Modulation of mammalian cell growth and death by prokaryotic and eukaryotic cytochrome c Yoshinori Hiraoka*†, Tohru Yamada*†, Masatoshi Goto*, Tapas K. Das Gupta‡, and Ananda M. Chakrabarty*§ Departments of *Microbiology and Immunology and ‡Surgical Oncology, University of Illinois College of Medicine, 835 South Wolcott Avenue, Chicago, IL 60612 Communicated by Emanuel Margoliash, Northwestern University, Evanston, IL, March 10, 2004 (received for review December 3, 2003)
C
ytochrome c in both prokaryotes and eukaryotes performs a major function: to transfer electrons from reduced substrates to physiological partners, thereby driving the pathway leading to oxidative phosphorylation and ATP synthesis. In eukaryotes, cytochrome c is mitochondrially located for supporting respiration, and a great deal is known about the functionality and structural requirement of cytochrome c for association with mitochondria (1, 2). Cytochrome c is also known to be critical in the induction of apoptosis in eukaryotic cells where, in the presence of certain death signals, it is released from the intermembrane space of the mitochondria to the cell cytosol. The released cytochrome c binds to a cytosolic protein Apaf-1 in a dATP兾ATP-dependent manner that eventually leads to an apoptosome formation and activation of the caspase cascade that triggers apoptosis (3–5). Cytochrome c preparations from various vertebrate species, but not from Saccharomyces cerevisiae, could trigger apoptosis in a Xenopus cell-free system. Interestingly, the electron transport activity of cytochrome c is not required for its proapoptotic function, as copper and zinc substituted cytochrome c, while lacking redox activity, demonstrated proapoptotic function (6). The inability of iso-1 and iso-2 cytochrome c from S. cerevisiae to activate the caspases has been shown to be due to posttranslational modification of lysine 72 trimethylation although iso-1 cytochrome c lacking the trimethylation modification was also reported to lack proapoptotic activity (7). In general, the intracellular release of mitochondrial cytochrome c to the cytosol leads to the apoptotic death of the cell, suggesting a vital function of cytochrome c in sustaining both the life and the death of the cells (3). We have recently reported that a clinical strain of Pseudomonas aeruginosa secretes two types of enzymes, ATP-using enzymes and redox enzymes such as azurin and cytochrome c551 (8). Azurin is a member of a family of blue copper proteins www.pnas.org兾cgi兾doi兾10.1073兾pnas.0401631101
known as cupredoxins that are small, soluble proteins (10–17 kDa) whose active site contains a type 1 copper (9, 10). Both azurin and cytochrome c551 can act as partners in electron transfer (ET) in in vitro systems whereas cd1 nitrite reductase is the physiological acceptor in ET. Zaborina et al. (8) demonstrated that a mixture of purified preparations of azurin and cytochrome c551 could enter mammalian cells such as J774 cell line-derived macrophages and induced apoptosis in such cells. More recently, Goto et al. (11) reported that cytochrome c551 had less cytotoxicity than azurin although its presence added to the cytotoxicity of azurin. Redox activity of azurin was not important in cytotoxicity (11). Yamada et al. (12) demonstrated that azurin enters J774 cells, forms a complex with the tumor suppressor protein p53, stabilizes it, and enhances the intracellular concentration of p53. The high intracellular level of p53 then allows higher level of Bax formation and its mobilization to the mitochondria of J774 cells, triggering apoptosis (12). The role of cytochrome c551 has remained undefined so far. We now report that purified cytochrome c551 from P. aeruginosa (termed Cyt c551) enters J774 cells and inhibits cell cycle progression through modulation of a tumor suppressor protein, p16Ink4a. Thus Cyt c551, similar to azurin, is considered a potential virulence factor elaborated by P. aeruginosa for evading host defense, a function unrelated to its hitherto known function in electron transport. Materials and Methods Preparation of WT and Mutant Cytochrome c551. Escherichia coli
JCB7120 was used as a host strain for expression of the WT and V23DI59E mutant cytochrome c551-encoding gene of P. aeruginosa. The hyperproduction and purification of the cytochrome c551 gene product has been reported (11). The recombinant strain JCB7120 was cultivated under anaerobic conditions at 37°C in the minimal medium as described by Hasegawa et al. (13). Site-Directed Mutagenesis of the Cytochrome c551 Gene. Sitedirected mutagenesis of the Cyt c551 gene was performed by using a QuickChange site-directed mutagenesis kit (Stratagene). A single set of oligonucleotides was designed for each mutation as follows: for V23D, 5⬘-GACACCAAGATGGACGGCCCGGCCTAC-3⬘ and 5⬘-GTAGGCCGGGCCGTCCATCTTGGTGTC-3⬘; for I59E, 5⬘-GTCTGGGGCCCGGAGCCGATGCCGCCGA AC-3⬘ and 5⬘-GT TCGGCGGCATCGGCTCCGGGCCCCAGAC-3⬘. Mutations were confirmed by DNA sequencing. The underlined codons encode the mutated amino acids. Cell Culture. The J774 cell line-derived murine macrophage cells were cultured in RPMI medium 1640 containing 2 mM Lglutamine, 10 mM Hepes, 10% (vol兾vol) heat-inactivated FBS,
Abbreviations: ET, electron transfer; Cyt c551, purified cytochrome c551 from P. aeruginosa; CDK, cyclin-dependent kinase. †Y.H. §To
and T.Y. contributed equally to this work.
whom correspondence should be addressed. E-mail:
[email protected].
© 2004 by The National Academy of Sciences of the USA
PNAS 兩 April 27, 2004 兩 vol. 101 兩 no. 17 兩 6427– 6432
BIOCHEMISTRY
Cytochrome c551, an 8,685-Da haem-containing protein, is known to be involved in electron transfer during dissimilative denitrification by Pseudomonas aeruginosa. Both cytochrome c551 and copper-containing redox protein azurin have been used in vitro as partners in electron transfer. Azurin has been reported to induce apoptosis in macrophages and cancer cells. We now report that, unlike azurin, cytochrome c551, termed Cyt c551, has very little ability to induce apoptosis in J774 cell line-derived macrophages but demonstrates significant inhibition of cell cycle progression in such cells. A mutant form of Cyt c551, V23DI59E, has significantly reduced ability to inhibit cell cycle progression but demonstrates a higher level of apoptosis-inducing activity in macrophages, compared with WT Cyt c551. Interestingly, the WT Cyt c551, but not the mutant form, significantly enhances the level of tumor suppressor protein p16Ink4a, a known inhibitor of cell cycle progression whereas the mutant form seems to form a complex with tumor suppressor protein p53, thereby enhancing its intracellular level to some extent. Eukaryotic cytochromes such as horse and bovine cytochrome c have also been shown to induce apoptosis but not inhibition of cell cycle progression in J774 cells, thus signifying a role of this redox protein in entry to, and in the induction of, cell death in mammalian cells.
100 units兾ml penicillin, and 100 g兾ml streptomycin at 37°C in a humidified incubator with 5% CO2 as described by Yamada et al. (12). The human adenocarcinoma cells of MCF-7 and MDD-2 were cultivated in MEM with Earle’s salts supplemented with 10% (vol兾vol) heat-inactivated FBS and 1 mM glutamic acid. Cell Cycle Analysis. The J774 cells (2 ⫻ 106 cells per well in six-well
plates) were treated with WT Cyt c551, mutant Cyt c551, yeast (S. cerevisiae), horse, and bovine Cyt c (50, 100, 200, 400, and 800 g兾ml) for 0, 4, 8, and 24 h. The cells were washed with PBS, fixed with 70% ethanol, and stored at ⫺20°C. Fixed cells were washed twice with PBS and stained with 50 g兾ml of propidium iodide in PBS containing 20 g兾ml RNase A for 30 min in the dark and analyzed by flow cytometry (Becton Dickinson). The percentage of cells in different phages of cell cycle was determined by MODFIT LT software. Quantification of Apoptosis. For quantification of apoptosis, the
J774 cells (2 ⫻ 106 cells per well in six-well plates) were treated with WT Cyt c551, mutant Cyt c551, yeast, horse, and bovine Cyt c (100, 200, 400, and 800 g兾ml) for 24 h. The cells were washed with PBS and fixed with 1% paraformaldehyde. Fixed cells were washed twice with PBS and suspended in 70% ethanol and stored at ⫺20°C. The cells were stained by using an APODIRECT apoptosis kit (Phoenix Flow Systems, San Diego) and analyzed by flow cytometry according to the manufacturer’s protocol. Similar experiments were also done with the human breast cancer cell line MCF-7 and its p53-negative derivative MDD-2. Intracellular Localization of Cytochrome c551. Subcellular fractionation.
The cells (107 cells) were treated with WT or mutant Cyt c551 (400 g兾ml) for 0, 4, 8, and 24 h. After treatment, cytosolic and nuclear fractions were isolated as described for subcellular localization of azurin (12). Each fraction (30 g of protein) was loaded on SDS兾PAGE and blotted on poly(vinylidene difluoride) membranes. The presence of Cyt c551 was confirmed by using anti-Cyt c551 antibody. Confocal microscopy. WT and mutant Cyt c551 proteins were conjugated with the fluorescent chemical Alexa Fluor 568 (Molecular Probes) and incubated with J774 cells for 1 h. Entry of fluorescent chemically labeled Cyt c551 into the cells was observed by a Carl Zeiss LSM 510 confocal microscope as described (12). Western Blot Analysis. After the isolation of various subcellular
fractions as described above, each fraction (30 g of protein) was separated by SDS兾PAGE and transferred to poly(vinylidene difluoride) membranes. The membranes were probed with primary antibodies followed by peroxidase-conjugated appropriate secondary antibody, and visualized by ECL detection system (Amersham Pharmacia). The following primary antibodies were used: p53, p15, p16, p18, and p19 (from Santa Cruz Biotechnology); p21, p27, cyclin D1, cyclin E, cyclin-dependent kinase (CDK) 2, CDK4, and CDK6 (BioSource International, Camarillo, CA); actin (from Sigma-Aldrich).
Glycerol Gradient Centrifugation Analysis. Complex formation be-
tween WT Cyt c551, mutant Cyt c551, and p53 was confirmed by glycerol gradient centrifugation analysis as described (12). Results and Discussion Entry of Cyt c551 in J774 Cells. Whereas mitochondrial cytochrome
c is well known to be released to the cytosol of mammalian cells, very little is known about the ability of purified cytochrome c proteins to enter mammalian cells and interfere in cellular functions. The results reported by Zaborina et al. (8) and Goto et al. (11) implied that Cyt c551 may also enter J774 macrophages, 6428 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0401631101
Fig. 1. (A) SDS兾PAGE of WT and V23DI59E double-mutant Cyt c551 proteins obtained after hyperexpression in E. coli. (B) Uptake at 1 h of fluorescent chemical Alexa Fluor 568-conjugated WT and V23DI59E mutant Cyt c551 in J774 cells as observed by confocal microscopy. The nucleus is stained blue with 4⬘,6-diamidino-2-phenylindole (DAPI). (C) The localization of WT and V23DI59E mutant Cyt c551 in the subcellular fractions of J774 cells either untreated (0 h) or treated with WT and mutant Cyt c551 for 4, 8, and 24 h. After each period, cell extracts were made, and different subcellular fractions were collected and monitored by immunoblotting by using anti-Cyt c551 antibody.
either singly or as a complex with azurin. As an ET partner of azurin in in vitro ET assays (11), Cyt c551 is known to form a complex with its ET partners through a hydrophobic surface patch (14). To determine whether Cyt c551 can enter mammalian cells such as J774 cells by itself, we hyperexpressed the P. aeruginosa Cyt c551 gene cloned in E. coli (13) as described by Goto et al. (11). We also replaced two hydrophobic residues at positions 23 (a valine to aspartic acid) and position 59 (an isoleucine to glutamic acid) in the surface hydrophobic patch of Cyt c551. These residues were shown to be important in the protein:protein interaction of Cyt c551 for its ET activity with its partners (14). The purified proteins migrated as 8-kDa proteins as expected although the Val23AspIle59Glu (V23DI59E) mutant Cyt c551 had an aberrant mobility (Fig. 1A). To determine the entry of Cyt c551 in J774 cells, we conjugated Alexa Fluor 568, a red fluorescent dye, with WT Cyt c551 and the V23DI59E mutant Cyt c551 and used confocal electron microscopy. Both the WT and the mutant Cyt c551 were found in the cytosol of J774 cells after 1 h of incubation (Fig. 1B). To determine whether a longer period of incubation will allow mobilization of Cyt c551 to the nucleus, as is seen with azurin (12, 15), we incubated J774 cells for 0 (no Cyt c551), 4, 8, and 24 h with 400 g兾ml WT and V23DI59E mutant Cyt c551, washed the cells to remove the proteins, made extracts, and fractionated the extracts to isolate cytosolic and nuclear fractions. Such fractions were then examined by Western blotting with anti-Cyt c551 antibody for the presence of Cyt c551. Even after 24 h incubation, Cyt c551 was found only in the cytosol but not in the nucleus (Fig. 1C). As seen with confocal microscopy, both the WT and the mutant Cyt c551 were found in the cytosol in 4 h. The absence of Cyt c551 in the nucleus contrasts with the presence of azurin in the nucleus soon after its entry into the cell, where azurin was seen to be present in the nucleus of p53-positive cells but not in the nucleus of p53-negative cells (15, 16), indicating a piggybacking of azurin to the nucleus by p53 (15, 16). Thus Cyt c551 seems to behave differently than azurin with regard to its complex formation with p53. WT Cyt c551 Inhibits Cell Cycle Progression at the G1 to S Phase. To examine whether Cyt c551 may demonstrate a similar mode of Hiraoka et al.
of treatment of J774 cells with 200 g兾ml WT Cyt c551 for varying time periods demonstrated very little inhibition at 4 h, but significant inhibition at 8 to 24 h (Fig. 3B). In contrast, treatment of J774 cells with varying concentrations of V23DI59E mutant Cyt c551 for 24 h demonstrated ⬍25% inhibition upto a concentration of 200 g兾ml (Fig. 3C). WT Cyt c551 Causes Inhibition of Cell Cycle Progression by Enhancing the Intracellular Level of p16Ink4a. We previously reported that Fig. 2. Apoptotic activity of WT (A) and V23DI59E mutant Cyt c551 (B) toward J774 cells. The cells were untreated (Control) or treated with WT and mutant Cyt c551 (100 – 800 g兾ml for 24 h), and analyzed by using APO-DIRECT apoptosis kit (Phoenix Flow Systems). Detailed methods are given under Materials and Methods. (C) The results of A and B are plotted graphically to demonstrate relative apoptotic activity of the WT and mutant Cyt c551.
BIOCHEMISTRY
action as azurin in inducing apoptosis in J774 cells (12), we treated J774 cells with different concentrations of WT and V23DI59E mutant Cyt c551 and measured the extent of apoptosis by using the APO-DIRECT apoptosis kit. WT Cyt c551 demonstrated a low level of apoptosis (Fig. 2 A and C) whereas the mutant Cyt c551 had a comparatively higher level of activity (Fig. 2 B and C). To examine whether either the WT or the V23DI59E mutant Cyt c551 may influence the cell cycle check points, we also determined the effect of WT and mutant Cyt c551 on the cell cycle progression in J774 cells. Treatment of J774 cells for 24 h with increasing concentrations of WT Cyt c551 led to increasing inhibition of cell cycle at the G1 to S phase reaching by ⬎75% at 200 g兾ml WT Cyt c551 concentration (Fig. 3A). A time curve
azurin induces apoptosis in J774 cells through stabilization of p53 and enhancing its intracellular level. p53 has two major functions in the cell; it is an inducer of apoptosis (17, 18), but it also causes growth arrest through inhibition of cell cycle progression (19, 20). As a sequence-specific transcriptional regulator, p53 can activate the promoters of the bax or the p21 genes, thereby modulating apoptosis or growth arrest, but its mode of activation of specific promoters differs with respect to cellular backgrounds, nature of stress, and a variety of other factors (21). To determine whether the strong inhibition of cell cycle progression in J774 cells might be due to modulation of the intracellular levels of cyclins and CDKs involved in cell cycle progression (19, 20) or the levels of p53 or other tumor suppressors, we treated J774 cells with WT Cyt c551 for 0, 4, 8, and 24 h, washed the cells, made extracts, fractionated the extracts, and determined the levels of various proteins involved in cell cycle progression. Commensurate with the pattern of the inhibition of cell cycle progression at the G1 to S phase, the level of cyclin D such as cyclin D1, but not that of cyclin E, was significantly inhibited by the WT Cyt c551 as a function of period of incubation (Fig. 4A).
Fig. 3. Effect of WT and V23DI59E mutant Cyt c551 on cell cycle in J774 cells. (A) The J774 cells were harvested either without treatment (0 g兾ml) for 24 h or after 24 h treatment with different concentrations of WT Cyt c551 (50, 100, 200 g兾ml). (B) The cells were treated with 200 g兾ml WT Cyt c551 and harvested at the indicated times. (C) The cells were harvested either without treatment (0 g兾ml) for 24 h or after 24 h treatment with V23DI59Emutant Cyt c551 (50, 100, 200 g兾ml). After DNA staining with propidium iodide (50 g兾ml), cells were analyzed by flow cytometry. Percentages of DNA content in cells at G0兾G1, S, and G2兾M phase were determined by using MODFIT LT software.
Hiraoka et al.
PNAS 兩 April 27, 2004 兩 vol. 101 兩 no. 17 兩 6429
Fig. 5. Complex formation between WT Cyt c551, V23DI59E mutant Cyt c551, and p53 as confirmed by glycerol gradient centrifugation analysis by using anti-Cyt c551 (A) and anti-p53 (B) antibodies. The detailed methods are described elsewhere (12, 15).
Fig. 4. Effect of WT (A) and V23DI59E mutant Cyt c551 (B) on cell cycle regulatory proteins in J774 cells. The J774 cells were treated with WT and mutant Cyt c551 for the indicated times, and cytoplasmic and nuclear protein fractions were prepared for Western blot analysis for the indicated proteins. (C) Densitometric scan of Western blot of p16 and p53 from A and B. The signals of the p16 and p53 bands were quantified and plotted relative to the values of control (0 h).
The levels of critical CDKs such as CDK2, CDK4 and also of CDK6, to some extent, were significantly reduced. The levels of p53, p21, p27 cip兾kip family of inhibitors, as well as Ink4 member proteins such as p15Ink4b, p18Ink4c, and p19Ink4d, were not significantly affected by WT Cyt c551. Most interestingly, however, the level of p16Ink4a, a tumor suppressor well known for its ability to inhibit cell cycle progression at the G1 to S phase (19), was significantly elevated in presence of Cyt c551 (Fig. 4A). The levels of actin as a control protein did not change. The Ink4 proteins, including p16, are known to sequester CDK4 and CDK6 into binary CDK-Ink4 complexes, thereby greatly reducing the intracellular levels of CDKs and inhibiting cell cycle progression (19, 20). To examine whether the V23DI59E mutant Cyt c551 [whose ability to inhibit cell cycle progression in J774 cells was greatly reduced (Fig. 3C) whereas its ability to induce apoptosis increased substantially (Fig. 2 B and C)] has any differential effect on the intracellular levels of tumor suppressors, we determined the intracellular levels of p53 and p16 in J774 cells treated for 0, 4, 8, and 24 h with 400 g兾ml the mutant Cyt c551 (Fig. 4B). Quantitation of such levels demonstrated that, whereas the p16 levels increased significantly in presence of WT Cyt c551, its level was unaltered or slightly went down in the presence of the V23DI59E mutant Cyt c551 (Fig. 4C). In contrast, whereas the level of p53 remained fairly unchanged in presence of WT Cyt c551 (Fig. 4C), the levels of p53 increased slightly but steadily in presence of the mutant Cyt c551 (Fig. 4C). Glycerol Gradient Centrifugation Indicates Association of Mutant but Not WT Cyt c551 with p53. The ability of the mutant Cyt c551, but
not the WT, to induce apoptosis in J774 cells, and its ability to enhance the intracellular p53 level, although only modestly, raised the question whether the mutant Cyt c551 may mimic azurin in its mode of action. We previously used glycerol gradient centrifugation to follow the association of azurin with p53. In this method, azurin and GST-p53 fusion proteins were sedimented 6430 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0401631101
singly or in combination in 5 to 25% glycerol gradients, various glycerol gradient fractions were collected, and the presence of azurin or p53 in various fractions was determined by immunoblotting with anti-azurin and anti-p53 monoclonal antibodies (12, 15). Because p53 forms various oligomeric complexes, such complexes sediment in different glycerol fractions. Because azurin formed complexes with various oligomeric forms of p53, when a combination of azurin and p53 were sedimented together, azurin was found in all of the fractions where GST-p53 sedimented (12, 15). A combination of GST and azurin had no such effect. Such complex formation was later verified by GST pull-down assays as well (22). We, therefore, centrifuged WT and V23DI59E mutant Cyt c551 singly or in combination with p53 or GST in a 5–25% glycerol gradient, collected various gradient fractions, and determined the presence of WT or mutant Cyt c551 by using anti-Cyt c551 antibody (Fig. 5A) or the presence of p53 by using anti-p53 antibodies (Fig. 5B). WT or V23DI59E mutant Cyt c551 sedimented at 5% glycerol. Whereas a mixture of WT Cyt c551 and p53 allowed sedimentation only at 5% glycerol, a mixture of V23DI59E mutant Cyt c551 and p53 allowed sedimentation of the mutant Cyt c551 at 5, 10, 15, and 20% glycerol gradients (Fig. 5A). p53 alone, or p53 with WT or mutant Cyt c551, sedimented as oligomers in 5, 10, 15, 20, and 25% glycerol (Fig. 5B). This result indicated that the mutant Cyt c551, but not the WT Cyt c551, formed complexes with oligomeric forms of p53. Whether such a complex formation leads to stabilization of p53, thereby enhancing its intracellular levels (Fig. 4 B and C) or its proapoptotic property (Fig. 2 B and C), is not clear. Do Other Eukaryotic Cytochromes c Enter J774 Cells and Exert Cytotoxicity? The ability of the bacterial WT Cyt c551 to induce
inhibition of cell cycle progression in J774 cells or the ability of the V23DI59E mutant Cyt c551 to induce apoptosis raised an interesting question: do eukaryotic cytochrome c proteins enter J774 cells and demonstrate either growth arrest or apoptotic properties? It has recently been shown that certain amino acids, viz., residues 7, 25, 39, 62–65, and 72 of horse or bovine cytochrome c are important in allowing cytochrome c’s interaction with Apaf-1, and these residues are altered in yeast cytochrome c, thereby accounting for the inability of yeast cytochrome c to exert its proapoptotic effect (23). It is, nevertheless, possible that yeast cytochrome c, similar to P. aeruginosa Cyt c551, may inhibit cell cycle progression, provided of course it can enter J774 cells. None of the three eukaryotic cytochromes c (yeast, horse, and bovine) had any effect on the inhibition of cell cycle progression when J774 cells were incubated with the purified proteins (400 g兾ml) for 24 h (Fig. 6A). It was deemed possible that none of these proteins could enter the J774 cells to exert any effect. Interestingly, however, when the J774 cells were incubated with various concentrations of these three proteins for 24 h and then examined for the induction of apoptosis, significant apoptosis was seen to be induced by bovine and horse cytoHiraoka et al.
Fig. 6. (A) Effect of eukaryotic cytochrome c on cell cycle in J774 cells. The cells were treated with 400 g兾ml indicated cytochrome c, harvested after 24 h, and analyzed by flow cytometry as described under Materials and Methods. The Control had no cytochrome c. (B) Apoptotic activity of eukaryotic cytochrome c toward J774 cells. The cells were treated with the indicated concentrations of cytochrome c for 24 h, and analyzed by using APO-DIRECT apoptosis kit. (C) Amino acid sequence comparison of cytochrome c from P. aeruginosa and eukaryotic sources. The multiple sequence alignment was produced with GENETYX software. Amino acid sequences are aligned by using residue numbers of PAc551. PA, P. aeruginosa.
Hiraoka et al.
tions (Fig. 7). This result seems to indicate that, whereas entry of vertebrate cytochrome c in the cytosol allowed induction of apoptosis irrespective of the p53 status of the cells, presumably through complex formation with Apaf-1, the bacterial V23DI59E mutant Cyt c551 could elicit apoptosis only when complexed with p53, which presumably allowed release of mitochondrial cytochrome c into the cytosol, thus triggering apoptosis (12, 15). In conclusion, P. aeruginosa Cyt c551 can enter mammalian cells such as J774 cell line-derived macrophages. J774 cells are murine reticulum cell sarcoma whose ascites form has the
BIOCHEMISTRY
chrome c but not by yeast cytochrome c (Fig. 6B). Thus, the vertebrate cytochrome c could enter J774 cells and induce apoptosis. Whether the yeast cytochrome c could enter J774 cells has not been looked at, but the fact that yeast cytochrome c has altered critical amino acids at positions 7, 25, 39, 62–65, and 72 (23) may explain its lack of apoptosis-inducing activity even if it entered the J774 cells. It is interesting to note that the bacterial WT Cyt c551 also lacks all of the critical amino acids required for Apaf-1 binding by the bovine and horse cytochrome c (Fig. 6C). Thus the WT Cyt c551 is presumably unable to form complexes with Apaf-1. Because the bovine or horse cytochrome c, when present in the cytosol, can induce apoptosis through complex formation with Apaf-1 whereas Cyt c551 presumably is unable to form complexes with Apaf-1 because of the loss of critical amino acids (Fig. 6C), an important question was whether the V23DI59E mutant Cyt c551 could induce apoptosis in J774 cells because of its complex-forming ability with p53 (Fig. 5A), similar to azurin (12). We therefore tested the effects of bovine and horse cytochrome c and V23DI59E mutant Cyt c551 on the induction of apoptosis in two cell lines derived from the mammary adenocarcinoma cells, the WT p53-positive MCF-7, and a dominant p53-negative mutant MDD-2 (24, 25). The bovine cytochrome c was effective in inducing significant apoptosis in both MCF-7 and MDD-2 cells, particularly at high concentrations (Fig. 7). At low concentrations, bovine cytochrome c was less effective in p53-deficient MDD-2 cells, but at or ⬎400 g兾ml, it was equally active in both p53-positive and p53negative cell lines. Similar p53-independent activity profile was noted for the horse cytochrome c (data not shown). In contrast, V23DI59E mutant Cyt c551 had a low level of apoptosis-inducing activity in p53-positive MCF-7 cells but demonstrated very little activity in p53-negative MDD-2 cells, even at high concentra-
Fig. 7. Apoptotic activity of bovine cytochrome c and bacterial V23DI59E mutant Cyt c551 toward MCF-7 and MDD-2 cells. The cells were treated with the indicated concentrations of cytochrome c for 24 h, and analyzed by using APO-DIRECT apoptosis kit as described under Materials and Methods. PNAS 兩 April 27, 2004 兩 vol. 101 兩 no. 17 兩 6431
macrophage properties of adherence, morphology, receptors for Ig, and antibody-dependent lysis of target cells (26). Entry of Cyt c551 in the cytosol of J774 cells allows enhanced accumulation of the tumor suppressor protein p16Ink4a, an inhibitor of cell cycle progression at the G1 to S phase because of its ability to sequester CDK4兾6 through complex formation, thereby negatively affecting phosphorylation of Rb (retinoblastoma) whose phosphorylation is important for S phase DNA synthesis and cell cycle progression (19, 20). When two hydrophobic amino acids in the hydrophobic patch of Cyt c551 are replaced by two polar amino acids, the resultant mutant V23DI59E Cyt c551 has greatly reduced ability to inhibit cell cycle progression in J774 cells but seems to gain an ability to
form a complex with the tumor suppressor p53 and demonstrates an ability to induce apoptosis in J774 cells. Most interestingly, vertebrate cytochrome c such as horse or bovine cytochrome c seems to mimic V23DI59E Cyt c551 in being able to enter and induce apoptosis in J774 cells although their dependence on p53 was different. Thus, not only bacterial azurin and mutant Cyt c551 but also mammalian cytochrome c seem to induce apoptosis in macrophages even when present outside the macrophages.
1. Moore, G. R. & Pettigrew, G. W. (1990) Cytochromes c: Evolutionary, Structural and Physicochemical Aspects (Springer, Berlin). 2. Margoliash, E. & Schejter, A. (1996) in Cytochrome c: A Multi-disciplinary Approach, eds. Scott, R. A. & Mauk, A. G. (University Science Books, Sausalito, CA), pp. 3–31. 3. Reed, J. C. (1997) Cell 91, 559–562. 4. Adrain, C. & Martin, S. J. (2001) Trends Biochem. Sci. 26, 390–397. 5. van Loo, G., Saelens, X., van Gurp, M., MacFarlane, M., Martin, S. J. & Vandenabeele, P. (2002) Cell Death Differ. 9, 1031–1042. 6. Kluck, R. M., Martin, S. J., Hoffman, B. M., Zhou, J. S., Green, D. R. & Newmeyer, D. D. (1997) EMBO J. 16, 4639–4649. 7. Kluck, R. M., Ellerby, L. M., Ellerby, H. M., Naiem, S., Yaffe, M. P., Margoliash, E., Bredesen, D., Mauk, A. G., Sherman, F. & Newmeyer, D. D. (2000) J. Biol. Chem. 275, 16127–16133. 8. Zaborina, O., Dhiman, N., Chen, M. L., Kostal, J., Holder, I. A. & Chakrabarty, A. M. (2000) Microbiology 146, 2521–2530. 9. de Rienzo, F., Gabdoulline, R. R., Menziani, M. C. & Wade, R. C. (2000) Protein Sci. 9, 1439–1454. 10. Murphy, L. M., Dodd, F. E., Yousafzai, F. K., Eady, R. R. & Hasnain, S. S. (2002) J. Mol. Biol. 315, 859–871. 11. Goto, M., Yamada, T., Kimbara, K., Horner, J., Newcomb, M., Das Gupta, T. K. & Chakrabarty, A. M. (2003) Mol. Microbiol. 47, 549 –559. 12. Yamada, T., Goto, M., Punj, V., Zaborina, O., Kimbara, K., Das Gupta, T. K. & Chakrabarty, A. M. (2002) Infect. Immun. 70, 7054–7062.
13. Hasegawa, J., Shimahara, H., Mizutani, M., Uchiyama, S., Arai, H., Ishii, M., Kobayashi, Y., Ferguson S. J., Sambongi Y. & Igarashi, Y. (1999) J. Biol. Chem. 274, 37533–37537. 14. Cutruzzola, F., Arese, M., Ranghino, G., van Pouderoyen, G., Canters, G. & Brunori, M. (2002) J. Inorg. Biochem. 88, 353–361. 15. Yamada, T., Goto, M., Punj, V., Zaborina, O., Chen, M. L., Kimbara, K., Majumdar, D., Cunningham, E., Das Gupta, T. K. & Chakrabarty, A. M. (2002) Proc. Natl. Acad. Sci. USA 99, 14098–14103. 16. Punj, V. & Chakrabarty, A. M. (2003) Cell. Microbiol. 5, 225–231. 17. Vogelstein, B., Lane, D. & Levine, A. J. (2000) Nature 408, 307–310. 18. Schuler, M. & Green, D. R. (2001) Biochem. Soc. Trans. 29, 684–688. 19. Sherr, C. J. (2000) Cancer Res. 60, 3689–3695. 20. Dotto, G. P. (2000) Biochim. Biophys. Acta 1471, M43–M56. 21. Zhao, R., Gish, K., Murphy, M., Yin, Y., Notterman, D., Hoffman, W. H., Tom, E., Mack, D. H. & Levine, A. J. (2000) Genes Dev. 14, 981–993. 22. Punj, V., Das Gupta, T. K. & Chakrabarty, A. M. (2003) Biochem. Biophys. Res. Commun. 312, 109–114. 23. Yu, T., Wang, X., Purring-Koch, C., Wei, Y. & McLendon, G. L. (2001) J. Biol. Chem. 276, 13034–13038. 24. Galmarini, C. M., Clarke, M. L., Falette, N., Puisieux, A., Mackey, J. R. & Dumonted, C. (2002) Int. J. Cancer 97, 439–445. 25. Punj, V., Bhattacharyya, S., Saint-Dic, D., Vasu, C., Cunningham, E. A., Graves, J., Yamada T., Constantinou A. I., Christov K., White B., et al. (2004) Oncogene 23, 2367–2378. 26. Ralph, P. & Nakoinz, I. (1975) Nature 257, 393–394.
6432 兩 www.pnas.org兾cgi兾doi兾10.1073兾pnas.0401631101
This research is funded by Public Health Service Grant ES-04050-17 from the National Institute of Environmental Health Sciences.
Hiraoka et al.