beta-Lapachone-induced apoptosis in human prostate cancer cells: involvement of NQO1/xip3

beta-Lapachone (beta-lap) induces apoptosis in various cancer cells, and its intracellular target has recently been elucidated in breast cancer cells. Here we show that NAD(P)H:quinone oxidoreductase (NQO1/xip3) expression in human prostate cancer cells is a key determinant for apoptosis and lethality after beta-lap exposures. beta-Lap-treated, NQO1-deficient LNCaP cells were significantly more resistant to apoptosis than NQO1-expressing DU-145 or PC-3 cells after drug exposures. Formation of an atypical 60-kDa PARP cleavage fragment in DU-145 or PC-3 cells was observed after 10 microM beta-lap treatment and correlated with apoptosis. In contrast, LNCaP cells required 25 microM beta-lap to induce similar responses. Atypical PARP cleavage in beta-lap-treated cells was not affected by 100 microM zVAD-fmk; however, coadministration of dicoumarol, a specific inhibitor of NQO1, reduced beta-lap-mediated cytotoxicity, apoptosis, and atypical PARP cleavage in NQO1-expressing cells. Dicoumarol did not affect the more beta-lap-resistant LNCaP cells. Stable transfection of LNCaP cells with NQO1 increased their sensitivity to beta-lap, enhancing apoptosis compared to parental LNCaP cells or vector-alone transfectants. Dicoumarol increased survival of beta-lap-treated NQO1-expressing LNCaP transfectants. NQO1 activity, therefore, is a key determinant of beta-lap-mediated apoptosis and cytotoxicity in prostate cancer cells.

Activation of a cysteine protease in MCF-7 and T47D breast cancer cells during beta-lapachone-mediated apoptosis

beta-Lapachone (beta-lap) effectively killed MCF-7 and T47D cell lines via apoptosis in a cell-cycle-independent manner. However, the mechanism by which this compound activated downstream proteolytic execution processes were studied. At low concentrations, beta-lap activated the caspase-mediated pathway, similar to the topoisomerase I poison, topotecan; apoptotic reactions caused by both agents at these doses were inhibited by zVAD-fmk. However at higher doses of beta-lap, a novel non-caspase-mediated "atypical" cleavage of PARP (i.e., an approximately 60-kDa cleavage fragment) was observed. Atypical PARP cleavage directly correlated with apoptosis in MCF-7 cells and was inhibited by the global cysteine protease inhibitors iodoacetamide and N-ethylmaleimide. This cleavage was insensitive to inhibitors of caspases, granzyme B, cathepsins B and L, trypsin, and chymotrypsin-like proteases. The protease responsible appears to be calcium-dependent and the concomitant cleavage of PARP and p53 was consistent with a beta-lap-mediated activation of calpain. beta-Lap exposure also stimulated the cleavage of lamin B, a putative caspase 6 substrate. Reexpression of procaspase-3 into caspase-3-null MCF-7 cells did not affect this atypical PARP proteolytic pathway. These findings demonstrate that beta-lap kills cells through the cell-cycle-independent activation of a noncaspase proteolytic pathway.

NAD(P)H:Quinone oxidoreductase activity is the principal determinant of beta-lapachone cytotoxicity

beta-Lapachone activates a novel apoptotic response in a number of cell lines. We demonstrate that the enzyme NAD(P)H:quinone oxidoreductase (NQO1) substantially enhances the toxicity of beta-lapachone. NQO1 expression directly correlated with sensitivity to a 4-h pulse of beta-lapachone in a panel of breast cancer cell lines, and the NQO1 inhibitor, dicoumarol, significantly protected NQO1-expressing cells from all aspects of beta-lapachone toxicity. Stable transfection of the NQO1-deficient cell line, MDA-MB-468, with an NQO1 expression plasmid increased apoptotic responses and lethality after beta-lapachone exposure. Dicoumarol blocked both the apoptotic responses and lethality. Biochemical studies suggest that reduction of beta-lapachone by NQO1 leads to a futile cycling between the quinone and hydroquinone forms, with a concomitant loss of reduced NAD(P)H. In addition, the activation of a cysteine protease, which has characteristics consistent with the neutral calcium-dependent protease, calpain, is observed after beta-lapachone treatment. This is the first definitive elucidation of an intracellular target for beta-lapachone in tumor cells. NQO1 could be exploited for gene therapy, radiotherapy, and/or chemopreventive interventions, since the enzyme is elevated in a number of tumor types (i.e. breast and lung) and during neoplastic transformation.

Coordinate modulation of Sp1, NF-kappa B, and p53 in confluent human malignant melanoma cells after ionizing radiation

Regulation of transcriptional responses in growth-arrested human cells under conditions that promote potentially lethal damage repair after ionizing radiation (IR) is poorly understood. Sp1/retinoblastoma control protein (RCP) DNA binding increased within 30 min and peaked at 2-4 h after IR (450-600 cGy) in confluent radioresistant human malignant melanoma (U1-Mel) cells. Increased phosphorylation of Sp1 directly corresponded to Sp1/RCP binding and immediate-early gene induction, whereas pRb remained hypophosphorylated. Transfection of U1-Mel cells with the human papillomavirus E7 gene abrogated Sp1/RCP induction and G(0)/G(1) cell cycle checkpoint arrest responses, increased apoptosis and radiosensitivity, and augmented genetic instability (i.e., increased polyploidy cells) after IR. Increased NF-kappaB DNA binding in U1-Mel cells after IR treatment lasted much longer (i.e., >20 h). U1-Mel cells overexpressing dominant-negative IkappaBalpha S32/36A mutant protein were significantly more resistant to IR exposure and retained both G(2)/M and G(0)/G(1) cell cycle checkpoint responses without significant genetic instability (i.e., polyploid cell populations were not observed). Nuclear p53 protein levels and DNA binding activity increased only after high doses of IR (>1200 cGy). Disruption of p53 responses in U1-Mel cells by E6 transfection also abrogated G(0)/G(1) cell cycle checkpoint arrest responses and increased polyploidy after IR, but did not alter radiosensitivity. These data suggest that abrogation of individual components of this coordinate IR-activated transcription factor response may lead to divergent alterations in cell cycle checkpoints, genomic instability, apoptosis, and survival. Such coordinate transcription factor activation in human cancer cells is reminiscent of prokaryotic SOS responses, and further elucidation of these events should shed light on the initial molecular events in the chromosome instability phenotype.-Yang, C.-R., Wilson-Van Patten, C., Planchon, S. M., Wuerzberger-Davis, S. M., Davis, T. W., Cuthill, C., Miyamoto, S., Boothman, D. A. Coordinate modulation of Sp1, NF-kappa B, and p53 in confluent human malignant melanoma cells after ionizing radiation.

Bcl-2 protects against beta-lapachone-mediated caspase 3 activation and apoptosis in human myeloid leukemia (HL-60) cells

We previously demonstrated that beta-lapachone (beta-lap) killed cancer cells solely by apoptosis. Beta-Lap induced apoptosis in HL-60 cells in a dose-dependent manner as measured by flow cytometry and DNA ladder formation. Cell cycle changes, such as accumulations in S and G2-phases, were not observed. Apoptosis was accompanied by activation of caspase 3 and concomitant cleavage of poly(ADP-ribose) polymerase (PARP) to an 89 kDa polypeptide. PARP cleavage was blocked by zDEVD-fmk or zVAD-fmk, caspase-specific cleavage site inhibitors. Retrovirally introduced bcl-2 prevented beta-lap-mediated caspase 3 activation and PARP cleavage and increased the viability of Bcl-2-expressing HL-60 cells compared to cells with vector alone. Various beta-lap-related analogs (e.g., dunnione and naphthoquinone derivatives) induced equivalent apoptosis in HL-60 cells, but no compound was more effective than beta-lap. These data provide further evidence that the primary mode of cell killing by beta-lap is by the initiation and execution of apoptosis in human cancer cells.

Induction of apoptosis in MCF-7:WS8 breast cancer cells by beta-lapachone

Beta-lapachone (beta-lap) affects a number of enzymes in vitro, including type I topoisomerase (Topo I); however, its exact intracellular target(s) and mechanism of cell killing remain unknown. We compared the cytotoxic responses of MCF-7:WS8 (MCF-7) human breast cancer cells after 4-h pulses of beta-lap or camptothecin (CPT), a known Topo I poison. A direct correlation between loss of survival and apoptosis was seen after beta-lap treatment (LD50 = 2.5 microM). A concentration-dependent, transient sub-2 N preapoptotic cell population was observed at 4-8 h. Estrogen deprivation-induced synchronization and bromodeoxyuridine-labeling studies revealed an apoptotic exit point near the G1-S border. Apoptosis activated by beta-lap was closely correlated with cleavage of lamin B but not with increases in p53/p21 or decreases in bcl-2. Loss of hyperphosphorylated forms of the retinoblastoma protein was observed within 5 h, but cyclins A, B1, and E levels were unaltered for up to 72 h after 5 microM beta-lap. Topo I and Topo IIalpha levels decreased at > 24 h. Logarithmic-phase MCF-7 cells were not affected by < or = 1 microM beta-lap. In contrast, dramatic and irreversible G2-M arrest with no apoptosis was observed in MCF-7 cells treated with 1 microM CPT, monitored for 6-10 days posttreatment. MCF-7 cells treated with supralethal doses of CPT (5 microM) resulted in only approximately 20% apoptosis. No correlation between apoptosis and loss of survival was observed. MCF-7 cells exposed to > 5 microM CPT arrested at key cell cycle checkpoints (i.e., G1, S, and G2-M), with little or no movement for 6 days. Ten-fold increases in p53/p21 and 2-5-fold decreases in bcl-2, Topo I, Topo IIalpha, and cyclins A and B1, with no change in cyclin E, were observed. Temporal decreases in bcl-2 and cleavage of lamin B corresponded to the minimal apoptotic response observed. Beta-lap activated apoptosis without inducing p53/p21 or cell cycle arrest responses and killed MCF-7 cells solely by apoptosis. In contrast, concentration-dependent increases in nuclear p53/p21 and various cell cycle checkpoint arrests were seen in MCF-7 cells after CPT. Despite dramatic p53/p21 protein induction responses, CPT-treated MCF-7 cells showed low levels of apoptosis, possibly due to protective cell cycle checkpoints or the lack of specific CPT-activated apoptotic pathways in MCF-7 cells.

Damage-sensing mechanisms in human cells after ionizing radiation

Human cells have evolved several mechanisms for responding to damage created by ionizing radiation. Some of these responses involve the activation or suppression of the transcriptional machinery. Other responses involve the downregulation of enzymes, such as topoisomerase I, which appear to be necessary for DNA repair or apoptosis. Over the past five years, many studies have established links between DNA damage, activation of transcription factors that are coupled to DNA repair mechanisms, increased gene transcription and altered cell cycle regulation to allow for repair or cell death via apoptosis or necrosis. Together these factors determine whether a cell will survive with or without carcinogenic consequences. The immediate responses of human cells to ionizing radiation, in terms of sensing and responding to damage, are therefore, critical determinants of cell survival and carcinogenesis.

Beta-lapachone-mediated apoptosis in human promyelocytic leukemia (HL-60) and human prostate cancer cells: a p53-independent response

beta-Lapachone and certain of its derivatives directly bind and inhibit topoisomerase I (Topo I) DNA unwinding activity and form DNA-Topo I complexes, which are not resolvable by SDS-K+ assays. We show that beta-lapachone can induce apoptosis in certain cells, such as in human promyelocytic leukemia (HL-60) and human prostate cancer (DU-145, PC-3, and LNCaP) cells, as also described by Li et al. (Cancer Res., 55: 0000-0000, 1995). Characteristic 180-200-bp oligonucleosome DNA laddering and fragmented DNA-containing apoptotic cells via flow cytometry and morphological examinations were observed in 4 h in HL-60 cells after a 4-h, > or = 0.5 microM beta-lapachone exposure. HL-60 cells treated with camptothecin or topotecan resulted in greater apoptotic DNA laddering and apoptotic cell populations than comparable equitoxic concentrations of beta-lapachone, although beta-lapachone was a more effective Topo I inhibitor. beta-Lapachone treatment (4 h, 1-5 microM) resulted in a block at G0/G1, with decreases in S and G2/M phases and increases in apoptotic cell populations over time in HL-60 and three separate human prostate cancer (DU-145, PC-3, and LNCaP) cells. Similar treatments with topotecan or camptothecin (4 h, 1-5 microM) resulted in blockage of cells in S and apoptosis. Thus, beta-lapachone causes a block in G0/G1 of the cell cycle and induces apoptosis in cells before, or at early times during, DNA synthesis. These events are p53 independent, since PC-3 and HL-60 cells are null cells, LNCaP are wild-type, and DU-145 contain mutant p53, yet all undergo apoptosis after beta-lapachone treatment. Interestingly, beta-lapachone treatment of p53 wild type-containing prostate cancer cells (i.e., LNCaP) did not result in the induction of nuclear levels of p53 protein, as did camptothecin-treated cells. Like other Topo I inhibitors, beta-lapachone may induce apoptosis by locking Topo I onto DNA, blocking replication fork movement, and inducing apoptosis in a p53-independent fashion. beta-Lapachone and its derivatives, as well as other Topo I inhibitors, have potential clinical utility alone against human leukemia and prostate cancers.

Regulation of intestinal epithelial barrier function by TGF-beta 1. Evidence for its role in abrogating the effect of a T cell cytokine

Maintenance of the integrity of the single-cell-thick intestinal epithelium as an in vivo barrier between environmental Ags and mucosal immunocytes is pivotal for health. The T cell cytokine IFN-gamma consistently disrupts this epithelial barrier in vitro, but the substances in mucosa that may be responsible for sustaining or enhancing barrier function have not been clearly identified. Therefore, we characterized the effect on the epithelial barrier of TGF-beta 1 and three prominent neuropeptides (VIP, substance P, somatostatin) by using a model system in which barrier function of a mature polar human colonic epithelial (T84) cell monolayer is reflected in 1) the electrical potential difference across the apical to basolateral surface of each cell, 2) the transmonolayer permeability to macromolecules such as horseradish peroxidase, and 3) lactate dehydrogenase release into the medium indicating epithelial cell cytolysis. Whereas T84 monolayers exposed to TGF-beta 1 alone demonstrated a modest increase in electrical resistance and barrier integrity, TGF-beta 1 showed a striking ability to reduce the capacity of IFN-gamma to disrupt epithelial barrier function. Characterization studies demonstrated that this TGF-beta 1 effect was prolonged (e.g., days) after a single exposure, progressive over the dose range 0.1 to 2.5 ng/ml, reversible with increased concentrations of IFN-gamma, and more pronounced when TGF-beta 1 exposure was to basolateral rather than to apical epithelial membranes. Macromolecular (horseradish peroxidase) penetration of epithelium was not simultaneously altered by TGF-beta 1 and epithelial cellular injury was minimal as gauged by lactate dehydrogenase release. Additional studies using a human pathogen demonstrated that TGF-beta 1 delayed and decreased the barrier disruption caused by exposure to Cryptosporidium parvum. TGF-beta 1 may be the first of a new class of cytokines that maintains and/or enhances barrier function of human enterocytes, in part by countering the effect of a T cell cytokine.

Regulation of intestinal epithelial barrier function by TGF-beta 1. Evidence for its role in abrogating the effect of a T cell cytokine

Maintenance of the integrity of the single-cell-thick intestinal epithelium as an in vivo barrier between environmental Ags and mucosal immunocytes is pivotal for health. The T cell cytokine IFN-gamma consistently disrupts this epithelial barrier in vitro, but the substances in mucosa that may be responsible for sustaining or enhancing barrier function have not been clearly identified. Therefore, we characterized the effect on the epithelial barrier of TGF-beta 1 and three prominent neuropeptides (VIP, substance P, somatostatin) by using a model system in which barrier function of a mature polar human colonic epithelial (T84) cell monolayer is reflected in 1) the electrical potential difference across the apical to basolateral surface of each cell, 2) the transmonolayer permeability to macromolecules such as horseradish peroxidase, and 3) lactate dehydrogenase release into the medium indicating epithelial cell cytolysis. Whereas T84 monolayers exposed to TGF-beta 1 alone demonstrated a modest increase in electrical resistance and barrier integrity, TGF-beta 1 showed a striking ability to reduce the capacity of IFN-gamma to disrupt epithelial barrier function. Characterization studies demonstrated that this TGF-beta 1 effect was prolonged (e.g., days) after a single exposure, progressive over the dose range 0.1 to 2.5 ng/ml, reversible with increased concentrations of IFN-gamma, and more pronounced when TGF-beta 1 exposure was to basolateral rather than to apical epithelial membranes. Macromolecular (horseradish peroxidase) penetration of epithelium was not simultaneously altered by TGF-beta 1 and epithelial cellular injury was minimal as gauged by lactate dehydrogenase release. Additional studies using a human pathogen demonstrated that TGF-beta 1 delayed and decreased the barrier disruption caused by exposure to Cryptosporidium parvum. TGF-beta 1 may be the first of a new class of cytokines that maintains and/or enhances barrier function of human enterocytes, in part by countering the effect of a T cell cytokine.

IFN-gamma modulation of epithelial barrier function. Time course, reversibility, and site of cytokine binding

The single cell-thick intestinal epithelium forms a crucial barrier between the host and environment, and is modeled in vitro by a monolayer of polarized, highly differentiated T84 epithelial cells impermeable to most macromolecules because of functional intercellular tight junctions. Absence of a permeability defect across the monolayer, either transcellular or paracellular, is indicated by development of a transepithelial electrical resistance of &gt; or = 1000 ohm-cm2, reported to be markedly diminished by exposure to a T lymphocyte cytokine, IFN-gamma. We sought to define this phenomenon in four ways by determining its duration and reversibility; the uniqueness of type II (gamma) IFN as opposed to type I (alpha) IFN; the surface of the polarized columnar epithelium likely involved in responding to IFN-gamma; and whether a specific surface membrane receptor on the epithelial cell participates in the response. Using a special apparatus that allows differential cytokine exposure of monolayer surfaces, our data demonstrate 1) only the monolayer's basolateral surface is IFN-gamma responsive, whereas the apical (microvillous) surface is no; 2) the alteration in electrical resistance of epithelium is prolonged (5 days), even after a single (24 h) exposure to IFN-gamma, but nevertheless is reversible; 3) the effect is likely receptor-ligand mediated, because it can be partially blocked by IFN-gamma receptor-specific monoclonal Ig; 4) an alteration in tight junction function (a paracellular pathway) rather than cell necrosis or a transcellular pathway is responsible for IFN-gamma-induced monolayer dysfunction because permeability to a 44,000-Da macromolecule (horseradish peroxidase) did not increase, and intracytoplasmic T84 cell enzymes were not released into the media; and 5) the biologic phenomenon could not be induced by a species (alpha) of class I IFN, making IFN-gamma reasonably unique in this regard. Given the proximity; activation status, and capacity of T lymphocytes for cytokine production in mucosa, we suggest that IFN-gamma-induced changes in epithelial permeability may be a major cause of altered intestinal barrier function in vivo.

IFN-gamma modulation of epithelial barrier function. Time course, reversibility, and site of cytokine binding

The single cell-thick intestinal epithelium forms a crucial barrier between the host and environment, and is modeled in vitro by a monolayer of polarized, highly differentiated T84 epithelial cells impermeable to most macromolecules because of functional intercellular tight junctions. Absence of a permeability defect across the monolayer, either transcellular or paracellular, is indicated by development of a transepithelial electrical resistance of > or = 1000 ohm-cm2, reported to be markedly diminished by exposure to a T lymphocyte cytokine, IFN-gamma. We sought to define this phenomenon in four ways by determining its duration and reversibility; the uniqueness of type II (gamma) IFN as opposed to type I (alpha) IFN; the surface of the polarized columnar epithelium likely involved in responding to IFN-gamma; and whether a specific surface membrane receptor on the epithelial cell participates in the response. Using a special apparatus that allows differential cytokine exposure of monolayer surfaces, our data demonstrate 1) only the monolayer's basolateral surface is IFN-gamma responsive, whereas the apical (microvillous) surface is no; 2) the alteration in electrical resistance of epithelium is prolonged (5 days), even after a single (24 h) exposure to IFN-gamma, but nevertheless is reversible; 3) the effect is likely receptor-ligand mediated, because it can be partially blocked by IFN-gamma receptor-specific monoclonal Ig; 4) an alteration in tight junction function (a paracellular pathway) rather than cell necrosis or a transcellular pathway is responsible for IFN-gamma-induced monolayer dysfunction because permeability to a 44,000-Da macromolecule (horseradish peroxidase) did not increase, and intracytoplasmic T84 cell enzymes were not released into the media; and 5) the biologic phenomenon could not be induced by a species (alpha) of class I IFN, making IFN-gamma reasonably unique in this regard. Given the proximity; activation status, and capacity of T lymphocytes for cytokine production in mucosa, we suggest that IFN-gamma-induced changes in epithelial permeability may be a major cause of altered intestinal barrier function in vivo.