Differential effects of ifosfamide on dendritic cell-mediated stimulation of T cell interleukin-2 production, natural killer cell cytotoxicity and interferon-gamma production.

Ifosfamide is a DNA-alkylating agent used frequently in chemotherapy of human malignancies. Ifosfamide and its major decomposition products deplete intracellular glutathione (GSH). Glutathione is the major intracellular thiol reductant that protects cells against oxidative injury. Ifosfamide depletion of intracellular GSH in human dendritic cells (DC), T cells and natural killer (NK) cells impairs their functional activity which can be restored by reconstituting GSH. Here we assessed the effect of ifosfamide on DC-mediated stimulation of NK cell proliferation via T cells and on direct DC stimulation of NK cell cytotoxicity and interferon (IFN)-gamma production. Indirect DC stimulation of NK cell proliferation via T cells and T cell-derived interleukin (IL)-2 were reduced by ifosfamide treatment of DC and reconstitution of GSH in DC restored both responses. When DC and NK cells were treated with ifosfamide, DC could overcome the negative effect of ifosfamide on NK cytotoxic function whereas NK cell IFN-gamma production was less efficiently restored. The ability of IL-2 activated NK cells to kill autologous immature DC or to induce DC maturation was reduced moderately by treatment of both cell types with ifosfamide. Overall, our results suggest that DC may stimulate anti-tumour effector cells in patients even if they had received treatment with chemotherapeutic agents such as ifosfamide.

The role of heat shock protein (hsp70) in dendritic cell maturation: hsp70 induces the maturation of immature dendritic cells but reduces DC differentiation from monocyte precursors.

Members of the heat shock protein (hsp70) family are either constitutively expressed (hsc70) or can be induced by hyperthermic stress (hsp70). Recombinant hsp70 (rhsp70) stimulates cytokine production from monocytes and enhances NK cell proliferation and cytotoxicity. Here we demonstrate that rhsp70 binds to immature dendritic cells (DC) derived from monocyte precursors and induces their maturation as evidenced by an increase in CD40, CD86 and CD83 expression. Immature DC stimulated to mature with rhsp70 show an enhanced ability to present tyrosinase peptide to specific CTL. Mature DC did not bind rhsp70, suggesting a down-regulation in the expression of its receptor. When rhsp70 was added to monocyte precursors at the same time as GM-CSF and IL-4 it reduced the differentiation of monocytes into DC as shown by a decrease in the level of CD40, CD83, CD86 and HLA-DR expression and an increase in CD14 expression. The constitutively expressed hsc70 had neither a stimulatory effect on the maturation of immature DC nor did it reduce the differentiation of monocytes into DC. These findings demonstrate the specific ability of rhsp70 to induce the maturation of immature DC. Therefore rhsp70 may be useful for its adjuvant like properties in DC based immunotherapy of certain tumors.

Cytokine regulation of granulocyte-macrophage colony-stimulating factor (GM-CSF) production by human retinal pigment epithelial cells.

GM-CSF is an important regulator of macrophage, granulocyte and dendritic cell behaviour and function. These cell types have been implicated in the retinal damage characteristic of endogenous posterior uveitis. Dendritic cells in the choroid have access to retinal antigens processed by the retinal pigment epithelial (RPE) cells of the blood-retinal barrier and are thought to be candidates for the presentation of antigen in uveoretinitis. We therefore investigated the production of GM-CSF and its regulation in human RPE cells. IL-1beta, tumour necrosis factor-alpha (TNF-alpha) and transforming growth factor-beta (TGF-beta) all stimulated GM-CSF production by RPE cells and a combination of these cytokines increased GM-CSF production over five-fold compared with that with the individual cytokines alone. Interferon-gamma (IFN-gamma) rapidly down-regulated these responses. IFN-gamma did not appear to be acting directly on IL-1beta or via the synthesis of another protein. GM-CSF mRNA expression showed the same pattern of response to these cytokines, indicating transcriptional or pre-transcriptional regulation, and there was no evidence that IFN-gamma was acting by destabilizing GM-CSF mRNA. These results are generally important in understanding the ways in which cytokine regulation differs between different cell types and also more specifically for determining ways in which a cytokine with a significant role in the development of autoimmune uveoretinitis may be manipulated.

Cytokine regulation of RANTES production by human retinal pigment epithelial cells.

The mechanism whereby inflammatory cells gain access to the retina in posterior intraocular inflammatory disease remains unclear. The chemokine RANTES has the potential to influence the migration of memory T cells and monocytes across the blood-retinal barrier during inflammatory eye disease. We have therefore examined the production of RANTES by cultured human retinal pigment epithelial cells (RPE), which form a part of the blood-retinal barrier, in response to cytokines likely to be present in the microenvironment. IL-1 beta and TNF alpha stimulated RANTES production by these cells. IFN-gamma acted synergistically with TNF alpha to increase RANTES production. In contrast, IL-4 downregulated RANTES production stimulated by TNF alpha. RT-PCR studies showed that RANTES mRNA from RPE followed the same pattern of expression in response to cytokines as did RANTES production indicating that RANTES production was controlled at, or prior to, transcription. RANTES is produced in vitro by RPE in response to the proinflammatory cytokines IL-1 beta, TNF alpha, and IFN-gamma and is therefore likely to play a role in the development of the inflammatory eye disease endogenous posterior uveitis.

TGF-beta and IL-1 beta act in synergy to enhance IL-6 and IL-8 mRNA levels and IL-6 production by human retinal pigment epithelial cells.

Cytokines produced by human retinal pigment epithelial (RPE) cells may function as important regulators of intraocular inflammation. We investigated the effect of transforming growth factor-beta (TGF-beta) on interleukin-1 beta (IL-1 beta) induction of IL-6 and IL-8 mRNA levels and protein production by human RPE cells. Both TGF-beta and IL-1 beta alone induce IL-6 mRNA and IL-6 production in human RPE cells and synergize to enhance IL-6 mRNA levels and IL-6 production over a range of TGF-beta (0.1-10 ng/ml) and IL-1 beta concentrations (5-500 U/ml). TGF-beta was also found to enhance IL-1 beta induction of IL-8 mRNA levels at lower IL-1 beta concentrations (50 U/ml) but had no effect at higher IL-1 beta concentrations (500 U/ml). However, TGF-beta had no synergistic effect on IL-1 beta induction of IL-8 secretion. These results suggest that expression of IL-6 and IL-8 in human RPE cells is regulated by different transcriptional and translational mechanisms, and that RPE cells are important regulators of cytokine production within the ocular microenvironment.

Adhesion molecule expression in acute and fibrotic sympathetic ophthalmia.

Samples of iris ciliary body, choroid and retina from normal eyes and from 2 cases of sympathetic ophthalmitis (one acute and one late stage fibrosis) were examined for the expression of the VLA integrins beta 1 and alpha 1-6, and the integrin beta 3, in addition to ICAM-1, VCAM-1, ELAM-1 and CD44 using an APAAP staining technique. The expression of VLA-4, VLA-5, VCAM-1, ICAM-1, and CD44 was significantly increased and ELAM-1 was slightly increased in acute sympathetic ophthalmitis in comparison to fibrotic and normal eyes. VLA-6 was moderately increased in acute and fibrotic cases and VLA-2 VLA-3 and beta 3 were moderately expressed on all tissues examined. The differential expression of molecules known to be involved in lymphocyte activation and adhesion in acute sympathetic ophthalmitis suggests that certain adhesion molecules play a role in the pathogenesis of intraocular inflammation and may be suitable targets for immunotherapy.

Adhesion molecules and malignant gliomas: implications for tumorigenesis.

Adhesion molecules, a family of cell-surface molecules, are likely to be of central importance in mediating cell-extracellular matrix and specific cell-cell interactions within both neoplastic and inflammatory sites. The recently discovered expression of adhesion molecules on glioma cells, tumor-infiltrating lymphocytes, and endothelial cells within the tumor offers insight into the molecular basis of the interactions both between the glioma cell and surrounding heterologous cell types within the tumor environment, and between the tumor cell and the extracellular matrix. Such interactions suggest that these molecules may play roles in the homing of immune cells to these tumors and in regulating the extent of local tumor invasion. The ability to modulate adhesion molecule expression on either immune cells or their respective ligands on gliomas provides an approach to modify cell-cell interactions that may be used to increase tumor kill by the immune system. A similar approach in the modulation of adhesion molecules involved in tumor cell adhesion to the extracellular matrix or endothelial cells may be a method to limit local invasion in these lesions.

Differential expression of the CD44 molecule in human brain tumours.

Expression of the CD44 molecule was examined in a variety of human brain tumours, brain metastases and normal brain. Immunohistological staining with several CD44 antibodies demonstrated differential expression of the CD44 molecule among different brain tumour types. CD44 was strongly expressed in high-grade gliomas and weakly expressed in meningiomas, medulloblastomas and normal brain. Northern blot analysis revealed the presence of 3 major CD44 mRNAs of 1.6, 2.2, and 5.0 kb in glioblastomas and a mRNA of 5.6 kb in meningiomas. CD44 expression was also detected by flow cytometric analysis on cultured cells derived from a variety of human brain tumours including glioblastomas and meningiomas.

Activation and adhesion molecule expression on lymphoid infiltrates in human glioblastomas.

Frozen tissue sections obtained from human glioblastomas, brain tumor metastases and normal brain were examined for the expression of molecules known to be involved in lymphocyte activation and/or adhesion and migration. The molecules studied included CD3, CD45R, UCHL-1 (CD45RO), lymphocyte function-associated antigen 1 (LFA-1) (CD11a, CD18), intercellular adhesion molecule 1 (ICAM-1) (CD54), 4B4 (CD29), CD44, CD2, and LFA-3 (CD58). CD3+ lymphocytes infiltrating human glioblastomas and brain tumor metastases expressed LFA-1 alpha and beta. Many cells were also UCHL-1+ whereas only a small percentage were CD45R+. CD2+ lymphocytes were also present. Tumor-infiltrating lymphocytes (TIL) were found to be negative for CD29, which was, however, expressed on intratumoral vessels in addition to vessels found in normal brain. Glioblastoma cells and intratumoral vessels expressed ICAM-1 whereas no ICAM-1 was found on TIL or on normal brain. Glioblastoma cells also expressed high levels of both CD44 and LFA-3 whereas TIL were negative for these antigens. CD44 was also expressed on certain regions of normal brain. Antibodies to LFA-1 alpha and -beta and ICAM-1 could significantly block the binding of lymphokine-activated killer (LAK) cells or TIL to human glioblastoma cells suggesting that these molecules play a role in the binding and subsequent migration of lymphocytes into brain tumor tissue.

Cytokine regulation of intercellular adhesion molecule-1 (ICAM-1) expression on human glioblastoma cells.

Intercellular adhesion molecule-1 (ICAM-1) has recently been identified as one of the ligands for lymphocyte function-associated antigen-1 (LFA-1). Immunohistochemical staining of frozen tissue sections using the ICAM-1 antibody RR1/1 demonstrated significant levels of ICAM-1 expression on human glioblastoma cells and on intratumoural vascular endothelial cells. ICAM-1 was weakly expressed or absent from low grade gliomas and absent from normal and fetal brain. ICAM-1 expression was similar to that of MHC class II. HLA-DR antigens. Glioblastoma cell lines constitutively expressed ICAM-1 to a minimal or moderate extent. Surface antigen expression of ICAM-1 and ICAM-1-specific mRNA could be significantly increased by incubating glioblastoma cells with interleukin-1 beta (IL-1 beta), tumour necrosis factor-alpha (TNF-alpha), and interferon-gamma (IFN-gamma). IL-2, IL-4, IL-6 and transforming growth factor beta 2 (TGF-beta 2) had no significant effect on surface antigen expression. Significant enhancement of ICAM-1 expression was obtained using TNF-alpha and IL-1 beta at 1-10 U/ml and at 500 U/ml of IFN-gamma. Induction of ICAM-1 specific mRNA was observed 4 h after cytokine treatment and decreased by 24 h. Surface antigen expression of ICAM-1 increased for up to 48 h after treatment.

Influence of PGE2- and cAMP-modulating agents on human glioblastoma cell killing by interleukin-2-activated lymphocytes.

Human glioblastoma cells secrete factors, such as prostaglandin E (PGE) and transforming growth factor beta type 2, which are capable of suppressing several immune functions. The present study investigated the effect of PGE2 and agents known to increase intracellular cyclic adenosine monophosphate (cAMP) levels on 1) the induction of lymphokine-activated killer (LAK) cell activity from the peripheral blood lymphocytes (PBL) of both normal and glioma patients and on 2) the cytolytic activities of tumor-infiltrating lymphocytes (TIL's) isolated from malignant gliomas after expansion in vitro with interleukin-2 (IL-2). Cytolytic activity was measured against autologous and allogeneic tumor cells and the natural killer-resistant Daudi cell line. The results demonstrate that PGE2 and agents known to increase intracellular cAMP levels can significantly suppress the IL-2-dependent generation of cytolytic activity from the PBL of normal and glioma patients and from glioblastoma-derived TIL's. The inhibitory effects of these agents could not be reduced by higher concentrations of IL-2 or by cyclic guanosine monophosphate. Although the suppressive effect of PGE2 was most significant during the early stages of LAK cell generation, an inhibitory effect was still evident when PGE2 was added directly to the cytotoxicity assay. Secretion of PGE2 by glioblastoma cells in vivo may regulate both the generation of an immune response and the effectiveness of adoptively transferred immune cells.

Immunohistochemical and in vitro functional analysis of pineal-germinoma infiltrating lymphocytes: report of a case.

The present study describes the phenotypic and functional analysis of tumor-infiltrating lymphocytes isolated from a germinoma located in the pineal region. Tumor-infiltrating lymphocytes were separated from the germinoma and cultured in medium containing IL-2 (1000 U/ml). An immunohistochemical analysis of frozen sections revealed that 90% of the germinoma-infiltrating lymphocytes were CD3-positive T cells expressing CD4, CD8, and HLA Class I and Class II antigens, but were negative for CD16, CD20, CD23, CD25 and CD14 antigens. After in vitro cultivation in the presence of high concentrations of IL-2, the lymphocytes proliferated for 2 weeks, showing marked DNA synthesis. In addition, the lymphocytes could lyse NK-resistant allogeneic target cells. These results provide evidence for a potential role of germinoma-infiltrating lymphocytes in vivo, and suggest that the lymphocytes may control the growth of autochthonous tumor cells by killing those that are not restricted to the major histocompatibility complex.

Inhibition of lymphocyte function by glioblastoma-derived transforming growth factor beta 2.

Human glioblastoma cells secrete an inhibitory factor termed "glioblastoma-derived T-cell suppressor factor" (G-TsF). A member of the transforming growth factor beta (TGF beta) family, G-TsF is identical to TGF beta 2. The present study investigated the effect of G-TsF/TGF beta 2 on the proliferative and cytotoxic properties of tumor-infiltrating lymphocytes (TIL's) isolated from malignant gliomas after expansion in vitro with interleukin-2 (IL-2). The results demonstrate that the IL-2 (5 to 20 U/ml)-dependent proliferative response of glioma-derived TIL's was inhibited 70% to 85% by G-TsF/TGF beta 2 and that the inhibitory effect could be reduced by using increasing concentrations of IL-2 (100 to 200 U/ml). Tumor necrosis factor alpha (TNF alpha) enhanced the IL-2-dependent proliferation of TIL's cultured in low concentrations of IL-2 (10 U/ml); however, neither TNF alpha nor interferon gamma was able to reduce the inhibitory effect of TGF beta 2 on TIL proliferation. In addition, TGF beta 2 suppressed 60% to 100% the cytotoxic response of glioma-derived TIL's against several tumor targets, including autologous glioma cells, and the suppressive effect was shown to be reduced by increasing concentrations of IL-2.

Antitumor activity and surface phenotypes of human glioma-infiltrating lymphocytes after in vitro expansion in the presence of interleukin 2.

The present study describes a method for in vitro expansion and characterization of antitumor-reactive lymphoid cells isolated from human malignant astrocytomas. Glioma-infiltrating lymphocytes were separated from 24 glioma specimens and cultured in medium containing interleukin 2 (50 to 2000 units/ml). Within 20 to 42 days after the initiation of culture, 20 of 24 cultures of glioma-derived lymphocytes expanded with a substantial increase in cell numbers, of at least 5 x 10(8) cells up to 5 x 10(9), with a simultaneous elimination of contaminating autologous glioma cells. The expanding glioma-derived lymphocytes consisted of 90 +/- 8% (SD) CD3+ T-cells including both CD4+ and CD8+ subpopulations. CD16 was expressed on 4 +/- 5% of the cells and three cultures studied exhibited 14% +/- 1 of Leu-19-positive cells. After 4 to 8 weeks of proliferation, interleukin 2 receptor expression decreased from 36 +/- 28% to less than 10% and the lymphocytes ceased to grow in all cultures. Glioma-derived effector lymphocytes could lyse almost all the autologous tumor targets as well as allogeneic glioma cells. The cytotoxic activity of long-term cultured peripheral blood lymphocytes obtained from the same patients appeared to be similar to that of glioma-derived lymphocytes in killing autologous tumor cells. In summary, glioma-derived lymphocytes expanded in bulk culture with high concentrations of interleukin 2 (2000 units/ml) consisted predominantly of T-lymphoblasts with the ability to kill autologous glioma cells. The tumor-infiltrating lymphocytes could be expanded to sufficient numbers for possible use in the adoptive immunotherapy of malignant gliomas.

Immunohistological and functional analyses of lymphoid infiltrates in human glioblastomas.

An immunohistological analysis of tumor tissue obtained from seven patients with malignant gliomas demonstrated varying levels of lymphoid cell infiltration. The tumor infiltrating lymphocytes obtained from each sample were cultured in vitro by a limiting dilution technique. In three of the cases studied many tumor infiltrating lymphocyte microcultures selectively lysed autologous glioblastoma cells but did not lyse allogeneic gliomas, natural killer-resistant fresh melanoma cells or K562 target cells. These cultures were found to consist of CD 3+ cells. In six cases studied a variable number of microcultures lysed both autologous tumor and K562 target cells only. A minority of the microcultures studied were cytolytic for allogeneic glioma cells and fresh melanoma target cells. The cytolytic activity expressed by tumor infiltrating lymphocytes against autologous tumor cells was significantly greater (P less than 0.001) than that obtained by the corresponding peripheral blood lymphocytes cultured in a similar manner. The present immunohistological and functional studies suggest that there is an immune response to human glioblastomas in vivo with an accumulation of cells with antitumor activity at the tumor site.

Transforming growth factor-beta 2 down-regulates HLA-DR antigen expression on human malignant glioma cells.

Transforming growth factor-beta (TGF-beta) is known to have a potent inhibitory influence on several immune functions. It has recently been demonstrated that TGF-beta 2 is identical to the glioblastoma-derived T cell suppressor factor (G-TsF). In the present study, human malignant glioma cell lines were incubated with various concentrations of TGF-beta 2. An optimal concentration of 1 ng/ml TGF-beta 2 produced a partial but significant decrease of HLA-DR (class II) surface antigen expression on glioma cells expressing this antigen, as well as decreased levels of HLA-DR-specific mRNA. The surface expression of other HLA-related molecules, such as HLA-ABC (class I) and beta 2-microglobulin, was not influenced by TGF-beta 2. The suppressive effect of TGF-beta 2 on HLA-DR expression, both at the surface antigenic and cytoplasmic mRNA levels, could be completely overcome by adding relatively high concentrations (500 U/ml) of interferon (IFN)-gamma to the culture system. However, TGF-beta 2 inhibited the enhancement of HLA-DR surface expression produced by low concentrations of IFN-gamma on some cells which initially did not express these antigens. These results show that TGF-beta 2 can act as a regulator of HLA-DR antigen expression on human glioma cells.

The glioblastoma-derived T-cell suppressor factor/transforming growth factor beta 2 inhibits the generation of lymphokine-activated killer (LAK) cells.

Glioblastoma cells release factors (G-TsF) which inhibit T-cell proliferation. The G-TsF is a novel member of the transforming growth factor beta family and is identical to TGF beta 2. The effect of G-TsF and TGF beta 2 on the induction of LAK cell activity was investigated by culturing PBL obtained from normal blood donors and brain tumour patients in varying concentrations (50-500 U/ml) of interleukin 2 (IL2) alone or IL2 plus G-TsF/TGF beta 2 (1 ng/ml) for 4 days. Subsequent cytolytic activity was measured against autologous and allogeneic glioblastoma targets, fresh NK-resistant melanoma cells and K562 cells. G-TsF/TGF beta 2 purified from glioblastoma cell cultures and TGF beta 2 isolated from porcine platelets significantly suppressed the generation of LAK cell activity, and the inhibitory effect could be reduced by higher concentrations of IL2. The suppressive effect of TGF beta 2 was most significant during the early stages of LAK cell generation and no inhibitory effect was seen when TGF beta 2 was added directly to the cytotoxicity assay. These results suggest that human glioblastomas may exert an inhibitory influence on the generation of an immune response in vivo through the production of G-TsF/TGF beta 2, and that the inhibitory effect may be modified by IL2.