ICAM1+ neutrophils promote chronic inflammation via ASPRV1 in B cell-dependent autoimmune encephalomyelitis.

Neutrophils contribute to demyelinating autoimmune diseases, yet their phenotype and functions have been elusive to date. Here, we demonstrate that ICAM1 surface expression distinguishes extra- from intravascular neutrophils in the mouse CNS during experimental autoimmune encephalomyelitis (EAE). Transcriptomic analysis of these 2 subpopulations indicated that neutrophils, once extravasated, acquire macrophage-like properties, including the potential for immunostimulation and MHC class II-mediated antigen presentation. In corroboration, super-resolution (3D stimulated emission-depletion [STED]) microscopy revealed neutrophils forming synapses with T and B cells in situ. Further, neutrophils specifically express the aspartic retroviral-like protease ASPRV1, which increases in the CNS during EAE and severe cases of multiple sclerosis. Without ASPRV1, mice immunized with a new B cell-dependent myelin antigen (but not with the traditional myelin oligodendrocyte glycoprotein peptide) develop a chronic phase of EAE that is less severe and even completely fades in many individuals. Therefore, ICAM1+ macrophage-like neutrophils can play both shared and nonredundant roles in autoimmune demyelination, among them perpetuating inflammation via ASPRV1.

Interleukin-36γ is expressed by neutrophils and can activate microglia, but has no role in experimental autoimmune encephalomyelitis.

BACKGROUND: Experimental autoimmune encephalomyelitis (EAE) is a model of inflammatory demyelinating diseases mediated by different types of leukocytes. How these cells communicate with each other to orchestrate autoimmune attacks is not fully understood, especially in the case of neutrophils, whose importance in EAE is newly established. The present study aimed to determine the expression pattern and role of different components of the IL-36 signaling pathway (IL-36α, IL-36ß, IL-36γ, IL-36R) in EAE. METHODS: EAE was induced by either active immunization with myelin peptide, passive transfer of myelin-reactive T cells or injection of pertussis toxin to transgenic 2D2 mice. The molecules of interest were analyzed using a combination of techniques, including quantitative real-time PCR (qRT-PCR), flow cytometry, Western blotting, in situ hybridization, and immunohistochemistry. Microglial cultures were treated with recombinant IL-36γ and analyzed using DNA microarrays. Different mouse strains were subjected to clinical evaluation and flow cytometric analysis in order to compare their susceptibility to EAE. RESULTS: Our observations indicate that both IL-36γ and IL-36R are strongly upregulated in nervous and hematopoietic tissues in different forms of EAE. IL-36γ is specifically expressed by neutrophils, while IL-36R is expressed by different immune cells, including microglia and other myeloid cells. In culture, microglia respond to recombinant IL-36γ by expressing molecules involved in neutrophil recruitment, such as Csf3, IL-1ß, and Cxcl2. However, mice deficient in either IL-36γ or IL-36R develop similar clinical and histopathological signs of EAE compared to wild-type controls. CONCLUSION: This study identifies IL-36γ as a neutrophil-related cytokine that can potentially activate microglia, but that is only correlative and not contributory in EAE.

A novel population of local pericyte precursor cells in tumor stroma that require Notch signaling for differentiation.

Pericytes are perivascular support cells, the origin of which in tumor tissue is not clear. Recently, we identified a Tie1(+) precursor cell that differentiates into vascular smooth muscle, in a Notch-dependent manner. To understand the involvement of Notch in the ontogeny of tumor pericytes we used a novel flow immunophenotyping strategy to define CD146(+)/CD45(-)/CD31(-/lo) pericytes in the tumor stroma. This strategy combined with ex vivo co-culture experiments identified a novel pericyte progenitor cell population defined as Sca1(hi)/CD146(-)/CD45(-)/CD31(-). The differentiation of these progenitor cells was stimulated by co-culture with endothelial cells. Overexpression of the Notch ligand Jagged1 in endothelial cells further stimulated the differentiation of Sca1(hi)/CD146(-)/CD45(-)/CD31(-) cells into pericytes, while inhibition of Notch signaling with a γ-secretase inhibitor reduced this differentiation. However, Notch inhibition specifically in Tie1-expressing cells did not change the abundance of pericytes in tumors, suggesting that the pericyte precursor is distinct from the vascular smooth muscle cell precursor. Transplant experiments showed that the bone marrow contributes minimally to tumor pericytes. Immunophenotyping revealed that Sca1(hi)/CD146(-)/CD45(-)/CD31(-) cells have greater potential to differentiate into pericytes and have increased expression of classic mesenchymal stem cell markers (CD13, CD44, Nt5e and Thy-1) compared to Sca1(-/lo)/CD146(-)/CD45(-)/CD31(-) cells. Our results suggest that a local Sca1(hi)/CD146(-)/CD45(-)/CD31(-) pericyte progenitor resides in the tumor microenvironment and requires Notch signaling for differentiation into mature pericytes.

Endothelial-specific Notch blockade inhibits vascular function and tumor growth through an eNOS-dependent mechanism.

Notch signaling is important for tumor angiogenesis induced by vascular endothelial growth factor A. Blockade of the Notch ligand Dll4 inhibits tumor growth in a paradoxical way. Dll4 inhibition increases endothelial cell sprouting, but vessels show reduced perfusion. The reason for this lack of perfusion is not currently understood. Here we report that inhibition of Notch signaling in endothelial cell using an inducible binary transgenic system limits VEGFA-driven tumor growth and causes endothelial dysfunction. Neither excessive endothelial cell sprouting nor defects of pericyte abundance accompanied the inhibition of tumor growth and functional vasculature. However, biochemical and functional analysis revealed that endothelial nitric oxide production is decreased by Notch inhibition. Treatment with the soluble guanylate cyclase activator BAY41-2272, a vasorelaxing agent that acts downstream of endothelial nitric oxide synthase (eNOS) by directly activating its soluble guanylyl cyclase receptor, rescued blood vessel function and tumor growth. We show that reduction in nitric oxide signaling is an early alteration induced by Notch inhibition and suggest that lack of functional vessels observed with Notch inhibition is secondary to inhibition of nitric oxide signaling. Coculture and tumor growth assays reveal that Notch-mediated nitric oxide production in endothelial cell requires VEGFA signaling. Together, our data support that eNOS inhibition is responsible for the tumor growth and vascular function defects induced by endothelial Notch inhibition. This study uncovers a novel mechanism of nitric oxide production in endothelial cells in tumors, with implications for understanding the peculiar character of tumor blood vessels.

Notch-dependent regulation of the ischemic vasodilatory response--brief report.

OBJECTIVE: We have recently described that Notch activates nitric oxide (NO) signaling in the embryonic endocardium. Both Notch signaling and NO signaling have been shown to be important during adult arteriogenesis. Notch has been shown to be required for remodeling of the collateral vessels, whereas NO is required for the initial vasodilatory response to ischemia. Whether Notch also has an impact on the vasodilatory phase of arteriogenesis after ischemia is not known. We tested the hypothesis that endothelial cell-Notch function is required for NO induction and vasodilation, in response to ischemia in the adult vasculature. METHODS AND RESULTS: We observed a significant decrease in NO levels in the dorsal aorta using a mouse model where Notch was inhibited in endothelial cell in a Tet-inducible fashion. In a femoral artery ligation model, inhibition of endothelial cell-Notch reduced reperfusion and NO generation, as quantified by laser Doppler perfusion imaging and by phosphoendothelial NO synthase, nitrotyrosine, and phosphovasodilator-stimulated phosphoprotein staining, respectively. CONCLUSIONS: Endothelial Notch activation is required for NO production and reactive vasodilation in a femoral artery ligation model.

Differentiation of vascular smooth muscle cells from local precursors during embryonic and adult arteriogenesis requires Notch signaling.

Vascular smooth muscle cells (VSMC) have been suggested to arise from various developmental sources during embryogenesis, depending on the vascular bed. However, evidence also points to a common subpopulation of vascular progenitor cells predisposed to VSMC fate in the embryo. In the present study, we use binary transgenic reporter mice to identify a Tie1(+)CD31(dim)vascular endothelial (VE)-cadherin(-)CD45(-) precursor that gives rise to VSMC in vivo in all vascular beds examined. This precursor does not represent a mature endothelial cell, because a VE-cadherin promoter-driven reporter shows no expression in VSMC during murine development. Blockade of Notch signaling in the Tie1(+) precursor cell, but not the VE-cadherin(+) endothelial cell, decreases VSMC investment of developing arteries, leading to localized hemorrhage in the embryo at the time of vascular maturation. However, Notch signaling is not required in the Tie1(+) precursor after establishment of a stable artery. Thus, Notch activity is required in the differentiation of a Tie1(+) local precursor to VSMC in a spatiotemporal fashion across all vascular beds.

Chloroethyl urea derivatives block tumour growth and thioredoxin-1 nuclear translocation.

Aryl chloroethyl ureas (CEUs) are new protein alkylating agents exhibiting anticancer activity both in vitro and in vivo. We report herein that 14C-labeled CEU derivatives, designated CEU-025 and CEU-027, covalently bind to thioredoxin-1 (TRX1). Covalent binding of these molecules slightly decreases the disulfide-reducing activity of recombinant TRX1, when compared with the effect of strong thioalkylating agents such as N-ethylmaleimide. Moreover, site-directed mutagenesis and diamide competition assays demonstrated that TRX1 cysteinyl residues are not the prime targets of CEUs. CEU-025 abrogates the nuclear translocation of TRX1 in human cancer cells. In addition, we show that CEU-025 can block TRX1 nuclear translocation induced by cisplatin. Unexpectedly, pretreatment with sublethal CEU-025 concentrations that block TRX1 nuclear translocation protected the cells against cisplatin cytotoxicity. Overexpression of TRX1 in HT1080 fibrosarcoma cells attenuated CEU-025 cytotoxicity, while its suppression using TRX1-specific siRNA increased the effects of CEU-025, suggesting that loss of function of TRX1 is involved, at least in part, in the cytotoxic activity of CEU-025. These results suggest that CEU-025 and CEU-027 exhibit anticancer activity through a novel, unique mechanism of action. The importance of TRX1 and the dependence of the cytotoxicity of CEU-025 and CEU-027 on TRX1 intracellular localization are also discussed.

ASK1-P38 pathway is important for anoikis induced by microtubule-targeting aryl chloroethylureas.

PURPOSE: We investigated the involvement of MAPK signaling in the cell death mechanisms of classical microtubule interfering agents (MIA) and aryl-3-(2-chloroethyl)ureas (CEU) acting as antimitotics, along with CEU that don't affect directly microtubules (non-MIA CEU). METHODS: To ascertain the activated signaling pathway profile of MIA and non-MIA CEU, Western blot, immunoprecipitation and transfection experiments were performed. RESULTS: Non-MIA CEU do not activate p38, as opposed to MIA, and the extent of ERK and JNK activation is lower than in response to MIA. The effect of MIA and non-MIA CEU on focal adhesion associated protein was also studied; MIA were shown to induce focal adhesion dismantlement associated with a sustained increase in paxillin phosphorylation and FAK cleavage, as opposed to non-MIA CEU. In addition, bcl-2 phosphorylation and AKT cleavage, induced by all MIA tested, was not observed in response to non-MIA CEU further emphasizing the differential cell death mechanisms induced by MIA and non-MIA CEU. Pharmacologic and genetic approaches emphasize that the ASK1-p38 pathway activation contributes to the cytotoxic mechanism of MIA, in contrast to non-MIA CEU. ASK1-p38 is important for increased paxillin phosphorylation and FAK cleavage, suggesting that ASK-1-p38 is an upstream event of FA structure dismantlement induced by MIA. Moreover, the endogen inhibitor of ASK-1, thioredoxin, is released from ASK-1 in response to MIA as opposed to non-MIA CEU. CONCLUSION: Our study supports that ASK1-p38 activation is an important signaling event, induced by MIA, which impairs focal adhesion structure and induces anchorage-dependent apoptosis or anoikis.

Involvement of endothelial progenitor cells in tumor vascularization.

The generation of new microvasculature progresses by the process of angiogenesis, which involves the invasion and proliferation of endothelial cells from existing blood vessels into the local environment. In recent years, de novo generation of endothelial cells from circulating or local precursor endothelial cells is thought to also contribute to the neovasculature, a process that is referred to as vasculogenesis. In the adult, endothelial progenitor cells (EPC) are believed to be recruited from the bone marrow, migrate to sites requiring neovascularization and participate in the assembly of newly-forming blood vessels. A growing number of studies report that EPC participate in tumor progression and influence the efficacy of anticancer chemotherapeutics, and thus are attractive targets for cancer treatments. However, recent evidence calls into question the ability of marrow-derived EPC to act as a bona fide precursor for adult vasculogenesis. This review focuses on studies reporting or precluding the importance of EPC in tumor vasculogenesis. The putative sources of these cells and difficulties associated with their detection are discussed.

Cycloalkyl-substituted aryl chloroethylureas inhibiting cell cycle progression in G0/G1 phase and thioredoxin-1 nuclear translocation.

1-(2-Chloroethyl)-3-(4-cyclohexylphenyl)urea (cHCEU) has been shown to abrogate the presence of thioredoxin-1 into the nucleus through its selective covalent alkylation. In the present letter we have evaluated the structure-activity relationships of the substituents at positions 3 and 4 of the phenyl ring of cHCEU derivatives on cell cycle progression and thioredoxin-1 nuclear translocation. Active CEU derivatives exhibited GI(50) ranging from 1.9 to 49muM on breast carcinoma MCF-7, skin melanoma M21, and colon carcinoma HT-29 cells. On one hand, compounds 1, 2, 9c, 10c, 13, and 14 arrested the cell cycle in G(2)/M phase while CEUs 3, 4, 5c, 6c, 11c, and 12c blocked the cell division in G(0)/G(1) phase. On the other hand, CEUs 2-4, 5c, 7c, 8c, 11c, and 12c abrogated the translocation of thioredoxin-1 while the other CEU derivatives were inactive in that respect. Our results suggest that CEU substituted on the phenyl ring at position 3 or 4 by lower cycloalkyl or cycloalkoxy groups arrest cell progression in G(0)/G(1) phase through mechanism of action different from their antimicrotubule counterparts, presumably via thioredoxin-1 alkylation and modulation of its activity. The mechanism of action of these new molecules is still undetermined. However, the significant accumulation of cells in G(0)/G(1) phase suggests that these molecules may act similarly to known chemopreventive agents against cancers. In addition, the inhibition of Trx-1 nuclear localization also suggests the abrogation of an important chemoresistance mechanism towards a variety of chemotherapeutic agents.

N-Phenyl-N'-(2-chloroethyl)ureas (CEUs) as potential antineoplastic agents. Part 3: role of carbonyl groups in the covalent binding to the colchicine-binding site.

In the course of the development of N-phenyl-N'-(2-chloroethyl)ureas (CEUs) as potential antineoplastic agents, we investigated the effect of carbonylated substituting chains of the aromatic ring of CEU on their covalent binding to the colchicine-binding site (C-BS). In this study, we found that CEU, 5e, 5f, 8e, and 8f substituted by either a methyl ester or a methyl ketyl group at the omega-position exhibited a significant antiproliferative activity on HT-29, M21, and MCF-7 tumor cells. SDS-PAGE assays and cell cycle analysis confirmed that 5e, 5f, 8e, and 8f covalently bind to the C-BS and arrest the cell division in G(2)/M phase. Surprisingly, the presence of omega-carboxyl, omega-ethyl esters or omega-amides decreased significantly both the antiproliferative activity and the specificity toward beta-tubulin.

Alkylation potency and protein specificity of aromatic urea derivatives and bioisosteres as potential irreversible antagonists of the colchicine-binding site.

A number of N-phenyl-N'-(2-chloroethyl)ureas (CEUs) have been shown to be potent antimitotics through their covalent binding to the colchicine-binding site on intracellular beta-tubulin. The present communication aimed to evaluate the role of the electrophilic 2-chloroethyl amino moiety of CEU on cell growth inhibition and the specificity of the drugs as irreversible antagonists of the colchicine-binding site. To that end, several N-phenyl-N'-(2-ethyl)urea (EU), N-phenyl-N'-(2-chloroethyl)urea (CEU), N-aryl amino-2-oxazoline (OXA), and N-phenyl-N'-(2-chloroacetyl)urea (CAU) derivatives were prepared and tested for their antiproliferative activity, their effect on the cell cycle, and their irreversible binding to beta-tubulin. EU derivatives were devoid of antiproliferative activity. CEUs (2h-2i, 2k, 2l, OXA 3e, 3h, 3i, 3k, 3l, tBCEU, and ICEU), OXA (3h, 3i, 3k, 3l, tBOXA, and IOXA), and CAU (4a-4m, tBCAU, and ICAU) had GI(50) between 1.7 and 10microM on three tumor cell lines. Cytotoxic CEU and OXA arrested the cell cycle in G(2)/M phase, while the corresponding CAU were not phase specific. Finally, Western blot analysis clearly showed that only CEUs 2h, 2k, 2l, tBCEU, ICEU and OXA 3h, 3i, 3k, 3l, tBOXA ,and IOXA were able to bind irreversibly to the colchicine-binding site. Our results suggest that increasing the potency of the electrophilic moiety of the aromatic ureas enhances their antiproliferative activity but decreases significantly their capacity to covalently bind to the colchicine-binding site.

New soft alkylating agents with enhanced cytotoxicity against cancer cells resistant to chemotherapeutics and hypoxia.

Chloroethylureas (CEU) are soft alkylating agents that covalently bind to beta-tubulin (betaTAC) and affect microtubule polymerization dynamics. Herein, we report the identification of a CEU subset and its corresponding oxazolines, which induce cell growth inhibition, apoptosis, and microtubule disruption without alkylating beta-tubulin (N-betaTAC). Both betaTAC and N-betaTAC trigger the collapse of mitochondrial potential (DeltaPsi(m)) and modulate reactive oxygen species levels, following activation of intrinsic caspase-8 and caspase-9. Experiments using human fibrosarcoma HT1080 respiratory-deficient cells (rho(0)) and uncoupler of the mitochondrial respiratory chain (MRC) showed that betaTAC and N-betaTAC impaired the MRC. rho(0) cells displayed an increased sensitivity toward N-betaTAC as compared with rho(+) cells but, in contrast, were resistant to betaTAC or classic chemotherapeutics, such as paclitaxel. Oxazoline-195 (OXA-195), an N-betaTAC derivative, triggered massive swelling of isolated mitochondria. This effect was insensitive to cyclosporin A and to Bcl-2 addition. In contrast, adenine nucleotide translocator (ANT) antagonists, bongkrekic acid or atractyloside, diminished swelling induced by OXA-195. The antiproliferative activities of the N-betaTACs CEU-025 and OXA-152 were markedly decreased in the presence of atractyloside. Conversely, pretreatment with cyclosporin A enhanced growth inhibition induced by betaTAC and N-betaTAC. One of the proteins alkylated by N-betaTAC was identified as the voltage-dependent anion channel isoform-1, an ANT partner. Our results suggest that betaTAC and N-betaTAC, despite their common ability to affect the microtubule network, trigger different cytotoxic mechanisms in cancer cells. The role of mitochondria in these mechanisms and the potential of N-betaTAC as a new therapeutic approach for targeting hypoxia-resistant cells are discussed.

Membrane composition modulates the interaction between a new class of antineoplastic agents deriving from aromatic 2-chloroethylureas and lipid bilayers: a solid-state NMR study.

We have investigated the interaction between a new class of antineoplastic agents derived from arylchloroethylureas (CEU) with three different model membranes by (31)P and (2)H solid-state NMR spectroscopy. First, we have prepared model membranes that mimic the mitochondrial inner (Mito IM) and outer (Mito OM) membranes and the endoplasmic reticulum membrane (End Ret). Our results indicate that the effects of the CEU derivatives on lipid bilayers are related to their cytotoxic activity. More specifically, a strong correlation is observed between the drug location in both the mitochondrial inner and outer membranes and its cytotoxicity. In addition, the results indicate that the lipid composition of the model membrane has a very important influence on the effects of CEUs. More specifically, a high proportion of cardiolipin in the mitochondrial inner membrane gives this system the highest fluidity and consequently, this model membrane is more rigidified by the presence of CEUs compared to the mitochondrial outer and endoplasmic reticulum membranes. Finally, the results propound a hypothesis for the location of CEUs in membranes.

N-Phenyl-N'-(2-chloroethyl)urea analogues of combretastatin A-4: Is the N-phenyl-N'-(2-chloroethyl)urea pharmacophore mimicking the trimethoxy phenyl moiety?

A series of novel N-phenyl-N'-(2-chloroethyl)urea derivatives potentially mimicking the structure of combretastatin A-4 were synthesized and tested for their cell growth inhibition and their binding to the colchicine-binding site of beta-tubulin. Compounds 2a, 3a, and 3b were found to inhibit cell growth at the micromolar level on four human tumor cell lines. Flow cytometric analysis indicates that the new compounds act as antimitotics and arrest the cell cycle in G(2)/M phase. Covalent binding of 2a, 3a, and 3b to the colchicine-binding site of beta-tubulin was confirmed also using SDS-PAGE and competition assays.

Microtubule-destabilizing agents induce focal adhesion structure disorganization and anoikis in cancer cells.

Microtubule disruption provokes cytoskeleton and cell adhesion changes whose importance for apoptosis induction remains unclear. The present study focuses on the functional and the molecular adhesion kinetics that are induced by microtubule disruption-mediated apoptosis. We showed that antimicrotubules induce a biphasic sequence of adhesion response that precedes the onset of apoptosis and focal adhesion kinase hydrolysis. Antimicrotubules first induced an increase of the cellular adhesion paralleled by the raise of focal adhesion sites and actin contractility, which was followed by a sharp decrease of cell adhesion and disorganization of focal adhesion and actin stress fibers. The latter sequence of events ends by cell rounding, detachment from the extracellular matrix, and cell death. Microtubule-disrupting agents induced a sustained paxillin phosphorylation, before the activation of apoptosis, that requires the prior activation of extracellular signal-regulated kinase and p38 but not c-Jun NH(2)-terminal kinase. Interestingly, integrin-linked kinase overexpression rescued the antimicrotubule-mediated loss of cell viability. Altogether, these results propound that antimicrotubule agents induce anoikis through the loss of focal adhesion structure integrity.

N-Phenyl-N'-(2-chloroethyl)ureas (CEU) as potential antineoplastic agents. Part 2: role of omega-hydroxyl group in the covalent binding to beta-tubulin.

Tubulin is the target of many anticancer drugs, including N-phenyl-N'-(2-chloroethyl)urea (CEU). Unlike most anti-beta-tubulin agents, CEUs are protein monoalkylating agents binding through their N'-(2-chloroethyl)urea moiety to an amino acid nearby the colchicine-binding site on beta-tubulin isoform-2. Following the previously synthesized and attractive N-(3-omega-hydroxyalkylphenyl)-N'-(2-chloroethyl)urea that exhibited growth inhibitory activity at the nanomolar level, we investigated the importance of lower alkyl and alkoxy groups to evaluate the effect of hydroxylated group and chain length on both cell growth inhibition and the mechanism of action of CEU. Here, we describe the preparation of two new series of CEU and show that the most potent CEU derivatives beside the omega-hydroxylated 1f were 2f and 3e, respectively. We have confirmed that the pentyl substituted CEUs 1f, 2f, and 3e are still covalently binding to beta-tubulin and still arrest cell division in G(2)/M phase.

Mitochondrial thioredoxin system: effects of TrxR2 overexpression on redox balance, cell growth, and apoptosis.

Thioredoxin-2 (Trx2) is a mitochondrial protein-disulfide oxidoreductase essential for control of cell survival during mammalian embryonic development. This suggests that mitochondrial thioredoxin reductase-2 (TrxR2), responsible for reducing oxidized Trx2, may also be a key player in the regulation of mitochondria-dependent apoptosis. With this in mind, we investigated the effects of overexpression of TrxR2, Trx2, or both on mammalian cell responses to various apoptotic inducers. Stable transfectants of mouse Neuro2A cells were generated that overexpressed TrxR2 or an EGFP-TrxR2 fusion protein. EGFP-TrxR2 was enzymatically active and was localized in mitochondria. TrxR2 protein level and TrxR activity could be increased up to 6-fold in mitochondria. TrxR2 and EGFP-TrxR2 transfectants showed reduced growth rates as compared with control cells. This growth alteration was not due to cytotoxic effects nor related to changes in basal mitochondrial transmembrane potential (DeltaPsi(m)), reactive oxygen species production, or to other mitochondrial antioxidant components such as Trx2, peroxyredoxin-3, MnSOD, GPx1, and glutathione whose levels were not affected by increased TrxR2 activity. In response to various apoptotic inducers, the extent of DeltaPsi(m) dissipation, reactive oxygen species induction, caspase activation, and loss of viability were remarkably similar in TrxR2 and control transfectants. Excess TrxR2 did not prevent trichostatin A-mediated neuronal differentiation of Neuro2A cells nor did it protect them against beta-amyloid neurotoxicity. Neither massive glutathione depletion nor co-transfection of Trx2 and TrxR2 in Neuro2A (mouse), COS-7 (monkey), or HeLa (human) cells revealed any differential cellular resistance to prooxidant or non-oxidant apoptotic stimuli. Our results suggest that neither Trx2 nor TrxR2 gain of function modified the redox regulation of mitochondria-dependent apoptosis in these mammalian cells.

Glutathione peroxidase-1 expression enhances recovery of human breast carcinoma cells from hyperoxic cell cycle arrest.

We previously reported that hyperoxia (95% O(2)) induces an S-phase cell cycle arrest in glutathione peroxidase-deficient human carcinoma cells T47D-H3 (Exp. Cell Res. 256:347-357; 2000). Here, we investigated whether increasing the peroxide scavenging capacity via glutathione peroxidase-1 (GPx1) expression can prevent cell cycle alterations induced by oxidative stress. We show that GPx1-proficient T47D-GPx-2 transfectant cells, in which GPx1 concentration is most elevated in mitochondria (Biochem. Biophys. Res. Commun. 272:416-422; 2000), are partially resistant to cell cycle inhibition induced by hyperoxia or menadione exposure. Transient cell growth resistance was observed at the level of cell cycle phase distribution, Cdk2 activity, and DNA synthesis after 40 h hyperoxia. This differential resistance was associated with an inhibition of ROS production and lipid peroxidation induced by hyperoxia. After 64 h hyperoxic exposure, cell growth was completely abolished in both cell lines, despite elevated glutathione levels. However, in contrast to the GPx1-deficient cells, T47D-GPx-2 cells showed an increased capacity to recover from a cell cycle arrest mediated by a 64 h hyperoxic stress. Differential recovery was also observed at the ultrastructural level between Gpx1-proficient and -deficient cells. These data indicate that GPx1 played an important role in the cell capacity to recover from hyperoxic insults. The limited protection conferred by GPx1 during hyperoxia suggests that the deleterious effects were partially mediated by peroxide-derived free radicals, but also involved the action of nonperoxide-derived reactive species.