Autocrine VEGF mediates the antiapoptotic effect of CD154 on CLL cells.

CD154 is an important regulator of chronic lymphocytic leukaemia (CLL)-cell survival. In CLL, high serum levels of VEGF are a feature of advanced disease, and we and others have previously shown that CLL cells produce and secrete this growth factor. Since CD154 stimulates VEGF production in other cell types, and VEGF is known to promote cell survival, we examined whether the cytoprotection of CLL cells by CD154 involves VEGF. We report for the first time that treatment of CLL cells with CD154 results in increased VEGF production and demonstrate involvement of NF-kappaB in this process. Moreover, we show that CD154-induced CLL-cell survival is reduced by anti-VEGF-neutralising antibody and by inhibiting VEGF receptor (VEGFR) signalling with SU5416. However, addition of exogenous VEGF alone or blocking secreted autocrine VEGF had little or no effect on CLL-cell survival. We therefore conclude that CLL-cell cytoprotection in the presence of CD154 requires combined signalling by both CD40 and VEGFR. This combined signalling and resulting cytoprotection were shown to involve NF-kappaB activation and increased survivin production. In conclusion, our findings identify autocrine VEGF as an important mediator of the antiapoptotic effect of CD40 ligation, and thus provide new insights into CLL-cell rescue by CD154 in lymphoreticular tissues.

Involvement of CD44-hyaluronan interaction in malignant cell homing and fibronectin synthesis in hairy cell leukemia.

The tissue homing of malignant hematic cells has both diagnostic and pathogenetic importance. Although such homing is incompletely understood, it generally involves cell adhesion and migration mediated by a number of adhesion receptors and cytokines. In this article, the potential importance of hyaluronan (HA) is examined for the tissue homing of hairy cells (HCs) in hairy cell leukemia (HCL). It is shown that HCs readily adhere to, and spontaneously move on, HA-coated surfaces using CD44. This indicates that activated CD44 and spontaneous movement on HA form part of the intrinsically activated phenotype of HCs. Interleukin-8 (IL-8) inhibited HC movement on HA, and this cell arrest was accompanied by increased actin polymerization and a more pronounced association of CD44 with the cytoskeleton. All of these findings are in sharp contrast to our previous observations with chronic lymphocytic leukemia cells, which are nonmotile on HA, but in response to IL-8 become polarized and motile using the receptor for HA-mediated motility rather than their apparently inactive CD44. Immunohistochemical examination of HCL tissues showed the ubiquitous presence of IL-8 and the prominence of HA in bone marrow stroma and hepatic portal tracts. This suggests that CD44-HA interactions are important in HC homing to these sites, but not to splenic red pulp or hepatic sinusoids, where HA is largely absent. Moreover, engagement of CD44 on HCs stimulates fibronectin synthesis, an observation that is likely to be relevant to the restriction of fibrosis in the disease to HC-infiltrated areas containing HA.

In vitro and in vivo production of vascular endothelial growth factor by chronic lymphocytic leukemia cells.

Expansion of primary solid tumors and their malignant dissemination are angiogenesis-dependent. Vascular endothelial growth factor (VEGF) is the key factor playing a pivotal role in solid tumor-induced angiogenesis. Recent studies indicate that angiogenesis may also be involved in the pathogenesis of certain hemic malignancies, including B-cell chronic lymphocytic leukemia (B-CLL). Mechanisms underlying angiogenesis in B-CLL and the role of VEGF in this process are incompletely understood. In this study, it was examined whether angiogenically functional VEGF is produced by B-CLL cells. Immunohistochemical staining with antibodies against VEGF and CD34, an endothelial cell marker, demonstrated the presence of VEGF protein and abundant blood vessels in infiltrated lymphoreticular tissues. Low levels of VEGF were detected by ELISA in the culture media of unstimulated cells; this was enhanced up to 7-fold by hypoxic stimulation. SDS-PAGE and Western blot analysis of the concentrated culture media showed 2 isoforms of VEGF protein with molecular weights of 28 and 42 kd, respectively. RNA hybridization showed that these cells expressed VEGF mRNA. Reverse transcription-polymerase chain reaction, combined with nucleotide sequence analysis, revealed that the predominantly expressed isoforms were VEGF121 and VEGF165. Moreover, (3)H-thymidine incorporation and an in vivo angiogenic assay demonstrated that the VEGF produced by CLL cells can induce angiogenesis by stimulating endothelial cell proliferation. In conclusion, this study shows that B-CLL cells produce VEGF and demonstrates the angiogenic effects of this growth factor, which may be relevant for the tissue phase of the disease.

The role of hyaluronan and interleukin 8 in the migration of chronic lymphocytic leukemia cells within lymphoreticular tissues.

Malignant lymphocyte migration into and within lymphoreticular tissue is an important aspect of chronic lymphocytic leukemia (CLL), yet little is known about the processes involved. Our previous studies of integrin expression and function in CLL have shown that the abnormal cells are relatively nonadhesive and nonmotile on the protein ligands of these receptors. Here we show that CLL cells adhere to a non-protein ligand, hyaluronan (HA), and become motile (as assessed by both Boyden chamber migration and time-lapse video microscopy) on this ligand when stimulated with interleukin (IL) 8. The combined presence of HA and IL-8 was essential for this motility because IL-8 did not stimulate movement on other surfaces. Blocking antibodies showed that this motility is mediated by the receptor for HA-mediated motility (RHAMM), without the involvement of CD44. Moreover, confocal microscopy showed a polarized distribution of RHAMM and F-actin, but not CD44, in cells which had become motile on HA in the presence of IL-8. Immunohistochemical studies of nodes and spleen demonstrated an abundant reticular network of HA-containing fibers throughout diseased nodes and in splenic white pulp. The splenic red pulp and the luminal surface of high endothelial venules lacked HA. IL-8 was ubiquitously present in these tissues. CLL cells were shown to move spontaneously on fibroblast monolayers derived from lymphoid tissue; this movement was largely blocked by hyaluronidase or anti-RHAMM or anti-IL-8 antibodies. These studies indicate that IL-8-induced motility on HA is likely to be important for CLL cell migration through lymphoid tissue.

Granulocyte-macrophage colony-stimulating factor receptor: stage-specific expression and function on late B cells.

Granulocyte-macrophage colony-stimulating factor (GM-CSF) receptors (GMR) are expressed on myeloid cells throughout their maturational sequence. During myelopoiesis, GM-CSF induces the proliferation of precursors and has multiple effects on more mature cells; such effects include induction of maturation and priming for subsequent stimulation. GMR is expressed on a range of other cell types including acute leukemic blasts of myeloid and lymphoid lineage, but has been little studied on more mature lymphoid cells. Using sensitive triple-layer immunophenotypic techniques, we show here that both the alpha and beta c chains of the GMR are expressed on hairy cells (HCs) and myelomatous plasma cells (PCs), but not on chronic lymphocytic leukemia (CLL) or prolymphocytic leukemia (PLL) lymphocytes. The receptor was demonstrable on normal PCs in tonsil, but not on either activated or resting tonsillar B cells or on circulating normal B lymphocytes. The expression of the receptor is therefore stage specific, rather than a feature of activation. Perhaps, surprisingly, in view of its effects on myeloid cells, GM-CSF did not stimulate the proliferation or differentiation of HCs and did not protect them from apoptosis. However, the cytokine had a profound effect on the interaction of the HC with its environment. Thus, the cytokine caused a major cytoskeletal reorganization resulting in the inhibition of motility and loss of adhesion to cellular and matrix ligands. These studies indicate the importance of GM-CSF outside myelopoiesis and demonstrate a previously unrecognized stage specific role for the cytokine in B-cell biology. Taken together with our previous report that M-CSF enhances B-cell motility, the present findings indicate that myeloid growth factors act in concert to facilitate the controlled migration of certain B cells into and within tissues.

Accessory cells mediate hairy-cell proliferation by mechanism(s) involving both adhesion and TNF alpha secretion.

The mechanisms responsible for hairy-cell (HC) growth both in vitro and in vivo are still unclear. In a recent study we showed that monocytes/macrophages induce HC proliferation in vitro. The purpose of the present paper is to examine the specificity of this accessory cell effect and to establish the mechanism(s) involved. We demonstrate that the effect is not confined to monocytes/macrophages but is also potentially seen with a range of other cell types. However, at low accessory cell:HC ratios (< 1:20) only human umbilical vein endothelial cells (HUVEC) and macrophages induce HC proliferation. We suggest that these observations are of pathophysiological significance in relation to the close association frequently observed between HCs and endothelial cells/macrophages in the liver and spleen of patients with hairy-cell leukaemia (HCL). Regarding the mechanisms of the accessory cell effect, we show that both soluble factors and cell contact are important. A blocking anti-TNF alpha antibody abrogated the HC proliferation induced by HUVEC supernatant, indicating the involvement of this cytokine. Interaction of HCs with HUVEC via CD11b and 11c leucocyte integrins was shown to be important in the contact effect. Our demonstration of the involvement of both cytokines and cell contact in HC proliferation is in accord with what is already known about the control of B-cell growth and differentiation. More specifically, our results suggest that TNF alpha and interaction with endothelial cells/macrophages via leucocyte integrins are involved in the proliferation of late B-cells of the maturational stage represented by HCs.

C-fms protein expression by B-cells, with particular reference to the hairy cells of hairy-cell leukaemia.

Although the hairy cells (HCs) of hairy cell leukaemia (HCL) are now thought to be a form of activated B cell, they have long been known to possess certain monocytoid characteristics. Since the proto-oncogene c-fms is a feature of cells of the monocyte/macrophage lineage, we examined HCs for c-fms expression. We found that approximately 80% of peripheral blood HCs expressed the c-fms protein (8/8 cases). Expression of the 150 kD protein by HCs was shown using three different techniques, APAAP, immunofluorescence and immunoprecipitation, using two different antibodies. Other mature B cell lymphoproliferative disorders examined (PLL, CLL and multiple myeloma) did not express c-fms. We also examined the c-fms expression of normal B-cells: both the in vivo activated (low density) fraction of tonsil B cells and tonsil B cells activated in vitro with SAC plus IL-2 expressed the c-fms protein. As in the case of monocytes c-fms expression by HCs was shown to be down regulated by its ligand M-CSF, and by TNF alpha, both caused a decrease in the receptor expression from 80% to 30% and in the intensity of staining from 6 to 3 x 10(4) molecules/cell. However, as for monocytes, GM-CSF treatment of HCs had no effect on the expression of c-fms; alpha IFN also had no effect. M-CSF treatment of HCs also induced phosphorylation of c-fms, and a number of other proteins, on tyrosine. However, M-CSF was unable to induce HC proliferation either alone or in combination with IL-2, IL-4 or IL-6; in addition it had no effect on HC proliferation induced by SAC, anti-mu or TNF alpha. In addition, M-CSF either alone, or in combination with the above cytokines, had no effect on the differentiated state of HCs as shown by both immunoglobulin secretion and surface antigen expression. M-CSF also had no effect on the morphology or long-term survival of HCs in culture. This study therefore demonstrates that both HCs and activated B-cells express c-fms, and that M-CSF binds to and activates its receptor on HCs. Although c-fms and several other proteins were shown to be phosphorylated in response to M-CSF, the functional consequences of this phosphorylation remain unclear.

Phenotypic changes on hairy cells exposed in vitro to interferons: a quantitative FACS study.

In vitro culture of hairy cells (HC) (five patients) with alpha IFN (100 U/ml) significantly enhanced MHC and CD22 antigen expression and reduced CD25, sIg and FMC7 positivity, together with consistent but not significant reductions in CD9, 19 and HC2. A sixth patient, who was refractory to the effects of alpha IFN in vivo, was also studied and none of these changes were seen in her HC when cultured with alpha IFN in vitro. gamma IFN (100 U/ml) produced a much less marked increase in MHC and reduced sIg. TPA (100 ng/ml) induced changes very similar to those observed with alpha IFN. alpha and gamma IFN had no effect on a range of other antigens in HCL (including CD23 and 38, and transferrin receptors) and also produced no significant antigenic changes in CLL cells. It is concluded that these findings are compatible with partial activation/maturation of HC by alpha IFN. It is suggested that the specificity of the phenotypic changes induced on HC by alpha IFN is a consequence of HC representing a stage of B-cell development at which alpha IFN has specific immunomodulatory effects.