheral blood lymphocytes. Moreover, cross-precipitation studies performed with mAb 4G4 and a CD73 workshop mAb 1E9 showed that these two antibodies recognize the same molecule. Since the tissue distribution and biochemical characteristics of the two molecules are also similar, we conclude that L-VAP-2 and CD73 are the same glycoprotein. Adhesion experiments showed significantly increased binding of freshly isolated lymphocytes to COS cells transfected with a CD73 cDNA, as compared to mock-transfected COS cells, and binding of lymphocytes to CD73-expressing COS cells was inhibited by the presence of mAb 4G4 in the adhesion assay. CD73 is a glycosyl phosphatidylinositol-linked molecule previously shown to have a cosignalling role in T lymphocyte proliferation. Our data suggest that it also has a function in mediating lymphocyte adhesion to the endothelium.

Different forms of human vascular adhesion protein-1 (VAP-1) in blood vessels in vivo and in cultured endothelial cells: implications for lymphocyte-endothelial cell adhesion models

Vascular endothelium plays a pivotal role in controlling leukocyte extravasation from the blood into the tissues. Vascular adhesion protein-1 (VAP-1) is a novel endothelial cell molecule which mediates lymphocyte binding to the vascular lining (Salmi, M., and Jalkanen, S., Science 1992. 257:1407). In this study, we analyzed endothelial cell type-specific differences of VAP-1. In vivo, VAP-1 is a 90/170-kDa molecule which is mainly expressed on the lumenal surface and in cytoplasmic granules of peripheral lymph node-type postcapillary venules (high endothelial venules, HEV). In tonsil HEV, VAP-1 is modified with abundant sialic acids. VAP-1 is also detectable in the cytoplasm of human umbilical vein endothelial cells (HUVEC) and in an endothelial cell hybrid EaHy-926, although both cell types lack detectable surface VAP-1. Cultured endothelial cells do not express MECA-79-defined peripheral lymph node addressins either. VAP-1 was not translocated onto the endothelial cell surface after stimulation with multiple cytokines, mitogens or secretagogues which induced expression of other known endothelial adhesion molecules. Biochemical analyses revealed that VAP-1 is a approximately 180-kDa protein in these endothelial cell types. Digestions with neuraminidase, O-glycanase and N-glycanase, as well as treatment of cells with tunicamycin and benzyl-N-acetylgalactosaminide, did not alter the molecular mass of VAP-1 in EaHy-926. Pulse-chase experiments showed that VAP-1 is directly synthesized as a 180-kDa molecule without any detectable precursors. Thus, in cultured endothelial cells, VAP-1 is a 180-kDa protein which is devoid of post-translational modifications, and in particular, lacks the sialic acids crucial for the function of VAP-1 in tonsil vessels. Notably, the endothelial cell types commonly used as a model in studying lymphocyte-endothelial cell interactions lack surface expression of VAP-1 and peripheral node addressins, and hence are inherently of limited use in analyses of the initial adhesion of lymphocytes.

Usage of beta 1 integrin ligands by B cells is developmentally regulated in avian bursa

Chicken B cell development takes place in a separate organ, the bursa of Fabricius, which provides the blood-borne stem cells, a microenvironment specialized for B cell maturation. Therefore, chicken can be used as a model to study specifically the molecules and interactions which control the development of the B cell compartment. In this work, we studied expression, localization and function of beta 1 integrins on maturing B cells and bursal stroma. The expression of beta 1 integrins on B cells increases during the embryonic development and beta 1 integrin-positive cells can be found both in the medulla and the cortex throughout the bursal development. The binding assays show that the attachment of B cells to stroma is mediated by beta 1 integrins. Binding to 10-day-old embryonic stroma is fibronectin-independent, whereas fibronectin-mediated binding takes place in an increasing manner during further embryonic maturation. After hatching fibronectin appears to be the main binding site for B cells. However, the epitope of beta 1 integrin which takes part in the adhesion of B cells to stroma and to fibronectin is not crucial for the homing of cells into the bursa in a cell transfer model. Our results indicate that the interaction of beta 1 integrins with their ligands is developmentally regulated in the bursa and suggest that B cell maturation may be partially controlled by this interaction.

Sialyl LewisX carbohydrate is expressed differentially during avian lymphoid cell development

Sialyl LewisX is a carbohydrate moiety involved in the regulation of white blood cell adhesion to endothelial cells. In this work, we have studied the expression, localization and function of sialyl LewisX carbohydrate on maturing B and T cells and the stroma of avian bursa, thymus and spleen as well as the role of sialyl LewisX in the generation of immune response and formation of germinal centers in the spleen. The expression of sialyl LewisX on bursal B cells decreases during the embryonic period, while on T cells it remains at a stable low level throughout their thymic development. Immediately after hatching, by which time the expression of sialyl LewisX on bursal B cells has already decreased, the stromal microenvironment of the bursa starts to express it. After hatching, sialyl LewisX is localized exclusively in the medullae of the follicles and the number of sialyl LewisX-positive follicles declines gradually during the following weeks. Thymic stroma does not express sialyl LewisX at any stage of development. In spleen, sialyl LewisX is expressed on a number of B and T cells as well as on other cell types in a way that seems not to be restricted to any structurally defined area. The cells or stroma of the germinal centers, however, do not express sialyl LewisX. When given simultaneously with an immunizing agent, anti-sialyl LewisX monoclonal antibody reduced the formation of germinal centers and suppressed antibody response in young birds. These results demonstrate that sialyl LewisX-bearing cell surface molecules are required during the generation of immune responses. They also suggest a role for sialyl LewisX in B cell maturation.

CD44v6 expression in non-Hodgkin's lymphoma: an association with low histological grade and poor prognosis

The family of CD44 glycoproteins has diverse functions in cell-cell and cell-matrix interactions. The standard form of CD44 is of importance in the dissemination of lymphoma, whereas the clinical significance of the variant exon v6-containing forms of CD44 (CD44v6) is not known. The expression of different forms of CD44 was investigated by using antibodies against the constant part of CD44 (CD44c) and CD44v6 in 56 primary and 17 recurrent non-Hodgkin's lymphomas and correlated with several clinicopathological parameters and with prognosis. Fifty-seven per cent of the primary non-Hodgkin's lymphomas expressed CD44v6 and 73 per cent expressed the constant epitope. Expression of both CD44c and CD44v6 was associated with low histological grade of malignancy. CD44c expression was associated with a low cellular proliferation rate as assessed by DNA flow cytometry. Of several factors tested, high expression of the variant from v6 was the only factor that was associated with unfavourable recurrence-free survival (P = 0.04). We conclude that CD44v6 is associated with a low histological grade, but, on the other hand, with an unfavourable outcome, which suggests that the combination of CD44v6 and histological grading may form a particularly strong prognostic parameter in non-Hodgkin's lymphoma.

Tumor endothelium selectively supports binding of IL-2-propagated tumor-infiltrating lymphocytes

Entrance of activated T cells into the tumor after adoptive transfer is a prerequisite for the efficacy of this form of immunotherapy. Because lymphocyte binding to vascular endothelium is the critical step in which lymphocyte extravasation into the tissue is controlled, we compared adhesion of tumor-infiltrating lymphocytes (TIL) to endothelial cells in tumors, peripheral lymph nodes, mucosa-associated lymphatic tissues, and inflamed synovium. Simultaneously, expression of the known homing-associated Ags both on TIL and tumor vasculature was analyzed. All TIL strongly expressed alpha 4-integrins, LFA-1 and CD44, whereas only a low level of L-selectin expression was detected. Tumor vasculature showed signs of activation in each patient on the basis of elevated levels of intercellular adhesion molecule-1, E-selectin, vascular cell adhesion molecule-1, and/or peripheral lymph node addressin (PNAd). TIL showed significantly enhanced binding to tumor vasculature in comparison with other endothelial specificities. Increased binding was not markedly due to up-regulation of the inflammation-induced endothelial cell adhesion molecules in tumors, because binding to inflamed synovium that expressed the same adhesion molecules was not enhanced. In summary, TIL show preferential binding to tumor vasculature and the binding is partially mediated by currently unknown mechanisms. In vitro analysis of endothelial cell binding properties may help to identify those TIL populations that will have the best potential to home back to tumor tissue after adoptive transfer.

Tumor endothelium selectively supports binding of IL-2-propagated tumor-infiltrating lymphocytes

Entrance of activated T cells into the tumor after adoptive transfer is a prerequisite for the efficacy of this form of immunotherapy. Because lymphocyte binding to vascular endothelium is the critical step in which lymphocyte extravasation into the tissue is controlled, we compared adhesion of tumor-infiltrating lymphocytes (TIL) to endothelial cells in tumors, peripheral lymph nodes, mucosa-associated lymphatic tissues, and inflamed synovium. Simultaneously, expression of the known homing-associated Ags both on TIL and tumor vasculature was analyzed. All TIL strongly expressed alpha 4-integrins, LFA-1 and CD44, whereas only a low level of L-selectin expression was detected. Tumor vasculature showed signs of activation in each patient on the basis of elevated levels of intercellular adhesion molecule-1, E-selectin, vascular cell adhesion molecule-1, and/or peripheral lymph node addressin (PNAd). TIL showed significantly enhanced binding to tumor vasculature in comparison with other endothelial specificities. Increased binding was not markedly due to up-regulation of the inflammation-induced endothelial cell adhesion molecules in tumors, because binding to inflamed synovium that expressed the same adhesion molecules was not enhanced. In summary, TIL show preferential binding to tumor vasculature and the binding is partially mediated by currently unknown mechanisms. In vitro analysis of endothelial cell binding properties may help to identify those TIL populations that will have the best potential to home back to tumor tissue after adoptive transfer.

Dual binding capacity of mucosal immunoblasts to mucosal and synovial endothelium in humans: dissection of the molecular mechanisms

Lymphocytes continuously migrate throughout the body in search of antigens. Virgin lymphocytes recirculate freely between the blood and different lymphatic organs, whereas immunoblasts extravasate preferentially into sites similar to those where they initially responded to antigen. Tissue-specific extravasation of lymphocytes is largely controlled by distinct lymphocyte surface receptors that mediate lymphocyte binding to high endothelial venules (HEV). In the present study, the molecular mechanisms determining the specificity of human mucosal (lamina propria) lymphocyte binding to different endothelial recognition systems were analyzed. Mucosal immunoblasts adhered five times better than small mucosal lymphocytes to mucosal HEV. Importantly, mucosal immunoblasts also bound to synovial HEV almost as efficiently as to mucosal HEV, but they did not adhere to peripheral lymph node HEV. To study the impact of different homing-associated molecules in this dual endothelial binding, we used a gut-derived T cell line and freshly isolated mucosal immunoblasts. Both cell types expressed integrins alpha 4, beta 1, beta 7, and lymphocyte function associated antigen 1 (LFA-1), and were CD44 positive, but practically L-selectin negative. Binding of mucosal immunoblasts to mucosal HEV was almost completely abolished by pretreatment with anti-beta 7 monoclonal antibodies, but it was independent of alpha 4/beta 1 function. In contrast, alpha 4/beta 1 partially mediated immunoblast adherence to synovial HEV, whereas alpha 4/beta 7 had only a minor role in adherence of blasts at this site. CD44 and LFA-1 contributed to HEV-binding both in mucosa and synovium. Taken together, this is the first report that demonstrates a critical role for alpha 4/beta 7 in the binding of gut lymphocytes to mucosal venules in humans. Moreover, a hitherto unknown interaction between mucosal effector cells and synovial endothelial cells was shown to be only partially mediated by the currently known homing receptors. The dual endothelial binding capacity of mucosal blasts may help to explain the pathogenesis of reactive arthritis not uncommonly associated with inflammatory and infectious bowel disease.

Effect of treatment on the prognostic value of S-phase fraction in non-Hodgkin's lymphoma

PURPOSE

To investigate the prognostic value of cell proliferation rate in non-Hodgkin's lymphoma, study its association with histologic classification, and investigate whether its predictive value is influenced by the type of treatment given.

PATIENTS AND METHODS

The S-phase fraction (SPF) size was determined by DNA flow cytometry from paraffin-embedded tissue obtained at diagnosis from 490 patients with non-Hodgkin's lymphoma, diagnosed in a defined geographic area from 1970 to 1991. Clinical data were collected from hospital records and the files of the Finnish Cancer Registry.

RESULTS

SPF size correlated well with histologic grading performed either according to the Working Formulation or Kiel classification (P < .0001 for both). The mean SPFs of low-, intermediate-, and high-grade malignant lymphomas were 4.9% (95% confidence interval [CI], 4.2% to 5.5%), 10.3% (95% CI, 9.3% to 11.4%), and 15.5% (95% CI, 14.0% to 16.9%), respectively. Lymphomas with an SPF lower than the median (7.9%) had a 58% 5-year and 44% 15-year survival rate, whereas those with an SPF larger than the median had a 44% 5-year and 40% 15-year survival rate (P < .0001). SPF size was not significantly associated with prognosis in some subgroups, such as among patients treated primarily with cyclophosphamide, doxorubicin, vincristine, and prednisone (CHOP) (n = 114) or cyclophosphamide, vincristine, and prednisone (COP) (n = 124) with or without radiotherapy (P > .05), whereas a stronger association was found among patients with stage I or II lymphoma treated with radiotherapy only (n = 100; P = .003) and among patients with stage III or IV lymphoma who did not receive chemotherapy (n = 44; P < .0001). In multivariate analyses that included the factors used to construct the International Prognostic Index, SPF had independent prognostic value both in low-grade and intermediate- or high-grade lymphomas, but not in the subset of patients treated with combination chemotherapy with or without radiotherapy.

CONCLUSION

Cell proliferation rate measured as SPF is closely associated with histologic grading in non-Hodgkin's lymphoma, and it has independent prognostic value. The treatment given influences considerably the prognostic value of SPF.

Serum CD44 in malignant lymphoma: an association with treatment response

CD44, a cell surface glycoprotein, is involved in lymphocyte trafficking from the blood to lymphatic tissues, and is of importance in dissemination of lymphoma. A variant form of CD44 that has additional amino acids in the common protein backbone (CD44v6) also seems to play a role in the metastatic dissemination of malignancies. We measured serum CD44 and CD44v6 in 34 patients with lymphoma and in healthy controls by dot blot assay. Small amounts of both CD44 (range, 10 to 80 ng/mL) and CD44v6 could be detected in sera of all controls. Serum CD44 was elevated in all patients with lymphoma before treatment (range, 70 to > 2,000 ng/mL, P < .0001), and CD44v6 was also slightly elevated. Serum CD44 levels correlated with response to treatment. Patients with complete response achieved similar CD44 levels as the controls, whereas those with progressive disease had increased serum CD44 levels. We conclude that both the standard and the variant form of CD44 are detectable in sera of healthy individuals, and that serum CD44 may be useful in monitoring treatment response in patients with lymphoma.

CD44-hyaluronate interaction mediates in vitro lymphocyte binding to the white matter of the central nervous system

The cell adhesion molecule CD44 is expressed in the central nervous system, especially on glial cells in the white matter, the extracellular matrix of which also contains one of its ligands, hyaluronate. We investigated the role of CD44 and hyaluronate in the adhesion of human peripheral blood lymphocytes to myelinated areas of cerebellum by an in vitro binding assay. Hermes-1 epitope, which recognizes the hyaluronate binding site of CD44, and Hermes-3 epitope, involved in lymphocyte binding to mucosal high endothelial venules, were both immunohistochemically expressed in the white matter. No immunoreactivity was observed with mAb Var3.1, which sees variant forms of CD44 containing the exon v6 encoding region. The molecular weight analysis showed that CD44 of the white matter was identical to the major 90 kD form of CD44 present on lymphocytes. The binding of both T and B lymphocytes was significantly inhibited by pretreatment of both cells and sections with mAb Hermes-1 but not with Hermes-3. Digestion of the sections and/or lymphocytes with hyaluronidase also reduced lymphocyte binding. These findings implicate that CD44-hyaluronate mediates lymphocyte adhesion to the white matter and this interaction may be involved in the pathogenesis of inflammations and lymphomas of the central nervous system.

YadA mediates specific binding of enteropathogenic Yersinia enterocolitica to human intestinal submucosa

The binding of live Yersinia enterocolitica to frozen sections of human intestine was investigated qualitatively by monitoring the binding of bacteria by using Gram or immunoperoxidase staining as well as quantitatively by a new enzyme immunoassay-on-slide method. We have demonstrated that the binding of various Y. enterocolitica serotypes and Escherichia coli clones to frozen sections of human intestine is mediated by the Yersinia adhesin, YadA. The YadA-mediated binding occurs mainly at the submucosal layer of the intestinal wall and only to a limited extent at the mucosal layer; there binding is mostly to the mucin threads. In addition, partially purified YadA binds to frozen sections with a pattern similar to that of intact bacteria. Collagen, laminin, or partially purified YadA only partially inhibited the YadA-mediated binding of bacteria, presumably because YadA is multifunctional. A combination of collagen and laminin inhibited the binding more efficiently. Therefore, YadA may be involved in the interactions with the extracellular matrix molecules after the invasion of the intestinal tissue.

Aberrant binding of lamina propria lymphocytes to vascular endothelium in inflammatory bowel diseases

BACKGROUND/AIMS

Lymphocyte recirculation between the blood and lymphoid tissues as well as lymphocyte migration into inflammatory sites are controlled by lymphocyte interaction with vascular endothelium. The aim of this study was to compare the endothelial binding mechanisms of lamina propria lymphocytes (LPL) isolated from normal and inflamed bowel.

METHODS

Binding of LPL from normal and inflamed (inflammatory bowel disease) bowel to endothelium was studied using an in vitro binding assay. Expression of adhesion molecules on LPL and on mucosal endothelial cells was analyzed by flow cytometry and immunohistochemistry.

RESULTS

In inflammatory bowel disease, the selectivity of lymphocyte-endothelial interaction was lost, i.e., immunoblasts bound well not only to mucosal vessels but also to peripheral lymph node venules. The set of homing receptors expressed on mucosal immunoblasts did not directly predict the endothelial binding properties of these cells; e.g., L-selectin-negative blasts adhered well to lymph node vessels using peripheral lymph node-specific endothelial adhesion antigen (PNAd). PNAd was aberrantly expressed in vessels in the inflamed mucosa, where it also mediated LPL binding.

CONCLUSIONS

These findings are important in understanding the physiology of lymphocyte homing into the gut and suggest an important role for PNAd in the pathogenesis of inflammatory bowel disease.

Induction and function of vascular adhesion protein-1 at sites of inflammation

Emigration of leukocytes from the blood into the tissues is critical in controlling lymphocyte patrolling in different lymphatic organs and in leukocyte accumulation at sites of inflammation. During the first stage of the extravasation process, leukocytes bind to the endothelial lining of vessels. At the molecular level, several adhesion molecules on leukocytes and endothelial cells function as receptor-ligand pairs in mediating this dynamic interaction. Recently, we have identified a novel human endothelial cell molecule, vascular adhesion protein 1 (VAP-1), that mediates lymphocyte binding (Salmi, M., and S. Jalkanen. 1992. Science [Wash. DC] 257:1407). VAP-1 was initially characterized by mAb 1B2 which inhibits lymphocyte adhesion to high endothelial venules (HEV) and to purified VAP-1 protein. Here we report the location and function of VAP-1 in normal and inflamed tissues in humans. VAP-1 is abundant in HEV of lymphatic organs belonging to the peripheral lymph node system, but considerably less is expressed in vessels of mucosa-associated lymphatic tissues. A subset of venules in most normal nonlymphatic tissues like skin, brain, kidney, liver, and heart is also VAP-1 positive. In addition to vessels, VAP-1 is distributed on a few other cell types, most notably in dendritic-like cells of germinal centers. At sites of inflammation, such as in inflammatory bowel diseases and chronic dermatoses, expression of VAP-1 is clearly increased. The induced VAP-1 is functional, since mAb 1B2 inhibits lymphocyte binding to inflamed lamina propria venules by approximately 60%. Thus VAP-1 is an endothelial adhesion molecule that under normal conditions is expressed mainly in HEV of lymphatic tissues. However, expression of functional VAP-1 in vivo is upregulated during an inflammatory reaction at other sites as well. Inducibility of VAP-1 suggests that it may play a significant role, not only in recirculation of lymphocytes, but also in controlling entry of leukocytes into sites of inflammation.

Yersinia lipopolysaccharide is modified by human monocytes

Reactive arthritis is usually self-limiting polyarthritis, which develops after certain gastrointestinal or urogenital tract infections, mostly in susceptible HLA B27-positive individuals. In the pathogenesis of this arthritis, it is probably important that structures of the causative bacteria are found in the affected joints. The structure found in the synovial fluid phagocytes of the patients with reactive arthritis after Yersinia, Salmonella, and Shigella infections has always been lipopolysaccharide (LPS) of the causative bacteria. It has been in a highly processed form but still immunoreactive. To follow the degradation process of LPS, we fed peripheral blood monocytes of healthy blood donors with heat-killed Yersinia enterocolitica O:3 bacteria in vitro and monitored the fate of LPS by immunofluorescence and immunoblotting methods. Heat-killed bacteria were used since Y. enterocolitica O:3 bacteria are able to live inside monocytes in vitro and dividing intracellular bacteria would have made it impossible to monitor the degradation process of LPS with these methods. Both the core region and the O-polysaccharide chain of LPS persisted in cytoplasmic vacuoles and on plasma membrane of monocytes through the 7-day follow-up time. Migration properties of processed LPS in sodium dodecyl sulfate-polyacrylamide gel electrophoresis suggested structural modifications of LPS. We also demonstrated that core epitopes appearing on the surface of Yersinia-fed monocytes on day 4 of incubation were processed intracellularly, suggesting that LPS-containing phagocytes are a constant source of membrane-active LPS in their microenvironment as well as in the joints of arthritic patients.

Endothelial cell-binding properties of lymphocytes infiltrated into human diabetic pancreas. Implications for pathogenesis of IDDM

In IDDM, mononuclear cells accumulate in the islets of Langerhans and destroy insulin-producing beta-cells. To study the mechanisms that control extravasation of circulating mononuclear cells into the pancreas, we examined the phenotype of vascular endothelium of the pancreas, propagated a T-cell line from pancreatic islets at the onset of the disease and compared endothelial binding of this cell line in vitro to vascular endothelium in different body regions. The adhesion molecules expressed on the resulting T-cell line and the functional binding capacity of these cells to the endothelium of the normal and diabetic pancreas, mucosa-associated lymphatic tissues, and regional and peripheral lymph nodes were studied. We present evidence of pancreatic endothelial activation in diabetes, leading to endothelial morphology typical for HEVs and accompanying local increase in extravasation of mononuclear cells into the pancreas. Endothelial-cell binding experiments with the T-cell line showed strong adherence of the cells to the endothelium of diabetic pancreas and mucosal lymphoid tissue. The cell line was uniformly CD4-positive, TCR V beta 5.1-positive, LFA-1-positive (CD 11a/CD18), VLA-4 alpha-positive (CD 49d), and CD 44-positive but negative for L-selectin (peripheral lymph node homing receptor). The pancreatic or control cell lines showed no binding to vessels of normal pancreas, and the binding of the pancreatic cell line to the endothelium of peripheral lymph node was weak. Our results suggest that lymphocyte-endothelial cell interactions are important for the accumulation of inflammatory mononuclear cells into the pancreas and imply that lymphocytes derived from the mucosal lymphoid tissue may be involved in the pathogenesis of IDDM.

Lymphocyte-vascular adhesion protein-2 is a novel 70-kDa molecule involved in lymphocyte adhesion to vascular endothelium

Extravasation of leukocytes from the blood is essential for normal lymphocyte recirculation as well as in mounting an adequate inflammatory response in different tissues. Leukocyte migration from the blood is controlled by sophisticated interactions between surface receptors on leukocytes and their corresponding endothelial cell ligands. Here we describe a novel adhesion molecule, lymphocyte-vascular adhesion protein-2 (L-VAP-2), recognized by 4G4 mAb that was produced by immunizing mice with an enriched endothelial cell preparation isolated from inflamed human synovium. mAb 4G4 stains a subpopulation of venules in lymphoid and nonlymphoid tissues as well as a few high endothelial venules in lymphoid tissues. L-VAP-2 is constitutively expressed on human umbilical vein endothelial cells, and its expression cannot be up-regulated by cytokines and mitogens such as TNF-alpha, IL-1, and LPS. The Ag is expressed on approximately 20% of PBL, whereas granulocytes and monocytes are negative. On lymphocytes, L-VAP-2 is preferentially expressed on B cells and CD8+ T cells. The molecular mass of L-VAP-2 is approximately 70 kDa as determined from immunoprecipitated, 125I-labeled endothelial cell lysates. The involvement of L-VAP-2 in lymphocyte binding to endothelium was tested in vitro using human umbilical vein endothelial cells. Both the intact antibody and F(ab')2 fragments of it consistently inhibited lymphocyte binding to human umbilical vein endothelial cells by approximately 25%. On the basis of the molecular mass estimation and the staining of tissue sections, leukocyte populations, and ICAM-1 and ICAM-2 transfectants, L-VAP-2 appears to be a novel Ag involved in the lymphocyte-endothelial cell interaction.

Lymphocyte-vascular adhesion protein-2 is a novel 70-kDa molecule involved in lymphocyte adhesion to vascular endothelium

Extravasation of leukocytes from the blood is essential for normal lymphocyte recirculation as well as in mounting an adequate inflammatory response in different tissues. Leukocyte migration from the blood is controlled by sophisticated interactions between surface receptors on leukocytes and their corresponding endothelial cell ligands. Here we describe a novel adhesion molecule, lymphocyte-vascular adhesion protein-2 (L-VAP-2), recognized by 4G4 mAb that was produced by immunizing mice with an enriched endothelial cell preparation isolated from inflamed human synovium. mAb 4G4 stains a subpopulation of venules in lymphoid and nonlymphoid tissues as well as a few high endothelial venules in lymphoid tissues. L-VAP-2 is constitutively expressed on human umbilical vein endothelial cells, and its expression cannot be up-regulated by cytokines and mitogens such as TNF-alpha, IL-1, and LPS. The Ag is expressed on approximately 20% of PBL, whereas granulocytes and monocytes are negative. On lymphocytes, L-VAP-2 is preferentially expressed on B cells and CD8+ T cells. The molecular mass of L-VAP-2 is approximately 70 kDa as determined from immunoprecipitated, 125I-labeled endothelial cell lysates. The involvement of L-VAP-2 in lymphocyte binding to endothelium was tested in vitro using human umbilical vein endothelial cells. Both the intact antibody and F(ab')2 fragments of it consistently inhibited lymphocyte binding to human umbilical vein endothelial cells by approximately 25%. On the basis of the molecular mass estimation and the staining of tissue sections, leukocyte populations, and ICAM-1 and ICAM-2 transfectants, L-VAP-2 appears to be a novel Ag involved in the lymphocyte-endothelial cell interaction.

Glycoprotein CD44 expression and its association with survival in breast cancer

To investigate the clinical significance of CD44 expression (lymphocyte-homing receptor) in adenocarcinoma, deparaffinized sections from 198 female breast carcinomas were stained with Hermes-3 MoAb for CD44 glycoprotein. In 16% of the cancers most (> or = 90%) of the cancer cells stained positively for CD44, whereas the rest of the cancers were either heterogenous (46%) or negative (38%) in CD44 staining. Cancers with > 50% CD44 positive cells were more often poorly differentiated (grade 3) than those with < or = 50% positive cells (38 vs. 19%, P = 0.006), they had higher mitotic counts (P = 0.04), and were more often estrogen receptor negative (52 vs. 31%, P = 0.01). Among ductal not otherwise specified cancers and node-positive cancers strong CD44 expression was associated with poor outcome (P = 0.05 and 0.02, respectively). However, CD44 expression was not an independent prognostic factor in these subgroups in a multivariate analysis. Unlike in lymphomas the unfavorable prognosis associated with CD44 expression may not be explained by the greater metastatic potential of CD44-positive cells, because the difference in mortality between the groups appeared to diminish with time, and CD44 positivity was associated with aggressive histological features.

Lymphocyte homing receptor (CD44) expression is associated with poor prognosis in gastrointestinal lymphoma

Lymphocyte homing receptor (CD44) is involved in lymphocyte adhesion to endothelial cells of high endothelial venules (HEVs) and lymphocyte exit from the blood circulation, and it may be involved also in hematogenous dissemination of malignant lymphoma. Prognostic significance of lymphocyte homing receptor expression defined by Hermes-3 antibody was studied among 27 gastrointestinal lymphomas followed up for 8 to 20 years after the diagnosis. Lymphomas lacking or with very weak homing receptor expression (n = 14, 52%) were associated with 57% 10-year survival rate as compared with only 15% among lymphomas that expressed CD44 more strongly (P = 0.02). We conclude that lack of lymphocyte homing receptor expression is common in gastrointestinal lymphoma, and that CD44 expression is associated with unfavourable prognosis.

A novel endothelial cell molecule mediating lymphocyte binding in humans

Leukocyte extravasation is critically dependent on proper interactions between leukocytes and vascular endothelial cells. Although several molecules mediating these interactions have been identified, all the homing phenomena observed in vivo cannot be explained on the basis of the known adhesion molecules. To discover novel molecules involved in leukocyte migration we have made monoclonal antibodies against human endothelial cells. With the help of one such monoclonal antibody a novel endothelial cell molecule called vascular adhesion protein-1 (VAP-1) was found. Uniqueness of VAP-1 is evident on the basis of its expression pattern, molecular mass, functional properties and N-terminal amino acid sequence. VAP-1 mediates lymphocyte binding to endothelium in peripheral lymph nodes, tonsil and inflamed synovium. Therefore, VAP-1 is relevant to understanding of the physiologic and pathologic lymphocyte migration in man, and may be valuable for dissecting the molecular events of tissue-selective lymphocyte homing.

Regulated expression of exon v6 containing isoforms of CD44 in man: downregulation during malignant transformation of tumors of squamocellular origin

CD44 is a family of glycoproteins involved in cell-cell and cell-matrix interactions. In addition to the major 90-kD form present on most hematopoietic cells, larger 140-230 kD forms are found on keratinocytes and carcinoma cell lines. These bigger isoforms of CD44 arise by alternative splicing that results in insertion of one or more of the "variant" exons into the extracellular part of the 90-kD constant form of the molecule. In rat, v6 (variant exon v6) containing form of CD44 confers metastatic potential to carcinoma cells, and therefore, it is of interest to study the distribution of this isoform in humans. We raised antibodies against a synthetic peptide containing a sequence encoded by the exon v6. A mAb thus obtained (designated Var3.1) strongly reacted with the plasma membranes of squamous cells in upper layers of skin and tonsil surface epithelia. Weaker staining was seen in germinal centers, vascular endothelia and enterocytes. Exon v6 containing forms of CD44 (CD44v6) were absent from tissue leukocytes and connective tissue components. In comparison, Hermes-3 epitope (on the constant part) containing forms of CD44 were preferentially localized in basal layers of epithelia, present on the surface on most leukocytes and connective tissue cells, and undetectable on the luminal surface of high endothelial venules. In benign neoplasms, epithelial cells stained with mAb Var3.1 like in normal tissues. In contrast, immunostaining of 30 squamous carcinoma specimens (both primary and metastatic lesions) revealed that malignant transformation resulted in downregulation or disappearance of Var3.1 epitope, but in majority of cases, not in diminished synthesis of the Hermes-3 epitope. Biochemical analyses showed that mAb Var3.1 recognized two major forms of CD44 (220 and 300 kD). In conclusion, epitopes on exon v6 and constant part of CD44 are differentially synthesized and regulated during normal and malignant growth of cells in man.

Chicken B cells adhere to the CS-1 site of fibronectin throughout their bursal and postbursal development

In chickens, the pre-B cell development takes place in a specialized organ, the bursa of Fabricius. In this study, we show that fibronectin can be detected in the bursa at least from the day 15 of embryonic incubation up to 10 weeks of age after which the bursa starts to involute. During the embryonic incubation, fibronectin localizes mainly in the areas between the follicles and later on also in the border zone between the cortical and medullary parts of the lymphoid follicles. By adhesion studies with synthetic peptides we show that immature bursal B cells bind in a dose-dependent manner to the CS-1 site of the type III connecting segment at the carboxy-terminal end of fibronectin throughout their bursal development. Postbursal B cells from chicken spleen also display a similar preference of binding to the CS-1 site of fibronectin, whereas peripheral blood B lymphocytes bind to the 40-kDa fragment of fibronectin, but do not adhere as strongly to the CS-1 site. Instead, they recognize to some extent also the heparin-binding YEKPGSPPREVVPRPRPGV peptide. The adhesion of the bursal B cells to the CS-1 site can partially be blocked by preincubation of cells with a synthetic CS-1 peptide and also by preincubation with a monoclonal anti-fibronectin receptor antibody. In contrast, the RGD motif present in several adhesion-associated stromal molecules did not promote any adherence of bursal lymphocytes. The bursal B cells which adhere to the fibronectin molecule in vitro form only a small fraction of the total number of B cells present in the bursa as only less than 4% of the cells remained bound to the surface coated with fibronectin in the adhesion studies. In conclusion, these results suggest that fibronectin may be involved in the differentiation of pre-B cells in the bursa of Fabricius and also in the adherence of postbursal B cells in the spleen.