Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 7-Day Trial for You or Your Team.

Learn More →

Depletion of the mitochondrial electron transport abrogates the cytotoxic and gene‐inductive effects of TNF.

Depletion of the mitochondrial electron transport abrogates the cytotoxic and gene‐inductive... The EMBO Journal vol.12 no.8 pp.3095-3104, 1993 Depletion of the mitochondrial electron transport abrogates the cytotoxic and gene-inductive effects of TNF 1991) or proteases (Ruggiero et al., 1987; Suffys et al., Klaus Schulze-Osthoff1, Rudi Beyaert, 1988) or DNA damage (Dealtry et al., 1987). Recently, we Vandevoorde, Guy Haegeman and Veronique have provided evidence that in the classical TNF-sensitive Walter Fiers2 fibrosarcoma cell lines L929 and WEHI 164 clone 13 a Laboratory of Molecular Biology, University of Ghent, Ghent, functional mitochondrial respiratory chain is crucial to the Belgium events leading to TNF cytotoxicity (Schulze-Osthoff et al., lPresent address: German Cancer Research Center, Department of 1992). Immunology and Genetics, Heidelberg, FRG While the cytotoxic activity of TNF seems to be rather author 2Corresponding restricted to tumor cells, nearly every cell type responds to Communicated by W.Fiers of different genes. TNF by the activation of a wide range Owing to this broad gene-activating effect, TNF is suggested Tumor necrosis factor (TNF?) has cytotoxic and gene- differentiation and as a cardinal mediator of several inductive activities on several cell types. Previous studies TNF has been immunoregulatory processes. Among others, on L929 fibrosarcoma cells have revealed that the of other such as shown to induce expression cytokines mitochondrial electron transport system plays a key role (IL-6) or GM-CSF, and interleukin-I (IL-1), interleukin-6 in inducing TNF cytotoxicity, presumably by the as receptors and cell surface-bound proteins such cytokine formation of reactive oxygen intermediates (ROI). Here in Furthermore, adhesion molecules (reviewed Fiers, 1991). we report that mitochondria-derived intermediates are of HIV-1 is also inducible transcriptional virus replication not only cytotoxic but, in addition, function as signal Osborn et al., 1989). by TNF (Duh et al., 1989; transducers of TNF-induced gene expression. The has been found that TNF gene- In the last few years, it activation of NFxB, which fulfills an important role in are mediated by certain transcriptional regulatory activities TNF-induced gene transcription, could be blocked by include the factors AP-1 activators, which transcription interference with the mitochondrial electron transport most importantly, (Brenner et al., 1989) and, probably system. Furthermore, antimycin A, a mitochondrial The latter was initially NFxB (Osborn et al., 1989). inhibitor that increases the generation of ROI, potentiated a factor of mature B cells, that identified as nuclear TNF-triggered NFxB activation. The dual role of with an enhancer element of the specifically interacts mitochondria-derived intermediates in cytotoxicity and chain in Lenardo immunoglobulin kappa light gene (reviewed immediate-early gene induction of TNF was further Baeuerle, 1991; Blank et al., 1992). and Baltimore, 1988; which lacked substantiated by isolating L929 subclones of a heterodimeric complex, which in its NFxB consists of the a functional respiratory chain. This depletion form the two proteins p50 and p65. It classical comprises resulted in resistance mitochondrial oxidative metabolism that or structurally related proteins now appears NFxB as in inhibition of towards TNF cytotoxicity, as well Rel are ubiquitously present in other belonging to the family NFxB activation and interleukin-6 gene induction by and are inducible a wide variety of extracellular cell types by TNF. These findings suggest that mitochondria are the NFxB controls the inducible expression of various stimuli. source of second messenger molecules and serve as involved in immune responses, as well as genes that are common mediators of the TNF-cytotoxic and gene- and cellular defense mechanisms. Target genes inflammatory regulatory signaling pathways. include receptors, MHC of NFxB cytokines, cytokine words: cytotoxicity/interleukin-6/NFxB/respiratory Key and several viral enhancers. antigens, acute-phase proteins factor chain/tumor necrosis A characteristic feature of immediate-early NFxB activation is that it occurs without new protein synthesis. In the uninduced cell, NFxB resides cytoplasmically as a Introduction latent form to the inhibitory protein IxB (Baeuerle complexed cell NFxB is tumor necrosis factor (TNF) elicits on a variety and 1988). Upon stimulation, The cytokine Baltimore, release from which allows of in Beutler activated by its IxB, of cells two major types responses (reviewed rapidly into the nucleus and Beutler, 1992). First, TNF migration of the p50/p65 complex and Cerami, 1989; Fiers, 1991; The dissociation of which exerts a rather selective activity on several tumor subsequent DNA binding. IxB, cytotoxic event of NFxB is still rather and transformed whereas normal cells are generally seems to be the key activation, cells, mechanism of TNF In it has been shown that kinase C towards the cytokine. The elusive. vitro, protein resistant understood. eIF-2 kinase can activate the transcription is currently rather incompletely or heme-regulated cytotoxicity it seems different cellular events and and However, on the cell type, factor (Ghosh Baltimore, 1990). Depending of these kinases or kinase A is to TNF-induced cell such that one protein may contribute killing, unlikely processes in vivo et of reactive intermediates involved in NFxB activation (Meichle al., 1990; as generation oxygen (Zimmerman Hohmann et Yamauchi et Schulze-Osthoff Bromsztyk et al., 1991; al., 1991). Quite et al., 1989; al., 1989, 1990; that activation of NFxB be activation of (Suffys et al., it was proposed may et al., 1992), phospholipases recently, Press Oxford University K.Schulze-Osthoff et al. NADH ..... ... .. .... ",;-. V. FMN - [F.S| Complex I Antimycin A Rotenone cyt.aa3 cyt.c, wt.b rN&* lo uQ - -_ _ cyt.c -4° Complex ill ..... Complex IV Complex 11 TTFA FAD -_ [Fe-S] SUCCINATE Fig. 1. Schematic of the electron carriers of the diagram mitochondrial respiratory chain. The sites of inhibition the inhibitors by respiratory rotenone, TTFA and A are indicated. antimycin generally controlled by reactive oxygen intermediates (ROI) and the cellular redox state (Staal et Mihm al., 1990; et al., .:; .. :: .,y;I 1991; Schreck et al., at in 1991, 1992). Firstly, least some cell types the transcription factor could be activated by i h LiI: dA ,. 4,.. exogenously applied hydrogen peroxide. Secondly, activation of NFxB by different such as stimuli, TNF, IL-1, phorbol esters, viral transactivator or proteins, cycloheximide, was commonly inhibited by antioxidant thiols or iron chelators. __ Thus, these observations suggest that oxygen radicals acting as second messengers the diverse of may integrate variety NFxB-inducing signals. In our previous studies, we have shown that mitochondrial intermediates, possibly oxygen radicals generated from the respiratory chain, are causally involved in the cytotoxic effects of TNF (Schulze-Osthoff et al., 1992). Presumably, TNF causes a short circuit of the electron flow, resulting in direct production of radicals from the ubiquinone site of the chain. Here respiratory we report that mitochondria are not in only implicated cytotoxicity, but may also participate in the gene-regulatory effects of TNF. Our data that suggest gene induction and TNF cytotoxicity by share an at least Fig. 2. Differential effects of mitochondrial inhibitors on NFxB common partially pathway. We show that elimination of the binding activity in L929 cells. (A) Competition analysis showing the mitochondrial oxidative metabolism not only inhibits TNF of specificity NFxB binding: 3 of nuclear extract obtained from jig L929 cells treated cytotoxicity, but also reduces without (lane 1) or with TNF (1000 IU/ml; lanes considerably TNF-mediated were 2-4) analyzed in electrophoretic mobility shift assays with a 32p- activation of NFxB and of expression IL-6 as an example labeled probe encompassing the xB-binding site of the mouse kappa of a TNF-inducible, NFxB-regulated gene. Oxygen inter- chain light gene. To the nuclear extracts a 100-fold excess of mediates may therefore act as signal transducers and have unlabeled mutated or (lane 3) wild-type oligonucleotide (lane 4) were a more widespread role than mixed with the radioactive xB enhancer currently appreciated. probe. (B) The effect of rotenone on NFxB activation in L929 cells. Cells were treated for 1 h with the indicated concentrations of rotenone prior to TNF addition (1000 IU/ml). Incubation was continued for 45 min before extraction Results of the cells. (C) The effect of rotenone on DNA binding of nuclear extracts in vitro. Identical nuclear extracts from cells treated with TNF Mitochondrial inhibitors modulate TNF-induced NFxB (1000 IU/ml, min) were incubated for 30 min with the indicated activation concentrations of rotenone, followed by the addition of the 32P-labeled The mitochondrial electron transport can be inhibited by xB The effect of probe. (D) antimycin A on NFxB activation in L929 specific inhibitors which prevent electron flow between cells. Cells were 1 pretreated for h with the indicated concentrations of antimycin followed the certain of the A, by addition of a suboptimal dose (10 complexes respiratory chain (Figure IU/ml) 1). of TNF (lanes or medium 18-20) (lanes 21-23). After 45 min of Previously, we have reported that inhibition of the electron further incubation, nuclear extracts were prepared and analyzed by entry from complex I to the I ubiquinone by complex electrophoretic mobility shift assays. inhibitors amytal and rotenone results in a strong reduction of TNF cytotoxicity (Schulze-Osthoff et al., 1992). However, when electron flow in L929 cells was inhibited largely mediated by the generation of ROI at the ubiquinone distal to the ubiquinone pool, by the addition of site of the mitochondrial chain. antimycin Early investigations with A, a marked of potentiation TNF-induced killing was isolated mitochondria have repeatedly shown that ubiquinone observed. As an for explanation these differential is the major radical source of ROI drug derived from the effects, it was that TNF-induced cell mitochondrial chain proposed killing is (Boveris and Chance, 1973; Turrens and 3096 Depletion of mitochondrial functions abolishes TNF effects I..% Boveris, 1980; Turrens et al., 1985). ROI generation in B NDGA isolated mitochondria can be inhibited by amytal and rotenone, whereas antimycin A leads to an increase in ROI formation (Boveris et al., 1976; Cadenas and Boveris, 1980; Konstantinov et al., 1987; Cino and Del Maestro, 1989). We were interested in analyzing the effects of mitochondrial inhibitors on early TNF effects such as activation of the transcription factor NFxB, which has been reported, at least in some cell types, to be additionally I .=. I. inducible by hydrogen peroxide (Schreck et al., 1991, 1992). L929 cells were preincubated for 1 h with different concentrations of mitochondrial inhibitors; 45 min after TNF stimulation cells were lysed, and cellular and nuclear lysates were analyzed in electrophoretic mobility shift assays (EMSA). TNF induced a xB binding activity in L929 cells which was not detectable in unstimulated cells (Figure 2A). Competition with a 100-fold molar excess of unlabeled oligonucleotide led to an inhibition of xB binding, thereby confirming the specificity of the activated NFxB. No reduction of DNA binding, however, was observed with an excess of mutated oligonucleotide. .1 -. 1 910 12 concentrations When cells were pretreated with non-toxic inhibition of of the complex I inhibitor rotenone, a strong Oxygen radical scavengers reduce activation of NFxB in TNF- Fig. 3. was observed (Figure 2B). Even a NFxB activation L929 cells. Cells were pretreated for 1 h with various stimulated 5 inhibition of -50% of concentration of led to an concentrations of BHA (A) or NDGA (B) prior to the addition of AM NFxB binding. There remained the possibility that decrease 1000 IU/ml TNF and further 45 min of incubation. Equal amounts of total cell lysates (10 yg) were analyzed by electrophoretic mobility due to a reduction of of xB binding in EMSA was either shift assays. NFxB activation in the cytoplasm or to a reduced DNA- of To binding affinity the released NFxB complex. answer Neither allopurinol, an inhibitor of cytoplasmic xanthine this question, rotenone was added directly to the lysates. of the inhibitor could be detected oxidase, nor diphenylene iodonium (DPI), an inhibitor of However, no interference 2C). reduction of xB binding by rotenone the plasma membrane-bound NADPH oxidase, interfered (Figure Therefore, (data not shown). was most caused by inhibition of TNF-induced with TNF-induced NFxB activation probably in TNF-sensitive cell line NFxB activation in the cytoplasm. Therefore, NFxB activation the specifically controlled by When Ill inhibitor antimycin A, L929 seems to be largely and we analyzed the complex from chain. which TNF cytotoxicity, contrasting effects on intermediates derived the respiratory potentiates of mitochondrial ROI for the NFxB activation became evident (Figure 2D). When a The specific involvement nuclear extracts TNF dose was applied, antimycin A activation of NFxB was tested by assaying suboptimal (10 IU/ml) of other factors. a of NFxB activation. for the DNA-binding activity transcription triggered strong potentiation of Oct-i and NF-IL6 are Sometimes A itself induced a faint xB band at DNA-binding activities AP-1, antimycin in cells and inducible the concentration. A effect of constitutively present L929 barely upon highest synergistic antimycin 4 shows that of L929 observed with TNF doses, which TNF stimulation. Figure pretreatment A was only suboptimal inhibited TNF activation of be indicative for a similar action mechanism of both cells with rotenone selectively may activities of AP-1, Oct-1 NFxB, whereas the DNA-binding agents. in the same nuclear In accordance with our previously published data regarding and NF-IL6 remained unchanged This that inhibition of mitochondrial ROI TNF-mediated cytotoxic effects on mitochondria (Schulze- extracts. suggests indicate an formation modulates the activation of NFxB. Osthoff et the observations specifically al., 1992), present involvement of mitochondria-derived intermediates, possibly of NFxB. The Depletion of the mitochondrial respiratory chain ROI, also in the activation process ROI was further results in acquired TNF resistance activation of NFxB by presumptive the dual involvement of In order to substantiate further using butylated hydroxyanisole supported by experiments and NFxB acid two mitochondria for both TNF effects, cytotoxicity (BHA) or nordihydroguaiaretic (NDGA), potent of L929 which were 3 shows that both we isolated subclones antioxidants Figure agents activation, (Niki, 1987). chain. We also the effect of two deficient in a functional respiratory Respiration reduced NFxB binding. analyzed culture of L929 cells have been described as efficient deficiency was generated by long-term antioxidants which recently bromide or of NFxB activation et 1991, 1992). in ethidium (EB)- chloramphenicol (CA)- inhibitors (Schreck al., which have been in our cell inhibition was only observed containing medium, procedures repeatedly However, system, used to cells from a functional mitochondrial electron rather concentrations of (30 mM), deplete at high N-acetylcysteine Wiseman and of reduced Attardi, 1978; radical and transport system (Nass, 1972; a scavenger precursor glutathione, Morais et and 100 an iron et al., 1988; King and of dithiocarbamate Desjardins al., 1985; pyrrolidine (PDTC, AM), Ethidium bromide has been described as a not Attardi, 1989). chelator (data shown). of mitochondrial DNA be different in several inhibitor (mtDNA) As ROI may generated by enzymes specific and et we also the contribution of (Wiseman Attardi, 1978; Desjardins al., cellular analyzed replication compartments, is known to block mitochondrial sources for the induction of NFxB. 1985). Chloramphenicol other ROI potential 3097 K.Schulze-Osthoff at a!. C ~ ~ v b~(-', b ll~ rotenone <>M - - - ~ C A\~~~~~1 C70 <4~0 N F S. Pt NFkB (V- MW _ il '""fl..;I, N4- ND)2 - N'11 - AP 1 I(1)1E11 (.\I I .. b _ a_ N(I) 0(CZ , gi,;-. , 0* 0 'l r <1- NF-IL6 ~~~~~~~~ ~l, r,~ .@ k~,'$<,~ ' ""'II| NO Fig. 4. The effect of rotenone on different DNA-binding activities. L929 cells were pretreated for 1 h with the indicated concentrations of rotenone, followed by a stimulation with TNF (1000 IU/ml) for 90 with 32P-labeled DNA min. Nuclear extracts were analyzed in EMSA of probes detecting binding activities NFxB, AP-1, octamer-binding the shifted bands are shown. proteins (Oct-1) and NF-IL6. Only 5. of mtDNA-encoded and mtDNA Fig. Analysis protein synthesis (A) oxidase and lactate in L929 levels in ethidium bromide- and L929 Table I. Cytochrome activity production (B) chloramphenicol-treated EB- and CA-treated subclones subclones. Cells were in 6 well dishes until confluent. cells and (A) grown 100 emetine was and after 30 Fresh medium containing ,ug/ml applied oxidasea Lactate min 100 ,uCi/ml [35S]methionine were added and the incubation Cytochrome (mM)b continued for 3 h. Cells were washed and in 25 mM Tris-HCl lysed SDS and 8 M urea. Extracts were on 12.5% (pH 6.8), 2% analyzed 65 16 2.26 L929 on were SDS-urea-PAGE. The radioactive bands the fluorogram 21 6 3.61 EB-2 identified to their molecular size et al., 1985). Cyt according (Chomyn EB-12 12 3 3.89 b: COIL subunits 3 of cytochrome b; COI, COIl, 1, 2, cytochrome EB-21 5 3 4.62 ND5: subunits 3, 4, 5 of oxidase; ND1, ND2, ND3, ND4, 1, 2, CA-22 29 8 3.92 NADH ATP8: subunit 8 of the H+-ATPase. CA and dehydrogenase, CA-34 16 4 3 4.43 EB denote different subclones which were isolated limiting dilution by in or ethidium selection chloramphenicol- bromide-containing medium, of c respectively. (B) Southern blot analysis of restriction endonuclease amnol cytochrome oxidized/min/mg protein. b106 1 24 cells were seeded in ml in a 6 well dish. After cells were obtained from DNA of the L929 cells and h, fragments wild-type FCS. After 12 lactate to fresh medium subclones. DNA (10 was digested with HindIH and changed containing dialyzed h, 1tg) was in the electrophoresed on a 0.7% agarose gel. The gel was transferred to a determined production supematant. nylon membrane and the separated DNA hybridized with a 32P-labeled fragment of the rat mitochondrial 12S rRNA gene. protein synthesis by binding to the 70S ribosome, while leaving cytoplasmic protein synthesis intact (Spolsky and of mtDNA was further Eisenstadt, 1972). From the EB- and CA-resistant cell confirmed by Southern blottings with a for the populations, several subclones were isolated by limiting probe coding mitochondrial 12S rRNA. In all dilution in subclones the amount of mtDNA and analyzed detail. Measurements of lactate tested, was considerably production and cytochrome c oxidase revealed that most cell reduced (Figure 5B). In accordance with Figure 5A, clones recruited most of their energy from anaerobic mtDNA was undetectable in EB-12 and EB-21 clones under glycolysis (Table I). To analyze further the presence of conventional hybridization conditions, whereas subclone mtDNA-encoded EB-2 a protein subunits in the subclones, cells gave weak hybridization signal. were labeled We first in with [35S]methionine in the presence of analyzed cytotoxicity assays the TNF sensitivity an inhibitor of of the subclones. In the absence of the transcription inhibitor emetine, cytoplasmic protein synthesis. By this of the different actinomycin D (Act D), all subclones were largely TNF procedure, synthesis mtDNA-encoded proteins could be visualized in SDS -urea gels. Figure 5A resistant in comparision with the sensitive wild-type cells shows that most of the cell clones had a reduced (Figure 6A). In some clones, no detrimental effects on cell mitochondrial In of protein synthesis. three the EB-resistant viability could be detected even at high TNF concentrations. EB-12 no clones (EB-2, and EB-21), mtDNA translation It is noteworthy that the simple addition of EB or CA to could be detected. A reduction or total untreated products depletion wild-type L929 cells did not affect TNF cytotoxic 3098 Depletion of mitochondrial functioixs abolishes TNF effects IJ5 V113-21 -( DOC 60 0 N- - ~~~~ Os'. 60~~~~~~~ 0~\ o ~o Y A \ 7 62 556 5000 1 6 56 500 TNF (IU/ml) Fig. 6. Comparison of the TNF sensitivity of the respiration-deficient subclones and wild-type L929 cells in the absence (A) and presence (B) of actinomycin D. Cells were seeded in 96 well plates at a density ^__ AL -. AL AL is-i- of 2 x 104 cells/well. After 16 h, serial concentrations of TNF " I ranging from 7 to 5000 IU/ml (A) or 0.7 to 500 IU/ml (B) were applied and incubated for 18 h. Survival was assayed by reduction of 1 2 3 4 5 7 8 9 10 11 12 13 14 MTT and calculated as the percentage of the staining value of untreated cultures. 0, L929 cells; V, EB-12; 0, EB-21; E, CA-22; Fig. 7. NFxB activation by TNF is reduced in respiration-deficient V, CA-34 subclone. L929 subclones. Wild-type L929 cells (lanes 1-6) and EB-21 cells, as a representative subclone (lanes were respiration-deficient 7-12), activity. We next analyzed the TNF sensitivity in the incubated for 45 min with the indicated concentrations of TNF. presence of Act D, which generally enhances TNF killing Nuclear extracts were prepared and tested for the presence of activated NFxB in electrophoretic mobility shift assays. In lanes 13-14, and additionally sensitizes some cell types which are resistant cytoplasmic extracts of both cell types were prepared and latent in the absence of the inhibitor. In the presence of Act D, NFxB-IxB complex activated by treatment with deoxycholate (DOC) TNF resistance of the cell clones was partially overcome, as described in Materials and methods. The EMSA reveals that although cells were still less sensitive than their wild-type TNF stimulation in EB-21 is reduction of NFxB activation after cells not due to diminished amounts of the inactive complex. homologues (Figure 6B). NFxB activation by TNF is strongly repressed in been identified in the IL-6 promoter, such as a cAMP- respiration-deficient cell clones responsive element and an AP-1 site, induction of IL-6 by Next to the cytotoxic effects, the respiration-deficient TNF has been reported to be largely controlled by a xB subclones of L929 were analyzed for the activation of NFxB element (Libermann et al., 1990; Shimizu et al., 1990; in response to TNF. Cells were treated with serial dilutions Zhang et al., 1990). of In TNF. the wild-type cells the addition of 200 IU/ml When wild-type and EB- and CA-treated cells were TNF already resulted in maximal activation of the stimulated with TNF, respiration-deficient cell clones transcription factor (Figure 7). The respiration-deficient produced remarkably less IL-6 than wild-type cells clones, however, only weakly responded to TNF, which was (Figure 8A). In some cell clones, IL-6 expression was obvious a by strongly reduced NFxB activation. Even at the reduced even up to 500- to 600-fold. To analyze whether TNF highest concentrations used, only light xB bands were reduction of IL-6 expression was specific for the TNF observed. stimulus, both cell types were further treated with dibutyryl To analyze whether the reduction of xB binding was really cyclic AMP (db-cAMP), which induces IL-6 expression by due to a or diminished cytoplasmic NFxB activation just the the elevation of cAMP levels. in these Indeed, experiments effect of decreased we further NFxB protein expression, no significant differences could be observed between L929 levels of latent in the cells and the that the compared NFxB-IxB complex subclones, suggesting IL-6 gene could respiration-deficient clones and L929 cells. When still be induced other in wild-type by signaling pathways the extracts of both cell were activated with respiration-deficient cell clones. The same differences in cytoplasmic types in order to dissociate the IxB no TNF-induced IL-6 were also observed at the deoxycholate subunit, expression differences could be observed in the EMSA level In in the (Figure 7). transcriptional (Figure 8B). fact, respiration- These results indicate that of a functional deficient clones IL-6 mRNA could be detected after depletion hardly TNF mitochondrial electron transport system largely abrogates stimulation, although comparable amounts of IL-6 mRNA found after cell activation a combination TNF-triggered NFxB activation. were by of and which db-cAMP, cycloheximide staurosporine, Inhibition of IL-6 in the IL-6 these results gene expression respiration- superinduces gene expression. Therefore, deficient subclones indicate that inhibition of the mitochondrial oxidative of TNF. The effects of of the mitochondrial electron metabolism also affects effects depletion gene-regulatory on TNF-induced NFxB activation us to transport prompted further TNF-induced in the L929 analyze gene expression Discussion which is subclones. IL-6 represents a cytokine rapidly TNF in cell et which radicals induced by ROI, comprise hydrogen peroxide, hydroxyl many types (Defilippi al., 1987). control elements have and are essential of an Although several other transcriptional superoxide anions, compounds 3099 K.Schulze-Osthoff et al. ?r il.H such as or adriamycin menadione, act as redox-cycling agents that continuously produce deleterious oxygen species. evidence now exists Accumulating that short-lived oxygen radicals a may provide versatile cellular control mechanism and additionally play important roles central to gene Pro-oxidant regulation. conditions can induce expression of .: immediate-early genes (reviewed in Holbrook and Fornance, in 1991), particular c-fos and (Crawford et al., 1988; c-jun Shibanuma et al., 1988; Stein et al., 1989; Amstad et al., 1992). Indeed, some gene-inductive effects of phorbol esters also be may independent of protein kinase C activation, since phorbol esters can generate a pro-oxidant cellular environment (Cerutti, 1985). Recently, activation and DNA of several binding transcription factors have been shown to depend on the cytoplasmic or nuclear redox state, including steroid receptors (Silva and Cidlowski, 1989), ISGF3 (Levy A | \i M0I 0s iK' et al., 1989), TFIHC (Cromlish and Roeder, 1989), c-myb protein (Guehmann et al., 1992) and the transcriptional activators AP-1 (Abate et al., 1990; Devary et al., 1991; Xanthoudakis et and al., 1992) NFxB (Mihm et al., 1991; Cl. 'K, B Schreck et mechanism al., 1991, 1992). Although the of .3. vC redox regulation in transcriptional activation is still rather a substantial unclear, body of evidence now exists that ROI may act as second in addition messengers and, to their harmful also exert effects, may signaling functions. In this study, we show that two diverse effects of TNF, , * Si i.e. cytotoxicity and transcriptional activation, require functional as an intracellular mitochondria, presumably source of ROI. TNF cytotoxicity and mitochondrial oxygen radicals In our previous studies (Schulze-Osthoff et al., 1992), we have reported that mitochondria play a crucial role in causing TNF-induced cytotoxicity in TNF-sensitive cells. Pre- sumably, upon TNF stimulation mitochondrial radicals are Fig. 8. TNF of IL-6 is defective in induction respiration-deficient L929 subclones. IL-6 106 cells were in well formed at the ubiquinone site of the respiratory where (A) bioassay: seeded 6 chain, dishes. After 12 medium was refreshed and cells h, were stimulated electrons are directly transferred from ubisemiquinone to with TNF (500 IU/ml, 2500 or db-cAMP 18 h later IU/nil) (1 mM); molecular oxygen. This one-electron transfer results in the supematants were harvested and for IL-6 analyzed bioactivity. (B) formation of superoxide anions as the primary reactive Analysis of IL-6 mRNA levels in L929 subclones. Confluent cell of oxygen species mitochondria (Loschen et al., 1974). monolayers were either untreated (lanes 1, 3, or treated with 5600 5) IU/ml TNF (lanes 2, 4, 6). In a second experiment, the cells were inhibition of Consequently, electron entry to the ubiquinone untreated (lanes 7, or stimulated with a combination of db- 9, 11) of the pool respiratory chain, by complex I inhibitors cAMP (2 mM), cycloheximide (50 and staurosporine (6 nM, Ag/ml) (rotenone, amytal) or complex II inhibitors (thenoyltrifluoro- added at -2 h). After each stimulation, total cytoplasmic RNA was reduced TNF acetone), greatly cytotoxicity (Schulze-Osthoff extracted and immobilized onto a nylon membrane in a dot-blot et al., 1992). On the other when the device. Consecutive lanes correspond to 20 ,g RNA and serial 1:3 hand, electron flow dilutions, respectively. The membrane was hybridized with a 32p was inhibited distal to the ubiquinone site by the addition labeled DNA fragment derived from mIL-6 cDNA and exposed to an of complex HI inhibitors (antimycin A), increased amounts X-ray film. of ROI may be generated, resulting in a marked potentiation of TNF cytotoxicity. oxidative metabolism (reviewed in Halliwell and This Gutteridge, study extends the former findings and provides further 1990). Generally, ROI are regarded as toxic and harmful supportive evidence for the involvement of the mitochondrial metabolites. In phagocytes, large amounts of ROI are electron in TNF transport cytotoxicity. Subclones of L929 generated the by membrane-bound NADPH oxidase, cells which were of depleted mitochondrial respiration providing an antimicrobial defense mechanism against revealed a clear-cut resistance TNF towards cytotoxicity. pathogens (reviewed in Rossi, 1986). Nonetheless, when Depletion of mitochondrial respiration was achieved by long- formation goes uncontrolled, ROI may be implicated in term treatment of the cells with either ethidium bromide or several diseases and autoimmune disorders, leading to chloramphenicol which block mitochondrial DNA replication cellular damage by peroxidizing lipids and disrupting and protein synthesis, respectively (Nass, 1972; Wiseman structural proteins, enzymes and nucleic acids. An important and Attardi, 1978; Desjardins et al., 1985; Morais et al., source of ROI are mitochondria (Boveris and Chance, 1973; 1988; King and Attardi, 1989). The isolated cell clones did Turrens and Boveris, 1980; Turrens et al., 1985). not reveal any significant morphological changes. However, Overproduction by these organelles may be toxic and harful the growth rate of the cells was reduced since the energy for the producing cell. Several cytostatic or antitumor drugs, was mostly derived from anaerobic glycolysis. Most 3100 Depletion of mitochondrial functions abolishes TNF effects or DNA-binding importantly, receptor binding assays with iodinated TNF did of the p55/p65 complex into the nucleus, It is remarkable that, not reveal alterations of TNF receptor numbers, thus ruling activity at the target sequences. contrary of the out reduced TNF binding as a possible explanation of the of NFxB in the cytoplasm, binding to activation acquired resistance (data not shown). Surprisingly, when the p50/p65 NFxB complex to its target sequences in purified respiration-deficient cells were incubated with TNF in the vitro is even promoted under reducing conditions (Toledano presence of the transcriptional inhibitor actinomycin D, 1991; Xanthoudakis et al., 1992). In the and Leonard, which generally increases sensitivity, TNF resistance was the target of the latent NFxB-complex seems to cytoplasm, overcome. An explanation for this which is released by a poorly defined mechanism. again partially be IxB may be the existence of an alternative TNF- can envision a direct effect of ROI by influencing protein sensitization One activated pathway that is less prominent in the absence of folding or damaging the IxB protein. However, purified transcriptional inhibition. Indicative of a second cytotoxic NFxB-IxB cannot be simply activated in vitro by there be a may be the observation that mitochondrial inhibitors exogenous hydrogen peroxide, although may pathway intermediates or radical scavengers which protect parental L929 cells were selective induction of NFxB by specific oxygen it has been that not or only marginally effective in the respiration-deficient et al., 1992). Previously, reported (Schreck R is inducible following L929 clones, when tested in the presence of actinomycin D. the bacterial transcription factor oxy Furthermore, overexpression of manganous superoxide exposure to H202, whereas the bacterial sox R regulon has the control dismutase (MnSOD), which efficiently reduces TNF been defined as being responsible for positive at least in some cell types, is rather inefficient of proteins induced by 02 and distinct from H202 (Storz cytotoxicity at protection against a combination of TNF and et al., 1990; Demple and Amabile-Cuevas, 1991). providing metabolic inhibitors (Wong et al., 1989). However, it seems to be more likely that radicals indirectly in NFxB activation. An indirect role participate Redox regulation of transcriptional activators of radicals in NFxB activation by TNF might be exerted This study shows that mitochondrial intermediates are not the activation of specific proteases or kinases. In through in vitro conditions only implicated in TNF cytotoxicity, but may also be this respect, it should be noted that under complex (Ghosh and involved in TNF-mediated activation of the transcription protein kinase C can activate the NFxB A role of protein kinase C factor NFxB. Firstly, the respiration-deficient L929 Baltimore, 1990). physiological as TNF-induced subclones revealed a strongly reduced NFxB activation by or A, however, seems to be rather unlikely PKC or PKA inhibitors TNF. Secondly, in the parental L929 cells mitochondrial NFxB activation is not abolished by However, inhibitors, that interfered with TNF cytotoxicity, apparently (Hohmann et al., 1991; Meichle et al., 1991). described in recent times, affected NFxB activation in similar ways. Inhibitors, which several novel kinases have been et were protective by blocking an early step in the such as a redox-activated tyrosine kinase (Bauskin al., which in NFxB activation should mitochondrial electron transfer, strongly reduced NFxB 1991), the possible role of controlling redox activation by TNF. Vice versa, the complex IH inhibitor be examined. Other candidate proteins acting similarly to the which confers potentiation to TNF cytotoxicity, regulation may be specific factors antimycin A, 1 which mediates the redox exerted a synergistic effect on NFxB activation by TNF. recently cloned REF- protein, other activators These observations suggest an at least partially common control of AP-1 activity and transcriptional et al., 1992). for both the cytotoxic and NFxB-inducing effects in the nuclear compartment (Xanthoudakis pathway that NFxB activation is of TNF. In accordance with this, it has been reported Furthermore, it may be possible such molecules, which recently that activation of NFxB by different agents particularly influenced by thiol-containing TNF stimulation. Although it as IL-I or phorbol myristate acetate could be prevented may become oxidized upon TNF, in a T cell line that TNF did not antioxidants and chelation of iron ions (which is required has recently been shown by and levels to convert oxygen intermediates to the highly reactive the total cellular thiol glutathione (Israel change be modified radical). In another study, it has been shown that et al., 1992), specific antioxidant systems may hydroxyl In this it has to be irradiation, by which ROI are continuously and affect NFxB activation. context, ionizing contain a number of different redox can induce NFxB activation (Brach et al., 1991). pointed out that cells generated, not linked to each other. With respect to TNF signal transduction, our results also systems which are necessarily metabolism has been also suggested that TNF stimulation implicate a role of the mitochondrial oxidative Recently, it that ROI formed by involves the activation of an acidic sphingomyel- in NFxB activation and suggest of NFxB in this et which is a compartmentalized of the electron transfer are responsible inase (Schutze al., 1992), perversion we tested a contribution of other with a of 4.5. At present it is process. Moreover, when enzyme pH optimum pH such as NADPH oxidase or unclear how ceramide, the reaction product of potential radical sources, no involvement of these enzymes in TNF- control NFxB activation. It should xanthine oxidase, sphingomyelinase, might with cell lines activation was detectable. We also analyzed be noted that these results were obtained induced NFxB the influence of ATP on the activation of NFxB. unlike L929, were not or only weakly susceptible synthesis which, of it is still an like TNF early NFxB activation was to the action TNF; open question However, cytotoxicity, cytotoxic oxidative a role in the TNF-mediated not influenced of phosphorylation whether sphingomyelinase plays by uncouplers in L929 cells if whether ceramide or a direct inhibition of ATPase by oligomycin (Schulze- NFxB activation and, so, and our ROI formation. Osthoff et unpublished results). influences al., 1992, molecular mechanism Our results do not allow the by activated to be identified. Intracellular IL-6 as an oxidative stress-responsive cytokine which NFxB becomes regulation Another remarkable difference between the respiration- early transcriptional by radicals could regulate was in of transduction, L929 subclones and their NFxB at several points the pathway signal deficient parental counterparts TNF induction of IL-6 subunit translocation the gene the release of the inhibitory IxB, observed by studying e.g. 3101 et al. K.Schulze-Osthoff of IL-6 in the subclones expression. Expression was to changed mitochondrial functions. Early investigations in considerably reduced at the level and 200- transcriptional Drosophila have shown that application of mitochondrial to 600-fold when the biological was measured. This activity inhibitors can influence the puff expression pattern in observation indicates that not only TNF cytotoxicity, but also polytene chromosomes (Leenders et al., 1974, and IL-6 induction, may be mediated by reactive oxygen references therein). Yeast cells respond to the complex Ill intermediates. The reduction of IL-6 expression be may inhibitor antimycin A by induction of cytoplasmic citrate explained by a decreased activation of NFxB to a binding synthase expression (Liao et al., 1991). Moreover, in yeast xB element (-73 to -64) which has been the level of suggested by expression of nuclear genes, not all of them several authors to be indispensible for TNF- or IL-1-induced related to directly mitochondrial functions, was found to IL-6 expression (Libermann et al., 1990; Shimizu et change in response to the loss of the mitochondrial al., genome 1990; Zhang et al., 1990). However, the of (Parikh et al., 1987). These observations therefore indeed importance NFxB in determining TNF-induced IL-6 should provide additional evidence for the existence of a expression signal not be overemphasized. We have recently described the communication between the mitochondrial and nuclear potentiating effect of lithium ions in TNF-induced IL-6 compartment. Oxygen radicals may provide a versatile et Vandevoorde et cellular control mechanism in this signaling system since expression (Beyaert al., 1991; al., 1991), they we could not observe contribution of NFxB comprise very short-lived molecules which can be tightly although any in these phenomena. it should be taken into regulated by a machinery of different enzymes. Furthermore, account that NFxB was A major question that has to be addressed in future activation, though seriously reduced, still detectable to a extent in the subclones. experiments is the mechanism of how TNF interferes significant Thus, with it is conceivable that a cooperative reduction of additional the mitochondrial electron transport. There has been the activators accounts for the decrease of suggestion that alterations in mitochondrial calcium transcriptional sharp IL-6 expression. Ray et al. (1989) have defined a 26 homeostasis affect the mitochondrial respiratory chain bp region of the IL-6 promoter, called (Richter and Frei, 1988). However, it is not clear whether multiple responsive element (MRE, -173 to -151), which is in TNF- changes in calcium metabolism precede the oxidative stress important induced transcription of the IL-6 gene. In human fibroblasts, or are just a consequence of it. Another interesting issue will it has been further suggested that TNF induction of IL-6 be whether phosphorylation events, induced by TNF, may may be mediated by a cellular increase of cAMP et regulate the electron transport. Up to now, there have been (Zhang al., no 1988, 1990) which may be additionally involved in the studies concerning phosphorylation of respiratory chain activation of NFxB (Shirakawa and Mizel, proteins. One could speculate that an increase of radical 1989). Although a cAMP-responsive element is contained in the IL-6 production at the ubiquinone site of the respiratory chain (CRE) - in promoter (-163 to 158), it seems to be dispensible for results a reduced electron transfer from the ubiquinone TNF induction of IL-6 expression. In L929 cells, cAMP and cycle to the cytochrome chain. Thus, a target protein between TNF in a these induce gene expression synergistic manner, components could be a ubiquinone-binding protein, of which would suggesting separate pathways IL-6 induction (Beyaert regulate the electron flow to complex HI, or et for a of certain subunits of of al., 1991). Supportive distinct regulation gene cytochrome b the respiratory chain (Yu TNF is the that in et expression by and cAMP finding the al., 1985). Interestingly, in plant photosynthesis it has L929 subclones IL-6 could still be been shown that the respiration-deficient redox-regulated LHC kinase can regulate TNF was the electron flow induced by cAMP, whereas ineffective. by phosphorylation (Gal et al., 1992). The dual role of mitochondrial intermediates in both the Future studies may therefore reveal whether these or similar induction of IL-6 and earlier mechanisms cytotoxicity may explain transmit the TNF signal into the mitochondria IL-6 and findings that expression generally parallels the outcome thereby regulate mitochondrial radical generation. of TNF In studies with several activators or cytotoxicity. it has been shown that under conditions where inhibitors, Materials and methods was either enhanced or cytotoxicity decreased, IL-6 expression was modulated et TNF and reagents equally (Vandevoorde al., TNFpreparation. Recombinant murine TNF was produced in Escherichia 1991, 1992). Thus, IL-6 expression may be determined at coli and purified to at least 99% et The homogeneity (Fransen al., 1985). least in part by the same cytoplasmic reactions and preparation had a specific activity of 3.7 x 107 and IU/mg protein metabolites responsible for cellular That ROI toxicity. may contained <4 ng endotoxin/mg protein. TNF was determined as activity be the common denominator in these is further processes described by Ostrove and Gifford (1979) an international standard using TNF the fact preparation (code no. 88/532, obtained from the Institute for Biological suggested by that some radical can inhibit scavengers Standards and Control, Potters Bar, UK) as a reference. IL-6 expression (our unpublished results). Reagents. The following reagents were purchased from Sigma Chemical Oxygen radicals as a putative signal transducers of Co. (Deisenhofen, FRG): N-acetyl-L-cysteine, actinomycin D, amytal, mitochondria - nucleus pathway butylated hydroxyanisole AMP (BHA), dibutyryl cyclic (db-cAMP), 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), The data presented here define a previously unexpected role nordihydroguaiaretic acid (NDGA), oligomycin, pyrrolidine dithiocarbamate for mitochondria in the control of nuclear gene expression. (PDTC), rotenone and thenoyltrifluoroactetone (TTFA). Antimycin A was Apparently, mitochondrial intermediates can control both the from Serva (Heidelberg, FRG). Stock solutions of the reagents were routinely cytotoxic and gene-regulatory effects of TNF, thus providing prepared in medium, dimethylsulfoxide or ethanol as appropriate. a basis for a mitochondria-nucleus signaling pathway. Up Cell culture to there is now, only limited information regarding a The standard medium for L929 cells was Dulbecco's modified Eagle's proposed retrograde signaling pathway (reviewed in Grivell, medium supplemented with 10% newborn calf serum, 100 U penicillin/mI 1989), although there are several reported instances of and 0.1 mg streptomycin/ml. Isolation of respiration-deficient cells was alterations in gene expression of nuclear genes in response carried out by long-term culture in medium containing 50 yg 3102 functions abolishes TNF effects of mitochondrial Depletion or 400 ng ethidium bromide/ml. The selection medium the rate of oxidation of ferrocytochrome c (Smith and Conrad, 1956). Cells chloramphenicol/ml further supplemented with 4.5 mg glucose/ml and 5 uridine/ml, were scraped off and lysed with 0.3% digitonin in 10 mM potassium was Ag have been reported to become pyrimidine phosphate (pH 7.0). Enzyme reactions were started by the addition of 50 since respiration-deficient cells Attardi, 1989). Addition of these agents did not, reduced ferrocytochrome c and recorded at 25°C every 15 s at 550 auxotrophs (King and lsM c in the blank cuvette was fully oxidized with 1 mM however, influence any of the assays. nM. Ferrocytochrome oxidation were determined from the initial ferricyanide. Rates of cytochrome coefficient of 21/mM/cm. TNF cytotoxicity assay linear rates using an extinction Cells were seeded in 96 microwell plates at 3 x 104 cells in 100 dl synthesis medium. Twelve to 16 h later, 50 of a drug solution were given. TNF Mitochondrial protein Ad in 6 well plates at 1 x 106 cells/well; 1 h before the with or without actinomycin D (1 was added 2 h later in a 50 Cells were seeded Ag/ml) A1 100 cytoplasmic protein synthesis was volume and a concentration range of 0.7-5000 IU/ml. After 18-24 h of addition of [35S]methionine, lsCi/ml with emetine at a concentration of 100 After further further incubation, cell viability was routinely determined via MTT staining inhibited Ag/ml. were obtained with crystal violet staining for 3 h, the adherent cells were scraped off and washed three (Tada et al., 1986). Similar results incubation in PBS. were prepared in 20 mM Tris-HC1 (pH 7.4) of attached cells. times Lysates 8 M urea and 2% SDS. Samples were loaded without prior heating containing lnterleukin-6 assay on 12.5% SDS-polyacrylamide gels containing 6 M urea (Chomyn et al., 6 well plates at 106 cells/well; 16-24 h later cells 1985). Gels were dried and exposed to X-ray films. Cells were seeded in twice and treated in 1 ml medium/well containing various TNF were washed h of incubation, the medium was Mitochondrial DNA analysis concentrations. Following another 16-24 IL-6. IL-6 was determined on mtDNA was prepared from the cells by the method of Wallace taken off and assayed in serial dilutions for Enriched mouse plasmocytoma cell line HindIl restriction fragments of the DNA were electrophoresed the basis of the proliferative response of the et al. (1988). in TAE buffer. After depurination and denaturation 7TD1 (Van Snick et al., 1986). in 0.7% agarose gels were blotted by capillarity in according to standard procedures, x onto N+ membranes (Amersham). DNA was fixed by 20 SSC Hybond Nuclear and whole cell extracts UV and in 5 x SSPE, 0.5% SDS and irradiation prehybridized Subconfluent cell cultures in a 6 well plate or 60 cm2 Petri dishes were 1 x For hybridization, an AccI insert, cloned in a Denhardt's solution. treated with TNF for 45 min. Inhibitors were applied 1 h before the onset vector and coding for the 12S rRNA of rat mtDNA (Gadaleta Bluescript of TNF addition. After incubation cells were rinsed in ice-cold phosphate- was labeled extension of random hexanucleotide primers. et al., 1989), by buffered saline, scraped off using a rubber policeman and collected by After at 45°C for 18 h, blots were washed in 0.5 x SSC, hybridization centrifugation. Nuclear extracts were prepared essentially as described by 0.1% SDS and exposed to X-ray films. Dignam et al. (1983). Cells were resuspended in 1 ml of a hypotonic lysis buffer (buffer A, Dignam et al., 1983). After 20 min, cells were RNA manipulations homogenized by 20 strokes with a loose-fitting Dounce. Nuclei were collected as previously Total cytoplasmic RNA was isolated essentially described in a microcentrifuge and proteins extracted in a high-ionic-strength buffer buffer. mRNA and Bancroft, 1982) using an NP40 lysis IL-6-specific (White was removed by centrifugation. (buffer C). After 30 min, nuclear debris a restriction fragment of the murine was quantitated in dot-blot analyses using The extracts were diluted with 4 vol of low-salt buffer (pH 7.5) (buffer Snick et al., 1986). IL-6 cDNA containing plasmid pUC8MIL6 (Van D; Dignam et al., 1983) containing 1% NP-40, and kept frozen until use. RNA were denatred with formaldehyde Appropriate amounts of the isolated cells were lysed in a high-salt buffer (20 mM For whole-cell extracts, A, Pall Biosupport, East and applied onto a nylon membrane (Pall Biodyne HEPES, pH 7.5, 400 mM NaCl, 1 mM MgCl2, 0.5 mM EDTA, 0.1 mM x SSC and immobilized by heat treatment Hills, NY) in the presence of 5 1% NP-40, 0.5 mM DTT, 10 leupeptin, 1 EGTA, 20% glycerol, h 42°C) and hybridization (overnight (1 h at 80°C). Prehybridization (1 at Ag/ml After 10 min of incubation, extracts were mM PMSF, 0.001% aprotinin). at carried out in a mixture containing 5 x Denhardt's solution, 42°C) were and kept frozen. cleared by centrifugation 5 x SSC, 50 mM sodium phosphate (pH 6.5), 0.1% SDS, 250 non- jig/ml DNA formamide. The washing steps were carried homologous and 50% mobility shift assays Electrophoretic in 2 x SSC/0. SDS x 5 min) and finally 0.1 x SSC/0. 1% SDS. out 1% (4 ( - 3 nuclear protein or -10 Itg crude Equal amounts of the extracts jg the Bio-Rad assay kit) were incubated with an NFxB- protein; determined by The sequence of the double-stranded specific 32P-labeled oligonucleotide. Acknowledaements NFxB site and flanking sequences from the 30 bp probe encompassing the is shown in Zabel et al. (1991). Binding mouse kappa light chain enhancer and di The authors wish to thank Drs C.Saccone E.Sbisa (Universita Bari) performed in a 20 1l volume containing 2-4 1l of extract, reactions were for DNA and Dr T.Wirth Heidelberg) kindly providing probes. (ZMBH, x buffer (20 mM HEPES pH 7.5, 50 mM KC1, 1mM DTT, 4 5 binding a from the Bundesministerium K.S.-O. acknowledges fellowship gratefully 2 as non-specific competitor DNA, 2.5 mM MgCl2, Ficoll), poly(dI-dC) R.B. and V.V. are fur und Technologie (AIDS-Stipendium). Forschung itg 2 serum albumin (BSA) and 3000-6000 c.p.m. Cerenkov of bovine with the Nationaal Fonds research assistants and G.H. a research director Ag oligonucleotide. After 30 min binding reaction the random primer-labeled The research was supported by voor Wetenschappelijk Onderzoek. project room were loaded on a 4% non-denaturing at temperature, samples from the Fonds voor Geneeskundig Wetenschappelijk Onderzoek, grants and run in 0.5 x TBE buffer (pH 8.0). As a negative polyacrylamide gel the en and Interuniversitaire Attractiepolen, Algemene Spaar- Lijfrentekas an mutated in the NFxB binding site by two point control, oligonucleotide Kanker. the Sportvereniging tegen mutations, but otherwise identical with the wild-type fragment, was used. In all cases, DNA-NFxB complexes were only observed using the wild- In some experiments, the DNA-binding activity of type oligonucleotide. References the extracts was by scintillation counting of the 32P-labeled quantified and Curran,T. (1990) Science, 249, NFxB-DNA complexes excised from the dried gels. In the case of cytosols, Abate,C., Patel,L., Rauscher,F.J,llI 1157-1161. was activated by incubating 2 of the cytosolic extract with 3 NFxB Al ILI Rahmsdorf,H.J., for 10 min before addition to a mixture of probe. Imagawa,M., Chiu,R., Stein,B., Imbra,R.J., 6.5% sodium deoxycholate Angel,P., and 729-739. For the of the DNA-binding activities of AP-1 and octamer binding Jonat,C. Herrlich,P. Karin,M. (1987) Cell, 49, assay and (1992) Cancer Res., 52, (Oct), the oligonucleotides used in this study have been described Amstad,P.A., Krupitza,G. Cerutti,P.A. proteins et The NF- 3952-3960. Angel et al., 1987; Oct, Wirth al., 1987). previously (AP-1, motif from the human Biochim. Biophys. Acta, 1072, 63-80. IL6 specific oligonucleotide containing the binding Baeuerle,P.A. (1991) was synthesized and Science, 242, 540-546. IL-6 (5'-GTACACATTGCACAATCTTA-3') Baeuerle,P.A. Baltimore,D. (1988) promoter 685-696. and Ben-Neriah,Y. (1991) Cell, 66, on an Applied Biosystems DNA Synthesizer. Bauskin,A.R., ALkalay,I. Necrosis Factors: Molecules and Their Tumor Ihe Beutler,B. (1992) Raven New York. Role in Medicine. Press, Anaerobic glycolysis Emerging Annu. Rev. 625-655. were washed twice with phosphate-buffered and Cerami,A. (1989) Immunol., 7, Exponentially growing cells Beutler,B. Van and incubated at time zero in DMEM with 10% dialyzed fetal Schulze-Osthoff,K., Roy,F. Fiers,W. (1991) Cytokine, saline (PBS) and Beyaert,R., 12 h, supernatants were assayed for lactate content 3, 284-291. calf serum (FCS). After and Trends Biochem. lactate Cellular Kourilsky,P. Isradl,A. (1992) Sci., 17, with a commercial kit (Sigma) using dehydrogenase. Blank,V., 135-140. measured a method analyzing oxidase was by spectrophotometric cytochrome 3103 K.Schulze-Osthoff et al. and Boveris,A. Chance,B. (1973) Biochem. J., 707-716. and Mol. 134, Ray,A., Sassone-Corsi,P. Sehgal,P.B. (1989) Cell. Biol., 9, Boveris,A., Cadenas,E. and Stroppani,A.O.M. (1976) Biochem. 5537-5547. J., 156, 435-445. and Free Rad. Biol. 365-375. Richter,C. Frei,B. (1988) Med., 4, Brach,M.A., Hass,R., Shernann,M.L., Gunji, H., and Biochim. 65-89. Weichselbaum,R. Rossi,F. (1986) Biophys. Acta, 853, Kufe,D. (1991) J. Clin. Invest., 88, 691-695. Ruggiero,V., Johnson,S.E. and Cell. Baglioni,C. (1987) Immunol., 107, Brenner,D.A., O'Hara,M., Angel,P., Choikier,M. and Karin,M. (1989) 317-325. Nature, 337, 661-663. Schreck,R., Rieber,P. and Baeuerle,P.A. EMBO 2247-2258. (1991) J., 10, Bromsztyk,K., Rooney,J.W., Iwasaki,T., Rachie,N.A., Dower,S.K. and Schreck,R., Meier,B., Mannel,D., and Dr6ge,W. Baeuerle,P.A. (1992) J. Sibley,C.H. (1991) Cell Reg., 2, 329-335. 1181-1194. Exp. Med., 175, Cadenas,E. and Boveris,A. (1980) Biochem. J., 188, 31-37. Schulze-Osthoff,K., Bakker,A.C., Vanhaesebroeck,B., Beyaert,R., Cerutti,P.A. (1985) Science, 227, 375-381. Jacobs,W. and J. Biol. 5317-5323. Fiers,W. (1992) Chem., 267, Chomyn,A., Mariottini,P., Cleeter,M.W.J., Ragan,C.I., Schuitze,S., and Matsuno-Yagi,A., Potthoff,K., Machleidt,T., Berkovic,D., Wiegmann,K. Hatefi,Y., Doolittle,R.F. and Attardi,G. (1985) Nature, 314, 592-597. 765-776. Kr6nke,M. (1992) Cell, 71, and Del Arch. Cino,M. Maestro,R.F. (1989) Biochem. Shibanuma,M., Kuroki,T. and Nose,K. (1988) Oncogene, 17-21. Biophys., 269, 3, 623-638. Shimizu,H., Mitomo,K., Watanabe,T., and Okamoto,S. Yamamoto,K.J. and Crawford,D., Zbinden,J., Amstad,P. Cerutti,P.A. (1990) Mol. Cell. Biol., 561-568. (1988) Oncogene, 10, 27-32. 3, Shirakawa,F. and Mol. Cell. 2424-2430. Mizel,S. (1989) Biol., 9, Cromlish,J.A. and J. Roeder,G. (1989) Biol. Chem., 264, 18100-18109. Silva,C.M. and Cidlowski,J.A. J. Biol. 6638-6647. (1989) Chem., 264, Dealtry,G.B., Naylor,M.S., Fiers,W. and Balkwill,F.R. Eur. J. Smith,L. and Conrad,H. (1956) Arch. Biochem. Biophys., 403 -410. (1987) 63, 689-693. and Immunol., 17, Spolsky,C.M. Eisenstadt,J.M. FEBS 319-324. (1972) Lett., 25, Defilippi,P., Poupart,P., Tavernier,J., Fiers,W. and Content,J. Proc. Staal,F.J.T., Roederer,M., and (1987) Herzenberg,L.A. Herzenberg,L.A (1990) Nati. Sci. 4557-4561. Proc. Natl. USA, 84, Acad. Sci. USA, 87, 9943-9947. Demple,B. and Amabile-Cuevas,C.F.(1991) Cell, 67, 837-839. Stein,B., Rahmsdorf,H.J., Steffen,A., Litfin,M. and Herrlich,P. (1989) Desjardins,P., and Mol. Cell. 5169-5181. Frost,E. Morais,R. (1985) Mol. Cell. Biol., 9, Biol., 5, 1163-1169. and Storz,G., Tartaglia,L.A. Ames,B.N. (1990) Science, 248, 189-194. Devary,Y., and Suffys,P., Van and Gotdieb,R.A., Laus,L.F. Karin,M. (1991) Mol. Cell. Beyaert,R., Roy,F. Fiers,W. (1987) Biochem. Biol., Biophys. 11, 2804-2811. Res. 7835-7843. Commun., 149, and Nucleic Van and Dignam,J.P., Lebovitz,R.M. Roeder,R.G. (1983) Acids Suffys,P., Beyaert,R., Roy,F. Fiers,W. (1988) Eur. J. Res., Biochem., 11, 1475-1489. 257-265. 178, Duh,E.J., and De Maury,W.J., Folks,T.M., Fauci,A.S. Suffys,P., Beyaert,R., Valck,D., Vanhaesebroeck,B., Van and Rabson,A. (1989) Roy,F. Proc. Natl. Acad. Sci. 5974-5978. Eur. J. USA, 86, Fiers,W. (1991) Biochem., 195, 465-475. Fiers,W. (1991) FEBS 199-212. Lett., 285, Tada,H., Shiho,O., Kuroshima,K., Koyama,M. and Tsukamoto,K. (1986) Van der J. Immunol. Methods, 157-165. Fransen,L., Muiller,R., Marmenout,A., Tavernier,J., 93, Heyden,J., Kawashima,E., Chollet,A., Van Van Toledano,M.B. and Proc. Tizard,R., Heuverswyn,H., Leonard,W.J. (1991) Natl. Acad. Sci. USA, 88, Vliet,A., and Nucleic Acids 4328-4332. Ruysschaert,M.R. Fiers,W. (1985) Res., 13, 4417-4429. Turrens,J.F. and Boveris,A. Biochem. (1980) J., 191, 421-427. Gadaleta,G., De and Turrens,J.F., Alexandre,A. and Pepe,G., Candia,G., Quagliarello,C., Sbisa,E. Lehninger,A.L. (1985) Arch. Biochem. Saccone,C. J. Mo. 497-516. Biophys., 408-414. (1989) Evol., 28, 237, Gal,A., Hermann,R.G., and Ohad,J. Vandevoorde,V., Haegeman,G. and Lottspeich,F. (1992) FEBSLett., 298, Fiers,W. (1991) Biochem. Biophys. 33-35. Res. Commun., 178, 993-1001. Ghosh,S. and Baltimore,D. 678-682. Vandevoorde,V., Haegeman,G. and Fiers,W. (1992) (1990) Nature, 344, FEBS Lett., 302, Eur. Grivell,L.A. (1989) J. 477-493. 235-238. Biochem., 183, Van Guehmann,S., and Snick,J., Cayphas,S., Vink,A., Uyttenhove,C., Vorbrueggen,G., Kalkbrenner,F. Moelling,K. (1992) Coulie,P.G., Nucleic Acids Res., 20, 2279-2286. Rubira,M.R. and Simpson,R.J. (1986) Proc. Natl. Acad. Sci. USA, 83, and Halliwell,B. Gutteridge,J.M.C. Methods 1-85. 9679-9683. (1990) Enzymol., 186, Hohmann,H.P., Kolbeck,R., Remy,R. and van Mol. Wallace,D.D., Zheng,X.X., Lott,M.T., Shoffner,J.M., Loon,A.P. (1991) Cell. Hodge,J.A., 2315-2318. Biol., 11, Kelley,R.I., Epstein,C.M. and Hopkins,L.C. (1988) Cell, 55, 601-610. and Holbrook,N.J. Fornance,A.J.,Jr New 825-833. White,B.A. and Bancroft,F.C. (1982) J. Biol. Chem., 257, 8569-8572. (1991) Biol., 3, Wirth,T., Israel,N., Gougerot-Pocidalo,M.-A., Aillet,F. and Staudt,L. and Baltimore,D. (1987) Nature, 329, Virelizier,J.-L. (1992) 174-178. J. Immunol., 149, 3386-3393. Wiseman,A. and Attardi,G. (1978) Mol. Gen. Genet., 167, 51-63. and Wong,G.H.W., King,M.E., Godman,G.C. King,D.W. (1972) J. Cell Elwell,J.H., Oberley,L.W. and Goeddel,D.V. (1989) Biol., 53, Cell, 127-142. 58, 923-931. King,M.P. and Attardi,G. Xanthoudakis,S., (1989) Science, 246, 500-503. Miao,G., Wang,F., Pan,Y.-C.E. and Curran,T. (1992) Konstantinov,A.A., EMBO J., 11, 3323-3335. Peskin,A.V., Popova,E.Y., Khomutov,G.B. and Ruuge,E.K. (1987) Biochim. Yamauchi,N., Kuriyama,Y.H., Biophys. Acta, 894, 1-10. Watanabe,N., Neda,H., Maeda,M. and Leenders,H.J., Berendes,H.D., Niitsu,Y. (1989) Cancer Res., Helmsing,P.J., Derksen,J. and 49, 1671-1675. Koninkx,J.F.J.G. (1974) Sub-Cell. Yamauchi,N., Watanabe,N., Biochem., 3, 119-147. Kuriyama,H., Neda,H., Maeda,M., and Lenardo,M. Baltimore,D. Himeno,T. and Tsuji,Y. (1990) Int. J. (1988) Cell, 58, 227-229. Cancer, 46, 884-888. Levy,D.E., Kessler,D.S., Pine,R. and Darnell,J.E., Jr (1989) Genes Yu,L., Yang,F.D. and Yu,C.A. (1985) Dev., J. Biol. Chem., 260, 963 -973. 3, 1362-1371. Zabel,U., Schreck,R. and Baeuerle,P.A. (1991) J. Biol. Chem., 266, Liao,X., Small,W.C., Srere,P.A. and Butow,R.A. (1991) Mol. Cell. 252-260 (1991). Biol., 11, 38-46. Zhang,Y., Lin,J.-X. and Vileek,J. (1988) Proc. Natl. Acad. Sci. USA, 85, Libermann,T.A. and Baltimore,D. (1990) Mol. Cell. Biol., 10, 2327-2334. 6802-6805. Loschen,G., Azzi,A., Richter,C. and Flohe,L. (1974) FEBS Lett., 42, Zhang,Y., Lin,J.-X. and Vileek,J. J. (1990) Biol. Chem., 263, 6177-6182. 68-72. Zimmerman,R.J., Chan,A. and Cancer Leadon,S.A. (1989) Res., 49, Meichle,A., Schiitze,S., Hensel,G., Brunsing,D. and Kronke,M. (1990) 1644-1648. J. Biol.Chem., 265, 8839-8843. Mihm,S., Ennen,J., Pessara,U., Kurth,R. and Droge,W. Received (1991) AIDS, 5, on February 15, 1993; revised on April 22, 497-503. Morais,R., Desjardins,P., Turmel,C. and Zinkewich-Peotti,K. (1988) In Vitro Cell. Dev. Biol., 24, 649-658. Nass,M.M.K. (1972) Exp. Cell Res., 72, 211-222. Niki,E. (1987) Chem. Phys. Lipids, 44, 227-232. Osbom,L., Kunkel,S. and Nabel, G.J. Proc. (1989) Natl. Acad. Sci. USA, 86, 5974-5978. and Ostrove,J. Proc. Soc. 354-358. Gifford,G. (1979) Exp. Med., 160, Parikh,V.S, and Morgan,M.M., Scott,R., Clements,L.S. Butow,R.A. 576-580. Science, 235, (1987) http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png The EMBO Journal Springer Journals

Depletion of the mitochondrial electron transport abrogates the cytotoxic and gene‐inductive effects of TNF.

Download PDF
10 pages

Loading next page...
 
/lp/springer-journals/depletion-of-the-mitochondrial-electron-transport-abrogates-the-AImxVMp9gP

References (3)

Publisher
Springer Journals
Copyright
Copyright © European Molecular Biology Organization 1993
ISSN
0261-4189
eISSN
1460-2075
DOI
10.1002/j.1460-2075.1993.tb05978.x
Publisher site
See Article on Publisher Site

Abstract

The EMBO Journal vol.12 no.8 pp.3095-3104, 1993 Depletion of the mitochondrial electron transport abrogates the cytotoxic and gene-inductive effects of TNF 1991) or proteases (Ruggiero et al., 1987; Suffys et al., Klaus Schulze-Osthoff1, Rudi Beyaert, 1988) or DNA damage (Dealtry et al., 1987). Recently, we Vandevoorde, Guy Haegeman and Veronique have provided evidence that in the classical TNF-sensitive Walter Fiers2 fibrosarcoma cell lines L929 and WEHI 164 clone 13 a Laboratory of Molecular Biology, University of Ghent, Ghent, functional mitochondrial respiratory chain is crucial to the Belgium events leading to TNF cytotoxicity (Schulze-Osthoff et al., lPresent address: German Cancer Research Center, Department of 1992). Immunology and Genetics, Heidelberg, FRG While the cytotoxic activity of TNF seems to be rather author 2Corresponding restricted to tumor cells, nearly every cell type responds to Communicated by W.Fiers of different genes. TNF by the activation of a wide range Owing to this broad gene-activating effect, TNF is suggested Tumor necrosis factor (TNF?) has cytotoxic and gene- differentiation and as a cardinal mediator of several inductive activities on several cell types. Previous studies TNF has been immunoregulatory processes. Among others, on L929 fibrosarcoma cells have revealed that the of other such as shown to induce expression cytokines mitochondrial electron transport system plays a key role (IL-6) or GM-CSF, and interleukin-I (IL-1), interleukin-6 in inducing TNF cytotoxicity, presumably by the as receptors and cell surface-bound proteins such cytokine formation of reactive oxygen intermediates (ROI). Here in Furthermore, adhesion molecules (reviewed Fiers, 1991). we report that mitochondria-derived intermediates are of HIV-1 is also inducible transcriptional virus replication not only cytotoxic but, in addition, function as signal Osborn et al., 1989). by TNF (Duh et al., 1989; transducers of TNF-induced gene expression. The has been found that TNF gene- In the last few years, it activation of NFxB, which fulfills an important role in are mediated by certain transcriptional regulatory activities TNF-induced gene transcription, could be blocked by include the factors AP-1 activators, which transcription interference with the mitochondrial electron transport most importantly, (Brenner et al., 1989) and, probably system. Furthermore, antimycin A, a mitochondrial The latter was initially NFxB (Osborn et al., 1989). inhibitor that increases the generation of ROI, potentiated a factor of mature B cells, that identified as nuclear TNF-triggered NFxB activation. The dual role of with an enhancer element of the specifically interacts mitochondria-derived intermediates in cytotoxicity and chain in Lenardo immunoglobulin kappa light gene (reviewed immediate-early gene induction of TNF was further Baeuerle, 1991; Blank et al., 1992). and Baltimore, 1988; which lacked substantiated by isolating L929 subclones of a heterodimeric complex, which in its NFxB consists of the a functional respiratory chain. This depletion form the two proteins p50 and p65. It classical comprises resulted in resistance mitochondrial oxidative metabolism that or structurally related proteins now appears NFxB as in inhibition of towards TNF cytotoxicity, as well Rel are ubiquitously present in other belonging to the family NFxB activation and interleukin-6 gene induction by and are inducible a wide variety of extracellular cell types by TNF. These findings suggest that mitochondria are the NFxB controls the inducible expression of various stimuli. source of second messenger molecules and serve as involved in immune responses, as well as genes that are common mediators of the TNF-cytotoxic and gene- and cellular defense mechanisms. Target genes inflammatory regulatory signaling pathways. include receptors, MHC of NFxB cytokines, cytokine words: cytotoxicity/interleukin-6/NFxB/respiratory Key and several viral enhancers. antigens, acute-phase proteins factor chain/tumor necrosis A characteristic feature of immediate-early NFxB activation is that it occurs without new protein synthesis. In the uninduced cell, NFxB resides cytoplasmically as a Introduction latent form to the inhibitory protein IxB (Baeuerle complexed cell NFxB is tumor necrosis factor (TNF) elicits on a variety and 1988). Upon stimulation, The cytokine Baltimore, release from which allows of in Beutler activated by its IxB, of cells two major types responses (reviewed rapidly into the nucleus and Beutler, 1992). First, TNF migration of the p50/p65 complex and Cerami, 1989; Fiers, 1991; The dissociation of which exerts a rather selective activity on several tumor subsequent DNA binding. IxB, cytotoxic event of NFxB is still rather and transformed whereas normal cells are generally seems to be the key activation, cells, mechanism of TNF In it has been shown that kinase C towards the cytokine. The elusive. vitro, protein resistant understood. eIF-2 kinase can activate the transcription is currently rather incompletely or heme-regulated cytotoxicity it seems different cellular events and and However, on the cell type, factor (Ghosh Baltimore, 1990). Depending of these kinases or kinase A is to TNF-induced cell such that one protein may contribute killing, unlikely processes in vivo et of reactive intermediates involved in NFxB activation (Meichle al., 1990; as generation oxygen (Zimmerman Hohmann et Yamauchi et Schulze-Osthoff Bromsztyk et al., 1991; al., 1991). Quite et al., 1989; al., 1989, 1990; that activation of NFxB be activation of (Suffys et al., it was proposed may et al., 1992), phospholipases recently, Press Oxford University K.Schulze-Osthoff et al. NADH ..... ... .. .... ",;-. V. FMN - [F.S| Complex I Antimycin A Rotenone cyt.aa3 cyt.c, wt.b rN&* lo uQ - -_ _ cyt.c -4° Complex ill ..... Complex IV Complex 11 TTFA FAD -_ [Fe-S] SUCCINATE Fig. 1. Schematic of the electron carriers of the diagram mitochondrial respiratory chain. The sites of inhibition the inhibitors by respiratory rotenone, TTFA and A are indicated. antimycin generally controlled by reactive oxygen intermediates (ROI) and the cellular redox state (Staal et Mihm al., 1990; et al., .:; .. :: .,y;I 1991; Schreck et al., at in 1991, 1992). Firstly, least some cell types the transcription factor could be activated by i h LiI: dA ,. 4,.. exogenously applied hydrogen peroxide. Secondly, activation of NFxB by different such as stimuli, TNF, IL-1, phorbol esters, viral transactivator or proteins, cycloheximide, was commonly inhibited by antioxidant thiols or iron chelators. __ Thus, these observations suggest that oxygen radicals acting as second messengers the diverse of may integrate variety NFxB-inducing signals. In our previous studies, we have shown that mitochondrial intermediates, possibly oxygen radicals generated from the respiratory chain, are causally involved in the cytotoxic effects of TNF (Schulze-Osthoff et al., 1992). Presumably, TNF causes a short circuit of the electron flow, resulting in direct production of radicals from the ubiquinone site of the chain. Here respiratory we report that mitochondria are not in only implicated cytotoxicity, but may also participate in the gene-regulatory effects of TNF. Our data that suggest gene induction and TNF cytotoxicity by share an at least Fig. 2. Differential effects of mitochondrial inhibitors on NFxB common partially pathway. We show that elimination of the binding activity in L929 cells. (A) Competition analysis showing the mitochondrial oxidative metabolism not only inhibits TNF of specificity NFxB binding: 3 of nuclear extract obtained from jig L929 cells treated cytotoxicity, but also reduces without (lane 1) or with TNF (1000 IU/ml; lanes considerably TNF-mediated were 2-4) analyzed in electrophoretic mobility shift assays with a 32p- activation of NFxB and of expression IL-6 as an example labeled probe encompassing the xB-binding site of the mouse kappa of a TNF-inducible, NFxB-regulated gene. Oxygen inter- chain light gene. To the nuclear extracts a 100-fold excess of mediates may therefore act as signal transducers and have unlabeled mutated or (lane 3) wild-type oligonucleotide (lane 4) were a more widespread role than mixed with the radioactive xB enhancer currently appreciated. probe. (B) The effect of rotenone on NFxB activation in L929 cells. Cells were treated for 1 h with the indicated concentrations of rotenone prior to TNF addition (1000 IU/ml). Incubation was continued for 45 min before extraction Results of the cells. (C) The effect of rotenone on DNA binding of nuclear extracts in vitro. Identical nuclear extracts from cells treated with TNF Mitochondrial inhibitors modulate TNF-induced NFxB (1000 IU/ml, min) were incubated for 30 min with the indicated activation concentrations of rotenone, followed by the addition of the 32P-labeled The mitochondrial electron transport can be inhibited by xB The effect of probe. (D) antimycin A on NFxB activation in L929 specific inhibitors which prevent electron flow between cells. Cells were 1 pretreated for h with the indicated concentrations of antimycin followed the certain of the A, by addition of a suboptimal dose (10 complexes respiratory chain (Figure IU/ml) 1). of TNF (lanes or medium 18-20) (lanes 21-23). After 45 min of Previously, we have reported that inhibition of the electron further incubation, nuclear extracts were prepared and analyzed by entry from complex I to the I ubiquinone by complex electrophoretic mobility shift assays. inhibitors amytal and rotenone results in a strong reduction of TNF cytotoxicity (Schulze-Osthoff et al., 1992). However, when electron flow in L929 cells was inhibited largely mediated by the generation of ROI at the ubiquinone distal to the ubiquinone pool, by the addition of site of the mitochondrial chain. antimycin Early investigations with A, a marked of potentiation TNF-induced killing was isolated mitochondria have repeatedly shown that ubiquinone observed. As an for explanation these differential is the major radical source of ROI drug derived from the effects, it was that TNF-induced cell mitochondrial chain proposed killing is (Boveris and Chance, 1973; Turrens and 3096 Depletion of mitochondrial functions abolishes TNF effects I..% Boveris, 1980; Turrens et al., 1985). ROI generation in B NDGA isolated mitochondria can be inhibited by amytal and rotenone, whereas antimycin A leads to an increase in ROI formation (Boveris et al., 1976; Cadenas and Boveris, 1980; Konstantinov et al., 1987; Cino and Del Maestro, 1989). We were interested in analyzing the effects of mitochondrial inhibitors on early TNF effects such as activation of the transcription factor NFxB, which has been reported, at least in some cell types, to be additionally I .=. I. inducible by hydrogen peroxide (Schreck et al., 1991, 1992). L929 cells were preincubated for 1 h with different concentrations of mitochondrial inhibitors; 45 min after TNF stimulation cells were lysed, and cellular and nuclear lysates were analyzed in electrophoretic mobility shift assays (EMSA). TNF induced a xB binding activity in L929 cells which was not detectable in unstimulated cells (Figure 2A). Competition with a 100-fold molar excess of unlabeled oligonucleotide led to an inhibition of xB binding, thereby confirming the specificity of the activated NFxB. No reduction of DNA binding, however, was observed with an excess of mutated oligonucleotide. .1 -. 1 910 12 concentrations When cells were pretreated with non-toxic inhibition of of the complex I inhibitor rotenone, a strong Oxygen radical scavengers reduce activation of NFxB in TNF- Fig. 3. was observed (Figure 2B). Even a NFxB activation L929 cells. Cells were pretreated for 1 h with various stimulated 5 inhibition of -50% of concentration of led to an concentrations of BHA (A) or NDGA (B) prior to the addition of AM NFxB binding. There remained the possibility that decrease 1000 IU/ml TNF and further 45 min of incubation. Equal amounts of total cell lysates (10 yg) were analyzed by electrophoretic mobility due to a reduction of of xB binding in EMSA was either shift assays. NFxB activation in the cytoplasm or to a reduced DNA- of To binding affinity the released NFxB complex. answer Neither allopurinol, an inhibitor of cytoplasmic xanthine this question, rotenone was added directly to the lysates. of the inhibitor could be detected oxidase, nor diphenylene iodonium (DPI), an inhibitor of However, no interference 2C). reduction of xB binding by rotenone the plasma membrane-bound NADPH oxidase, interfered (Figure Therefore, (data not shown). was most caused by inhibition of TNF-induced with TNF-induced NFxB activation probably in TNF-sensitive cell line NFxB activation in the cytoplasm. Therefore, NFxB activation the specifically controlled by When Ill inhibitor antimycin A, L929 seems to be largely and we analyzed the complex from chain. which TNF cytotoxicity, contrasting effects on intermediates derived the respiratory potentiates of mitochondrial ROI for the NFxB activation became evident (Figure 2D). When a The specific involvement nuclear extracts TNF dose was applied, antimycin A activation of NFxB was tested by assaying suboptimal (10 IU/ml) of other factors. a of NFxB activation. for the DNA-binding activity transcription triggered strong potentiation of Oct-i and NF-IL6 are Sometimes A itself induced a faint xB band at DNA-binding activities AP-1, antimycin in cells and inducible the concentration. A effect of constitutively present L929 barely upon highest synergistic antimycin 4 shows that of L929 observed with TNF doses, which TNF stimulation. Figure pretreatment A was only suboptimal inhibited TNF activation of be indicative for a similar action mechanism of both cells with rotenone selectively may activities of AP-1, Oct-1 NFxB, whereas the DNA-binding agents. in the same nuclear In accordance with our previously published data regarding and NF-IL6 remained unchanged This that inhibition of mitochondrial ROI TNF-mediated cytotoxic effects on mitochondria (Schulze- extracts. suggests indicate an formation modulates the activation of NFxB. Osthoff et the observations specifically al., 1992), present involvement of mitochondria-derived intermediates, possibly of NFxB. The Depletion of the mitochondrial respiratory chain ROI, also in the activation process ROI was further results in acquired TNF resistance activation of NFxB by presumptive the dual involvement of In order to substantiate further using butylated hydroxyanisole supported by experiments and NFxB acid two mitochondria for both TNF effects, cytotoxicity (BHA) or nordihydroguaiaretic (NDGA), potent of L929 which were 3 shows that both we isolated subclones antioxidants Figure agents activation, (Niki, 1987). chain. We also the effect of two deficient in a functional respiratory Respiration reduced NFxB binding. analyzed culture of L929 cells have been described as efficient deficiency was generated by long-term antioxidants which recently bromide or of NFxB activation et 1991, 1992). in ethidium (EB)- chloramphenicol (CA)- inhibitors (Schreck al., which have been in our cell inhibition was only observed containing medium, procedures repeatedly However, system, used to cells from a functional mitochondrial electron rather concentrations of (30 mM), deplete at high N-acetylcysteine Wiseman and of reduced Attardi, 1978; radical and transport system (Nass, 1972; a scavenger precursor glutathione, Morais et and 100 an iron et al., 1988; King and of dithiocarbamate Desjardins al., 1985; pyrrolidine (PDTC, AM), Ethidium bromide has been described as a not Attardi, 1989). chelator (data shown). of mitochondrial DNA be different in several inhibitor (mtDNA) As ROI may generated by enzymes specific and et we also the contribution of (Wiseman Attardi, 1978; Desjardins al., cellular analyzed replication compartments, is known to block mitochondrial sources for the induction of NFxB. 1985). Chloramphenicol other ROI potential 3097 K.Schulze-Osthoff at a!. C ~ ~ v b~(-', b ll~ rotenone <>M - - - ~ C A\~~~~~1 C70 <4~0 N F S. Pt NFkB (V- MW _ il '""fl..;I, N4- ND)2 - N'11 - AP 1 I(1)1E11 (.\I I .. b _ a_ N(I) 0(CZ , gi,;-. , 0* 0 'l r <1- NF-IL6 ~~~~~~~~ ~l, r,~ .@ k~,'$<,~ ' ""'II| NO Fig. 4. The effect of rotenone on different DNA-binding activities. L929 cells were pretreated for 1 h with the indicated concentrations of rotenone, followed by a stimulation with TNF (1000 IU/ml) for 90 with 32P-labeled DNA min. Nuclear extracts were analyzed in EMSA of probes detecting binding activities NFxB, AP-1, octamer-binding the shifted bands are shown. proteins (Oct-1) and NF-IL6. Only 5. of mtDNA-encoded and mtDNA Fig. Analysis protein synthesis (A) oxidase and lactate in L929 levels in ethidium bromide- and L929 Table I. Cytochrome activity production (B) chloramphenicol-treated EB- and CA-treated subclones subclones. Cells were in 6 well dishes until confluent. cells and (A) grown 100 emetine was and after 30 Fresh medium containing ,ug/ml applied oxidasea Lactate min 100 ,uCi/ml [35S]methionine were added and the incubation Cytochrome (mM)b continued for 3 h. Cells were washed and in 25 mM Tris-HCl lysed SDS and 8 M urea. Extracts were on 12.5% (pH 6.8), 2% analyzed 65 16 2.26 L929 on were SDS-urea-PAGE. The radioactive bands the fluorogram 21 6 3.61 EB-2 identified to their molecular size et al., 1985). Cyt according (Chomyn EB-12 12 3 3.89 b: COIL subunits 3 of cytochrome b; COI, COIl, 1, 2, cytochrome EB-21 5 3 4.62 ND5: subunits 3, 4, 5 of oxidase; ND1, ND2, ND3, ND4, 1, 2, CA-22 29 8 3.92 NADH ATP8: subunit 8 of the H+-ATPase. CA and dehydrogenase, CA-34 16 4 3 4.43 EB denote different subclones which were isolated limiting dilution by in or ethidium selection chloramphenicol- bromide-containing medium, of c respectively. (B) Southern blot analysis of restriction endonuclease amnol cytochrome oxidized/min/mg protein. b106 1 24 cells were seeded in ml in a 6 well dish. After cells were obtained from DNA of the L929 cells and h, fragments wild-type FCS. After 12 lactate to fresh medium subclones. DNA (10 was digested with HindIH and changed containing dialyzed h, 1tg) was in the electrophoresed on a 0.7% agarose gel. The gel was transferred to a determined production supematant. nylon membrane and the separated DNA hybridized with a 32P-labeled fragment of the rat mitochondrial 12S rRNA gene. protein synthesis by binding to the 70S ribosome, while leaving cytoplasmic protein synthesis intact (Spolsky and of mtDNA was further Eisenstadt, 1972). From the EB- and CA-resistant cell confirmed by Southern blottings with a for the populations, several subclones were isolated by limiting probe coding mitochondrial 12S rRNA. In all dilution in subclones the amount of mtDNA and analyzed detail. Measurements of lactate tested, was considerably production and cytochrome c oxidase revealed that most cell reduced (Figure 5B). In accordance with Figure 5A, clones recruited most of their energy from anaerobic mtDNA was undetectable in EB-12 and EB-21 clones under glycolysis (Table I). To analyze further the presence of conventional hybridization conditions, whereas subclone mtDNA-encoded EB-2 a protein subunits in the subclones, cells gave weak hybridization signal. were labeled We first in with [35S]methionine in the presence of analyzed cytotoxicity assays the TNF sensitivity an inhibitor of of the subclones. In the absence of the transcription inhibitor emetine, cytoplasmic protein synthesis. By this of the different actinomycin D (Act D), all subclones were largely TNF procedure, synthesis mtDNA-encoded proteins could be visualized in SDS -urea gels. Figure 5A resistant in comparision with the sensitive wild-type cells shows that most of the cell clones had a reduced (Figure 6A). In some clones, no detrimental effects on cell mitochondrial In of protein synthesis. three the EB-resistant viability could be detected even at high TNF concentrations. EB-12 no clones (EB-2, and EB-21), mtDNA translation It is noteworthy that the simple addition of EB or CA to could be detected. A reduction or total untreated products depletion wild-type L929 cells did not affect TNF cytotoxic 3098 Depletion of mitochondrial functioixs abolishes TNF effects IJ5 V113-21 -( DOC 60 0 N- - ~~~~ Os'. 60~~~~~~~ 0~\ o ~o Y A \ 7 62 556 5000 1 6 56 500 TNF (IU/ml) Fig. 6. Comparison of the TNF sensitivity of the respiration-deficient subclones and wild-type L929 cells in the absence (A) and presence (B) of actinomycin D. Cells were seeded in 96 well plates at a density ^__ AL -. AL AL is-i- of 2 x 104 cells/well. After 16 h, serial concentrations of TNF " I ranging from 7 to 5000 IU/ml (A) or 0.7 to 500 IU/ml (B) were applied and incubated for 18 h. Survival was assayed by reduction of 1 2 3 4 5 7 8 9 10 11 12 13 14 MTT and calculated as the percentage of the staining value of untreated cultures. 0, L929 cells; V, EB-12; 0, EB-21; E, CA-22; Fig. 7. NFxB activation by TNF is reduced in respiration-deficient V, CA-34 subclone. L929 subclones. Wild-type L929 cells (lanes 1-6) and EB-21 cells, as a representative subclone (lanes were respiration-deficient 7-12), activity. We next analyzed the TNF sensitivity in the incubated for 45 min with the indicated concentrations of TNF. presence of Act D, which generally enhances TNF killing Nuclear extracts were prepared and tested for the presence of activated NFxB in electrophoretic mobility shift assays. In lanes 13-14, and additionally sensitizes some cell types which are resistant cytoplasmic extracts of both cell types were prepared and latent in the absence of the inhibitor. In the presence of Act D, NFxB-IxB complex activated by treatment with deoxycholate (DOC) TNF resistance of the cell clones was partially overcome, as described in Materials and methods. The EMSA reveals that although cells were still less sensitive than their wild-type TNF stimulation in EB-21 is reduction of NFxB activation after cells not due to diminished amounts of the inactive complex. homologues (Figure 6B). NFxB activation by TNF is strongly repressed in been identified in the IL-6 promoter, such as a cAMP- respiration-deficient cell clones responsive element and an AP-1 site, induction of IL-6 by Next to the cytotoxic effects, the respiration-deficient TNF has been reported to be largely controlled by a xB subclones of L929 were analyzed for the activation of NFxB element (Libermann et al., 1990; Shimizu et al., 1990; in response to TNF. Cells were treated with serial dilutions Zhang et al., 1990). of In TNF. the wild-type cells the addition of 200 IU/ml When wild-type and EB- and CA-treated cells were TNF already resulted in maximal activation of the stimulated with TNF, respiration-deficient cell clones transcription factor (Figure 7). The respiration-deficient produced remarkably less IL-6 than wild-type cells clones, however, only weakly responded to TNF, which was (Figure 8A). In some cell clones, IL-6 expression was obvious a by strongly reduced NFxB activation. Even at the reduced even up to 500- to 600-fold. To analyze whether TNF highest concentrations used, only light xB bands were reduction of IL-6 expression was specific for the TNF observed. stimulus, both cell types were further treated with dibutyryl To analyze whether the reduction of xB binding was really cyclic AMP (db-cAMP), which induces IL-6 expression by due to a or diminished cytoplasmic NFxB activation just the the elevation of cAMP levels. in these Indeed, experiments effect of decreased we further NFxB protein expression, no significant differences could be observed between L929 levels of latent in the cells and the that the compared NFxB-IxB complex subclones, suggesting IL-6 gene could respiration-deficient clones and L929 cells. When still be induced other in wild-type by signaling pathways the extracts of both cell were activated with respiration-deficient cell clones. The same differences in cytoplasmic types in order to dissociate the IxB no TNF-induced IL-6 were also observed at the deoxycholate subunit, expression differences could be observed in the EMSA level In in the (Figure 7). transcriptional (Figure 8B). fact, respiration- These results indicate that of a functional deficient clones IL-6 mRNA could be detected after depletion hardly TNF mitochondrial electron transport system largely abrogates stimulation, although comparable amounts of IL-6 mRNA found after cell activation a combination TNF-triggered NFxB activation. were by of and which db-cAMP, cycloheximide staurosporine, Inhibition of IL-6 in the IL-6 these results gene expression respiration- superinduces gene expression. Therefore, deficient subclones indicate that inhibition of the mitochondrial oxidative of TNF. The effects of of the mitochondrial electron metabolism also affects effects depletion gene-regulatory on TNF-induced NFxB activation us to transport prompted further TNF-induced in the L929 analyze gene expression Discussion which is subclones. IL-6 represents a cytokine rapidly TNF in cell et which radicals induced by ROI, comprise hydrogen peroxide, hydroxyl many types (Defilippi al., 1987). control elements have and are essential of an Although several other transcriptional superoxide anions, compounds 3099 K.Schulze-Osthoff et al. ?r il.H such as or adriamycin menadione, act as redox-cycling agents that continuously produce deleterious oxygen species. evidence now exists Accumulating that short-lived oxygen radicals a may provide versatile cellular control mechanism and additionally play important roles central to gene Pro-oxidant regulation. conditions can induce expression of .: immediate-early genes (reviewed in Holbrook and Fornance, in 1991), particular c-fos and (Crawford et al., 1988; c-jun Shibanuma et al., 1988; Stein et al., 1989; Amstad et al., 1992). Indeed, some gene-inductive effects of phorbol esters also be may independent of protein kinase C activation, since phorbol esters can generate a pro-oxidant cellular environment (Cerutti, 1985). Recently, activation and DNA of several binding transcription factors have been shown to depend on the cytoplasmic or nuclear redox state, including steroid receptors (Silva and Cidlowski, 1989), ISGF3 (Levy A | \i M0I 0s iK' et al., 1989), TFIHC (Cromlish and Roeder, 1989), c-myb protein (Guehmann et al., 1992) and the transcriptional activators AP-1 (Abate et al., 1990; Devary et al., 1991; Xanthoudakis et and al., 1992) NFxB (Mihm et al., 1991; Cl. 'K, B Schreck et mechanism al., 1991, 1992). Although the of .3. vC redox regulation in transcriptional activation is still rather a substantial unclear, body of evidence now exists that ROI may act as second in addition messengers and, to their harmful also exert effects, may signaling functions. In this study, we show that two diverse effects of TNF, , * Si i.e. cytotoxicity and transcriptional activation, require functional as an intracellular mitochondria, presumably source of ROI. TNF cytotoxicity and mitochondrial oxygen radicals In our previous studies (Schulze-Osthoff et al., 1992), we have reported that mitochondria play a crucial role in causing TNF-induced cytotoxicity in TNF-sensitive cells. Pre- sumably, upon TNF stimulation mitochondrial radicals are Fig. 8. TNF of IL-6 is defective in induction respiration-deficient L929 subclones. IL-6 106 cells were in well formed at the ubiquinone site of the respiratory where (A) bioassay: seeded 6 chain, dishes. After 12 medium was refreshed and cells h, were stimulated electrons are directly transferred from ubisemiquinone to with TNF (500 IU/ml, 2500 or db-cAMP 18 h later IU/nil) (1 mM); molecular oxygen. This one-electron transfer results in the supematants were harvested and for IL-6 analyzed bioactivity. (B) formation of superoxide anions as the primary reactive Analysis of IL-6 mRNA levels in L929 subclones. Confluent cell of oxygen species mitochondria (Loschen et al., 1974). monolayers were either untreated (lanes 1, 3, or treated with 5600 5) IU/ml TNF (lanes 2, 4, 6). In a second experiment, the cells were inhibition of Consequently, electron entry to the ubiquinone untreated (lanes 7, or stimulated with a combination of db- 9, 11) of the pool respiratory chain, by complex I inhibitors cAMP (2 mM), cycloheximide (50 and staurosporine (6 nM, Ag/ml) (rotenone, amytal) or complex II inhibitors (thenoyltrifluoro- added at -2 h). After each stimulation, total cytoplasmic RNA was reduced TNF acetone), greatly cytotoxicity (Schulze-Osthoff extracted and immobilized onto a nylon membrane in a dot-blot et al., 1992). On the other when the device. Consecutive lanes correspond to 20 ,g RNA and serial 1:3 hand, electron flow dilutions, respectively. The membrane was hybridized with a 32p was inhibited distal to the ubiquinone site by the addition labeled DNA fragment derived from mIL-6 cDNA and exposed to an of complex HI inhibitors (antimycin A), increased amounts X-ray film. of ROI may be generated, resulting in a marked potentiation of TNF cytotoxicity. oxidative metabolism (reviewed in Halliwell and This Gutteridge, study extends the former findings and provides further 1990). Generally, ROI are regarded as toxic and harmful supportive evidence for the involvement of the mitochondrial metabolites. In phagocytes, large amounts of ROI are electron in TNF transport cytotoxicity. Subclones of L929 generated the by membrane-bound NADPH oxidase, cells which were of depleted mitochondrial respiration providing an antimicrobial defense mechanism against revealed a clear-cut resistance TNF towards cytotoxicity. pathogens (reviewed in Rossi, 1986). Nonetheless, when Depletion of mitochondrial respiration was achieved by long- formation goes uncontrolled, ROI may be implicated in term treatment of the cells with either ethidium bromide or several diseases and autoimmune disorders, leading to chloramphenicol which block mitochondrial DNA replication cellular damage by peroxidizing lipids and disrupting and protein synthesis, respectively (Nass, 1972; Wiseman structural proteins, enzymes and nucleic acids. An important and Attardi, 1978; Desjardins et al., 1985; Morais et al., source of ROI are mitochondria (Boveris and Chance, 1973; 1988; King and Attardi, 1989). The isolated cell clones did Turrens and Boveris, 1980; Turrens et al., 1985). not reveal any significant morphological changes. However, Overproduction by these organelles may be toxic and harful the growth rate of the cells was reduced since the energy for the producing cell. Several cytostatic or antitumor drugs, was mostly derived from anaerobic glycolysis. Most 3100 Depletion of mitochondrial functions abolishes TNF effects or DNA-binding importantly, receptor binding assays with iodinated TNF did of the p55/p65 complex into the nucleus, It is remarkable that, not reveal alterations of TNF receptor numbers, thus ruling activity at the target sequences. contrary of the out reduced TNF binding as a possible explanation of the of NFxB in the cytoplasm, binding to activation acquired resistance (data not shown). Surprisingly, when the p50/p65 NFxB complex to its target sequences in purified respiration-deficient cells were incubated with TNF in the vitro is even promoted under reducing conditions (Toledano presence of the transcriptional inhibitor actinomycin D, 1991; Xanthoudakis et al., 1992). In the and Leonard, which generally increases sensitivity, TNF resistance was the target of the latent NFxB-complex seems to cytoplasm, overcome. An explanation for this which is released by a poorly defined mechanism. again partially be IxB may be the existence of an alternative TNF- can envision a direct effect of ROI by influencing protein sensitization One activated pathway that is less prominent in the absence of folding or damaging the IxB protein. However, purified transcriptional inhibition. Indicative of a second cytotoxic NFxB-IxB cannot be simply activated in vitro by there be a may be the observation that mitochondrial inhibitors exogenous hydrogen peroxide, although may pathway intermediates or radical scavengers which protect parental L929 cells were selective induction of NFxB by specific oxygen it has been that not or only marginally effective in the respiration-deficient et al., 1992). Previously, reported (Schreck R is inducible following L929 clones, when tested in the presence of actinomycin D. the bacterial transcription factor oxy Furthermore, overexpression of manganous superoxide exposure to H202, whereas the bacterial sox R regulon has the control dismutase (MnSOD), which efficiently reduces TNF been defined as being responsible for positive at least in some cell types, is rather inefficient of proteins induced by 02 and distinct from H202 (Storz cytotoxicity at protection against a combination of TNF and et al., 1990; Demple and Amabile-Cuevas, 1991). providing metabolic inhibitors (Wong et al., 1989). However, it seems to be more likely that radicals indirectly in NFxB activation. An indirect role participate Redox regulation of transcriptional activators of radicals in NFxB activation by TNF might be exerted This study shows that mitochondrial intermediates are not the activation of specific proteases or kinases. In through in vitro conditions only implicated in TNF cytotoxicity, but may also be this respect, it should be noted that under complex (Ghosh and involved in TNF-mediated activation of the transcription protein kinase C can activate the NFxB A role of protein kinase C factor NFxB. Firstly, the respiration-deficient L929 Baltimore, 1990). physiological as TNF-induced subclones revealed a strongly reduced NFxB activation by or A, however, seems to be rather unlikely PKC or PKA inhibitors TNF. Secondly, in the parental L929 cells mitochondrial NFxB activation is not abolished by However, inhibitors, that interfered with TNF cytotoxicity, apparently (Hohmann et al., 1991; Meichle et al., 1991). described in recent times, affected NFxB activation in similar ways. Inhibitors, which several novel kinases have been et were protective by blocking an early step in the such as a redox-activated tyrosine kinase (Bauskin al., which in NFxB activation should mitochondrial electron transfer, strongly reduced NFxB 1991), the possible role of controlling redox activation by TNF. Vice versa, the complex IH inhibitor be examined. Other candidate proteins acting similarly to the which confers potentiation to TNF cytotoxicity, regulation may be specific factors antimycin A, 1 which mediates the redox exerted a synergistic effect on NFxB activation by TNF. recently cloned REF- protein, other activators These observations suggest an at least partially common control of AP-1 activity and transcriptional et al., 1992). for both the cytotoxic and NFxB-inducing effects in the nuclear compartment (Xanthoudakis pathway that NFxB activation is of TNF. In accordance with this, it has been reported Furthermore, it may be possible such molecules, which recently that activation of NFxB by different agents particularly influenced by thiol-containing TNF stimulation. Although it as IL-I or phorbol myristate acetate could be prevented may become oxidized upon TNF, in a T cell line that TNF did not antioxidants and chelation of iron ions (which is required has recently been shown by and levels to convert oxygen intermediates to the highly reactive the total cellular thiol glutathione (Israel change be modified radical). In another study, it has been shown that et al., 1992), specific antioxidant systems may hydroxyl In this it has to be irradiation, by which ROI are continuously and affect NFxB activation. context, ionizing contain a number of different redox can induce NFxB activation (Brach et al., 1991). pointed out that cells generated, not linked to each other. With respect to TNF signal transduction, our results also systems which are necessarily metabolism has been also suggested that TNF stimulation implicate a role of the mitochondrial oxidative Recently, it that ROI formed by involves the activation of an acidic sphingomyel- in NFxB activation and suggest of NFxB in this et which is a compartmentalized of the electron transfer are responsible inase (Schutze al., 1992), perversion we tested a contribution of other with a of 4.5. At present it is process. Moreover, when enzyme pH optimum pH such as NADPH oxidase or unclear how ceramide, the reaction product of potential radical sources, no involvement of these enzymes in TNF- control NFxB activation. It should xanthine oxidase, sphingomyelinase, might with cell lines activation was detectable. We also analyzed be noted that these results were obtained induced NFxB the influence of ATP on the activation of NFxB. unlike L929, were not or only weakly susceptible synthesis which, of it is still an like TNF early NFxB activation was to the action TNF; open question However, cytotoxicity, cytotoxic oxidative a role in the TNF-mediated not influenced of phosphorylation whether sphingomyelinase plays by uncouplers in L929 cells if whether ceramide or a direct inhibition of ATPase by oligomycin (Schulze- NFxB activation and, so, and our ROI formation. Osthoff et unpublished results). influences al., 1992, molecular mechanism Our results do not allow the by activated to be identified. Intracellular IL-6 as an oxidative stress-responsive cytokine which NFxB becomes regulation Another remarkable difference between the respiration- early transcriptional by radicals could regulate was in of transduction, L929 subclones and their NFxB at several points the pathway signal deficient parental counterparts TNF induction of IL-6 subunit translocation the gene the release of the inhibitory IxB, observed by studying e.g. 3101 et al. K.Schulze-Osthoff of IL-6 in the subclones expression. Expression was to changed mitochondrial functions. Early investigations in considerably reduced at the level and 200- transcriptional Drosophila have shown that application of mitochondrial to 600-fold when the biological was measured. This activity inhibitors can influence the puff expression pattern in observation indicates that not only TNF cytotoxicity, but also polytene chromosomes (Leenders et al., 1974, and IL-6 induction, may be mediated by reactive oxygen references therein). Yeast cells respond to the complex Ill intermediates. The reduction of IL-6 expression be may inhibitor antimycin A by induction of cytoplasmic citrate explained by a decreased activation of NFxB to a binding synthase expression (Liao et al., 1991). Moreover, in yeast xB element (-73 to -64) which has been the level of suggested by expression of nuclear genes, not all of them several authors to be indispensible for TNF- or IL-1-induced related to directly mitochondrial functions, was found to IL-6 expression (Libermann et al., 1990; Shimizu et change in response to the loss of the mitochondrial al., genome 1990; Zhang et al., 1990). However, the of (Parikh et al., 1987). These observations therefore indeed importance NFxB in determining TNF-induced IL-6 should provide additional evidence for the existence of a expression signal not be overemphasized. We have recently described the communication between the mitochondrial and nuclear potentiating effect of lithium ions in TNF-induced IL-6 compartment. Oxygen radicals may provide a versatile et Vandevoorde et cellular control mechanism in this signaling system since expression (Beyaert al., 1991; al., 1991), they we could not observe contribution of NFxB comprise very short-lived molecules which can be tightly although any in these phenomena. it should be taken into regulated by a machinery of different enzymes. Furthermore, account that NFxB was A major question that has to be addressed in future activation, though seriously reduced, still detectable to a extent in the subclones. experiments is the mechanism of how TNF interferes significant Thus, with it is conceivable that a cooperative reduction of additional the mitochondrial electron transport. There has been the activators accounts for the decrease of suggestion that alterations in mitochondrial calcium transcriptional sharp IL-6 expression. Ray et al. (1989) have defined a 26 homeostasis affect the mitochondrial respiratory chain bp region of the IL-6 promoter, called (Richter and Frei, 1988). However, it is not clear whether multiple responsive element (MRE, -173 to -151), which is in TNF- changes in calcium metabolism precede the oxidative stress important induced transcription of the IL-6 gene. In human fibroblasts, or are just a consequence of it. Another interesting issue will it has been further suggested that TNF induction of IL-6 be whether phosphorylation events, induced by TNF, may may be mediated by a cellular increase of cAMP et regulate the electron transport. Up to now, there have been (Zhang al., no 1988, 1990) which may be additionally involved in the studies concerning phosphorylation of respiratory chain activation of NFxB (Shirakawa and Mizel, proteins. One could speculate that an increase of radical 1989). Although a cAMP-responsive element is contained in the IL-6 production at the ubiquinone site of the respiratory chain (CRE) - in promoter (-163 to 158), it seems to be dispensible for results a reduced electron transfer from the ubiquinone TNF induction of IL-6 expression. In L929 cells, cAMP and cycle to the cytochrome chain. Thus, a target protein between TNF in a these induce gene expression synergistic manner, components could be a ubiquinone-binding protein, of which would suggesting separate pathways IL-6 induction (Beyaert regulate the electron flow to complex HI, or et for a of certain subunits of of al., 1991). Supportive distinct regulation gene cytochrome b the respiratory chain (Yu TNF is the that in et expression by and cAMP finding the al., 1985). Interestingly, in plant photosynthesis it has L929 subclones IL-6 could still be been shown that the respiration-deficient redox-regulated LHC kinase can regulate TNF was the electron flow induced by cAMP, whereas ineffective. by phosphorylation (Gal et al., 1992). The dual role of mitochondrial intermediates in both the Future studies may therefore reveal whether these or similar induction of IL-6 and earlier mechanisms cytotoxicity may explain transmit the TNF signal into the mitochondria IL-6 and findings that expression generally parallels the outcome thereby regulate mitochondrial radical generation. of TNF In studies with several activators or cytotoxicity. it has been shown that under conditions where inhibitors, Materials and methods was either enhanced or cytotoxicity decreased, IL-6 expression was modulated et TNF and reagents equally (Vandevoorde al., TNFpreparation. Recombinant murine TNF was produced in Escherichia 1991, 1992). Thus, IL-6 expression may be determined at coli and purified to at least 99% et The homogeneity (Fransen al., 1985). least in part by the same cytoplasmic reactions and preparation had a specific activity of 3.7 x 107 and IU/mg protein metabolites responsible for cellular That ROI toxicity. may contained <4 ng endotoxin/mg protein. TNF was determined as activity be the common denominator in these is further processes described by Ostrove and Gifford (1979) an international standard using TNF the fact preparation (code no. 88/532, obtained from the Institute for Biological suggested by that some radical can inhibit scavengers Standards and Control, Potters Bar, UK) as a reference. IL-6 expression (our unpublished results). Reagents. The following reagents were purchased from Sigma Chemical Oxygen radicals as a putative signal transducers of Co. (Deisenhofen, FRG): N-acetyl-L-cysteine, actinomycin D, amytal, mitochondria - nucleus pathway butylated hydroxyanisole AMP (BHA), dibutyryl cyclic (db-cAMP), 3-(4,5-dimethylthiazol-2-yl)-2,5-diphenyltetrazolium bromide (MTT), The data presented here define a previously unexpected role nordihydroguaiaretic acid (NDGA), oligomycin, pyrrolidine dithiocarbamate for mitochondria in the control of nuclear gene expression. (PDTC), rotenone and thenoyltrifluoroactetone (TTFA). Antimycin A was Apparently, mitochondrial intermediates can control both the from Serva (Heidelberg, FRG). Stock solutions of the reagents were routinely cytotoxic and gene-regulatory effects of TNF, thus providing prepared in medium, dimethylsulfoxide or ethanol as appropriate. a basis for a mitochondria-nucleus signaling pathway. Up Cell culture to there is now, only limited information regarding a The standard medium for L929 cells was Dulbecco's modified Eagle's proposed retrograde signaling pathway (reviewed in Grivell, medium supplemented with 10% newborn calf serum, 100 U penicillin/mI 1989), although there are several reported instances of and 0.1 mg streptomycin/ml. Isolation of respiration-deficient cells was alterations in gene expression of nuclear genes in response carried out by long-term culture in medium containing 50 yg 3102 functions abolishes TNF effects of mitochondrial Depletion or 400 ng ethidium bromide/ml. The selection medium the rate of oxidation of ferrocytochrome c (Smith and Conrad, 1956). Cells chloramphenicol/ml further supplemented with 4.5 mg glucose/ml and 5 uridine/ml, were scraped off and lysed with 0.3% digitonin in 10 mM potassium was Ag have been reported to become pyrimidine phosphate (pH 7.0). Enzyme reactions were started by the addition of 50 since respiration-deficient cells Attardi, 1989). Addition of these agents did not, reduced ferrocytochrome c and recorded at 25°C every 15 s at 550 auxotrophs (King and lsM c in the blank cuvette was fully oxidized with 1 mM however, influence any of the assays. nM. Ferrocytochrome oxidation were determined from the initial ferricyanide. Rates of cytochrome coefficient of 21/mM/cm. TNF cytotoxicity assay linear rates using an extinction Cells were seeded in 96 microwell plates at 3 x 104 cells in 100 dl synthesis medium. Twelve to 16 h later, 50 of a drug solution were given. TNF Mitochondrial protein Ad in 6 well plates at 1 x 106 cells/well; 1 h before the with or without actinomycin D (1 was added 2 h later in a 50 Cells were seeded Ag/ml) A1 100 cytoplasmic protein synthesis was volume and a concentration range of 0.7-5000 IU/ml. After 18-24 h of addition of [35S]methionine, lsCi/ml with emetine at a concentration of 100 After further further incubation, cell viability was routinely determined via MTT staining inhibited Ag/ml. were obtained with crystal violet staining for 3 h, the adherent cells were scraped off and washed three (Tada et al., 1986). Similar results incubation in PBS. were prepared in 20 mM Tris-HC1 (pH 7.4) of attached cells. times Lysates 8 M urea and 2% SDS. Samples were loaded without prior heating containing lnterleukin-6 assay on 12.5% SDS-polyacrylamide gels containing 6 M urea (Chomyn et al., 6 well plates at 106 cells/well; 16-24 h later cells 1985). Gels were dried and exposed to X-ray films. Cells were seeded in twice and treated in 1 ml medium/well containing various TNF were washed h of incubation, the medium was Mitochondrial DNA analysis concentrations. Following another 16-24 IL-6. IL-6 was determined on mtDNA was prepared from the cells by the method of Wallace taken off and assayed in serial dilutions for Enriched mouse plasmocytoma cell line HindIl restriction fragments of the DNA were electrophoresed the basis of the proliferative response of the et al. (1988). in TAE buffer. After depurination and denaturation 7TD1 (Van Snick et al., 1986). in 0.7% agarose gels were blotted by capillarity in according to standard procedures, x onto N+ membranes (Amersham). DNA was fixed by 20 SSC Hybond Nuclear and whole cell extracts UV and in 5 x SSPE, 0.5% SDS and irradiation prehybridized Subconfluent cell cultures in a 6 well plate or 60 cm2 Petri dishes were 1 x For hybridization, an AccI insert, cloned in a Denhardt's solution. treated with TNF for 45 min. Inhibitors were applied 1 h before the onset vector and coding for the 12S rRNA of rat mtDNA (Gadaleta Bluescript of TNF addition. After incubation cells were rinsed in ice-cold phosphate- was labeled extension of random hexanucleotide primers. et al., 1989), by buffered saline, scraped off using a rubber policeman and collected by After at 45°C for 18 h, blots were washed in 0.5 x SSC, hybridization centrifugation. Nuclear extracts were prepared essentially as described by 0.1% SDS and exposed to X-ray films. Dignam et al. (1983). Cells were resuspended in 1 ml of a hypotonic lysis buffer (buffer A, Dignam et al., 1983). After 20 min, cells were RNA manipulations homogenized by 20 strokes with a loose-fitting Dounce. Nuclei were collected as previously Total cytoplasmic RNA was isolated essentially described in a microcentrifuge and proteins extracted in a high-ionic-strength buffer buffer. mRNA and Bancroft, 1982) using an NP40 lysis IL-6-specific (White was removed by centrifugation. (buffer C). After 30 min, nuclear debris a restriction fragment of the murine was quantitated in dot-blot analyses using The extracts were diluted with 4 vol of low-salt buffer (pH 7.5) (buffer Snick et al., 1986). IL-6 cDNA containing plasmid pUC8MIL6 (Van D; Dignam et al., 1983) containing 1% NP-40, and kept frozen until use. RNA were denatred with formaldehyde Appropriate amounts of the isolated cells were lysed in a high-salt buffer (20 mM For whole-cell extracts, A, Pall Biosupport, East and applied onto a nylon membrane (Pall Biodyne HEPES, pH 7.5, 400 mM NaCl, 1 mM MgCl2, 0.5 mM EDTA, 0.1 mM x SSC and immobilized by heat treatment Hills, NY) in the presence of 5 1% NP-40, 0.5 mM DTT, 10 leupeptin, 1 EGTA, 20% glycerol, h 42°C) and hybridization (overnight (1 h at 80°C). Prehybridization (1 at Ag/ml After 10 min of incubation, extracts were mM PMSF, 0.001% aprotinin). at carried out in a mixture containing 5 x Denhardt's solution, 42°C) were and kept frozen. cleared by centrifugation 5 x SSC, 50 mM sodium phosphate (pH 6.5), 0.1% SDS, 250 non- jig/ml DNA formamide. The washing steps were carried homologous and 50% mobility shift assays Electrophoretic in 2 x SSC/0. SDS x 5 min) and finally 0.1 x SSC/0. 1% SDS. out 1% (4 ( - 3 nuclear protein or -10 Itg crude Equal amounts of the extracts jg the Bio-Rad assay kit) were incubated with an NFxB- protein; determined by The sequence of the double-stranded specific 32P-labeled oligonucleotide. Acknowledaements NFxB site and flanking sequences from the 30 bp probe encompassing the is shown in Zabel et al. (1991). Binding mouse kappa light chain enhancer and di The authors wish to thank Drs C.Saccone E.Sbisa (Universita Bari) performed in a 20 1l volume containing 2-4 1l of extract, reactions were for DNA and Dr T.Wirth Heidelberg) kindly providing probes. (ZMBH, x buffer (20 mM HEPES pH 7.5, 50 mM KC1, 1mM DTT, 4 5 binding a from the Bundesministerium K.S.-O. acknowledges fellowship gratefully 2 as non-specific competitor DNA, 2.5 mM MgCl2, Ficoll), poly(dI-dC) R.B. and V.V. are fur und Technologie (AIDS-Stipendium). Forschung itg 2 serum albumin (BSA) and 3000-6000 c.p.m. Cerenkov of bovine with the Nationaal Fonds research assistants and G.H. a research director Ag oligonucleotide. After 30 min binding reaction the random primer-labeled The research was supported by voor Wetenschappelijk Onderzoek. project room were loaded on a 4% non-denaturing at temperature, samples from the Fonds voor Geneeskundig Wetenschappelijk Onderzoek, grants and run in 0.5 x TBE buffer (pH 8.0). As a negative polyacrylamide gel the en and Interuniversitaire Attractiepolen, Algemene Spaar- Lijfrentekas an mutated in the NFxB binding site by two point control, oligonucleotide Kanker. the Sportvereniging tegen mutations, but otherwise identical with the wild-type fragment, was used. In all cases, DNA-NFxB complexes were only observed using the wild- In some experiments, the DNA-binding activity of type oligonucleotide. References the extracts was by scintillation counting of the 32P-labeled quantified and Curran,T. (1990) Science, 249, NFxB-DNA complexes excised from the dried gels. In the case of cytosols, Abate,C., Patel,L., Rauscher,F.J,llI 1157-1161. was activated by incubating 2 of the cytosolic extract with 3 NFxB Al ILI Rahmsdorf,H.J., for 10 min before addition to a mixture of probe. Imagawa,M., Chiu,R., Stein,B., Imbra,R.J., 6.5% sodium deoxycholate Angel,P., and 729-739. For the of the DNA-binding activities of AP-1 and octamer binding Jonat,C. Herrlich,P. Karin,M. (1987) Cell, 49, assay and (1992) Cancer Res., 52, (Oct), the oligonucleotides used in this study have been described Amstad,P.A., Krupitza,G. Cerutti,P.A. proteins et The NF- 3952-3960. Angel et al., 1987; Oct, Wirth al., 1987). previously (AP-1, motif from the human Biochim. Biophys. Acta, 1072, 63-80. IL6 specific oligonucleotide containing the binding Baeuerle,P.A. (1991) was synthesized and Science, 242, 540-546. IL-6 (5'-GTACACATTGCACAATCTTA-3') Baeuerle,P.A. Baltimore,D. (1988) promoter 685-696. and Ben-Neriah,Y. (1991) Cell, 66, on an Applied Biosystems DNA Synthesizer. Bauskin,A.R., ALkalay,I. Necrosis Factors: Molecules and Their Tumor Ihe Beutler,B. (1992) Raven New York. Role in Medicine. Press, Anaerobic glycolysis Emerging Annu. Rev. 625-655. were washed twice with phosphate-buffered and Cerami,A. (1989) Immunol., 7, Exponentially growing cells Beutler,B. Van and incubated at time zero in DMEM with 10% dialyzed fetal Schulze-Osthoff,K., Roy,F. Fiers,W. (1991) Cytokine, saline (PBS) and Beyaert,R., 12 h, supernatants were assayed for lactate content 3, 284-291. calf serum (FCS). After and Trends Biochem. lactate Cellular Kourilsky,P. Isradl,A. (1992) Sci., 17, with a commercial kit (Sigma) using dehydrogenase. Blank,V., 135-140. measured a method analyzing oxidase was by spectrophotometric cytochrome 3103 K.Schulze-Osthoff et al. and Boveris,A. Chance,B. (1973) Biochem. J., 707-716. and Mol. 134, Ray,A., Sassone-Corsi,P. Sehgal,P.B. (1989) Cell. Biol., 9, Boveris,A., Cadenas,E. and Stroppani,A.O.M. (1976) Biochem. 5537-5547. J., 156, 435-445. and Free Rad. Biol. 365-375. Richter,C. Frei,B. (1988) Med., 4, Brach,M.A., Hass,R., Shernann,M.L., Gunji, H., and Biochim. 65-89. Weichselbaum,R. Rossi,F. (1986) Biophys. Acta, 853, Kufe,D. (1991) J. Clin. Invest., 88, 691-695. Ruggiero,V., Johnson,S.E. and Cell. Baglioni,C. (1987) Immunol., 107, Brenner,D.A., O'Hara,M., Angel,P., Choikier,M. and Karin,M. (1989) 317-325. Nature, 337, 661-663. Schreck,R., Rieber,P. and Baeuerle,P.A. EMBO 2247-2258. (1991) J., 10, Bromsztyk,K., Rooney,J.W., Iwasaki,T., Rachie,N.A., Dower,S.K. and Schreck,R., Meier,B., Mannel,D., and Dr6ge,W. Baeuerle,P.A. (1992) J. Sibley,C.H. (1991) Cell Reg., 2, 329-335. 1181-1194. Exp. Med., 175, Cadenas,E. and Boveris,A. (1980) Biochem. J., 188, 31-37. Schulze-Osthoff,K., Bakker,A.C., Vanhaesebroeck,B., Beyaert,R., Cerutti,P.A. (1985) Science, 227, 375-381. Jacobs,W. and J. Biol. 5317-5323. Fiers,W. (1992) Chem., 267, Chomyn,A., Mariottini,P., Cleeter,M.W.J., Ragan,C.I., Schuitze,S., and Matsuno-Yagi,A., Potthoff,K., Machleidt,T., Berkovic,D., Wiegmann,K. Hatefi,Y., Doolittle,R.F. and Attardi,G. (1985) Nature, 314, 592-597. 765-776. Kr6nke,M. (1992) Cell, 71, and Del Arch. Cino,M. Maestro,R.F. (1989) Biochem. Shibanuma,M., Kuroki,T. and Nose,K. (1988) Oncogene, 17-21. Biophys., 269, 3, 623-638. Shimizu,H., Mitomo,K., Watanabe,T., and Okamoto,S. Yamamoto,K.J. and Crawford,D., Zbinden,J., Amstad,P. Cerutti,P.A. (1990) Mol. Cell. Biol., 561-568. (1988) Oncogene, 10, 27-32. 3, Shirakawa,F. and Mol. Cell. 2424-2430. Mizel,S. (1989) Biol., 9, Cromlish,J.A. and J. Roeder,G. (1989) Biol. Chem., 264, 18100-18109. Silva,C.M. and Cidlowski,J.A. J. Biol. 6638-6647. (1989) Chem., 264, Dealtry,G.B., Naylor,M.S., Fiers,W. and Balkwill,F.R. Eur. J. Smith,L. and Conrad,H. (1956) Arch. Biochem. Biophys., 403 -410. (1987) 63, 689-693. and Immunol., 17, Spolsky,C.M. Eisenstadt,J.M. FEBS 319-324. (1972) Lett., 25, Defilippi,P., Poupart,P., Tavernier,J., Fiers,W. and Content,J. Proc. Staal,F.J.T., Roederer,M., and (1987) Herzenberg,L.A. Herzenberg,L.A (1990) Nati. Sci. 4557-4561. Proc. Natl. USA, 84, Acad. Sci. USA, 87, 9943-9947. Demple,B. and Amabile-Cuevas,C.F.(1991) Cell, 67, 837-839. Stein,B., Rahmsdorf,H.J., Steffen,A., Litfin,M. and Herrlich,P. (1989) Desjardins,P., and Mol. Cell. 5169-5181. Frost,E. Morais,R. (1985) Mol. Cell. Biol., 9, Biol., 5, 1163-1169. and Storz,G., Tartaglia,L.A. Ames,B.N. (1990) Science, 248, 189-194. Devary,Y., and Suffys,P., Van and Gotdieb,R.A., Laus,L.F. Karin,M. (1991) Mol. Cell. Beyaert,R., Roy,F. Fiers,W. (1987) Biochem. Biol., Biophys. 11, 2804-2811. Res. 7835-7843. Commun., 149, and Nucleic Van and Dignam,J.P., Lebovitz,R.M. Roeder,R.G. (1983) Acids Suffys,P., Beyaert,R., Roy,F. Fiers,W. (1988) Eur. J. Res., Biochem., 11, 1475-1489. 257-265. 178, Duh,E.J., and De Maury,W.J., Folks,T.M., Fauci,A.S. Suffys,P., Beyaert,R., Valck,D., Vanhaesebroeck,B., Van and Rabson,A. (1989) Roy,F. Proc. Natl. Acad. Sci. 5974-5978. Eur. J. USA, 86, Fiers,W. (1991) Biochem., 195, 465-475. Fiers,W. (1991) FEBS 199-212. Lett., 285, Tada,H., Shiho,O., Kuroshima,K., Koyama,M. and Tsukamoto,K. (1986) Van der J. Immunol. Methods, 157-165. Fransen,L., Muiller,R., Marmenout,A., Tavernier,J., 93, Heyden,J., Kawashima,E., Chollet,A., Van Van Toledano,M.B. and Proc. Tizard,R., Heuverswyn,H., Leonard,W.J. (1991) Natl. Acad. Sci. USA, 88, Vliet,A., and Nucleic Acids 4328-4332. Ruysschaert,M.R. Fiers,W. (1985) Res., 13, 4417-4429. Turrens,J.F. and Boveris,A. Biochem. (1980) J., 191, 421-427. Gadaleta,G., De and Turrens,J.F., Alexandre,A. and Pepe,G., Candia,G., Quagliarello,C., Sbisa,E. Lehninger,A.L. (1985) Arch. Biochem. Saccone,C. J. Mo. 497-516. Biophys., 408-414. (1989) Evol., 28, 237, Gal,A., Hermann,R.G., and Ohad,J. Vandevoorde,V., Haegeman,G. and Lottspeich,F. (1992) FEBSLett., 298, Fiers,W. (1991) Biochem. Biophys. 33-35. Res. Commun., 178, 993-1001. Ghosh,S. and Baltimore,D. 678-682. Vandevoorde,V., Haegeman,G. and Fiers,W. (1992) (1990) Nature, 344, FEBS Lett., 302, Eur. Grivell,L.A. (1989) J. 477-493. 235-238. Biochem., 183, Van Guehmann,S., and Snick,J., Cayphas,S., Vink,A., Uyttenhove,C., Vorbrueggen,G., Kalkbrenner,F. Moelling,K. (1992) Coulie,P.G., Nucleic Acids Res., 20, 2279-2286. Rubira,M.R. and Simpson,R.J. (1986) Proc. Natl. Acad. Sci. USA, 83, and Halliwell,B. Gutteridge,J.M.C. Methods 1-85. 9679-9683. (1990) Enzymol., 186, Hohmann,H.P., Kolbeck,R., Remy,R. and van Mol. Wallace,D.D., Zheng,X.X., Lott,M.T., Shoffner,J.M., Loon,A.P. (1991) Cell. Hodge,J.A., 2315-2318. Biol., 11, Kelley,R.I., Epstein,C.M. and Hopkins,L.C. (1988) Cell, 55, 601-610. and Holbrook,N.J. Fornance,A.J.,Jr New 825-833. White,B.A. and Bancroft,F.C. (1982) J. Biol. Chem., 257, 8569-8572. (1991) Biol., 3, Wirth,T., Israel,N., Gougerot-Pocidalo,M.-A., Aillet,F. and Staudt,L. and Baltimore,D. (1987) Nature, 329, Virelizier,J.-L. (1992) 174-178. J. Immunol., 149, 3386-3393. Wiseman,A. and Attardi,G. (1978) Mol. Gen. Genet., 167, 51-63. and Wong,G.H.W., King,M.E., Godman,G.C. King,D.W. (1972) J. Cell Elwell,J.H., Oberley,L.W. and Goeddel,D.V. (1989) Biol., 53, Cell, 127-142. 58, 923-931. King,M.P. and Attardi,G. Xanthoudakis,S., (1989) Science, 246, 500-503. Miao,G., Wang,F., Pan,Y.-C.E. and Curran,T. (1992) Konstantinov,A.A., EMBO J., 11, 3323-3335. Peskin,A.V., Popova,E.Y., Khomutov,G.B. and Ruuge,E.K. (1987) Biochim. Yamauchi,N., Kuriyama,Y.H., Biophys. Acta, 894, 1-10. Watanabe,N., Neda,H., Maeda,M. and Leenders,H.J., Berendes,H.D., Niitsu,Y. (1989) Cancer Res., Helmsing,P.J., Derksen,J. and 49, 1671-1675. Koninkx,J.F.J.G. (1974) Sub-Cell. Yamauchi,N., Watanabe,N., Biochem., 3, 119-147. Kuriyama,H., Neda,H., Maeda,M., and Lenardo,M. Baltimore,D. Himeno,T. and Tsuji,Y. (1990) Int. J. (1988) Cell, 58, 227-229. Cancer, 46, 884-888. Levy,D.E., Kessler,D.S., Pine,R. and Darnell,J.E., Jr (1989) Genes Yu,L., Yang,F.D. and Yu,C.A. (1985) Dev., J. Biol. Chem., 260, 963 -973. 3, 1362-1371. Zabel,U., Schreck,R. and Baeuerle,P.A. (1991) J. Biol. Chem., 266, Liao,X., Small,W.C., Srere,P.A. and Butow,R.A. (1991) Mol. Cell. 252-260 (1991). Biol., 11, 38-46. Zhang,Y., Lin,J.-X. and Vileek,J. (1988) Proc. Natl. Acad. Sci. USA, 85, Libermann,T.A. and Baltimore,D. (1990) Mol. Cell. Biol., 10, 2327-2334. 6802-6805. Loschen,G., Azzi,A., Richter,C. and Flohe,L. (1974) FEBS Lett., 42, Zhang,Y., Lin,J.-X. and Vileek,J. J. (1990) Biol. Chem., 263, 6177-6182. 68-72. Zimmerman,R.J., Chan,A. and Cancer Leadon,S.A. (1989) Res., 49, Meichle,A., Schiitze,S., Hensel,G., Brunsing,D. and Kronke,M. (1990) 1644-1648. J. Biol.Chem., 265, 8839-8843. Mihm,S., Ennen,J., Pessara,U., Kurth,R. and Droge,W. Received (1991) AIDS, 5, on February 15, 1993; revised on April 22, 497-503. Morais,R., Desjardins,P., Turmel,C. and Zinkewich-Peotti,K. (1988) In Vitro Cell. Dev. Biol., 24, 649-658. Nass,M.M.K. (1972) Exp. Cell Res., 72, 211-222. Niki,E. (1987) Chem. Phys. Lipids, 44, 227-232. Osbom,L., Kunkel,S. and Nabel, G.J. Proc. (1989) Natl. Acad. Sci. USA, 86, 5974-5978. and Ostrove,J. Proc. Soc. 354-358. Gifford,G. (1979) Exp. Med., 160, Parikh,V.S, and Morgan,M.M., Scott,R., Clements,L.S. Butow,R.A. 576-580. Science, 235, (1987)

Journal

The EMBO Journal – Springer Journals

Published: Aug 1, 1993

There are no references for this article.