hFis1, a Novel Component of the Mammalian Mitochondrial Fission Machinery
2003; Elsevier BV; Volume: 278; Issue: 38 Linguagem: Inglês
10.1074/jbc.m303758200
ISSN1083-351X
AutoresDominic I. James, Philippe A. Parone, Yves Mattenberger, Jean‐Claude Martinou,
Tópico(s)Ubiquitin and proteasome pathways
ResumoThe balance between the fission and fusion mechanisms regulate the morphology of mitochondria. In this study we have identified a mammalian protein that we call hFis1, which is the orthologue of the yeast Fis1p known to participate in yeast mitochondrial division. hFis1, when overexpressed in various cell types, localized to the outer mitochondrial membrane and induced mitochondrial fission. This event was inhibited by a dominant negative mutant of Drp1 (Drp1(K38A)), a major component of the fission apparatus. Fragmentation of the mitochondrial network by hFis1 was followed by the release of cytochrome c and ultimately apoptosis. Bcl-xL was able to block cytochrome c release and apoptosis but failed to prevent mitochondrial fragmentation. Our studies show that hFis1 is part of the mammalian fission machinery and suggest that regulation of the fission processes might be involved in apoptotic mechanisms. The balance between the fission and fusion mechanisms regulate the morphology of mitochondria. In this study we have identified a mammalian protein that we call hFis1, which is the orthologue of the yeast Fis1p known to participate in yeast mitochondrial division. hFis1, when overexpressed in various cell types, localized to the outer mitochondrial membrane and induced mitochondrial fission. This event was inhibited by a dominant negative mutant of Drp1 (Drp1(K38A)), a major component of the fission apparatus. Fragmentation of the mitochondrial network by hFis1 was followed by the release of cytochrome c and ultimately apoptosis. Bcl-xL was able to block cytochrome c release and apoptosis but failed to prevent mitochondrial fragmentation. Our studies show that hFis1 is part of the mammalian fission machinery and suggest that regulation of the fission processes might be involved in apoptotic mechanisms. hFis1, a novel component of the mammalian mitochondrial fission machinery. Vol. 278 (2003) 36373–36379Journal of Biological ChemistryVol. 279Issue 34PreviewPage 36378:Fig. 5C was inadvertently omitted. Panel C is shown below. Full-Text PDF Open Access Mitochondria play a key role in many cellular processes ranging from apoptosis (1Ravagnan L. Roumier T. Kroemer G. J. Cell. Physiol. 2002; 192: 131-137Crossref PubMed Scopus (436) Google Scholar) to aging (2Nicholls D.G. Int. J. Biochem. Cell Biol. 2002; 34: 1372-1381Crossref PubMed Scopus (191) Google Scholar), and it is possible that the processes that determine their morphology also modulate their function (3Desagher S. Martinou J.C. Trends Cell Biol. 2000; 10: 369-377Abstract Full Text Full Text PDF PubMed Scopus (1693) Google Scholar,4Osiewacz H.D. Gene (Amst.). 2002; 286: 65-71Crossref PubMed Scopus (58) Google Scholar). The morphology of mitochondria is regulated by the controlled action of fusion and fission mechanisms, which give rise to, in many cases, a branched tubular network extending throughout the cell (5Westermann B. EMBO Rep. 2002; 3: 527-531Crossref PubMed Scopus (121) Google Scholar). In Saccharomyces cerevisiae, the fission of the outer mitochondrial membrane (OMM) 1The abbreviations used are: OMM, outer mitochondrial membrane; ORF, open reading frame; YFP, yellow fluorescent protein; CFP, cyan fluorescent protein; GFP, green fluorescent protein; HA, hemagglutinin; TBS, Tris-buffered saline; PBS, phosphate-buffered saline; CGI, comparative gene identification; aa, amino acid(s); PTP, permeability transition pore; Z-VAD-fmk, N-benzyloxycarbonyl-Val-Ala-Asp(O-Me) fluoromethyl ketone. is regulated by three proteins: Dnm1p, Fis1p, and Mdv1p (6van der Bliek A.M. J. Cell Biol. 2000; 151: F1-F4Crossref PubMed Scopus (56) Google Scholar,7Shaw J.M. Nunnari J. Trends Cell Biol. 2002; 12: 178-184Abstract Full Text Full Text PDF PubMed Scopus (308) Google Scholar), whereas Mdm33 is involved in the fission of the inner mitochondrial membrane (8Messerschmitt M. Jakobs S. Vogel F. Fritz S. Dimmer K.S. Neupert W. Westermann B. J. Cell Biol. 2003; 160: 553-564Crossref PubMed Scopus (92) Google Scholar). Deletion of any of these genes abrogates the fission mechanism, and fusion continues unabated leading to a fused mitochondrial network. Dnm1p is a dynamin-related GTPase that assembles on the fission points and is thought to form a circular structure that "pinches" or constricts the outer membrane. The constriction and complete fission of the OMM requires Mdv1p, which is thought to act as an adaptor molecule between Dnm1p and Fis1p, the latter being an integral outer mitochondrial membrane protein. In the absence of Fis1p, Dnm1p can no longer associate with the OMM and is distributed in a diminished number of punctate spots in the cytosol. Further delineation of this pathway has shown that Fis1p not only allows initial assembly of Dnm1p but also the final scission process, which is dependent on the presence of Mdv1p. The mechanisms of mitochondrial fission in other eukaryotic species are less clearly understood. Orthologues of Dnm1p have been characterized in Caenorhabditis elegans (DRP1 (9Labrousse A.M. Zappaterra M.D. Rube D.A. van der Bliek A.M.C Mol. Cell. 1999; 4: 815-826Abstract Full Text Full Text PDF PubMed Scopus (517) Google Scholar)) and mammals (Drp1 (10Smirnova E. Shurland D.L. Ryazantsev S.N. van der Bliek A.M. J. Cell Biol. 1998; 143: 351-358Crossref PubMed Scopus (581) Google Scholar)), but they are different in their functionality. Expression of DRP1 in C. elegans promotes mitochondrial fission, but expression of mammalian Drp1 has no effect on the morphology of mitochondria. In addition Drp1 has recently been implicated in the division of peroxisomes (11Koch A. Thiemann M. Grabenbauer M. Yoon Y. McNiven M.A. Schrader M. J. Biol. Chem. 2003; 278: 8597-8605Abstract Full Text Full Text PDF PubMed Scopus (277) Google Scholar). Furthermore, the regulation of mitochondrial morphology is a key factor in the process of apoptosis. Exposure of cells to apoptotic inducers leads to fragmentation of their mitochondria and the release of apoptogenic proteins from the intermembrane space (3Desagher S. Martinou J.C. Trends Cell Biol. 2000; 10: 369-377Abstract Full Text Full Text PDF PubMed Scopus (1693) Google Scholar). Recently it was shown that expression of a mutant Drp1 (Drp1(K38A)) blocked the fragmentation of mitochondria induced by either treatment with the kinase inhibitor staurosporine, or expression of Bax, the pro-apoptotic Bcl-2 family member. Importantly, expression of Drp1(K38A) reduced the staurosporine-induced release of cytochrome c and inhibited apoptosis (12Frank S. Gaume B. Bergmann-Leitner E.S. Leitner W.W. Robert E.G. Catez F. Smith C.L. Youle R.J. Dev. Cell. 2001; 1: 515-525Abstract Full Text Full Text PDF PubMed Scopus (1453) Google Scholar). As yeast Dnm1p requires Fis1p for complete fission of the mitochondria, we identified and cloned the human orthologue of Fis1p (hFis1) and studied its function in mammalian mitochondrial morphology. Isolation and Expression of hFis1—The human homologue of Fis1 (NP057152) was shown to be CGI-135-like (BC009428) after performing a Fasta 3.3 homology search (EBI, Hinxton, UK (13Pearson W.R. Lipman D.J. Proc. Natl. Acad. Sci. U. S. A. 1988; 85: 2444-2448Crossref PubMed Scopus (9381) Google Scholar)). Oligonucleotides derived from the 5′ and 3′ untranslated region of CGI-135-like were used to amplify, by PCR (Pwo or Tgo polymerase, Roche Applied Science, Rotkreuz, Switzerland), CGI-135-like from a human liver library (Serono, Geneva, Switzerland). This template was used in a further PCR step to add His (MAHHHHHH) or HA (MQDLPGNDNSTAGL) tags 5 prime to the ORF of hFis1 lacking the initial methionine or a deletion mutant that lacked the last 28 amino acids (ΔTM). The PCR products were ligated into pCI (Promega, Madison, WI), and all positive clones were sequenced. YFP/CFP/GFP-hFis1 was generated by PCR with the appropriate oligonucleotides to add restriction enzyme sites to the ORF or ORF-ΔTM of hFis1, which was then subcloned into pECFP-CI, pEGFP-CI, pEYFP-CI, or pEGFP-N1 (Clontech, Palo Alto, CA). Expression plasmids for HA-Drp1 and HA-Drp1 K38A were obtained from A van der Bliek and subcloned using PCR into pCI with an alternative HA tag (as above) or as YFP/CFP fusions. Recombinant hFis1 protein was generated by subcloning His-hFis1ΔTM into pTYB1 (New England BioLabs, Beverly, MA). The resultant plasmid was transformed into Escherichia coli BL21 cells for production of the recombinant protein. The intein-His-hFis1 fusion protein was purified over a chitin column, cleaved by dithiothreitol, further purified by affinity to Nickel resin (Qiagen, Basel, Switzerland) and subsequently dialyzed against 30% glycerol, 25 mm Hepes, 1 mm dithiothreitol. Immunizations and antisera collection were performed by Sigma-Genosys (Pampisford, UK). Cell Culture and Reagents—HEK 293, COS-7, and HeLa cells were cultured in Dulbecco's modified Eagle's medium plus 10% fetal bovine serum, 100 units/ml penicillin, 0.1 mg/ml streptomycin, and 2 mm glutamine and maintained in 5% CO2 at 37 °C. Tissue culture plates were obtained from Nunc (Roskilde, Denmark), and all other cell culture reagents were obtained from Sigma (Buchs, Switzerland). The mitochondrial specific dye MitoTracker™ Red CMXRos, was obtained from Molecular Probes (Eugene, OR), and the caspase inhibitor Z-VAD-fmk was from Enzyme Systems Products (Livermore, CA). Fluorescent images were visualized using a Zeiss Axiovert 135TV with a 100× objective, and images were captured using a charge-couple device camera (photometric CE200A) with IP Lab software. Alternatively, for confocal microscopy, images were visualized using a Zeiss inverted 200M microscope with a scanhead QLC100 Nipkow disk (Yokogawa) equipped with a 3-line argon laser 457/488/514 (Laser Physics) with AOTF control of laser lines (Visitech). Pictures were acquired with a Coolsnap HQ camera (Roper Scientific). Image acquisition and analysis was performed with Metamorph/Metafluor 4.1.2 software (Universal Imaging). The resultant images were processed using Adobe Photoshop version 6. Subcellular Fractionation—HeLa cells transfected with His-hFis1 for 24 h (or untransfected for endogenous hFis1) were fractionated as described previously (14Antonsson B. Montessuit S. Sanchez B. Martinou J.C. J. Biol. Chem. 2001; 276: 11615-11623Abstract Full Text Full Text PDF PubMed Scopus (592) Google Scholar). 100 μg of the mitochondrial fraction was incubated in 200 μl of MB buffer (210 mm mannitol, 70 mm sucrose, 10 mm Hepes, pH 7.5, 1 mm EDTA) with 20 μg of proteinase K at room temp for the indicated times, subsequently pelleted at 10,000 g and lysed in 50 μlof1× SDS loading buffer and separated by SDS-PAGE for analysis by Western blotting. The anti-prohibitin antibody was from Neo Markers (Fremont, CA), anti-human Bcl-xL from Chemicon (Temecula, CA), and anti-His (H15) from Santa Cruz Biotechnologies (Santa Cruz, CA). Live Cell Fluorescence—COS-7 or HeLa cells were plated at 2.5 × 105 cells/3-cm glass bottomed dish (Mattek) 6 h prior to transfection. Cells were then transfected using FuGENE (Roche Applied Science) with a 1:1 ratio of target DNA to YFP-mito (Clontech) and a total of 1 μg DNA/dish and observed after a minimum of 10 h post-transfection. Luciferase Assay—COS-7 cells in 3-cm dishes were transfected using Calcium phosphate with up to 2.0 μg of target DNA (pCI-GFP was used to make up any difference) together with 2.0 μg of pCI-Luc. All transfections were performed in triplicate. 16 h after transfection cells were washed once in TBS and cultured for a further 20 h. Cells were then washed once in PBS and lysed in 500 μl of luciferase assay buffer (Promega) for 15 min at room temp. 10 μl of the extract was mixed with 50 μl of LDBII and immediately measured using a Turner TD-20e luminometer. All transfections were performed in triplicate. Immunocytochemistry—HeLa or COS-7 cells were plated on glass coverslips at a confluency of 50–80% in 6-well plates and maintained in culture medium for 2–3 h before transfection with 5 μg of DNA using a standard calcium Phosphate transfection procedure. 16 h after transfection the cells were washed in TBS for 10 min, followed by up to 20-h incubation in culture medium supplemented with 50 μm Z-VAD-fmk. Cells were fixed in 4% paraformaldehyde/PBS for 20 min at room temperature followed by PBS washes. The cells were then permeabilized with 0.1% saponin in PBS for 15 min at room temperature and, after PBS washes were blocked for 1 h at 4 °C, with PBS containing 0.1% saponin and 5% bovine serum albumin. The cells were then incubated with primary antibodies diluted in PBS/0.1% saponin/5% bovine serum albumin for 2 h at room temperature followed by washes in PBS/0.1% saponin. Immunoreactive proteins were visualized by incubating the cells with fluorescein isothiocyanate-coupled mouse secondary and Texas Red-coupled rabbit secondary antibodies (Vector Laboratories, Burlingame, CA) in PBS containing 0.1% saponin for 1 h at room temperature, followed by PBS washes. During the last PBS wash, the cells were co-stained with Hoechst 33258 (25 μg/ml) to visualize the nucleus. The coverslips were then mounted using Vectashield H-100 fluorescent mounting medium (Vector Laboratories). To visualize endogenous hFis1, untransfected cells, which had been plated on glass coverslips, were fixed for 5 min in 100% methanol and processed as above. For UV-induced apoptosis, cells were plated on glass coverslips as described above and transfected, using FuGENE (Roche Applied Science) with 0.2 μg YFP-mito together with 1.0 μg of Bcl-xL or HA-Drp1(K38A) or pCI-Luciferase. 48 h after transfection 100 μm Z-VAD-fmk was added to the media, and the cells were irradiated at 90 mJ/cm2 using a UV Stratalinker 2400 (Stratagene), cultured for a further 16 h, and processed for immunofluorescence as above. To assay cell death, prior to UV irradiation Z-VAD-fmk was omitted from the culture medium. 16 h following UV irradiation, cells were incubated for 30 min with Hoechst 33258, as described above, and transfected cells were scored for apoptotic nuclei. Cells that had a diffuse, non-mitochondrial immunofluorescence using the cytochrome c antibody were scored as having released cytochrome c. Cells were scored as having a fragmented mitochondrial network when >50% of the mitochondria appeared punctate. The following primary antibodies were used: monoclonal anti-cytochrome c antibody (BD Pharmingen, Germany), anti-His polyclonal antibodies (Santa Cruz Biotechnologies), anti-mHsp70 (ABR, Golden, CO). Immunoprecipitation—Cells were seeded in 14-cm dishes 2 h prior to transfection and transfected with a total of 20 μg of DNA using standard calcium phosphate techniques. 16 h after transfection, cells were washed once with 1× TBS, cultured for a further 8–20 h, washed once with 1× PBS, and lysed in 1 ml of immunoprecipitation buffer (10 mm Hepes, 143 mm KCl, 5 mm MgCl2, 1 mm EGTA, 0.5% (v/v) IGEPAL CA 630 supplemented with 1× proteinase inhibitor mixture (Roche Applied Science)) for 20 min on ice. The lysate was cleared by centrifugation at 10,000 × g and incubated with anti-HA antibody (D. Picard, University of Geneva) for 16 h at 4 °C, and the immunocomplexes were captured on 20 μl of protein A/G beads (Amersham Biosciences) for 3 h at 4 °C. Beads were subsequently washed four or five times with 1 ml of lysis buffer and finally resuspended in 40 μl of 1× loading buffer, separated by SDS-PAGE, transferred to Nitrocellulose, and probed with monoclonal anti-Bcl-xL (BD Pharmingen) or anti-HA antibody. The Human Homologue of Yeast Fis1p—Recent advances made in yeast have shed more light on the mechanisms that serve to divide or fuse mitochondria (6van der Bliek A.M. J. Cell Biol. 2000; 151: F1-F4Crossref PubMed Scopus (56) Google Scholar) and the possibility that these may play a part in the apoptotic demise of a cell (12Frank S. Gaume B. Bergmann-Leitner E.S. Leitner W.W. Robert E.G. Catez F. Smith C.L. Youle R.J. Dev. Cell. 2001; 1: 515-525Abstract Full Text Full Text PDF PubMed Scopus (1453) Google Scholar). Although yeast Fis1p has been shown to be localized in the outer membrane of yeast mitochondria and to be necessary for yeast mitochondrial fission (15Mozdy A.D. McCaffery J.M. Shaw J.M. J. Cell Biol. 2000; 151: 367-380Crossref PubMed Scopus (537) Google Scholar), the function of this protein in higher order eukaryotes is unknown. A search of the EMBL data base confirmed the human homologue of yeast Fis1p as CGI-135-like (accession number Q9BTB3), which was similar to a transcript identified using comparative gene identification (CGI) of the human and C. elegans databases (16Lai C.H. Chou C.Y. Ch'ang L.Y. Liu C.S. Lin W. Genome Res. 2000; 10: 703-713Crossref PubMed Scopus (325) Google Scholar). CGI-135-like, which we call hFis1, is 30% homologous to Fis1p at the protein level and 23% at the nucleotide level. Further analysis revealed an orthologue in the mouse (accession number Q9CQ92), which was 96% homologous to hFis1 and orthologues in C. elegans (Q19383) and Drosophila melanogaster (AAL48886), which show 41 and 42% similarity to hFis1, respectively (Fig. 1A). Interestingly, there are two proteins in C. elegans (Q19383 and Q20291) and two in plants (Arabidopsis thaliana, Q94CK3 and Q9M1J1), which are homologous to hFis1 but as yet, only one Fis1p homologue has been found in humans. Two regions of high similarity are apparent from the alignment: the region spanning amino acids (aa) 71–87 from hFis1, which contains no known motif, and the region spanning aa 127–144, which Prosite (EBI, Hinxton, UK) identifies as a putative transmembrane domain. Prosite additionally identified a possible leucine zipper (aa 77–98), a coiled-coil (aa 108–120), and a TPR repeat, which suggest that hFis1 preferentially interacts with proteins containing WD40 repeats and could aggregate in multiprotein complexes. hFis1 Localizes to the Mitochondria—A polyclonal antibody was raised against recombinant hFis1ΔTM and used to assess the subcellular localization of hFis1 in HeLa cells. The cells were found to display a typical mitochondrial staining that was confirmed by co-staining with the mitochondrial marker mHsp70 (Fig. 1C). We also generated expression constructs (Fig. 1B) to confirm the localization of hFis1, and tested the expression in other cell types, including COS-7 and HEK-293 cells (data not shown). Immunofluorescence performed on these cells expressing His-hFis1 confirmed a mitochondrial staining pattern (Fig. 1D). In contrast, a mutant of hFis1 lacking the C terminus (His-hFis1ΔTM) was found to be cytosolic, indicating that the C terminus is required for the mitochondrial localization. To define the submitochondrial localization of endogenous hFis1, mitochondria from HeLa cells were isolated and treated with proteinase K, which digested the parts of the outer membrane proteins exposed to the cytosol. Incubation with proteinase K reduced endogenous hFis1, whereas it did not reduce the levels of prohibitin, which is localized on the inner mitochondrial membrane, indicating that the outer mitochondrial membrane remained intact during the treatment with proteinase K (Fig. 1E). We furthermore showed that the overexpressed His-tagged hFis1 was also sensitive to proteinase K digestion (Fig. 1F). The proteolytic treatment significantly reduced the levels of another mitochondrial membrane protein: Bcl-xL. We therefore concluded that both endogenous and overexpressed hFis1 were localized to the outer mitochondrial membrane and that the presence of the N-terminal tag did not affect the localization of hFis1. hFis1 Promotes Fragmentation of the Mitochondrial Network—During these studies we noticed that the morphology of mitochondria in cells overexpressing hFis1 was abnormal (Fig. 1D) and, therefore, decided to assess the role of hFis1 in mitochondrial morphology. To clearly visualize the changes induced by hFis1 on mitochondrial morphology we fused hFis1 to the C terminus of YFP. Fig. 2 shows representative fluorescent images of YFP-hFis1-expressing cells. At early time points (approximately 10 h, top panel) the mitochondrial network started to bud or fragment, and some mitochondrial swelling was apparent. Later time points (middle panel) were typified by mitochondria, which appear banded or striped and later still (approximately 16 h, bottom panel) the network had degenerated into the punctiform phenotype, which appeared to be identical to the His-tagged hFis1 (Fig. 1D and data not shown). We furthermore confirmed that the structures were part of a mitochondrial network by loading the cells with MitoTracker™ Red CMX Ros (data not shown). The deletion mutant of hFis1 (hFis1ΔTM) was fused to the N terminus of GFP, and its expression pattern showed that this mutant remained localized to the cytosol and did not affect mitochondrial morphology (data not shown). To rule out the possibility that insertion of a protein with a transmembrane domain into the outer membrane of mitochondria was sufficient to trigger fragmentation, we expressed a GFP protein fused to the transmembrane domain of Bax. This construct localized to mitochondria but did not induce fragmentation (data not shown). Therefore the effects seen with hFis1 appear to be specific. hFis1-induced Fragmentation Is Inhibited by Mutant Drp1— Changes in mitochondrial morphology in mammals is known to involve Drp1, which can relocalize from the cytosol to the points of mitochondrial division (17Smirnova E. Griparic L. Shurland D.L. van der Bliek A.M. Mol. Biol. Cell. 2001; 12: 2245-2256Crossref PubMed Scopus (1336) Google Scholar). Expression of a dominant negative mutant of Drp1 (Drp1(K38A)) has also been shown to hinder the changes in mitochondrial morphology induced by apoptotic inducers such as staurosporine (12Frank S. Gaume B. Bergmann-Leitner E.S. Leitner W.W. Robert E.G. Catez F. Smith C.L. Youle R.J. Dev. Cell. 2001; 1: 515-525Abstract Full Text Full Text PDF PubMed Scopus (1453) Google Scholar). We therefore examined the possibility that Drp1(K38A) might prevent the fragmentation induced by hFis1. Drp1 and Drp1(K38A) were subcloned and expressed as HA or CFP fusions and subsequently co-transfected with YFP-hFis1 to assess their effect on hFis1 induced changes in mitochondrial morphology. Expression of CFP-Drp1 had no discernable effect on hFis1-induced fragmentation (Fig. 3A). However, co-expression of CFP-Drp1(K38A) altered the morphology induced by hFis1 (Fig. 3B), and the mitochondrial network appeared as a number of long filaments as well as cisternae-like structures. A quantitative analysis revealed that expression of Drp1(K38A) was able to reduce the number of hFis1-expressing cells with fragmented mitochondria by 31 and 24% at 24 and 36 h, respectively, as compared with cells transfected with hFis1 alone. These results prompted us to investigate how the dominant negative of Drp1 inhibited hFis1 induced fragmentation. In yeast, it has been shown that, in the absence of Fis1p, Dnm1p (the yeast orthologue of Drp1) is not recruited to mitochondria (15Mozdy A.D. McCaffery J.M. Shaw J.M. J. Cell Biol. 2000; 151: 367-380Crossref PubMed Scopus (537) Google Scholar). Therefore we tested if Drp1(K38A) blocked the recruitment of Drp1 to the mitochondria. HeLa cells were co-transfected with hFis1, CFP-mito, and YFP-Drp1 in the presence or absence of HA-Drp1(K38A). Thirty-six hours after transfection the cells were fixed and examined using confocal microscopy. Fig. 3C shows that Drp1 (shown in red) localized principally to the ends of the mitochondria (shown in green). Expression of Drp1(K38A) dramatically altered the localization of Drp1, sequestering it into a number of (usually 3–5) discrete spots (Fig. 3D). We also verified that expression of the dominant negative Drp1 was able to alter localization of Drp1 in the absence of transfected hFis1 using YFP-Drp1 and CFP-Drp1(K38A) (see Supplemental Data). In addition, Western blot analysis of sodium carbonate-treated mitochondria showed that Drp1(K38A) diminished the amount of Drp1 found strongly associated with the outer mitochondrial membrane (Fig. 3E). Taken together, these data indicate that Drp1(K38A) prevented hFis1-induced mitochondrial fragmentation by sequestering Drp1, thereby preventing its localization to the mitochondria. Expression of hFis1 Induced a Caspase-dependent Reduction in Cell Viability—Our data so far showed that tagged hFis1 was effective in promoting the fragmentation of mitochondria and that its localization to the mitochondrial membrane was necessary for its function. However, we had noted that expression of hFis1 for more than 24 h correlated with an increased number of detached cells. We therefore investigated if fragmentation of mitochondria by hFis1 could affect cell viability. We utilized a luciferase assay where a reduction in the translation of a transfected luciferase construct correlates with a decrease in cellular viability after co-transfection with a target gene. Co-transfection with increasing amounts of HA-hFis1 reduced the amount of luciferase in a dose-dependent manner (Fig. 4A). As a positive control we used a plasmid encoding His-tagged Bax, which has previously been shown to induce cell death and observed a pronounced decrease in luciferase activity (data not shown). The addition of the pan-caspase inhibitor Z-VAD-fmk (100 μm) effectively blocked the reduction in the luciferase activity at all concentrations of HA-hFis1 transfected (Fig. 4A and data not shown) indicating that a caspase-dependent process was leading to the changes in viability as assayed by luciferase activity. hFis1 Induces Cytochrome c release—Mitochondria are considered as pivotal organelles in many apoptotic responses. Perturbation of the outer membrane often leads to the release of apoptogenic factors, including cytochrome c. We therefore investigated if the expression of hFis1 led to cytochrome c release (Fig. 4B). HeLa cells were transiently transfected with His-tagged hFis1 and were fixed and stained for His expression and cytochrome c. Expression of hFis1 promoted the release of cytochrome c as shown by the diffuse staining of the cells. Identical results were also obtained using YFP-hFis1. Expression of the ΔTM mutant (His-hFis1ΔTM), which does not localize to mitochondria, did not alter the distribution of cytochrome c (Fig. 4C). These results therefore suggested that misregulated expression of hFis1 was sufficient to promote the redistribution of cytochrome c from the mitochondria to the cytosol. Bcl-xL and Drp1(K38A) Inhibit the Cytochrome c Release Induced by hFis1—Anti-apoptotic proteins of the Bcl-2 family, as well as Drp1(K38A) (12Frank S. Gaume B. Bergmann-Leitner E.S. Leitner W.W. Robert E.G. Catez F. Smith C.L. Youle R.J. Dev. Cell. 2001; 1: 515-525Abstract Full Text Full Text PDF PubMed Scopus (1453) Google Scholar), are known to inhibit the apoptotic process at the mitochondrial membrane. We therefore tested if the co-expression of Bcl-xL and Drp1(K38A) could block the hFis1-induced cytochrome c release. Cells were transiently transfected with His-hFis1 and Bcl-xL or Drp1(K38A) and stained for cytochrome c 36 h later (Fig. 5A). Although 50% of cells expressing hFis1 had released cytochrome c, only 10% of cells co-expressing Bcl-xL and hFis1 and 15% of cells expressing Drp1(K38A) displayed a diffuse cytosolic cytochrome c staining (Fig. 5B). Interestingly, the mitochondria in all cells co-expressing Bcl-xL and hFis1 were punctiform (Fig. 5A), whereas cells co-expressing Drp1(K38A) and hFis1 principally displayed elongated mitochondria. Bcl-xL but Not Drp1(K38A) Prevents hFis1-induced Apoptosis—We also investigated if Bcl-xL and Drp1(K38A) could also prevent apoptosis-induced by hFis1. We therefore co-transfected His-tagged hFis1 with Bcl-xL or Drp1(K38A) and luciferase (Fig. 5C). Bcl-xL but not Drp1(K38A) was able to inhibit the reduction in luciferase activity induced by hFis1. To ensure that the changes in luciferase activity were mirrored by changes in cell viability, cells were incubated with Hoechst and apoptotic nuclei were counted. As shown in Fig. 5C Bcl-xL significantly reduced the number of apoptotic nuclei induced by hFis1 expression, whereas Drp1(K38A) had no effect. hFis1 Triggers the Fragmentation of Mitochondria—In this study we have identified the human orthologue of Fis1p (hFis1) and show that its expression in cultured cells induces fragmentation of the mitochondrial network. The morphology of mitochondria resulting from the overexpression of hFis1 in mammalian cells is indicative of a role in mitochondrial fission and is consistent with previous data obtained in yeast (15Mozdy A.D. McCaffery J.M. Shaw J.M. J. Cell Biol. 2000; 151: 367-380Crossref PubMed Scopus (537) Google Scholar, 18Tieu Q. Nunnari J. J. Cell Biol. 2000; 151: 353-366Crossref PubMed Scopus (288) Google Scholar). Although this finding would argue that the mitochondrial division machinery is conserved during evolution, nevertheless some differences are notable between species. Overexpression of YFP-hFis1 in various mammalian cells (HEK, COS-7, and HeLa) clearly resulted in mitochondrial fragmentation (Fig. 2A), whereas yeast cells that expressed GFP-Fis1p displayed no apparent fragmentation effect yet there was clear mitochondrial localization (15Mozdy A.D. McCaffery J.M. Shaw J.M. J. Cell Biol. 2000; 151: 367-380Crossref PubMed Scopus (537) Google Scholar). Moreover, overexpressed Drp1 did not lead to fragmentation of mammalian mitochondria (data not shown and Ref. 10Smirnova E. Shurland D.L. Ryazantsev S.N. van der Bliek A.M. J. Cell Biol. 1998; 143: 351-358Crossref PubMed Scopus (581) Google Scholar), as also shown in S. cerevisiae (Dnm1p (15Mozdy A.D. McCaffery J.M. Shaw J.M. J. Cell Biol. 2000; 151: 367-380Crossref PubMed Scopus (537) Google Scholar)), whereas the orthologue in C. elegans (DRP (9Labrousse A.M. Zappaterra M.D. Rube D.A. van der Bliek A.M.C Mol. Cell. 1999; 4: 815-826Abstract Full Text Full Text PDF PubMed Scopus (517) Google Scholar)) localized to sites of mitochondrial division and promoted fission. Our data also showed that overexpressed Drp1 was insensitive to sodium carbonate treatment, which suggested that Drp1 was strongly attached to the mitochondrial membrane. In our experiments, overexpressed hFis1 in the outer mitochondrial membrane was sufficient to trigger a complete mitochondrial fission. Although no increase in the amount of overexpressed Drp1 was observed, it is possible that the conformation of Drp1 is altered upon hFis1 expression. One hypothesis is that high hFis1 levels overcome an endogenous inhibitor that would normally maintain the endogenous hFis1 in an inactive state and would impede the recruitment of the division apparatus. It is possible that in mammalian cells the signal that triggers fission of the OMM, as in yeast (8Messerschmitt M. Jakobs S. Vogel F. Fritz S. Dimmer K.S. Neupert W. Westermann B. J. Cell Biol. 2003; 160: 553-564Crossref PubMed Scopus (92) Google Scholar), initiates from the inner membrane fission apparatus. The inhibition of hFis1-induced fragmentation by Drp1(K38A) also places hFis1 within the mitochondrial fission apparatus, upstream of Drp1. However, we have been unable to co-immunoprecipitate endogenous Drp1 and hFis1, and therefore the mechanism by which hFis1 recruits Drp1 is still unclear. The fact that this inhibition was only partial (on average 30%) suggests that Drp1(K38A) is unable to sequester all of the endogenous Drp1 from points of mitochondrial fission or that Drp1-independent fragmentation also occurred. Mitochondrial Fragmentation and Apoptosis—A link between mitochondrial fission and apoptosis has been recently made by Frank et al. (12Frank S. Gaume B. Bergmann-Leitner E.S. Leitner W.W. Robert E.G. Catez F. Smith C.L. Youle R.J. Dev. Cell. 2001; 1: 515-525Abstract Full Text Full Text PDF PubMed Scopus (1453) Google Scholar, 19Karbowski M. Lee Y.J. Gaume B. Jeong S.Y. Frank S. Nechushtan A. Santel A. Fuller M. Smith C.L. Youle R.J. J. Cell Biol. 2002; 159: 931-938Crossref PubMed Scopus (674) Google Scholar) who reported that a mutant of Drp1, DrpK38A, was able to prevent apoptosis triggered by many stimuli. In addition, it was found that the pro-apoptotic Bcl-2 family member Bax colocalized with members of both the fusion (Mfn2) and fission (Drp1) apparatus (12Frank S. Gaume B. Bergmann-Leitner E.S. Leitner W.W. Robert E.G. Catez F. Smith C.L. Youle R.J. Dev. Cell. 2001; 1: 515-525Abstract Full Text Full Text PDF PubMed Scopus (1453) Google Scholar, 19Karbowski M. Lee Y.J. Gaume B. Jeong S.Y. Frank S. Nechushtan A. Santel A. Fuller M. Smith C.L. Youle R.J. J. Cell Biol. 2002; 159: 931-938Crossref PubMed Scopus (674) Google Scholar). Here we show that prolonged misregulation of the mitochondrial fission apparatus was lethal for mammalian cells. Overexpression of hFis1 rapidly fragmented the mitochondrial network (Figs. 1D and 2) into clusters of punctate mitochondria that surrounded the nucleus. This striking change in morphology was accompanied 16 h later by a release of cytochrome c from the mitochondria and subsequent cell death. Overexpression of Bcl-xL was able to inhibit cytochrome c release and cell death, although this failed to prevent mitochondrial fragmentation. This suggests that Bcl-xL does not interfere with the fission machinery per se. However, overexpressed Bcl-xL was found to co-immunoprecipitate with overexpressed hFis1 (see Supplemental Data), but this interaction was not detected between the endogenous proteins. In contrast to Bcl-xL, Drp1(K38A) was able to prevent mitochondrial fission but not cell death. This suggested that the mechanism by which hFis1 triggered apoptosis did not involve proapoptotic Bcl-2 family members, because Drp1(K38A) is known to inhibit many death stimuli involving these proteins (12Frank S. Gaume B. Bergmann-Leitner E.S. Leitner W.W. Robert E.G. Catez F. Smith C.L. Youle R.J. Dev. Cell. 2001; 1: 515-525Abstract Full Text Full Text PDF PubMed Scopus (1453) Google Scholar). Indeed, we found that neither Bax nor Bak were activated in this process, because no conformational change in the N terminus of these proteins was detected using specific antibodies (data not shown). In addition, the cell death process appeared to be independent of the opening of the permeability transition pore (PTP), which Bcl-xL is known to prevent (20Gottlieb E. Vander Heiden M.G. Thompson C.B. Mol. Cell. Biol. 2000; 20: 5680-5689Crossref PubMed Scopus (310) Google Scholar). Indeed, treatment of hFis1-expressing cells with inhibitors of the PTP, namely bongkrekic acid and cyclosporin A, did not reduce the amount of cell death (data not shown). Taken together these results suggest that Bcl-xL inhibit hFis1-induced apoptosis by a mechanism that is Bax-, Bak-, and PTP-independent (21Cheng E.H. Levine B. Boise L.H. Thompson C.B. Hardwick J.M Nature. 1996; 379: 554-556Crossref PubMed Scopus (444) Google Scholar). How prolonged expression of hFis1 triggers apoptosis requires further investigation. These data are important in light of findings that have shown that a disease linked to mutations in the OPA1 gene, which encodes a protein involved in the fusion of the inner mitochondrial membrane (22Olichon A. Emorine L.J. Descoins E. Pelloquin L. Brichese L. Gas N. Guillou E. Delettre C. Valette A. Hamel C.P. Ducommun B. Lenaers G. Belenguer P. FEBS Lett. 2002; 523: 171-176Crossref PubMed Scopus (333) Google Scholar, 23Delettre C. Lenaers G. Griffoin J.M. Gigarel N. Lorenzo C. Belenguer P. Pelloquin L. Grosgeorge J. Turc-Carel C. Perret E. Astarie-Dequeker C. Lasquellec L. Arnaud B. Ducommun B. Kaplan J. Hamel C.P. Nat. Genet. 2000; 26: 207-210Crossref PubMed Scopus (1161) Google Scholar, 24Alexander C. Votruba M. Pesch U.E. Thiselton D.L. Mayer S. Moore A. Rodriguez M. Kellner U. Leo-Kottler B. Auburger G. Bhattacharya S.S. Wissinger B. Nat. Genet. 2000; 26: 211-215Crossref PubMed Scopus (1066) Google Scholar), leads to degeneration of the retinal ganglion cells. Monocytes from patients with OPA1 mutations displayed an abnormal mitochondrial network, with clumps of punctate mitochondria (23Delettre C. Lenaers G. Griffoin J.M. Gigarel N. Lorenzo C. Belenguer P. Pelloquin L. Grosgeorge J. Turc-Carel C. Perret E. Astarie-Dequeker C. Lasquellec L. Arnaud B. Ducommun B. Kaplan J. Hamel C.P. Nat. Genet. 2000; 26: 207-210Crossref PubMed Scopus (1161) Google Scholar), therefore suggesting that mutations in OPA1 lead to a more fragmented mitochondrial network. Our results suggest that the deregulated fission of mitochondria in the ganglion cells of the retina could be responsible for their degeneration through apoptosis. The characterization of a second component of the mammalian mitochondrial fission machinery provides another piece of the jigsaw of this complex process. It will be interesting to see, if like OPA1, defects in the regulation hFis1 are involved in human pathologies. We thank Bernard Corfe for critical reading of the manuscript and Sylvie Montessuit and Monique Fornallaz for technical assistance. In addition we are grateful to Nicolas Demaurex for help with confocal microscopy and to Serono Pharmaceutical Research Institute for providing the liver library used to isolate hFis1 and A. van der Bliek for providing YFP-Drp1, pCDNA-HA-Drp1, and pCDNA-HA-Drp1(K38A). Download .pdf (.13 MB) Help with pdf files
Referência(s)