Catherine Fady

Apoptotic vs. Nonapoptotic Cytotoxicity Induced by Hydrogen Peroxide

The regulation of cellular cytotoxicity induced by hydrogen peroxide (H2O2) over a wide concentration range was assessed. Three distinct patterns were detected: the highest concentrations (> 10 mM) rapidly induced a necrotic form of death characterized by smeared patterns of DNA digestion and morphological evidence of primary cytoplasm and plasma membrane damage; In contrast, 10 and 5 mM H2O2 induced endonucleosomal DNA digestion concurrently with cytotoxicity and target cell death was associated with morphologic evidence of apoptosis. Apoptosis was inhibited by cycloheximide, emetine, aminobenzamide (ABA), aurintricarboxylic acid, and calcium depletion. The lowest concentrations of H2O2 (0.5 and 0.1 mM)-induced delayed cytotoxicity (at 24 or 48 hr), which was not associated with DNA ladder formation or morphologic evidence of apoptosis, but was inhibited by ABA. Enforced expression of BCL-2 induced resistance to 0.5 and 0.1 mM H2O2 but had no effect on cytotoxicity induced by 5 and 10 mM. Exposure of isolated nuclei to H2O2 in the absence of calcium or magnesium failed to induce endonucleosomal fragmentation. These data indicate that distinct pathways of H2O2-induced cytotoxicity can be distinguished by their different concentration dependences, and that BCL-2 can protect against some forms of H2O2-induced cytotoxicity.

Evidence against the hypothesis that BCL-2 inhibits apoptosis through an anti-oxidant effect

We contrasted possible protection against apoptosis afforded by either BCL-2 expression or anti-oxidant inhibitors in the same tumor target challenged by two distinct triggers of apoptosis. Exposure of L929 fibroblasts to tumor necrosis factor (TNF) or etoposide (VP-16) induced apoptotic death with similar kinetics. Enforced expression of BCL-2 significantly protected against apoptosis induced by VP-16 but had no effect against TNF-induced apoptosis. In contrast, the anti-oxidants desferrioxamine, butylated hydroxyanisol and N-acetyl cysteine all inhibited TNF-induced apoptosis in a concentration-dependent fashion. Although exposure to VP-16 resulted in a significant generation of intracellular oxyradicals, the above three anti-oxidant inhibitors had no effect on VP-16-induced apoptotic death. Interestingly, enforced expression of BCL-2 also inhibited the ability of VP-16 to generate oxy-radicals and to depress intracellular glutathione levels. These results indicate that BCL-2 can exert anti-oxidant effects but argue against the hypothesis that these effects are critical to its protection against apoptosis.

Effects of β-2 microglobulin anti-sense oligonucleotides on sensitivity of HER2/neu oncogene-expressing and nonexpressing target cells to lymphocyte-mediated lysis

The mechanism by which HER2/neu overexpressing tumor cells resist NK, LAK, and LDCC cytotoxic lymphocytes was investigated. Resistance was not explained by a delay in kinetics of lysis, concurrent resistance to TNF, or a diminished expression of the transferrin receptor. HLA-class I expression, however, was markedly elevated compared to HER2 nonexpressing targets suggesting a reason for resistance. To test the role of class I, we selectively decreased expression by incubation of targets with beta-2 microglobulin anti-sense oligonucleotides. Anti-sense-treated HER2+ targets, displaying levels of class I comparable to HER2- targets, were still markedly resistant to cytotoxic effectors. Down-regulation of class I expression in HER2- carcinoma cells also had no effect on sensitivity to cytotoxicity by anti-sense treatment of Raji and U937 targets resulted in enhanced sensitivity to NK and LAK effectors but not to T cells mediating LDCC. These data indicate resistance to cytotoxicity in HER2-expressing targets cannot be solely explained by heightened expression of class I. The data also support the concept that class I expression regulates sensitivity to NK and LAK cells (but not LDCC effectors) in selected targets.

Interferon-γ-induced increased sensitivity ofHER2/neu-overexpressing tumor cells to lymphokine-activated killer cell lysis: importance of ICAM-1 in binding and post-binding events

Treatment ofHER2/neu-overexpressing target cells with interferon γ(IFNγ) (200–2000 U/ml for 3 days) markedly enhances their sensitivity to lymphokine-activated killer (LAK) cell lysis. Increased sensitivity is associated with an up-regulation of intercellular adhesion molecule ICAM-1 determinants and a down-regulation ofHER2/neu expression. In the present study, we show that exposure to another cytokine, tumor necrosis factor α (200 U/ml for 3 days), also decreasedHER2/neu expression but had no effect on LAK cell lysis and ICAM-1 expression. This suggests that down-regulation of oncogene expression is not sufficient by itself to induce an enhanced sensitivity to LAK cell lysis. IFN-induced enhanced lysis was associated with an increased binding between effectors and targets, and antibodies to ICAM-1 as well as its counter-receptor LFA-1, blocked the increased binding and lysis. Treatment with IFNγ still significantly enhanced lysis even when concanavalin A was added to the assay to induce maximal binding, indicating that a post-binding effect also participated in enhanced cytotoxicity. These post-binding alterations, were also sensitive to blocking with anti-ICAM-1 and anti-LFA-1 antibodies. Treatment with IFN also sensitized targets to lysis by T cells in the presence of lectin but had no effect on the relative resistance of HER2+ cells to lysis mediated by perforin or TNF. Together these data demonstrate the importance of ICAM-1 determinants in binding and post-binding events in the IFN-induced increased lysis ofHER2/neu+ targets.

Resistance of HER2/neu-overexpressing tumor targets to lymphokine-activated-killer-cell-mediated lysis: evidence for deficiency of binding and post-binding events

HER2/neu-overexpressing tumor cell lines are relatively resistant to lymphokine-activated killer (LAK) cell cytotoxicity when compared toHER2/neu-nonexpressing lines.HER2/neu+ targets were also resistant to binding by LAK large granular lymphocytes (LGL) as shown by visualization at the single-cell level, a target monolayer binding assay and in “cold” target inhibition experiments.HER2/neu+ LAK-resistant ovarian cell lines demonstrated an absence of ICAM-1 expression while expression of LFA-3, N-CAM, laminin and β1 integrins was comparable to that ofHER2/neu− targets. In contrast, theHER2/neu+ breast cell line, SKBR-3, which was also resistant to lysis and binding by LAK LGL, demonstrated normal expression of ICAM-1. Anti-ICAM-1 antibodies blocked binding and lysis ofHER2/neu− carcinoma targets by LAK cells, further supporting the notion that lack of ICAM-1 expression onHER2/neu+ cells contributes to their resistance. The modest binding and lysis ofHER2/neu+ targets by LAK cells was significantly inhibited by anti-LFA-1 antibodies, suggesting the existence of another counter-receptor for LFA-1 onHER2/neu+ targets. The following also supported deficiencies in post-binding events whenHER2/neu+ cells resisted the lytic activity of LAK cells: (a) when the relative resistance to effector cell binding was overcome by exogenous lectin,HER2/neu+ cell lines were still resistant to LAK cytolysis, and (b)HER2/neu+ targets were resistant to perforin-containing granule extracts obtained from the CTLL-R8 cytotoxic lymphocyte cell line. These results indicate that deficiency in effector binding as well as post-binding events contributes to the resistance ofHER2/neu-overexpressing tumor targets to LAK-cell-mediated lysis.

Inhibition of Head and Neck Squamous Cell Carcinoma Cell Lines by Transforming Growth Factor-β

Transforming growth factor-beta is known to be a potent autocrine growth inhibitor produced by a wide variety of cells, including cells of the immune system. Other investigators have noted that the growth of nontransformed keratinocytes is inhibited by transforming growth factor-beta, whereas various carcinoma cell lines are resistant to these effects. Head and neck squamous cell carcinoma cells are known to have surface receptors for this cytokine. We thus assessed the effect of transforming growth factor-beta on the growth of head and neck squamous cell carcinoma cell lines. Four head and neck squamous cell carcinoma cell lines were incubated with varying concentrations of transforming growth factor-beta, and cytotoxicity was evaluated with a methylene blue colorimetric assay. After culturing in transforming growth factor-beta for 4 days, inhibition of growth was detected in CAL-27 (maximal inhibition at 5.0 ng/ml), UMSCC-1, and UMSCC-19 (maximal inhibition at 50 ng/ml) cell lines. One other cell line, UMSCC-8 was found resistant to the inhibitory effects of transforming growth factor-beta. Kinetics analysis experiments revealed minimal inhibition before day 2 of incubation, at which time inhibition increased linearly to day 4. Assessment of double-stranded DNA fragmentation suggested that DNA fragmentation occurs before significant cytotoxicity. Electron microscopic analysis and gel electrophoresis of extracted DNA revealed morphologic features consistent with apoptotic cell death. Our findings indicate that transforming growth factor-beta significantly inhibits the growth of head and neck squamous cell carcinoma cell lines by inducing apoptotic cell death.