Galectin-1, an endogenous lectin produced by thymic epithelial cells, induces apoptosis of human thymocytes.

Galectin-1, a beta-galactoside binding protein, is produced by thymic epithelial cells and binds to human thymocytes. We have previously reported that galectin-1 induces the apoptosis of activated T lymphocytes. Because the majority of thymocytes die via apoptosis while still within the thymus, we tested whether galectin-1 could induce the apoptosis of these cells. We now report that in vitro exposure to galectin-1 induced apoptosis of two subsets of CD4(lo) CD8(lo) thymocytes. The phenotypes of susceptible thymocytes were consistent with that of both negatively selected and nonselected cells. Galectin-1-induced apoptosis was enhanced by preexposure of thymocytes to antibody to CD3, suggesting that galectin-1 may be a participant in T-cell- receptor mediated apoptosis. In contrast, pretreatment of thymocytes with dexamethasone had no effect on galectin-1 susceptibility. We noted that 71% of the cells undergoing apoptosis after galectin-1 treatment had a DNA content greater than 2N, indicating that proliferating thymocytes were most sensitive to galectin-1. We propose that galectin-1 plays a role in the apoptosis of both negatively selected and nonselected thymocytes, and that the susceptibility of thymocytes to galectin-1 is regulated, in part, by entry or exit from the cell cycle.

Human thymic epithelial cells express an endogenous lectin, galectin-1, which binds to core 2 O-glycans on thymocytes and T lymphoblastoid cells.

Thymic epithelial cells play a crucial role in the selection of developing thymocytes. Thymocyte-epithelial cell interactions involve a number of adhesion molecules, including members of the integrin and immunoglobulin superfamilies. We found that human thymic epithelial cells synthesize an endogenous lectin, galectin-1, which binds to oligosaccharide ligands on the surface of thymocytes and T lymphoblastoid cells. Binding of T lymphoblastoid cells to thymic epithelial cells was inhibited by antibody to galectin-1 on the epithelial cells, and by two antibodies, T305 and 2B11, that recognize carbohydrate epitopes on the T cell surface glycoproteins CD43 and CD45, respectively. T lymphoblastoid cells and thymocytes bound recombinant galectin-1, as demonstrated by flow cytometric analysis, and lectin binding was completely inhibited in the presence of lactose. The degree of galectin-1 binding to thymocytes correlated with the maturation stage of the cells, as immature thymocytes bound more galectin-1 than did mature thymocytes. Preferential binding of galectin-1 to immature thymocytes may result from regulated expression of preferred oligosaccharide ligands on those cells, since we found that the epitope recognized by the T305 antibody, the core 2 O-glycan structure on CD43, was expressed on cortical, but not medullary cells. The level of expression of the UDP-GlcNAc:Gal beta 1,3GalNAc-R beta 1, 6GlcNAc transferase (core 2 beta 1, 6 GlcNAc transferase, or C2GnT), which creates the core 2 O-glycan structure, correlated with the glycosylation change between cortical and medullary cells. Expression of mRNA encoding the C2GnT was high in subcapsular and cortical thymocytes and low in medullary thymocytes, as demonstrated by in situ hybridization. These results suggest that galectin-1 participates in thymocyte-thymic epithelial cell interactions, and that this interaction may be regulated by expression of relevant oligosaccharide ligands on the thymocyte cell surface.

Galectins: versatile modulators of cell adhesion, cell proliferation, and cell death.

Lectins, or carbohydrate binding proteins, recognize specific oligosaccharide structures on glycoproteins and glycolipids. Several families of animal lectins have been identified; for some of these lectins, functions such as leukocyte adhesion and microbial opsonization have been described. The galectins are a family of lectins found in species ranging from sponges and nematodes to humans. Members of the galectin family have been proposed to mediate cell adhesion, to regulate cell growth, and to trigger or inhibit apoptosis. The expression pattern of different galectins changes during development, and this pattern is also altered at sites of inflammation and in breast, colon, prostate, and thyroid carcinomas. In addition, the level of expression of some galectins by tumor cells has been shown to be correlated with metastatic potential. The mechanisms by which galectins exert these diverse effects remain largely unknown. Some glycoprotein counterreceptors recognized by certain galectins have been identified; this is an important first step in understanding the cell-type specific effects of different galectins. This review discusses the way in which the modulation of galectin activity may affect strategies for treatment of a variety of human diseases, including autoimmunity and cancer.

The in vitro senescence of human T lymphocytes: failure to divide is not associated with a loss of cytolytic activity or memory T cell phenotype.

Normal human T lymphocytes, activated in vitro and cultured in the continuous presence of the growth factor interleukin 2 (IL2), have a limited proliferative potential. Senescent T cell cultures will not proliferate, even if restimulated by the original allogeneic stimulator cells. However, we have now observed that such restimulation induces an increase in the percentage of cells expressing the 55 kDa chain of the IL2 receptor (IL2R alpha, CD25) without any associated increase in cell number. A younger culture, which showed a comparable increase in CD25, underwent two population doublings in the same time period after restimulation. The senescent cultures, (primarily of the CD8+, cytotoxic/suppressor, phenotype), were also found to be highly potent and specific effector cells in a 51chromium release assay for cytolytic activity. Furthermore, senescent cultures maintain the surface phenotype of memory T cells. These findings demonstrate that while senescent T cells are unable to proliferate in response to restimulation or to IL2, they are able to recognize the foreign stimulator cells and to initiate an otherwise normal T cell response. Our results lend support to the hypothesis that in vitro senescence is not associated with a generalized decline in functional activity in a differentiated cell type, but with a specific event which limits cell division. Thus, the long term T lymphocyte culture system will be useful for studying the mechanism by which proliferation is blocked in these, apparently, post-mitotic cells.

Human T lymphocytes possess a limited in vitro life span.

The T lymphocyte offers certain theoretical advantages over other available cell types for the study of aging. Immunosenescence is a well-established part of, and may be directly relevant to, mammalian aging, and the T lymphocyte is well-characterized as to function, cell-surface antigen make-up, and other factors. However, prior efforts at studying in vitro aging of T cells have been hampered by poor reproducibility in doubling potential and the occurrence of a peculiar type of crisis. We have improved the culture conditions for long-term in vitro propagation of normal human T lymphocytes so that previously described variability between identically manipulated cultures and the crisis period have been eliminated. Analysis of the growth patterns of 109 individual cultures revealed a limited proliferative life span, with the number of cumulative population doublings corresponding to that reported for adult human fibroblasts. This accord between the in vitro life spans of two vastly different cell types lends further support to the concept of the Hayflick Limit as a general biological phenomenon.

Alpha-actin synthesis changes in cultured cardiac myocytes: relationship to anthracycline structure.

Four anthracyclines (AAs) were examined comparatively to determine effects on actin isoform synthesis in vitro. Cultured cardiac myocytes (CMCs) were incubated for 24 hours with 35 microCi sulfur 35-labeled methionine and 10(-10) to 10(-5) mol/L doxorubicin hydrochloride (Adriamycin) (ADR), daunomycin (DM), 5-iminodaunorubicin IDR), or 3'-deamino-3'-(3-cyano-4-morpholinyl)doxorubicin (MRA-CN). CMCs were harvested in buffered Triton X-100 and homogenized. Proteins in the extracts were fractionated by centrifugation. Equal protein quantities were subjected to sodium dodecyl sulfate-polyacrylamide gel electrophoresis, to two-dimensional electrophoresis, and to autoradiography. AAs caused a dose-dependent decrease in radiolabeling of CMC proteins in Triton-soluble and -insoluble fractions of the extracts. ADR and DM (10(-6) mol/L each) and IDR (10(-5) mol/L) decreased radiolabeling of CMC polypeptides including alpha-actin. In polypeptides extracted in the Triton X-100-soluble pool, the effect on CMC alpha-actin synthesis was greater than the effect on beta- or gamma-actin. CMC alpha-actin synthesis was susceptible to ADR and to DM in a dose-dependent fashion. Contrastingly, alpha-actin radiolabeling was not altered by exposing CMCs to 10(-5) mol/L MRA-CN. Decreased sarcomeric actin isoform synthesis in vitro may reflect forms of subcellular damage in this model of anthracycline cardiomyopathy.

Apoptosis of T cells mediated by galectin-1.

Galectin-1, a member of the family of beta-galactoside binding proteins, has growth regulatory and immunomodulatory activities. We report here that galectin-1, expressed by stromal cells in human thymus and lymph nodes, is present at sites of cell death by apoptosis during normal T-cell development and maturation. Galectin-1 induced apoptosis of activated human T cells and human T leukaemia cell lines. Resting T cells also bound galectin-1, but did not undergo apoptosis. Human endothelial cells that expressed galectin-1 induced apoptosis of bound T cells. Galectin-1-induced apoptosis required expression of CD45, and was decreased when N-glycan elongation was blocked by treatment of the cells by swainsonine, whereas inhibition of O-glycan elongation potentiated the apoptotic effect of galectin-1. Induction of apoptosis by an endogenous mammalian lectin represents a new mechanism for regulating the immune response.

Effects of the specific cAMP antagonist, (Rp)-adenosine cyclic 3',5'-phosphorothioate, on the cAMP-dependent protein kinase-induced activity of hepatic glycogen phosphorylase and glycogen synthase.

The cAMP-dependent protein kinase-induced effects on phosphorylase and glycogen synthase activities and glucose production were studied in hepatocytes isolated from fed rats in the presence of the diastereomers of adenosine cyclic 3',5'-phosphorothioate, (Sp)-cAMPS and (Rp)-cAMPS. Incubation of hepatocytes with (Sp)-cAMPS or glucagon, both of which lead to cAMP-dependent protein kinase activation, resulted in a concentration-dependent increase in glycogen phosphorylase activity and a decrease in glycogen synthase activity. Incubation of hepatocytes with the cAMP-dependent protein kinase antagonist, (Rp)-cAMPS, in the absence of an agonist, had no significant effect on phosphorylase or glycogen synthase activities. Incubation of hepatocytes with a half-maximally inhibitory concentration of (Rp)-cAMPS shifted the agonist-induced activation curves for phosphorylase and the agonist-induced inhibition curves for glycogen synthase to 5-fold higher concentrations for both (Sp)-cAMPS and glucagon. Phosphorylase activity was very sensitive to the rapid, concentration-dependent inhibition by (Rp)-cAMPS of agonist-induced activation of cAMP-dependent protein kinase. The effects on phosphorylase activity were observable in 30 s and were concentration-dependent with half-maximal inhibition at 10 microM, similar to that observed for cAMP-dependent protein kinase. In contrast, glycogen synthase activity was less sensitive to (Rp)-cAMPS inhibition of agonist-induced activation of cAMP-dependent protein kinase. The effects on glycogen synthase activity lagged behind those on phosphorylase activity and the concentration dependence did not parallel the cAMP-dependent protein kinase effect, but was shifted to higher concentrations of (Rp)-cAMPS with half-maximal inhibition at 60 microM. Glucose (10 to 40 mM) increased the sensitivity of glycogen synthase to (Rp)-cAMPS inhibition of cAMP-dependent protein kinase over a narrow range of agonist concentration, but had no significant effect throughout most of the agonist-induced activation range. Thus, the diastereomers, (Sp)- and (Rp)-cAMPS, influence glycogen metabolism and the glycogenolytic enzymes through their modulation of cAMP-dependent protein kinase levels.

In vitro cellular aging in T-lymphocyte cultures: analysis of DNA content and cell size.

Like other normal diploid mammalian cells, human T-lymphocytes display a limited in vitro lifespan. Long-term cultures of normal adult peripheral blood T cells, activated in vitro and passaged in the presence of interleukin-2 undergo 23 +/- 7 cumulative population doublings. Cell cycle analysis revealed that in senescent T cell cultures, restimulation with the original antigen resulted in 16-22% of the cells entering S phase, compared to 60% entering cycle in young cultures. In addition, within 1 week of restimulation, the senescent cultures do not increase in cell number, but 93% of the cells return to the G1/G0 DNA content, a proportion typically seen in quiescent (nonrestimulated) cultures. Coculture of early- and late-passage cells at two different ratios excluded a putative inhibitory factor produced by the old cells and similarly eliminated a possible stimulatory product in early cultures. Flow cytometry measure of forward angle light scatter revealed no difference in cell size between early- and late-passage cells, in contrast to the findings with senescent fibroblasts. Thus, while increasing cell size may contribute to the senescent phenotype of fibroblast cultures, it is not a factor in the senescence of human T-lymphocytes, and it is therefore doubtful that alterations in cell size are fundamental to in vitro cellular aging.

Characterization of terminal sialic acid linkages on human thymocytes. Correlation between lectin-binding phenotype and sialyltransferase expression.

T cell surface sialylation changes during maturation in the thymus. We have previously demonstrated increased expression of mRNA encoding the Gal beta 1, 3GalNAc alpha 2,3-sialyltransferase in mature medullary human thymocytes, compared with immature cortical thymocytes. For this enzyme, increased expression of transferase mRNA correlated with increased sialylation of O-glycans. We have now examined the pattern of expression in the human thymus of two additional sialyltransferases, the Gal beta 1,4GlcNAc alpha 2,6-sialyltransferase (ST6N) and the Gal beta 1,3/4GlcNAc alpha 2,3-sialyltransferase (ST3N). The patterns of mRNA expression were compared with the pattern of binding of two sialic acid-specific plant lectins, Sambucus nigra agglutinin and Maackia amurensis agglutinin, which preferentially recognize alpha 2,6- and alpha 2,3-linked sialic acids, respectively, on N-glycans. By in situ hybridization, mRNA encoding ST3N was detected uniformly throughout the thymus. All thymocytes bound M. amurensis agglutinin, demonstrating a direct correlation between the level of ST3N mRNA expression and cell-surface glycosylation. In contrast, mRNA encoding ST6N was also expressed uniformly throughout the thymus; however, only mature (CD3hi) medullary thymocytes bound S. nigra agglutinin. On mature thymocytes, S. nigra agglutinin appeared to bind primarily to the cell-surface glycoprotein CD45; since only the mature thymocytes expressed the CD45RA isoform, while both mature and immature populations expressed the CD45R0 isoform, CD45RA may be a preferred substrate for ST6N. These results demonstrate that glycoprotein sialylation is tightly regulated during T cell development and that the developmentally regulated expression of specific oligosaccharide structures on the cell surface may be influenced by expression of both the relevant glycosyltransferase and specific acceptor substrates.