Upregulation of galectin-7 in murine lymphoma cells is associated with progression toward an aggressive phenotype.

We have previously shown that ICAM-1-deficient mice were resistant to lymphoma dissemination of intravenously injected 164T2 lymphoma cells. Highly aggressive variants of this cell line, however, could overcome this resistance. To discern the complex pattern of gene expression involved in the evolution of aggressiveness in lymphoma cells, we compared the transcriptome of 164T2 cells with that of their aggressive variants using cDNA arrays. We identified several genes that were differentially expressed in nonmetastatic lymphoma cells and their metastatic variants. Galectin-7, associated with the development of chemically induced mammary carcinoma, was one such gene whose expression was significantly upregulated. We showed that it was constitutively expressed in aggressive variants, at both mRNA and protein levels. Galectin-7 expression in aggressive lymphoma cells was induced upon in vivo selection in several organs, including the thymus, the spleen and kidneys. We also showed that treatment of nonaggressive lymphoma cells with 5-aza-2'-deoxycytidine was sufficient to induce galectin-7 gene expression. This report is the first to show that galectin-7 is expressed in aggressive lymphoma.

Leukocyte elastase in murine and human non-Hodgkin lymphomas.

Extracellular proteases play a crucial role in the invasive behavior of normal and transformed leukocytes. Thus far, however, most of the attention has been focused on members of the family of matrix metalloproteinases. In this work, we show that lymphoma cells can express leukocyte elastase (LE) and recruit the enzyme at their surface via ICAM-1. The expression of LE by lymphoma cells was augmented significantly by stimulation with IL-6 and IL-13, both of which also induced the expression of MMP-9. Although LE and IL-13 transcripts were detected in several non-Hodgkin's lymphomas, immunohistochemical analysis of lymphoma tissues also showed that LE was strongly expressed in infiltrating leukocytes. Given the spectrum of key molecules that can be cleaved by LE and that LE and MMP-9 are involved in the invasive behavior of normal or transformed leukocytes, our results raise the hypothesis that LE plays a crucial role in the multistep processes of inflammation and lymphoma metastasis.

UEA-I-binding to thymic medullary epithelial cells selectively reduces numbers of cortical TCRalphabeta+ thymocytes in FTOCs.

Thymic medullary epithelial cells (TMECs) constitute a major stromal cell type, the function of which is incompletely understood. Some TMECs express L-fucose-glycosylated proteins on their plasma membrane; these have been shown to specifically bind the lectin UEA-I. We exploited this observation to investigate the consequences of in situ blockage of TMECs in FTOCs by UEA-I. In UEA-I-treated FTOCs, we noted a decreased cellularity among TCRalphabeta+ but not TCRgammadelta+ cells. In fact, CD3- and CD3lo cortical cells were markedly depleted, while CD3hi cells were unaffected. Since the affected cell subsets are in a different compartment from that where UEA-I binding occurs, it is likely that the effect is mediated through a soluble factor. Two possible mechanisms are proposed: a reduced activation of either TMECs or of medullary thymocytes which normally bind to them, results in lowered production of soluble factors responsible for cortical thymocyte proliferation. Alternately, the binding of UEA-I to TMECs could activate the latter to produce signals inhibitory to cortical thymocytes.

Immunosuppression in mice fed on diets containing beluga whale blubber from the St Lawrence estuary and the Arctic populations.

In order to assess the immunotoxic potential of naturally relevant mixtures of PCBs and other organohalogens, C57Bl/6 mice were fed on diets in which lipids were replaced by blubber of beluga whales from the highly contaminated population of the Saint-Lawrence River, and the less contaminated population from the Arctic. Different ratios of blubber from both sources were mixed in order to allow a dose-response study. Mice were fed for a period of 90 days at the end of which their immunological status was monitored. For general parameters such as body weight, weight of the spleen and the thymus no significant effect of diets were observed. The immunological endpoints such as the blastic transformation of splenocytes and the spleen NK cell activity were not significantly affected by any of the diets compared to control diets. While the different cell subpopulations of peripheral blood and thymus were not affected by the diets, a significant decrease was noted in the CD8+ T cell population in the spleen of mice fed with most of the diets containing beluga blubber. Moreover, the ability of splenic cells to elicit humoral response against sheep red blood cells as well as the potential of peritoneal macrophages to perform phagocytosis were suppressed by all diets containing beluga blubbers. In summary, there was no differences between the groups fed with a blubber diet with low and high organochlorine contamination. However, a clear immunosuppression was demonstrated when these groups were compared to the group fed with beef oil. Despite the fact that we cannot exclude a possible contribution of the fatty acid composition of the beluga blubber to the immunosupression, these results suggest the sensitivity of mouse immune system towards organohalogens, and point out the toxic potential of contaminant mixtures as found in the less contaminated Arctic population.

Resistance of ICAM-1-deficient mice to metastasis overcome by increased aggressiveness of lymphoma cells.

Our recent finding that resistance to lymphoma cell metastasis in intercellular adhesion molecule-1-(ICAM-1)-deficient mice was manifested after homing suggested that the mechanism could involve the capacity of ICAM-1 to induce, via leukocyte function-associated antigen-1 (LFA-1) signaling, the expression of new genes necessary for migration and survival of lymphoma cells after homing. This hypothesis would imply that lymphoma cells, on repeated metastatic cycles, would acquire such a highly aggressive phenotype that they no longer require contact with ICAM-1 at later stages of metastasis. We addressed this question by generating highly aggressive lymphoma variants to determine if increased tumorigenicity would allow lymphoma cells to grow into tumors in ICAM-1-deficient mice. We found that on repeated in vivo passages, a selective pressure favored the lymphoma cells that constitutively express high levels of matrix metalloproteainse-9 (MMP-9), a gene associated with a poor clinical outcome in non-Hodgkins's lymphoma. We further found that although the parent lymphoma cells could not grow tumors in ICAM-1-deficient mice, the aggressive lymphoma variants could. This indicates that, at late stages of the disease, tumor cells with a high metastatic efficiency, encoded by the repertoire of selected genes, no longer require some of the signals normally delivered by cell adhesion molecules. In light of these findings, the possibility of inhibiting dissemination of lymphoma cells at the late stage of the disease by acting against cell adhesion molecules must be reconsidered. (Blood. 2000;95:314-319)

Gelatinase B (MMP-9), but not its inhibitor (TIMP-1), dictates the growth rate of experimental thymic lymphoma.

Dysregulation of metalloproteinase production at tumor sites contributes to the modification of local stromal tissue necessary for tumor development. Gelatinase B (matrix metalloproteinase-9, MMP-9) is one of the key enzymes that have been associated with the progression of several tumors. Paradoxically, MMP-9 expression by tumor cells, most notably by lymphoma cells, is concomitant with the expression of its physiological inhibitor, TIMP-1. Not only are both genes often co-expressed in the most aggressive forms of lymphomas but also both are up-regulated upon contact with stromal cells. Since TIMP-1 is known to regulate growth in several cell types and some aggressive lymphoma cells express TIMP-1 constitutively without MMP-9, it is unclear whether the over-expression of MMP-9 is counterbalanced by TIMP-1 and whether TIMP-1 expression alone could favor the development of lymphoma. To gain further insight into the respective roles of MMP-9 and TIMP-1 in lymphoma, we generated lymphoma cell lines expressing constitutively high levels of MMP-9 or TIMP-1 and compared these cells for the ability to form thymic lymphoma in vivo. Moreover, we generated lymphoma cell lines expressing constitutively high levels of both MMP-9 and TIMP-1 to reproduce the net physiological balance resulting from the expression of both genes simultaneously and to determine which gene overrides the other. Our results show that mice injected with lymphoma cells expressing MMP-9 constitutively developed thymic lymphoma more rapidly than those injected with control lymphoma cells. Over-expression of TIMP-1 alone did not significantly influence tumor progression of lymphoma nor did it delay the capacity of MMP-9 to accelerate the development of thymic lymphoma.

Dissemination of T cell lymphoma to target organs: a post-homing event implicating ICAM-1 and matrix metalloproteinases.

There is increasing evidence that metastasis of a tumor cell (its ability to induce the "development of a tumor" at distant sites following intravasation) is manifested only after homing to distant site(s). All tumor cells, however, do not necessarily undergo uncontrolled cellular division to form secondary tumors once they have "homed" to a target site. One of the major rate-limiting steps in metastasis is in fact related to the ability of the extravasated tumor cells to find an appropriate "nest", where favorable growth conditions will allow them to form a secondary tumor upon massive cell division (1). But to establish such a favorable nest (referred herein as the "nidification" process), tumor cells must penetrate deep into the stroma of the target tissue. This process is facilitated when tumor cells produce of specific proteases, which degrade structural proteins of the extracellular matrix (2,3). The production of proteases by stromal cells can also occur; these enzymes will degrade stroma surrounding the tumor cells, resulting in a massive remodeling of the local parenchyma that may interfere with the vital functions of a target organ as well as help nidification (4). In this review, we focus our attention on post-extravasation events involving adhesion molecules and MMP in the metastatic process of lymphoma cells. We propose that during dissemination of LFA-1-positive lymphoma cells to peripheral organs, the interaction between lymphoma cells and vascular endothelial cells upregulates the local expression of MMP and TIMPs. Since control of lymphoma metastasis appears to occur at the post-extravasation level, we hypothesize that in addition to extravasation, adhesion molecules are implicated in the control of post-extravasation events.

Lack of suppressive effects of mixtures containing low levels of methylmercury (MeHg), polychlorinated dibenzo-p-dioxins (PCDDS), polychlorinated dibenzofurans (PCDFS), and aroclor biphenyls (PCBS) on mixed lymphocyte reaction, phagocytic, and natural killer cell activities of rat leukocytes in vitro.

Rat splenocyte mixed leukocyte reaction (MLR), splenic natural killer (NK) cell activity, and phagocytic activities of splenic, peritoneal, and peripheral blood leukocytes (PBLs) were evaluated in vitro to determine the immunotoxicity of mixtures containing low levels of methylmercury (MeHg), polychlorinated dibenzo-p-dioxins (PCDDs), polychlorinated dibenzofurans (PCDFs), and Aroclor polychlorinated biphenyls (PCBs). The mixtures were based on the concentrations of the chemicals in fish flesh. Leukocytes from male Fischer rats were exposed to MeHg (0.1-2 microg/ml), PCDD/PCDF mixtures (1-15 pg/ml) of three PCDDs (2,3,7,8-tetrachlorodibenzo-p-dioxin, 1,2,3,7,8-pentachlorodibenzo-p-dioxin, and 1,2,3,4,7,8-hexachlorodibenzo-p-dioxin) and two PCDFs (2,3,7,8-tetrachlorodibenzofuran and 1,2,3,7,8-pentachlorodibenzofuran), three Aroclor PCB (Aroclor 1242, 1254, and 1260) mixtures (0.01-0.5 microg/ml), or combinations of MeHg/PCB/PCDD/PCDF mixtures for 24 or 72 h before immunological assays. Phagocytosis and NK cell cytotoxicity were evaluated with a flow cytometer, and MLR of Fischer rat responder splenocytes cultured with mitomycin C-treated Long-Evans splenocytes by [3H]thymidine uptake. Exposure to MeHg (2 microg/ml) alone or with PCB/ PCDD/PCDF resulted in significant cytolethality in rat splenocytes, peritoneal leukocytes, and PBLs at 24 h exposure. Treatment with Aroclor PCB mixtures, PCDD/PCDF mixtures, 0.1 microg MeHg/ml (noncytolethal), or PCB/PCDD/PCDF mixtures with 0.1 microg MeHg/ml caused no suppression of splenocyte MLR response, splenic NK cell-mediated lysis of Yac-l cells, or phagocytosis of fluorescent beads by splenic, peritoneal, and peripheral blood phagocytic cells. The results indicate that in vitro exposure of rat leukocytes to low levels of MeHg, Aroclor PCB mixtures, PCDD/PCDF mixtures, or MeHg/PCB/PCDD/PCDF mixtures had no suppressive effects on the immune functions assayed, and thus produced no additive immunotoxicity. However, in order to predict the potential risk of these chemical mixtures to the human immune system, in vivo animal studies with blood (tissue) levels compatible with the levels of MeHg, PCBs, and PCDDs/PCDFs in exposed human populations should be evaluated.

Protection from lymphoma cell metastasis in ICAM-1 mutant mice: a posthoming event.

It has been hypothesized that the intercellular adhesion receptors used by normal cells could also be operative in the spreading of circulating malignant cells to target organs. In the present work, we show that genetic ablation of the ICAM-1 gene confers resistance to T cell lymphoma metastasis. Following i.v. inoculation of LFA-1-expressing malignant T lymphoma cells, we found that ICAM-1-deficient mice were almost completely resistant to the development of lymphoid malignancy compared with wild-type control mice that developed lymphoid tumors in the kidneys, spleen, and liver. Histologic examinations confirmed that ICAM-1-deficient mice, in contrast to wild-type mice, had no evidence of lymphoid infiltration in these organs. The effect of ICAM-1 on T cell lymphoma metastasis was observed in two distinct strains of ICAM-1-deficient animals. Nonetheless, lymphoma cells migrated with the same efficiency to target organs in both normal and ICAM-1-deficient mice, indicating not only that ICAM-1 expression by the host is essential in lymphoma metastasis, but also that this is so at stages subsequent to homing and extravasation into target organs. These results point to posthoming events as a focus of future investigation on the control of metastasis mediated by ICAM-1.

TL antigen is not linked to radioinduced thymic lymphoma.

X irradiation of C57BL/Ka mice induces thymic lymphoma after a period of 8 to 36 weeks. This latency period represents an ideal time window in which to follow the development of prelymphoma cells that give rise to overt thymic lymphoma. Several attempts have been made to identify an unequivocal prelymphoma cell marker but these efforts have so far been unsuccessful. We monitored the evolution of thymocyte populations containing prelymphoma cells during the latency period, using CD3 and TL as markers, in a transfer assay. We demonstrated that: (1) particular cell populations could appear or disappear; (2) there were at least two prelymphoma phenotypes: CD3loTL+ and CD3hiTL-; (3) TL could be present transiently; and (4) TL could be absent throughout the latency period. We conclude that split-dose irradiation may induce both TL gene expression and a prelymphoma state but that the two are not necessarily related.

The fate of thymocytes labeled in vivo with CFSE.

The fate of thymic emigrants had so far been studied using a variety of markers, each of which had inherent limitations as to stability, toxicity, or selectivity. We describe a new technique which relies on the in vivo injection of CFSE, an esterified vital dye hitherto used at 80 times lower concentrations for in vitro cell labeling. We show that CFSE labels a representative sample of all thymocyte subsets and that these migrate at a rate of approximately 2-3 x 10(6) cells/day to peripheral lymphoid organs. We show that they enter lymph nodes at day 1 postinjection and stay for at least 21 days, whereas the turnover in the spleen is more rapid. We also show by immunohistochemistry, using peroxidase-labeled anti-FITC antibodies, that CFSE-labeled thymic emigrants are confined to T-dependent areas of peripheral lymphoid organs.

Bi-directional induction of matrix metalloproteinase-9 and tissue inhibitor of matrix metalloproteinase-1 during T lymphoma/endothelial cell contact: implication of ICAM-1.

The mechanisms that lead to the expression of matrix metalloproteinases (MMP) and tissue inhibitors of MMP (TIMPs) during the invasive process of normal and transformed T cells remain largely unknown. Since vascular cells form a dynamic tissue capable of responding to local stimuli and activating cells through the expression of cytokine receptors and specific cell adhesion molecules, we hypothesized that the firm adhesion of T lymphoma cells to endothelial cells is a critical event in the local production of MMP and TIMP. In the present work, we show that adhesion of lymphoma cells to endothelial cells induced a transient and reciprocal de novo expression of MMP-9 mRNA and enzymatic activity by both cell types. Up-regulation of MMP-9 in T lymphoma cells was concomitant to that of TIMP-1, and required direct contact with endothelial cells. Induction of MMP-9, but not of TIMP-1, was blocked by anti-LFA-1 and anti-intercellular adhesion molecule-1 Abs, indicating that induction of MMP-9 and TIMP-1 in lymphoma cells required direct, yet distinct, intercellular contact. In contrast, the induction of MMP-9 in endothelial cells by T lymphoma cells did not necessitate direct contact and could be achieved by exposure to IL-1 and TNF, or to the supernatant of T lymphoma cell culture. Together, these results demonstrate that firm adhesion of T lymphoma cells to endothelial cells participates in the production of MMP-9 in both cell types through bi-directional signaling pathways, and identify intercellular adhesion molecule-1/LFA-1 as a key interaction in the up-regulation of MMP-9 in T lymphoma cells.

The metastatic characteristics of murine lymphoma cell lines in vivo are manifested after target organ invasion.

The ability of a tumor cell to survive is critical for successful dissemination to sites distant from the primary tumor. Tumor cells must enter blood circulation, resist hemodynamic shear stress of the blood circulation, successfully extravasate, and then migrate through dense tissue stroma to a site favorable for tumor growth. Some tumor cells must therefore be endowed with peculiar abilities to successfully metastasize, whereas others, although capable of forming tumor in specific organs, cannot metastasize. This property has often been associated with the homing ability of a given tumor cell, likely through the expression of organ-specific homing receptors that are critical for the extravasation process. The present work was aimed at establishing the point at which metastatic and nonmetastatic lymphoma cells diverge. Although 164T2 and 267T2 lymphoma cell lines can successfully form thymic lymphoma when injected intrathymically, only the 164T2 clone can efficiently form tumor in kidneys, spleen, and liver after intravenous inoculation. Using the indium-labeling technique to monitor the homing kinetic of both cell lines, we showed that the critical step for the successful metastasis of the lymphoma cell was determined in the final steps of the disseminating process, namely after homing. These results indicate that, whereas binding of tumor cells to vascular endothelium through specific adhesion mechanisms is a prerequisite for dissemination of tumor cells, the resistance of a tumor cell to the antagonist action of the host and/or its ability to grow tumor occurs only after homing to the target organ.

Gelatinase B (MMP-9) production and expression by stromal cells in the normal and adult thymus and experimental thymic lymphoma.

Correlative and functional evidence support a crucial role for metalloproteinase (MMP) activity in tumor progression. Dysregulation of MMP production at local tumor sites is thought to participate in the remodeling of the local stromal tissue necessary for tumor growth. The extent of damages in local tissues is often reflected by the high concentration of MMP released in the bloodstream of cancer patients. The integrity of the thymic architecture plays a crucial role in the development of mature T cells, but it is compromised by extensive remodeling occurring during the development of thymic lymphomas. In the present work, we have used an experimental thymic lymphoma model to investigate the regulation of MMP-9 (gelatinase B) production in animals bearing large thymic lymphomas. We show a 3-fold increase in serum gelatinase B (Gel B) levels in animals bearing thymic lymphoma compared with those found in normal animals and a correlation between these levels and the size of the tumor. Although Gel B was found within the thymic tumor, lymphoma cells did not express it in vivo, indicating that Gel B expression was associated with thymic stromal cells rather than lymphoma cells. This was corroborated by evidence that lymphoma cells have the capacity to stimulate Gel B gene expression in stromal cells. Our results suggest that lymphoma cells can exert a significant control over Gel B expression by local stromal cells, thereby inducing the extensive remodeling necessary for tumor growth.

Dendritic cells prevent radiation-induced thymic lymphoma.

Thymic lymphomas develop in C57BL/Ka mice within 36 weeks after split-dose X-irradiation. Lymphoma development can be abrogated in such mice by the injection of syngeneic bone marrow from healthy donors. The abrogation mechanism is unknown, but since bone marrow supplies the thymus with precursors of thymocytes and of dendritic cells, we tested the ability of early thymocytes and of immortalized thymic dendritic cells to abrogate lymphomagenesis. Fifteen weeks after irradiation, mice which had received bone marrow or dendritic cells had an equally low incidence of lymphoma, whereas mice which had received thymocytes or which had been only irradiated developed equally high levels of lymphomas, indicating that thymic dendritic cells played a key role in the prevention of lymphoma development. When thymuses from 15-week survivors were tested for pre-lymphoma cells, those from dendritic cell-treated mice proved to be endowed with a level of lymphomagenic potential intermediate between that from bone marrow-treated mice (nonlymphomagenic) and that from untreated or thymocyte-treated mice (highly lymphomagenic). These data indicate that lymphoma abrogation by bone marrow cells involves the participation of marrow-derived thymic dendritic cells.

Modulation of integrin-mediated intercellular adhesion during the interaction of thymocytes with stromal cells expressing VLA-4 and LFA-1 ligands.

Mature peripheral T cells closely regulate their intercellular interactions by modulating integrin adhesion functions. The ability of members of the integrin family to mediate intercellular adhesion is dependent on signals from within the cells (inside-out signaling) that increase the avidity of integrins for their ligands. These changes in avidity are independent of the quantitative changes on the number of receptors, and there is evidence to suggest that phosphorylation events play a predominant role in the regulation of the avidity state of the integrins. Whether such regulatory mechanisms are operative during T cell development had hitherto been an opened question. In the present work, we have used an in vitro adhesion assay between thymocytes and target cells expressing VLA-4 and LFA-1 counter ligands to determine how thymocytes can discriminate between integrin-specific signals during T cell development. Our findings are that VLA-4, but not LFA-1, is constitutively expressed in its high-avidity state during the early stages of T cell development, and that the high-avidity state of thymocytes for VCAM-1-expressing cells is closely regulated by signaling through protein kinase C and protein tyrosine kinase pathways. At later stages of development, mature thymocytes prior to leaving the thymus turn off both VLA-4 and LFA-1 adhesion functions. Our results show that the low-affinity state of integrins on peripheral mature T cells is established before mature thymocytes leave the thymus. Only when mature T cells recognize antigenic peptides in the context of major histocompatibility complex in the periphery will they turn on the adhesion function of VLA-4 and/or LFA-1 integrins.

Expression of protein tyrosine kinases in the murine thymus stroma.

Thymocytes not only receive signals from thymic epithelial cells but can also activate the latter, at least in the medulla. We have previously reported tyrosine phosphorylation of medullary epithelial cell substrates, after co-culture with thymocytes, and identified a number of protein tyrosine kinases in a line of thymic epithelial cells. We report here the in situ localisation by immunohistochemistry of JAK2 in medullary epithelial cells, of PDGF-R in medullary vascular endothelium, of FGF-R in Hassall's corpuscles, and the weak expression of JAK1 and RYK throughout the thymus.

Cloning of a thymic stromal cell capable of protecting thymocytes from apoptosis.

A clone of thymic stromal cells, namely 2BH4, was established by primary culture, cellular transfection and limiting dilution. Morphological analysis by transmission electron microscopy revealed that these cells grow as multilayers, producing a well-defined basement membrane to which they attach and frequently form structures similar to hemidesmosomes. The adjoining cells are connected by intercellular junctions, as tight junctions, intermediate junctions, and desmosome-like junctions, as well as interdigitations. Their cytoplasm contains microtubules, strands of actin filaments, and scarce intermediate filaments. Fluorescence microscopy revealed that 2BH4 cells stain with anti-cytokeratin antibodies, the majority of them giving a faint reaction. In addition, they express Thy-1.1, LFA-1, ICAM-1, and the gp23 epithelial antigen, and synthesize laminin. They have a doubling time of 16 hr and are able to bind thymocytes. Thymocytes cultured in the presence of 2BH4 cells are partially protected from both spontaneous and PMA- or dexamethasone-induced apoptosis. This protection is conferred neither by soluble factors normally produced by the 2BH4 cells nor by the sole contact with fixed 2BH4 cells. Rather, thymocytes must interact with metabolically active 2BH4 cells in order to receive the protective signal(s).

Protein tyrosine kinases transcribed in a murine thymic medullary epithelial cell line.

Medullary epithelial cells of the thymus can be activated by contact with thymocytes. We identified a number of protein tyrosine kinases (PTKs) that could be responsible for the tyrosine phosphorylation observed in a murine thymic medullary epithelial cell line (E-5) following complex formation with thymocytes. Degenerate oligodeoxyribonucleotides (oligos) derived from the amino acid (aa) sequence motifs of PTK catalytic domains were used as oligo primers for PCR amplification to determine the PTK genes which are normally transcribed in the E-5 cell line. Amplicons were cloned, sequenced and the deduced aa sequences were compared to known PTK sequences. Among the 13 distinct PTK catalytic domains identified in E-5 cells, two were novel: they were encoded by eteck, a member of the eph sub-family of PTKs, and thy, a member of the src sub-family of PTKs.