Tankyrase 1 Interacts with Mcl-1 Proteins and Inhibits Their Regulation of Apoptosis*

Mcl-1L (myeloid cell leukemia-1 long) is an antiapoptotic Bcl-2 family protein discovered as an early induction gene during leukemia cell differentiation. Previously, we identified Mcl-1S (short) as a short splicing variant of theMcl-1 gene with proapoptotic activity. To identify Mcl-1-interacting proteins, we performed yeast two-hybrid screening and found cDNAs encoding tankyrase 1. This protein possesses poly(ADP-ribose) polymerase activity and presumably facilitates the turnover of substrates following ADP-ribosylation. In yeast and mammalian cells, tankyrase 1 interacts with both Mcl-1L and Mcl-1S, but does not bind to other Bcl-2 family proteins tested. Analysis of truncated tankyrase 1 mutants indicated that the first 10 ankyrin repeats are involved in interaction with Mcl-1. In the N terminus of Mcl-1, a stretch of 25 amino acids is sufficient for binding to tankyrase 1. Overexpression of tankyrase 1 antagonizes both Mcl-1L-mediated cell survival and Mcl-1S-induced cell death. Furthermore, coexpression of tankyrase 1 with Mcl-1L or Mcl-1S decreased the levels of Mcl-1 proteins. Although tankyrase 1 down-regulates Mcl-1 protein expression, no ADP-ribosylation of Mcl-1 was detected. In contrast, overexpression of Mcl-1 proteins suppressed the ADP-ribosylation of the telomeric repeat binding factor 1, another tankyrase 1-interacting protein. Thus, interaction of Mcl-1L and Mcl-1S with tankyrase 1 could serve as a unique mechanism to decrease the expression of these Bcl-2 family proteins, thereby leading to the modulation of the apoptosis pathway.


Mcl-1L (myeloid cell leukemia-1 long) is an antiapo-
Phylogenetically conserved Bcl-2 family proteins play a pivotal role in the regulation of apoptosis from virus to human (1). Members of the Bcl-2 family consist of antiapoptotic proteins such as Bcl-2, Bcl-x L , and Bcl-w, and proapoptotic proteins such as BAD, Bax, BOD, and Bok. It has been proposed that anti-and proapoptotic Bcl-2 proteins regulate cell death by binding to each other and forming heterodimers (2). A delicate balance between anti-and proapoptotic Bcl-2 family members exists in each cell and the relative concentration of these two groups of proteins determines whether the cell survives or undergoes apoptosis.
In the present study, we found tankyrase 1 as a specificbinding protein of Mcl-1. Overexpression of tankyrase 1 led to the inhibition of both the survival action of Mcl-1L and the apoptotic activity of Mcl-1S in mammalian cells. Unlike other known tankyrase 1-interacting proteins, tankyrase 1 did not poly(ADP-ribosyl)ate either of the Mcl-1 proteins despite its ability to decrease Mcl-1 protein levels following coexpression. Therefore, tankyrase 1 could regulate Mcl-1-modulated apo-ptosis by down-regulating the expression of Mcl-1 proteins without the involvement of its ADP-ribosylation activity.

EXPERIMENTAL PROCEDURES
Yeast Two-hybrid Screening-The open reading frame of human Mcl-1S cDNA was fused in-frame with the GAL4-binding domain into the pGBT9 yeast shuttle vector (Clontech, Palo Alto, CA). This vector was used to identify Mcl-1S-interacting proteins by screening 1.5 million transformants from a GAL4 activation domain-tagged ovarian fusion cDNA library prepared from rats primed with equine chorionic gonadotropin (17). Yeast cells were cotransformed with pGBT9-Mcl-1S and cDNAs from the ovarian library, and colonies were selected in plates deficient in tryptophan, leucine, and histidine but containing 30 mM 3-amino-1,2,4-triazole (Sigma). Plasmids were isolated from positive colonies following transformation of Escherichia coli cells and then sequenced. Nine independent clones encoded the rat ortholog of human tankyrase 1. Full-length cDNA coding human tankyrase 1 was fused with the activation domain of GAL4 in a yeast shuttle vector, pGADGH. Subsequently, the specific interaction of tankyrase 1 with Mcl-1S was confirmed based on the activation of the GAL1-HIS3 reporter gene.
Assessment of Tankyrase 1 Interactions with Bcl-2 Family Proteins in the Yeast Two-hybrid System-Complementary DNAs encoding tankyrase 1 and various pro-or antiapoptotic Bcl-2 family proteins were subcloned into pGADGH and pGBT9 vectors, respectively. Specific binding of different protein pairs was evaluated based on the activation of the GAL1-HIS3 reporter gene. At least 10 different colonies were tested for each reaction.
In Vivo Binding of Tankyrase 1 with Mcl-1 Proteins-Chinese hamster ovary (CHO) cells (1.5 ϫ 10 6 ) were cultured in Dulbecco's modified Eagle's medium/F-12 supplemented with 10% fetal bovine serum, 100 units/ml penicillin, 100 g/ml streptomycin, and 2 mM glutamine. After 24 h of culture, cells were transfected with 1.8 g each of pcDNA3-FLAG-epitope-tagged Mcl-1L, Mcl-1S, or BAD alone or together with 4.2 g of either HA-epitope-tagged tankyrase 1 or an empty vector using LipofectAMINE (Invitrogen, Carlsbad, CA). CHO cells were harvested at 24 h post-transfection and lysed in prechilled 1% Nonidet P-40 lysis buffer containing 10% of a protease inhibitor mixture (Sigma). Following 30 min incubation in the lysis buffer, lysates were centrifuged at 10,000 ϫ g for 10 min at 4°C and supernatants were collected. Aliquots of the lysates were precleared by incubation with mouse IgG and Protein-A-agarose (Santa Cruz Biotechnology, Santa Cruz, CA) for 30 min at 4°C. The precleared lysates were then incubated with 1 g/ml anti-mouse FLAG M2 monoclonal antibody (Sigma) for 1 h followed by Protein-A-agarose for an additional 1 h at 4°C. The immune complexes were centrifuged for 5 min at 2,000 ϫ g and washed five times with 1% Nonidet P-40 lysis buffer.
The immunoprecipitates and aliquots of total lysates were boiled in SDS sample buffer for 5 min, subjected to SDS-PAGE, and electroblotted onto 0.22-m nitrocellulose membranes. The membranes were blocked in 5% nonfat dry milk in Tris-buffered saline solution with 1% Tween 20 for 1 h followed by incubation with 0.2 g/ml of the anti-HA monoclonal antibody (Sigma) for 1 h at room temperature. The blot was then incubated for 10 min with 0.1 g/ml anti-rabbit or mouse IgGhorseradish peroxidase conjugate (Promega, Madison, WI) as a secondary antibody before visualization by enhanced chemiluminescence (Amersham Biosciences). The same membrane was stripped and incubated with 0.2 g/ml mouse anti-FLAG M2 monoclonal antibody (Sigma), rabbit anti-Mcl-1 polyclonal antibody (Santa Cruz Biotechnology), and rabbit anti-BAD polyclonal antibody (Santa Cruz Biotechnology) for 1 h at room temperature. The blot was exposed to the same secondary antibody and visualized by enhanced chemiluminescence.
Binding of Endogenous Tankyrase 1 and Mcl-1L-Human myeloid leukemia cell line K562 (ATCC, Manassas, VA) was cultured in RPMI 1640 medium with 10% fetal bovine serum, 100 units/ml penicillin, 100 g/ml streptomycin, and 2 mM glutamine. Cells were lysed and precleared before incubation with 2 g/ml mouse anti-Mcl-1 monoclonal antibody (BD Pharmingen, San Diego, CA) or mouse IgG for 2 h, followed by incubation with Protein-A-agarose for 2 h at 4°C. The immune complexes were centrifuged for 5 min at 2,000 ϫ g and washed five times with 1% Nonidet P-40 lysis buffer. Immunoblotting was performed using either the anti-Mcl-1 polyclonal antibody or the antitankyrase 1 antibody.
Regions Responsible for Interactions between Mcl-1 and Tankyrase 1-To construct various truncated mutants of Mcl-1 and tankyrase 1, PCR amplifications were performed using Pfu DNA polymerase (Stratagene, La Jolla, CA). Truncated Mcl-1 and tankyrase 1 cDNAs were generated using primers as described in Table I. Wild type and truncated Mcl-1 cDNAs were subcloned into the pGBT9 vector whereas wild type and mutant tankyrase 1 cDNAs were inserted into the pGADGH vector. Specific interactions between these proteins were determined by the activation of the GAL1-HIS3 reporter gene.
Assessment of Apoptosis in Transfected CHO Cells and Down-regulation of Mcl-1 Proteins by Tankyrase 1-Apoptosis was monitored following transfection of different cDNAs as described previously (18). CHO cells (2 ϫ 10 5 cells/35-mm well) were cultured for 24 h and transfected with the pcDNA3 expression vector with or without different cDNA inserts using LipofectAMINE. For all experiments, the indicator plasmid pCMV␤ (Clontech) was included to allow the identification of transfected cells. Inclusion of a 10-fold excess of expression vectors as compared with the pCMV␤ reporter plasmid ensured that most of the ␤-galactosidase-expressing cells also expressed the protein(s) under investigation. After a 4-h incubation of cells with the transfection mixtures in a serum-free medium, cells were incubated with fresh medium containing 10% fetal bovine serum. Twenty-four hours after transfection, cells were fixed in 0.3% glutaraldehyde and stained with 0.4 mg/ml 5-bromo-4-chloro-3-indolyl-D-galactoside (Invitrogen) to detect ␤-galactosidase expression. The number of blue cells was counted by microscopic examination. Data were expressed as the percentage (mean Ϯ S.E.) of viable cells as compared with the control group.
To assess Mcl-1 protein levels following tankyrase 1 coexpression, CHO cells were transfected with different cDNAs and cell lysates were prepared at 24 h after transfection. Equal amounts of the lysates were subjected to SDS-PAGE. The same gel was immunoblotted with the anti-FLAG M2, the anti-tankyrase 1, and the anti-␤-actin monoclonal antibodies (Sigma).
Purification of Recombinant Mcl-1L, Mcl-1S, and Tankyrase 1 Proteins-N-terminal FLAG-tagged Mcl-1 proteins were expressed in E. coli (BL21 codon plus) using the pET-21a Vector (Novagen, Madison, WI) and cell lysates were prepared by a French presser. Recombinant Mcl-1 proteins were purified using an anti-FLAG M2 affinity gel and eluted using the 3ϫ FLAG peptide (Sigma). Purified Mcl-1 proteins were monitored using SDS-PAGE followed by Coomassie Blue staining. Baculovirus-derived tankyrase 1 and TRF1 were prepared as described (9).
In Vitro Interactions between Recombinant Tankyrase 1 and Mcl-1 Proteins-Tankyrase 1 (100 ng) was incubated with either 100 ng of Mcl-1L or Mcl-1S in 100 l of 1% Nonidet P-40 lysis buffer. After 1 h at 4°C, the mixtures were subjected to immunoprecipitation using the anti-FLAG M2 affinity gel. The immune complexes were centrifuged for 5 min at 2,000 ϫ g and washed five times with the lysis buffer and boiled in the SDS sample buffer. Immunoblotting was done using rabbit antibodies against tankyrase 1 or Mcl-1.

Specific Interactions of Tankyrase 1 with Mcl-1L and Mcl-1S but Not Other Bcl-2 Family Proteins in Yeast and Mammalian
Cells-We performed a yeast two-hybrid screening using fulllength Mcl-1S as bait to identify interacting proteins from the ovary cDNA library. Sequence analysis revealed that nine strongly interacting clones encode a rat ortholog of human tankyrase 1 with 330 amino acids truncated at the C terminus. The yeast two-hybrid system was used to determine the interactions of tankyrase 1 with Mcl-1 and other Bcl-2 family members. Human tankyrase 1 interacted strongly with both Mcl-1L and Mcl-1S (Fig. 1). In contrast, no interaction was detectable between tankyrase 1 and different Bcl-2 family proteins tested including antiapoptotic Bcl-2, Bcl-w, Bcl-x L , Bfl-1, Ced-9, and BHRF-1, as well as proapoptotic Bax, Bak, Bik, Bod-L, BAD, Diva, Nix, and Bok-L (Fig. 1).
In vivo interactions of tankyrase 1 with Mcl-1 proteins were confirmed in CHO cells transiently transfected with tankyrase 1 cDNA and Mcl-1 cDNAs. Cell lysate was precipitated using the M2 antibody against the FLAG epitope, followed by blotting using the HA antibodies. As shown in Fig. 2A (panel a, lanes 4  and 6), HA-tagged tankyrase 1 was coimmunoprecipitated with FLAG-tagged Mcl-1L or Mcl-1S. In contrast, tankyrase 1 failed to coprecipitate with FLAG-tagged BAD protein (Fig. 2A, panel  a, lane 2), consistent with a lack of interaction observed in the In vivo interactions between Mcl-1 and tankyrase 1 were further determined in CHO cells following immunoprecipitation of tankyrase 1. Mcl-1L was coprecipitated with HA-tagged tankyrase 1 using the HA antibody, followed by immunoblotting using the Mcl-1 antibody (Fig. 2B, panel a, lane 3). Because a nonspecific protein reacted to the Mcl-1 antibody and migrated to the same position as the Mcl-1S in the SDS-PAGE gel (data not shown), coprecipitation of Mcl-1S with tankyrase 1 could not be tested.
To test interactions between endogenous Mcl-1 and tankyrase 1 proteins, we used the human leukemia cell line K562. As shown in Fig. 2C (left panel), endogenous Mcl-1L in these cells was precipitated by an Mcl-1 antibody but not by nonimmune IgG. In addition, tankyrase 1 was found to be coprecipitated with Mcl-1L as evidenced by an immunoreactive band above 150 kDa (Fig. 2C, right panel). The detection of endogenous Mcl-1L and tankyrase 1 as complexes supports the physiological interactions between these two proteins.
Tankyrase 1 Interacts with the Mcl-1 Proteins through Its Ankyrin Domain-To determine the region of tankyrase 1 responsible for interaction with Mcl-1 proteins, truncated mutants of tankyrase 1 (T1 to T10) were generated and their binding to Mcl-1L and Mcl-1S was tested in the yeast twohybrid system. Tankyrase 1 protein has four major domains (Fig. 3) including the HPS sequence, 24 ankyrin repeats, SAM motif, and the catalytic domain of PARP (9). The two Mcl-1 proteins did not interact with the HPS (T1), SAM (T2), or PARP (T2) motif of tankyrase 1, whereas the ankyrin domain of tankyrase 1 alone (T3) exhibited a strong interaction to both Mcl-1 proteins (Fig. 3). We further truncated the ankyrin repeats of tankyrase 1 to determine the minimal region responsible for binding to Mcl-1. As shown in Fig. 3, the region spanning from amino acid 181 to 586 of tankyrase 1 (T5), containing 12 ankyrin repeats exhibited the same degree of interaction as wild type tankyrase 1. Additional truncation of  (Fig. 4A, M1). In addition, the BH3 domain, the only conserved region found in all Bcl-2 family members, was not involved in tankyrase 1 binding (M2). Deletion of either one or both of the known PEST sequences of Mcl-1 (M4 and M5) also did not alter tankyrase 1 binding; the N-terminal region of Mcl-1 interacted with tankyrase 1 as strongly as the full-length Mcl-1 (M5). However, truncation of the N-terminal region of Mcl-1 from amino acid 76 to 100 (M7) abolished its interaction with tankyrase 1 (Fig. 4B). The importance of this region was further supported by a lack of  (Fig. 6, panels A and B). In contrast, tankyrase 1 did not alter the expression of BAD. Furthermore, mutant tankyrase 1 devoid of the PARP domain (mutant Tanky) also effectively decreased the expression of both Mcl-1L and Mcl-1S. These data indicate that tankyrase 1-mediated Mcl-1 down-regulation does not require the PARP enzymatic activity of tankyrase 1. Equal loading of samples was demonstrated by Western blotting of the same membrane using the ␤-actin antibody (Fig. 6, panel C), and the nonspecific 30-kDa band that reacted with the FLAG antibody served as an internal control (Fig. 6, panel A). Several lower bands found in lanes containing either wild type (lanes 3-5, 7-9, and 11-13) or  Fig. 7A  (lower panel, lanes 2 and 3), the recombinant Mcl-1 proteins efficiently pulled down tankyrase 1, indicating that the purified Mcl-1 proteins are functionally active. The purity of the Mcl-1 proteins used in this test were analyzed by SDS-PAGE followed by Coomassie Blue staining (Fig. 7B).
Tankyrase 1 contains the catalytic domain of PARP, and ADP-ribosylates itself (9) and its interacting proteins such as TRF1 (9), IRAP (11), and TAB182 (12). However, our studies using PARP-deleted tankyrase 1 suggest that ADP-ribosylation is not essential for the modulation of Mcl-1 activities and levels. We further tested whether tankyrase 1 could ADPribosylate Mcl-1 proteins. An in vitro poly(ADP-ribosyl)ation assay was performed using purified recombinant Mcl-1 proteins and tankyrase 1. Incubation of tankyrase 1 in the presence of [ 32 P]NAD ϩ resulted in auto-ADP-ribosylation (Fig. 8A, lane 1) consistent with previous findings (9). Coincubation of increasing amounts of either Mcl-1L or Mcl-1S with tankyrase 1 did not lead to ADP-ribosylation of Mcl-1L or Mcl-1S (Fig. 8A,  lanes 5-10). In contrast, tankyrase 1 ADP-ribosylated its known substrate, TRF1, in a concentration-dependent manner (Fig. 8A, lanes 2-4)  in a concentration-dependent manner. In addition, auto-ADPribosylation of tankyrase 1 was also decreased in the presence of increasing amounts of Mcl-1 protein (Fig. 8B; shorter exposure). These data suggest that tankyrase 1-mediated ADPribosylation of Mcl-1 is unlikely to be the mechanism underlying the down-regulation of the Mcl-1 proteins, and the Mcl-1 proteins could affect tankyrase 1 function by modulating its ADP-ribosylation of itself and its binding protein, TRF1.

DISCUSSION
Mcl-1, a unique Bcl-2 family protein, is transiently induced by various growth factors and shows a rapid turnover rate (20,21). Earlier studies demonstrated that Mcl-1 expression is upregulated by epidermal growth factor, elk-1, granulocyte-macrophage colony-stimulating factor, gonadotropins, stem cell factor, interferon-␣, and different interleukins in diverse cell lineages (22)(23)(24)(25)(26). Despite numerous studies on the induction of Mcl-1, no information is available regarding the suppression of the rapidly induced Mcl-1 proteins. Based on yeast twohybrid screening, we identified an Mcl-1-interacting protein, tankyrase 1, that could provide a novel mechanism to decrease the levels and to suppress the actions of both Mcl-1L and its splicing variant Mcl-1S.
We demonstrated that overexpression of tankyrase 1 suppressed cell survival induced by Mcl-1L. In addition, tankyrase 1 also effectively blocked apoptosis induced by Mcl-1S. In contrast, overexpression of tankyrase 1 itself exhibited no significant effect on CHO cell viability, consistent with a recent observation (34).  More than 20 known mammalian Bcl-2 family proteins share common regulatory mechanisms to modulate apoptosis. They dimerize with other Bcl-2 family members (1) or interact with specific interacting proteins. For instance, BAD binding to 14-3-3 proteins suppresses the apoptotic activity of BAD (18,27), whereas Bim/BOD binding to the LC8 dynein light chain leads to its sequestration to the microtubule-associated dynein motor complex (28). Likewise, Bmf interacts with the dynein light chain 2 before being sequestered to the myosin V actin motor complex (29), leading to apoptosis suppression. Here, we have demonstrated that both Mcl-1L and Mcl-1S are strongly associated with tankyrase 1 in yeast and mammalian cells. Interactions of tankyrase 1 with Mcl-1 proteins were specific because none of the other members of the Bcl-2 family that were tested bound to tankyrase 1.
The antiapoptotic Mcl-1L protein consists of two consensus PEST sequences as well as the BH3, BH1, BH2, and a transmembrane domain. However, the proapoptotic splicing variant, Mcl-1S, possesses only the PEST sequences and the BH3 domain. The common N-terminal end of both Mcl-1 proteins contain the two PEST motifs and four pairs of arginine residues found in rapidly degraded proteins such as c-Myc and p53 (30). Although the deletion of these two PEST sequences in Mcl-1 does not affect the half-life of the mutant protein (20), the extreme N-terminal end of Mcl-1 has a stretch of residues with a weak PEST homology. We demonstrated that the binding of Mcl-1 proteins to tankyrase 1 involves this short stretch of 25 amino acids (76 to 100). Of interest, the same region also contains a RPPPIG sequence that resembles the consensus tankyrase-binding motif (RXXPDG) found in other tankyrase 1-interacting proteins (31).
Tankyrase 1 possesses four distinct motifs: the HPS module, 24 ankyrin motifs, SAM, and the catalytic domain of PARP (9). Tankyrase 1 is a unique protein with structural features found in both ankyrin and PARP family genes. Ankyrin family proteins are linkers that couple diverse membrane proteins via ankyrin motifs to the underlying cytoskeleton (32, 33). All known tankyrase 1-interacting proteins (TRF1, IRAP, and TAB182) bind to tankyrase 1 through ankyrin repeat regions (9,11,12). Likewise, our data indicated that interactions between Mcl-1 proteins and tankyrase 1 also involved the ankyrin motifs, and the first half of the ankyrin repeats (1 to 12) was sufficient for dimerization. Of particular interest, TRF1 and Mcl-1 are capable of binding to overlapping ankyrin repeats in tankyrase 1, suggesting potential competition between these tankyrase 1-interacting proteins.
Tankyrase 1 contains the catalytic domain of PARP and is known to poly(ADP-ribosyl)ate its interacting proteins, TRF1, IRAP, and TAB182 (9,11,12). PARP family proteins catalyze the attachment of the poly(ADP-ribose) moiety onto a protein acceptor using the substrate NAD ϩ (nicotinamide adenine dinucleotide), and ADP-ribosylation of proteins usually leads to protein inactivation (13)(14)(15)(16). However, our study demonstrated that tankyrase 1 does not poly(ADP-ribosyl)ate either of the Mcl-1 proteins. Furthermore, a tankyrase 1 mutant with the PARP domain deleted is still capable of decreasing the Mcl-1 protein levels and blocking the actions of the Mcl-1 proteins. These data suggest that the ability of tankyrase 1 to suppress both anti-and pro-apoptotic actions of the Mcl-1 proteins is not mediated by ADP-ribosylation of these proteins. The findings also suggest that tankyrase 1-interacting proteins are not always the substrates for tankyrase 1. Consistent with the unique structural features and subcellular localization of tankyrase 1, the current study suggests that tankyrase 1 is different from other PARP enzymes (9, 11, 35, 36) and could have nonenzymatic functions. Of interest, tankyrase 2, a gene with 83% sequence identity to tankyrase 1, has recently been isolated (37,38) and found to induce cell death in different cell lines (34).
In addition to the tankyrase 1 regulation of Mcl-1 functions, Mcl-1 proteins could reciprocally regulate the function of tankyrase 1. When Mcl-1L or Mcl-1S are coincubated with tankyrase 1, auto-ADP-ribosylation of tankyrase 1 decreases in a Mcl-1 concentration-dependent manner. Although the physiological significance of the apparent decrease in poly(ADPribosyl)ation of tankyrase 1 is still unclear and one cannot completely rule out the nonspecific effects of unknown contaminants in the purified Mcl-1 preparation, Mcl-1 proteins caused a more profound suppression of the poly(ADP-ribosyl)ation of TRF1 in the same in vitro test. Because both Mcl-1 proteins and TRF1 interact with the ankyrin repeats in tankyrase 1, Mcl-1 proteins could compete with TRF1 for binding to tankyrase 1, leading to lower ADP-ribosylation of TRF1. Because tankyrase 1-mediated telomere extension is dependent on ADP-ribosylation of TRF1 (10), inhibition of the ADP-ribosylation of TRF1 by Mcl-1 could prevent telomere elongation and facilitate cell senescence. This is consistent with the ability of tankyrase 1 to suppress the survival action of Mcl-1L, thus leading to the prevention of cell immortalization.
Recent studies indicated that Mcl-1L is localized to the nucleus (39,40). and is involved in cell cycle regulation by interacting with the proliferating cell nuclear antigen (39). Mcl-1 also serves as a nuclear chaperone for another antiapoptotic protein, fortilin (40). Together with our observation of interactions between Mcl-1 and tankyrase 1, these findings suggest an important role of Mcl-1 in the regulation of diverse nuclear events including cell cycle regulation and telomere elongation. In conclusion, the observed interactions between Mcl-1 proteins and tankyrase 1 represent a novel mechanism for the regulation of the apoptosis function of Mcl-1 proteins. Because Mcl-1 could also regulate the function of tankyrase 1 and other nuclear proteins, Mcl-1 could play an important role in the coordinated regulation of cell survival, proliferation, and immortalization.